10.07.2015 Views

ACCP Pulmonary Medicine Board Review: 25th Edition - Joan C ...

ACCP Pulmonary Medicine Board Review: 25th Edition - Joan C ...

ACCP Pulmonary Medicine Board Review: 25th Edition - Joan C ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>ACCP</strong> <strong>Pulmonary</strong><strong>Medicine</strong><strong>Board</strong> <strong>Review</strong>:<strong>25th</strong> <strong>Edition</strong>The American <strong>Board</strong> of Internal <strong>Medicine</strong> (ABIM) is not affiliated with, nor does itendorse, preparatory examination review programs or other continuing medical education.The content of the <strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> isdeveloped independently by the American College of Chest Physicians (<strong>ACCP</strong>), whichhas no knowledge of or access to ABIM examination material.The views expressed herein are those of the authors and do not necessarily reflect theviews of the <strong>ACCP</strong>. Use of trade names or names of commercial sources is for informationonly and does not imply endorsement by the <strong>ACCP</strong>. The authors and the publisherhave exercised great care to ensure that drug dosages, formulas, and other informationpresented in this book are accurate and in accord with the professional standardsin effect at the time of publication. However, readers are advised to always check themanufacturer’s product information sheet packaged with the respective products to befully informed of changes in recommended dosages, contraindications, etc., before prescribingor administering any drug.ACC-PULM-09-0302-0FM.indd i7/11/09 1:23:06 AM


Copyright © 2009 by the AMERICAN COLLEGE OF CHEST PHYSICIANS.Copyright not claimed on material authored by the US Government. All rights reserved.No part of this book may be reproduced in any manner without permission of the publisher.Published by theAmerican College of Chest Physicians3300 Dundee RoadNorthbrook, IL 60062-2348Telephone: (847) 498-1400; Fax: (847) 498-5460<strong>ACCP</strong> Web site: www.chestnet.orgPrinted in the United States of AmericaFirst PrintingISBN 978-0-916609-77-1ACC-PULM-09-0302-0FM.indd ii7/11/09 1:23:06 AM


ContentsUnusual Lung Infection, Bronchiectasis, and Cystic Fibrosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1Lisa K. Moores, MD, FCCP<strong>Pulmonary</strong> Vascular Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Lisa K. Moores, MD, FCCPLung Cancer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39W. Michael Alberts, MD, MBA, FCCPHypersensitivity Pneumonitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63W. Michael Alberts, MD, MBA, FCCPEosinophilic Lung Diseases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73W. Michael Alberts, MD, MBA, FCCPAsthma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Sidney S. Braman, MD, FCCPCardiopulmonary Exercise Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113Darcy D. Marciniuk, MD, FCCPHypercapnic Respiratory Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129Mark J. Rosen, MD, FCCPThoracic Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137Rakesh D. Shah, MD, FCCPEthics in <strong>Pulmonary</strong> and Critical Care <strong>Medicine</strong> . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147Mark J. Rosen, MD, FCCPChronic Obstructive <strong>Pulmonary</strong> Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153Sidney S. Braman, MD, FCCP<strong>Pulmonary</strong> Function Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187Darcy D. Marciniuk, MD, FCCPBronchoscopy and Interventional Pulmonology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205David Feller-Kopman, MD, FCCP<strong>Pulmonary</strong> Complications of HIV Infection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219Mark J. Rosen, MD, FCCP<strong>Pulmonary</strong> Complications of Cardiothoracic Surgery and Trauma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233Bruce P. Krieger, MD, FCCPSleep Physiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245Teofilo L. Lee-Chiong, Jr., MD, FCCPSleep-Related Breathing Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253Teofilo L. Lee-Chiong, Jr., MD, FCCPNonrespiratory Sleep Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261Teofilo L. Lee-Chiong, Jr., MD, FCCP<strong>Pulmonary</strong> Fungal Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273George A. Sarosi, MD, FCCP<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong>iiiACC-PULM-09-0302-0FM.indd iii7/11/09 1:23:06 AM


Other <strong>Pulmonary</strong> Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285Mark J. Rosen, MD, FCCPMedical Statistics/Test-Taking Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295David W. Kamp, MD, FCCPAcid-Base Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309Janice L. Zimmerman, MD, FCCPDrug-Induced Lung Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319David W. Kamp, MD, FCCPHemodynamic Monitoring and Shock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345Janice L. Zimmerman, MD, FCCPCommunity-Acquired Pneumonia: Advances in Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359Ronald R. Grossman, MD, FCCPPneumoconiosis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369David W. Kamp, MD, FCCPHospital-Acquired and Ventilator-Associated Pneumonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391Ronald R. Grossman, MD, FCCPHypoxemic Respiratory Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401Curtis N. Sessler, MD, FCCPSymptoms of Respiratory Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415Richard S. Irwin, MD, FCCPMechanical Ventilatory Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457Curtis N. Sessler, MD, FCCPUnusual and Uncommon <strong>Pulmonary</strong> Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479Jay H. Ryu, MD, FCCPMediastinal and Other Neoplasms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487Jay H. Ryu, MD, FCCPPathology of Airway Disease and Organizing Pneumonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497Henry D. Tazelaar, MD, FCCPPathology of Diffuse and Neoplastic Disease. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501Henry D. Tazelaar, MD, FCCPPleural Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513Steven A. Sahn, MD, FCCPPleural Pearls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 547Steven A. Sahn, MD, FCCPLung Transplantation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563Stephanie M. Levine, MD, FCCPRare Interstitial Lung Diseases: <strong>Pulmonary</strong> Langerhans Cell Histiocytosis,Lymphangioleiomyomatosis, and Cryptogenic Organizing Pneumonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585Joseph P. Lynch III, MD, FCCPTuberculosis and Other Mycobacterial Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619David Ashkin, MDivContentsACC-PULM-09-0302-0FM.indd iv7/11/09 1:23:06 AM


Idiopathic <strong>Pulmonary</strong> Fibrosis, Nonspecific Interstitial Pneumonia/Fibrosis, and Sarcoidosis . . . . . . . . . . . . . . . 635Joseph P. Lynch III, MD, FCCP<strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage Syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687Ulrich Specks, MDWomen’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong> . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705Stephanie M. Levine, MD, FCCPOccupational Asthma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 723William S. Beckett, MD, FCCPDiseases Related to High Altitude and Diving and Near-Drowning: Basics of Hyperbaric <strong>Medicine</strong> . . . . . . . . . . 729Igor Aksenov, MD, PhD; and Michael Strauss, MD<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong>vACC-PULM-09-0302-0FM.indd v7/11/09 1:23:07 AM


AuthorsIgor Aksenov, MD, PhD<strong>Pulmonary</strong> and Critical Care <strong>Medicine</strong> FellowDivision of <strong>Pulmonary</strong>, Critical Care andSleep <strong>Medicine</strong>University of FloridaGainesville, FLW. Michael Alberts, MD, MBA, FCCPChief Medical OfficerH. Lee Moffitt Cancer CenterProfessor of Oncology and <strong>Medicine</strong>Department of Interdisciplinary OncologyUniversity of South Florida College of<strong>Medicine</strong>Tampa, FLDavid Ashkin, MDMedical Executive Director, A. G. HolleyState TB HospitalFlorida State TB Health Officer, FloridaDepartment of HealthAdjunct Assistant Professor, Division of<strong>Pulmonary</strong> and Critical Care <strong>Medicine</strong>University of Florida College of<strong>Medicine</strong>Gainesville, FLVisiting Associate Professor, Division of<strong>Pulmonary</strong> and Critical Care<strong>Medicine</strong>University of Miami School of <strong>Medicine</strong>Miami, FLWilliam S. Beckett, MD, FCCPAssociate Professor of <strong>Medicine</strong>Harvard Medical SchoolPrimary Care CenterMount Auburn HospitalCambridge, MASidney S. Braman, MD, FCCPProfessor of <strong>Medicine</strong>Director, Division of <strong>Pulmonary</strong> andCritical Care <strong>Medicine</strong>The Alpert Medical School of BrownUniversity and Rhode Island HospitalProvidence, RIDavid Feller-Kopman, MD, FCCPDirector, Interventional PulmonologyAssociate Professor of <strong>Medicine</strong>Johns Hopkins HospitalBaltimore, MDRonald F. Grossman, MD, FCCPStaff RespirologistUniversity of TorontoChief of <strong>Medicine</strong>Credit Valley HospitalMississauga, ON, CanadaRichard S. Irwin, MD, FCCPProfessor of <strong>Medicine</strong> and NursingUniversity of MassachusettsChair, Critical CareUMass Memorial Medical CenterEditor in Chief, CHESTWorcester, MAviDavid W. Kamp, MD, FCCPProfessor of <strong>Medicine</strong>Associate Chief, Division of <strong>Pulmonary</strong> andCritical Care <strong>Medicine</strong>Northwestern University FeinbergSchool of <strong>Medicine</strong>Chicago, ILBruce P. Krieger, MD, FCCPProfessor of <strong>Medicine</strong>Miller School of <strong>Medicine</strong>University of MiamiAssociate Medical DirectorCritical Care CenterMemorial HospitalJacksonville, FLTeofilo L. Lee-Chiong, Jr., MD, FCCPAssociate Professor of <strong>Medicine</strong>Head, Division of Sleep <strong>Medicine</strong>Department of <strong>Medicine</strong>National Jewish HealthAssociate Professor of <strong>Medicine</strong>University of Colorado Denver Schoolof <strong>Medicine</strong>Denver, COStephanie M. Levine, MD, FCCPProfessor of <strong>Medicine</strong>Division of <strong>Pulmonary</strong> Diseases andCritical Care <strong>Medicine</strong><strong>Pulmonary</strong> and Critical Care FellowshipProgram DirectorUniversity of Texas Health Science CenterSan Antonio, TXJoseph P. Lynch III, MD, FCCPProfessor of Clinical <strong>Medicine</strong>Associate Chief<strong>Pulmonary</strong>, Critical Care <strong>Medicine</strong>, andHospitalistsDavid Geffen School of <strong>Medicine</strong> atUCLALos Angeles, CADarcy D. Marciniuk, MD, FCCPProfessor and Vice-Chair, Department of<strong>Medicine</strong>University of Saskatchewan, RoyalUniversity Hospital,Saskatoon, SK, CanadaLisa K. Moores, MD, FCCPAssistant Dean for Clinical SciencesProfessor of <strong>Medicine</strong>The Uniformed Services University ofthe Health SciencesBethesda, MDMark J. Rosen, MD, FCCPChief, Division of <strong>Pulmonary</strong>, Critical Care,and Sleep <strong>Medicine</strong>North Shore University Hospital andLong Island Jewish Medical CenterProfessor of <strong>Medicine</strong>Albert Einstein College of <strong>Medicine</strong>New Hyde Park, NYJay H. Ryu, MD, FCCPProfessor of <strong>Medicine</strong>Mayo Clinic College of <strong>Medicine</strong>Consultant, Division of <strong>Pulmonary</strong> andCritical Care <strong>Medicine</strong>Director, Interstitial Lung Disease ClinicMayo ClinicRochester, MNSteven A. Sahn, MD, FCCPProfessor of <strong>Medicine</strong>Director, Division of <strong>Pulmonary</strong>, CriticalCare <strong>Medicine</strong>,Allergy and Sleep <strong>Medicine</strong>Medical University of South CarolinaCharleston, SCGeorge A. Sarosi, MD, FCCPProfessor of <strong>Medicine</strong>University of MinnesotaStaff PhysicianMinneapolis VA Medical CenterMinneapolis, MNCurtis N. Sessler, MD, FCCPOrhan Muren Professor of <strong>Medicine</strong>Division of <strong>Pulmonary</strong> and Critical Care<strong>Medicine</strong>Virginia Commonwealth UniversityHealth SystemMedical Director of Critical CareMedical College of Virginia HospitalsRichmond, VARakesh D. Shah, MD, FCCPChief of Thoracic RadiologyDepartment of RadiologyNorth Shore University HospitalManhasset, NYUlrich Specks, MDProfessor of <strong>Medicine</strong>Division of <strong>Pulmonary</strong> and Critical Care<strong>Medicine</strong>Mayo ClinicRochester, MNMichael Strauss, MDDirectorDepartment of Hyperbaric <strong>Medicine</strong>Long Beach Memorial Medical CenterLong Beach, CAHenry D. Tazelaar, MD, FCCPConsultantMayo Clinic ArizonaProfessor of PathologyMayo Clinic College of <strong>Medicine</strong>Scottsdale, AZJanice L. Zimmerman, MD, FCCPProfessor of Clinical <strong>Medicine</strong>Weill Cornell Medical CollegeHead, Critical Care DivisionDepartment of <strong>Medicine</strong>Director of MICUThe Methodist HospitalHouston, TXACC-PULM-09-0302-0FM.indd Sec1:vi7/11/09 1:23:07 AM


DISCLOSURE OF AUTHORS’ CONFLICTS OF INTERESTThe American College of Chest Physicians (<strong>ACCP</strong>) remains strongly committed to providing the best available evidence-basedclinical information to participants of this educational activity and requires an open disclosure of any potential conflict ofinterest identified by our committee members. It is not the intent of the <strong>ACCP</strong> to eliminate all situations of potential conflict ofinterest, but rather to enable those who are working with the <strong>ACCP</strong> to recognize situations that may be subject to question byothers. All disclosed conflicts of interest are reviewed by the educational activity course director/chair, the Education Committee,or the Conflict of Interest <strong>Review</strong> Committee to ensure that such situations are properly evaluated and, if necessary,resolved. The <strong>ACCP</strong> educational standards pertaining to conflict of interest are intended to maintain the professional autonomyof the clinical experts inherent in promoting a balanced presentation of science. Through our review process, all <strong>ACCP</strong>CME activities are ensured of independent, objective, scientifically balanced information. Disclosure of any or no relationshipsis made available for all educational activities.The following authors of the <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> have disclosed to the <strong>ACCP</strong> that a relationshipdoes exist with the respective company/organization as it relates to their presentation of material and should be communicatedto the participants of this educational activity:AuthorsRonald Grossman, MD, FCCPTeofilo L. Lee-Chiong, Jr., MD, FCCPDarcy D. Marciniuk, MD, FCCPUlrich Specks, MDRelationshipAdvisory committee: GSK, Bayer, Ortho McNeil, Sanofi Aventis, WyethGrant Monies (from sources other than industry): NIHGrant Monies (from industry-related sources): Respironics, Restore,Takeda; Consultant fee, speaker bureau, advisory committee, etc:Consultant for Sleep <strong>Medicine</strong> Clinics (Elsevier)Grant Monies (from sources other than industry): Canadian Agency forDrugs and Technologies in Health, Canadian Institutes of Health Research,Lung Association of Saskatchewan, Public Health Agency of Canada, RoyalUniversity Hospital Foundation, Saskatoon Health Region, SaskatchewanMinistry of HealthGrant Monies (from industry-related sources): AstraZeneca, Boehringer-Ingelheim, GlaxoSmithKline, Forest Research, Novartis, PfizerFiduciary Position (of any other organization, association, society, etc,other than <strong>ACCP</strong>): Canadian Thoracic SocietyConsultant fee, speaker bureau, advisory committee, etc: AstraZeneca,Boehringer-Ingelheim, GlaxoSmithKline, Novartis, Nycomed, PfizerProduct/procedure/technique that is considered research and is NOT yetapproved for any purpose: All medications/interventions used for thetreatment of ANCA-associated vasculitis, including cyclophosphamide,methotrexate, azathioprine, mycophenolate mofetil, lefluomide, infliximab,etanercept, abatacept, rituximab, and plasma-exchange represent off-labeluse for this indication.The following authors of the <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> have indicated to the <strong>ACCP</strong> that no potentialconflict of interest exists with any respective company/organization, and this should be communicated to the participants ofthis educational activity:W. Michael Alberts, MD, FCCPDavid Ashkin, MDWilliam Beckett, MD, FCCPSidney Braman, MD, FCCPRichard Irwin, MD, FCCPDavid Kamp, MD, FCCPJoseph P. Lynch III, MD, FCCPLisa Moores, MD, FCCPMark Rosen, MD, FCCPJay Ryu, MD, FCCPSteven Sahn, MD, FCCPGeorge Sarosi, MD, FCCPCurtis Sessler, MD, FCCPUlrich Specks, MDHenry Tazelaar, MD, FCCPJanice Zimmerman, MD, FCCP<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong>viiACC-PULM-09-0302-0FM.indd Sec1:vii7/11/09 1:23:07 AM


Needs AssessmentRely on the <strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> to review the type of information you should know forthe <strong>Pulmonary</strong> Disease Subspecialty <strong>Board</strong> Examination of the American <strong>Board</strong> of Internal <strong>Medicine</strong> (ABIM). Designed as thebest preparation for anyone taking the exam, this comprehensive, exam-focused review will cover current critical care literatureand management strategies for critically ill patients.The ABIM <strong>Pulmonary</strong> Disease Subspecialty <strong>Board</strong> Examination tests knowledge and clinical judgment in crucial areas of criticalcare medicine. This premier course will review the information you should know for the exam. Course content mirrors thecontent of the exam, as outlined by the ABIM, and includes the following topics:Asthma 7%Cell/Molecular biology 3.5%Congenital/neuromuscular/Skeletal 2%COPD 7%Critical care. Non-lung 4%Critical care, lung 9%ILD-related disorders 6%Infections 13%Neoplasms 9%Obstructive airways disease, other 2%Occupational and environmental disease 5%Physiology and metabolism 4%Pleural disease 5%Quality, safety and complications 5%Sleep, Non-respiratory 2%Sleep, Respiratory 8%Transplantation 2%Vascular diseases 2.5%Total 100%Target AudiencePhysicians in critical care pulmonary medicineFellows in critical care and pulmonary medicineAdvanced critical care nurse practitionersAdvanced respiratory therapy practitionersPhysician assistantsGeneral Publications DisclaimerThe American College of Chest Physicians (“<strong>ACCP</strong>”) and its officers, regents, executive committee members, members, relatedentities, employees, representatives and other agents (collectively, “<strong>ACCP</strong> Parties”) are not responsible in any capacity for, donot warrant and expressly disclaim all liability for, any content whatsoever in any <strong>ACCP</strong> publication or other product (in anymedium) and the use or reliance on any such content, all such responsibility being solely that of the authors or the advertisers,as the case may be. By way of example, without limiting the foregoing, this disclaimer of liability applies to the accuracy,completeness, effectiveness, quality, appearance, ideas, or products, as the case may be, of or resulting from any statements,references, articles, positions, claimed diagnosis, claimed possible treatments, services, or advertising, express or implied,contained in any <strong>ACCP</strong> publication or other product. Furthermore, the content should not be considered medical advice andis not intended to replace consultation with a qualified medical professional. Under no circumstances, including negligence,shall any of the <strong>ACCP</strong> Parties be liable for any DIRECT, INDIRECT, INCIDENTAL, SPECIAL or CONSEQUENTIAL DAM-AGES, or LOST PROFITS that result from any of the foregoing, regardless of legal theory and whether or not claimant wasadvised of the possibility of such damages.viiiACC-PULM-09-0302-0FM.indd Sec1:viii7/11/09 1:23:07 AM


<strong>ACCP</strong> Member BenefitsSo Many Reasons To Join.Find Yours Today.Communications• CHEST, for specialists in pulmonology, critical care,sleep medicine, thoracic surgery, cardiorespiratoryinteractions, and related disciplinesNamed one of the 100 most influentialjournals over the last 100 years in medicineand biology.Available at www.chestjournal.org• CHEST Physician, the <strong>ACCP</strong> monthly newspublication, featuring current chest medicinenews from around the globe, plus updates on<strong>ACCP</strong> matters and events.Health-care Advocacy• A unified voice to policymakers on medicaland payment issues• Access to electronic tools for contactingCongress about issues affecting your patients,practice, and profession• Timely alerts on legislation that impact thepractice of medicinePractice Resources• Coding and reimbursement education• Business of medicine resourcesEducational Resources• Discounted tuition for all CME courses andeducational products<strong>Board</strong> review courses and preparation materialsHands-on clinical learning at the <strong>ACCP</strong>Simulation Center for Advanced ClinicalEducationSelf-study tools, including <strong>ACCP</strong>-SEEK• Discounted tuition for the annual CHEST meeting,offering essential updates in pulmonary, criticalcare, and sleep medicineCHEST benefits for <strong>ACCP</strong> members only:Free abstract submissionFree topic proposal submission• Interactive online resources, including freeCME via <strong>Pulmonary</strong> and Critical Care Update(PCCU) articles• Evidence-based practice guidelines and clinicalresources, outlining new protocols in chestmedicine• Patient education tools and teaching materials• Tools for making presentations to communitiesabout lung health and smoking• Patient education tobacco cessation products• <strong>ACCP</strong> Career Connection (online career service)Join the <strong>ACCP</strong> TodayLearn more about membership and apply online.www.chestnet.org/membership/join(800) 343-2227 or (847) 498-1400ACC-PULM-09-0302-0FM.indd Sec1:ix7/11/09 1:23:08 AM


Point. Click. Access.Link to the resources you needwww.chestnet.orgParticipation Opportunities• <strong>ACCP</strong> NetWorks—special interest groups withinthe <strong>ACCP</strong> that focus on particular areas of chestmedicine• The <strong>ACCP</strong> Critical Care and Sleep Institutes arecenters of excellence merging all programsand resources into central organizational units,providing the <strong>ACCP</strong> with a strong voice in thefuture of critical care and sleep medicine.• Committee and leadership positions• CHEST Foundation awards in clinical research,leadership in end-of-life care, and humanitarianservice• The CHEST Foundation’s Ambassador’s GroupFinancial Benefits• Discounts for <strong>ACCP</strong> courses and products• Credit card programs• Discounts for Apple and Dell computersand products.® ®Education• Calendar of upcoming courses, includingthe annual CHEST meeting and hands-oneducation opportunities in the <strong>ACCP</strong>Simulation Center for Advanced ClinicalEducation• Online education opportunities andinteractive resources, including A Physician’sPerspective®• Self-study products• Evidence-based clinical practice guidelinesMembership and NetWorks• Membership information and applications• <strong>ACCP</strong> NetWorks and their activities• Join NetWorks onlineOther Resources• Downloads of consensus statements,<strong>ACCP</strong> publications, podcasts, and more• Practice management information• Access to health-care advocacy updateson socioeconomic and political issuesaffecting practice and research• Career Connection employment services• <strong>ACCP</strong> product catalogwww.chestfoundation.org• Award and grant information andapplications• Professional resources• Patient education productswww.chestjournal.org• Full articles online• Numerous tools and features, includingInteractive Physiology Grand Rounds• Article submission information• Subscription informationACC-PULM-09-0302-0FM.indd Sec1:x7/11/09 1:23:08 AM


The CHEST Foundation is the philanthropicarm of the American College of Chest Physicians(<strong>ACCP</strong>), a 17,500-member international medicalspecialty society.The CHEST Foundation mission is to provideresources to advance the prevention and treatmentof diseases of the chest. In order to fulfill its missionto advance patient care in cardiopulmonary andcritical care medicine, The CHEST Foundation hastargeted the following four focus areas:• Tobacco Prevention• Humanitarian Service• Clinical Research• Critical Care/End-of-Life CareTobacco Prevention Education• The 4th <strong>Edition</strong> CD-ROM: Make the Choice:Tobacco or Health? Speaker’s Kit for presentationsto health professionals and patients• Lung Lessons SM curriculum teaches elementaryschool children the negative health effects ofsmoking• Lung Lessons SM : A Presenter’s Guide DVDdemonstrates how to teach the Lung Lessons SMcurriculum to children• Evils of Tobacco CD-ROM and video for childrenand women in IndiaCritical Care and End-of-Life Care• The Critical Care Family Assistance Program andreplication tool kit to improve coordination of careand communication with ICU staff, patients, andfamilies• ICU – Frequently Asked Questions in the ICU booklet,includes responses to questions that familymembers have when a family member is admittedto a hospital ICU• Stories at the End of Life booklet series to comfortpatients and their family membersHumanitarian AwardsNearly $1.4 million awarded from 1998 to 2009to recognize and support volunteer servicein over 180 projects/services of <strong>ACCP</strong> membersworldwideACC-PULM-09-0302-0FM.indd Sec1:xi7/11/09 1:23:08 AM


Clinical Research Awards• Over $5 million conferred from 1998 to 2009 tosupport promising clinical research• Distinguished Scholar awards to foster innovationin clinical care to address public health related tochest and critical care medicineEli Lilly and Company Distinguished Scholarin Critical Care <strong>Medicine</strong>GlaxoSmithKline Distinguished Scholarin Respiratory HealthGlaxoSmithKline Distinguished Scholarin Thrombosis• Clinical research awards in asthma, COPD,critical care, pulmonary fibrosis, thrombosis,and women’s health• The CHEST Foundation and the LUNGevityFoundation Clinical Research Award in LungCancer• The CHEST Foundation California Chapter ClinicalResearch/Medical Education Award• Scientific abstract-related awardsYou can support these programs and projectsthrough a tax-deductible donation to The CHESTFoundation. Your gift will help support theseprograms and resources that help you help yourpatients live and breathe easier.Donate online atwww.chestfoundation.org.• Roger C. Bone Award for Advances inEnd-of-Life Care• The CHEST Foundation and <strong>ACCP</strong> Grantin Venous Thromboembolism• Alpha-1 Foundation and The CHEST FoundationClinical Research Award in COPDand Alpha-1 Antitrypsin Deficiency• The American Lung Association and CHESTFoundation Clinical Investigator Award• The American Society of Transplantation andThe CHEST Foundation Clinical Research Awardin Lung Transplantation• Association of Specialty Professors and The CHESTFoundation of the <strong>ACCP</strong> Geriatric DevelopmentResearch AwardACC-PULM-09-0302-0FM.indd Sec1:xii7/11/09 1:23:09 AM


Unusual Lung Infection, Bronchiectasis, andCystic FibrosisLisa K. Moores, MD, FCCPObjectives:• Compare actinomycosis and nocardiosis infections of thelung• Discuss the treatment of actinomycosis and nocardiosisinfection in the lung• List the causes of bronchiectasis• Discuss the therapeutic options for the treatment ofbronchiectasis• <strong>Review</strong> the genetic aspects of cystic fibrosis• Discuss the newer therapeutic approaches to the treatmentof pulmonary disease in patients with cystic fibrosisKey words: actinomycosis; bronchiectasis; cystic fibrosis;cystic fibrosis transmembrane regulator; nocardiosisUnusual Lung InfectionsNocardiosisNocardiosis refers to invasive disease causedby members of the genus Nocardia. Respiratorytract disease and extrapulmonary disseminationare the most common manifestations. <strong>Pulmonary</strong>infection is increasingly seen in immunosuppressedpatients, particularly those with defects incellular immunity. Other presentations includecellulitis, lymphocutaneous syndrome, actinomycetoma,and keratitis. Nocardia species are aerobic,nonmobile, and non-spore-forming organisms thatlive as soil saprophytes. In tissue specimens, theorganisms reveal delicate branching filamentousforms that are Gram-positive and usually acid fastif weak decolorizing agents are used for the stains.Seven species have been associated with humandisease. Nocardia asteroides is the most commonspecies associated with invasive disease. Nocardiafarcinica is less common and is associated with dissemination.Other species known to cause humaninfection are Nocardia pseudobrasiliensis and Nocardiabrasiliensis.Epidemiology/Pathogenesis: Nocardia speciesare common natural inhabitants of the soilthroughout the world. Epidemics within the hospitalenvironment are rare, and person-to- persontransmission has only rarely been suggested.Although Nocardia can occur as a primary pulmonarypathogen in patients with no underlyingdisease and can complicate surgery or trauma,it is frequently recognized as an opportunisticdisease, especially among patients with cellmediatedimmune deficiencies, including transplantation(greatest frequency in those who haveundergone lung transplantation), lymphoma,and AIDS. High-dose corticosteroids, cytomegalovirusinfection in the past 6 months, and highcalcineurin inhibitor levels (cyclosporine or tacrolimus)are independent risk factors for Nocardiainfection in organ transplant recipients. InHIV-positive persons, nocardiosis most often presentswith a CD4 lymphocyte concentration of 100 cells/μL but can occur in patients withcounts 250 cells/μL.Nocardiosis has also been reported to be associatedwith pulmonary alveolar proteinosis, mycobacterialdiseases, and chronic granulomatousdisease. Finally, nocardiosis is not uncommon inpatients with COPD receiving long-term corticosteroidtreatment, but it is also increasingly beingseen in patients with COPD who have receivedonly a recent short course of corticosteroids for anexacerbation. The disease has been reported worldwideand is more common in men than women(approximately 3:1). Most infections result fromthe inhalation of bacilli. Lesions are characterizedby necrotizing abscesses that are not well encapsulatedand spread easily. Granuloma formationand fibrosis are infrequent. Because of its propensityfor hematogenous dissemination, Nocardiaoften is associated with metastatic spread,especially to the brain (in up to one third of cases).Chest wall invasion may occur but is uncommon.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 1ACC-PULM-09-0302-001.indd 17/10/09 8:00:02 PM


Radiographic Manifestations: The chest radiographicpatterns are variable. The most frequentmanifestation is an airspace consolidation, usuallyhomogeneous, but occasionally patchy. Nodules,either single or multiple, may be confusedwith metastatic carcinoma. The most commonradiographic manifestation is cavitation, which isfound in both consolidations and nodules. Pleuralinvolvement with an empyema is present inapproximately one third of cases.Clinical Manifestations: Important features ofNocardia infection include diverse clinical andradiographic presentation, the ability to disseminateto any organ system in the body, and thetendency to relapse or remain unresponsive totherapy. The nonspecific features make diagnosischallenging, and the disease often is not suspectedearly in the patient’s course of symptoms.The median time to diagnosis ranges from 40 to55 days in reported series. Nocardial pneumoniais the most common respiratory tract presentation.Although the clinical course may be acutein immunosuppressed patients, typically thepatients have a subacute presentation consistingof several weeks of symptoms. Cough, purulentsputum, occasional blood-streaked sputum, nightsweats, and pleuritic pain are the most common.Superior vena caval syndrome, mediastinitis, andpericarditis have been reported from direct spreadfrom the lungs. Nocardia rarely involves the chestwall. In approximately 50% of pulmonary cases,extrapulmonary dissemination occurs. In a significantnumber of cases of disseminated disease,the initial respiratory tract involvement does notelicit symptoms. As noted previously, nocardiosishas the propensity for dissemination to the brain,but other extrapulmonary sites include the skin,bone, and muscle. In the CNS, Nocardia brainabscesses may be single or multiple. Nocardiais not usually recovered from the cerebrospinalfluid.Diagnosis: In most cases of pneumonia, sputumsmear findings are negative. Bronchoscopymay be necessary to obtain an adequate specimenfor the characteristic Gram-positive filaments thatmay be acid fast and often take up silver stains.Cultures for Nocardia require special handlingbecause colonies may not appear for 2 to 4 weeks.Blood cultures require incubation aerobically forup to 4 weeks. Advances in DNA extraction andreal-time polymerase chain reaction assays mayallow for identification of most Nocardia specieswithin hours. The isolation of Nocardia fromthe sputum in a non-immunosuppressed patientwithout radiographic abnormalities may representcolonization. However, a sputum culture thatis positive for Nocardia in an immunosuppressedpatient more often indicates disease. A number ofserologic tests have been evaluated, but the diversityof species known to cause disease and thepotential lack of sensitivity for detecting an antibodyresponse in immunocompromised patientshave limited their practical use.Treatment: Sulfonamide agents remain thedrug of choice for nocardiosis (sulfadiazine or sulfisoxazole,6 to 8 g/d, then decreasing to 4 g/d asthe disease is controlled). The combination of trimethoprimand sulfamethoxazole is thought to bean equally effective alternate choice. Minocyclineis an alternative choice for an oral medication inthose patients who have sulfa allergies. IV regimensinclude amikacin, ceftriaxone, cefotaxime,ceftizoxime, and imipenem. Linezolid has nowbeen shown in vitro to be highly effective againstmost strains of Nocardia. However, the high costand serious potential toxicities currently relegatethis agent to refractory cases. Because of the riskof relapse, patients who have intact host defensesare generally treated for 6 to 12 months, whereasimmunodeficient hosts and those with CNSinvolvement are treated for 12 months. Whenthe CNS is involved, treatment with cefotaximeor ceftriaxone in addition to the trimethoprimand sulfamethoxazole is recommended. Surgicaldrainage should be considered for patients withbrain abscesses, empyema, and subcutaneousabscesses. The rate of mortality caused by pulmonarynocardiosis is high (15 to 40%) and increasessignificantly with CNS involvement. Delay indiagnosis and treatment affects prognosis. It istherefore imperative that practitioners have a highindex of suspicion in evaluating patients whoare immunosuppressed or who have significantchronic lung disease.ActinomycosisActinomycosis is a slowly progressive infectiousdisease that is caused by anaerobic or microaerophilicbacteria from the genus Actinomyces.2 Unusual Lung Infection, Bronchiectasis, and Cystic Fibrosis (Moores)ACC-PULM-09-0302-001.indd 27/10/09 8:00:03 PM


When they were first described, they were misclassifiedas fungi. The word actinomycosis is derivedfrom the Greek terms aktino (the radiating appearanceof the sulphur granule) and mykos (mycoticdisease). The classic clinical picture is a cervicofacialdisease in which the patient presents with alarge mass on the jaw. It is now recognized that theorganisms colonize in the mouth, colon, andvagina. Infection results from mucosal disruptionand can occur at any site in the body. Infection ischaracterized by a pyogenic response and necrosis,followed by intense fibrosis. Actinomycosis is difficultto diagnose and is often confused with tuberculosis,lung abscess, or malignancy.Epidemiology/Pathogenesis: The organisms area normal inhabitant of the human oropharynxand frequently are found in dental caries and atthe gingival margins of persons with poor oralhygiene. Actinomycosis is most commonly causedby Actinomyces israelii. Actinomycotic infectionsin the lung are, however, usually polymicrobial.In tissue, actinomycotic infection grows in microcoloniesor granules. Because these granulesare yellow, they are often called sulfur granules,although they contain minimal amounts of sulfa.Actinomycosis of the respiratory tract is acquiredmost commonly by aspiration, but direct extensionof the disease from the head and neck orabdominal cavity can occur. The peak incidenceis reported in the fourth and fifth decades of life;nearly all series have reported a male predominance(3:1). The pulmonary form typically constitutesonly 15% of reported cases but has been ashigh as 50% in some series. Most infections occurin individuals who are immunocompetent. However,some cases have been reported in patientswith impaired host defenses. Patients with alcoholismand poor dental hygiene are at increasedrisk. The presentation of pulmonary actinomycosishas changed in recent years to a less aggressiveinfection, which is likely related to improved oralhygiene and increased use of penicillins, evenwhen the specific diagnosis is not suspected.Radiographic Manifestations: The usual patternof acute pulmonary actinomycosis consists of airspaceconsolidation, commonly in the peripheryof the lung and often in the lower lung fields. Thetypical CT scan feature is a chronic segmentalairspace consolidation containing necrotic lowattenuationareas with frequent cavity formation.Other findings include hilar or mediastinal adenopathy,bronchiectasis within the consolidation,and localized pleural thickening and/or effusion.If not treated with appropriate antibiotics, alung abscess may develop, and the infiltrate mayextend into the pleura with an associated empyema.Subsequently, actinomycosis will extendinto the chest wall with osteomyelitis of the ribs,abscess formation, and draining sinus formation.CT of the chest may be a more sensitive imagingmodality, especially if bone windows are obtained(which might reveal early rib erosion). Actinomycosisoften is mistaken for pulmonary carcinoma;the presence of an air bronchogram in the masslesion should suggest the possibility of a nonneoplasticprocess. Actinomycosis also can present asan endobronchial infection, which is often associatedwith a broncholith or other foreign body.Both actinomycosis and nocardiosis often are confusedwith tuberculosis, cryptococcosis, anaerobicpulmonary infection, bronchogenic carcinoma,and lymphoma.Clinical Manifestations: Actinomyces most commonlypresents as a disease of the cervicofacialregion after dental extraction, with osteomyelitisof the mandible or a soft-tissue abscess that drainsthrough the skin. <strong>Pulmonary</strong> actinomycosis usuallypresents with an indolent, progressive course.The initial manifestations include a nonproductivecough and low-grade fever, subsequently followedby a productive cough, which can be associatedwith hemoptysis. With chest wall involvement,pleuritic pain will develop in the patient. Rarely,a sinus tract may appear as a bronchocutaneousfistula. When this occurs, it is highly suggestiveof actinomycosis. Late in the disease, the patientmay present with weight loss, anemia, and clubbingof the digits.Diagnosis: A diagnosis of actinomycosis israrely suspected; in one series, it was suspectedon hospital admission in 7% of the patients inwhom it was ultimately diagnosed. Because actinomycosisoften mimics malignancy, diagnosismay not be made until surgical resection. Becausethese organisms are normal oropharyngeal flora,isolation in specimens of sputum or bronchialwashings is not considered significant, unlesssulfa granules are found. Actinomyces are fastidiousbacteria that are difficult to culture and, thus,correlation with the clinical and radiographic<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 3ACC-PULM-09-0302-001.indd 37/10/09 8:00:03 PM


presentation is essential. Bronchoscopy is usuallynot diagnostic unless endobronchial disease ispresent, and samples must be obtained anaerobicallywith a protected specimen brush and deliveredto the laboratory under anaerobic conditions.Histologic examination of lung tissue obtained byeither transbronchial lung biopsy or open lungbiopsy may be necessary to confirm the diagnosis.There is no reliable serologic test.Treatment: Untreated, actinomycosis is ultimatelyfatal, but early treatment can result in curerates of 90%. Prolonged treatment with highdoses of antimicrobial agents is necessary. Penicillinis the drug of choice. Regimens include IVadministration of 18 to 24 million units of penicillinfor 2 to 6 weeks, followed by oral penicillin foran additional 6 to 12 months. Tetracyclines, erythromycin,clindamycin, imipenem, and chloramphenicolare acceptable alternatives. Theorganisms generally are not sensitive to the fluoroquinolones,metronidazole, or the aminoglycosides.Whether patients should be treated for thecopathogens usually associated with actinomycesis not resolved, but most experts do not recommendthe administration of additional antibiotics.Patients with actinomycosis have a tendencyto relapse, and prolonged therapy optimizes thelikelihood of a cure. However, small trials haveshown success with relatively brief courses oftherapy (6 weeks). Some cases may require bothmedical and surgical therapies. Patients withbulky disease should probably not receive shortcourses of therapy unless surgical debulking isalso performed. A comparison of the main featuresof both actinomycosis and nocardiosis isshown in Table 1.BronchiectasisBronchiectasis is a syndrome, with manyunderlying etiologies and associations, that hasbeen defined as an irreversible dilation anddestruction of one or more bronchi, and inadequateclearance and pooling of mucus in the airways.Bronchiectasis is also characterized by persistentmicrobial infection and inflammation with releaseof microbial toxins and immune mediators. Bronchiectasisoften is divided into a form associatedwith cystic fibrosis (CF) and a non-CF form. Non-CF bronchiectasis will be reviewed first.ClassificationA classification system has been devised byReed. This system classifies bronchiectasis accordingto anatomic and morphologic patterns ofairway dilatation as follows: (1) cylindrical bronchiectasis,in which there is uniform dilatation ofthe bronchi which are thick walled and extendto the lung periphery without normal taperingTable 1. Distinguishing Features of Nocardiosis and ActinomycosisNocardiosisGram-positive aerobic*Incidence increasingMale 3:1 predominanceOccurs primarily in immunocompromised hosts<strong>Pulmonary</strong> manifestations predominateChest wall involvement is uncommonMetastatic spread (especially to the brain) iscommonGranuloma formation and fibrosis are rareDiagnosis can usually be made on sputum,BAL fluid, or pleural fluid culturesTreatment with sulfonamidesSurgical drainage often neededActinomycosisGram-positive anaerobic*Incidence decreasingMale 3:1 predominanceOccurs primarily in immunocompetent hosts; alcoholism and poordental hygiene are risks<strong>Pulmonary</strong> manifestations in a minority of patients (approximately15%)Chest wall involvement and bony erosion are commonMetastatic spread is uncommon; spread by direct contiguous invasionGranuloma and intense fibrosis are common; form the characteristicsulfur granuleDiagnosis often requires cytologic or histologic examinationTreatment with penicillinOften treated successfully with antibiotics alone*On Gram stain, the two organisms are indistinguishable as branching, Gram-positive filamentous organisms. However,Nocardia species will stain weakly with modified acid-fast stain, whereas actinomycosis will not.4 Unusual Lung Infection, Bronchiectasis, and Cystic Fibrosis (Moores)ACC-PULM-09-0302-001.indd 47/10/09 8:00:03 PM


(on a high-resolution CT scan, it has parallel “tramtrack” lines or “signet ring” appearance); (2)varicose bronchiectasis, which has an irregularand beaded outline of bronchi with alternatingareas of constriction and dilatation similar inappearance to saphenous varicosities; (3) cysticbronchiectasis, which is the most severe form andis common in patients with CF, is characterized bybronchial dilation and clusters of round air-filledand fluid-filled cysts, with a honeycomb appearance;and (4) follicular bronchiectasis, which hasextensive lymphoid nodules and follicles withinthickened bronchial walls. Cystic, cylindrical, andvaricose forms may coexist in the same patient.The fourth pattern, follicular bronchiectasis, usuallyoccurs after the occurrence of childhoodpneumonia, measles, pertussis, or adenovirusinfection. The term traction bronchiectasis is a radiologicdescription of widened airways withoutthickening, which is in response to surroundinginterstitial lung disease. These patients often donot have symptoms related to the bronchiectasis.Although insightful, these definitions are not particularlyhelpful from a clinical or therapeuticstandpoint.EtiologyThe most common causes of non-CF bronchiectasisare listed in Table 2. In general, the etiologiescan be categorized as idiopathic, postinfectious,or the result of an underlying anatomic or systemicdisease. The number of patients in each categoryhas changed since the introduction of modernantibiotic therapy. Previously, untreated infectionwas the leading cause of bronchiectasis, but withprompt treatment of infection, it is becoming muchless common. In developed countries, the conditionis now most often idiopathic in adults. The prevalenceof undiagnosed CF is not known, but studieshave suggested it is very low. Patients with focalbronchiectasis, which is localized to a segment orlobe, should undergo bronchoscopy to evaluate forand eliminate an obstructing bronchial lesion.The factor most commonly associated withbronchiectasis in childhood is infection, althoughit is being seen more commonly in adults now,especially with the increased use of HRCT scanning.Radiographic findings obtained with anHRCT scan of the chest have been described inpatients with Mycobacterium avium complex (MAC).Table 2. Predisposing Factors for BronchiectasisCauseLocalizedMechanical airway obstructionNecrotizing pneumoniaDiffuseCongenitalCFMucociliary clearance defectsImmune disordersPostinfectious complicationsRheumatologic diseasesSequelae of aspiration or toxic inhalationCOPDLung fibrosisMiscellaneousDisease ExampleForeign-body aspirationExternal compressionStenosisLung tumorsIntraluminal websAbsent cartilage<strong>Pulmonary</strong> sequestrationsKartagener syndromePrimary cilia dyskinesiaYoung syndromeHypogammaglobulinemia, IgG subclass deficiency, HIV, allergicbronchopulmonary aspergillosis, post-lung transplantBacteriaMycobacterial infections (tuberculous and NMTb*)Whooping coughViral (measles, adenovirus, influenza virus)Rheumatoid arthritis, Sjögren syndrome, inflammatory bowel diseaseSarcoidosis, after radiation therapyα 1-Antitrypsin deficiency, yellow nail syndrome*NMTb = non-tuberculosis mycobacteria.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 5ACC-PULM-09-0302-001.indd 57/10/09 8:00:03 PM


TherapyFigure 1. Large internal diameter of the bronchi (greaterthan the accompanying vessel), which is diagnostic of bronchiectasis(large arrows). From the author’s personal files.Differential DiagnosisGiven the list of possible etiologies, the followinginformation should be obtained in the evaluationof patients with suspected bronchiectasis: ageof onset of symptoms; history of frequent upperairway infections; smoking history; history of previouslung surgery; childhood infections (pertussis,pneumonia, measles); pulmonary tuberculosis oratypical mycobacterial infections; atopy or asthma;connective tissue disorders; infertility; reflux symptoms;and risk factors for HIV infection. Bronchiectasisshould be distinguished from COPD(particularly chronic bronchitis). Both diseasespresent with cough, sputum production, wheezing,and dyspnea. Exacerbations are common in bothdisorders, although the volume of sputum productionis greater in patients with bronchiectasis.Recurrent fever and hemoptysis are less likely tobe found in patients with chronic bronchitis. Thepresence of Pseudomonas in the sputum may behelpful to the diagnosis. The incidence of Pseudomonasaeruginosa is approximately 31% in patientswith bronchiectasis, but only 2 to 4% in patientswith COPD. Bronchiectasis also can be confusedwith interstitial fibrosis, especially in patients withend-state fibrosis who have a honeycomb-likeappearance seen on a chest radiograph. This parenchymalhoneycomb appearance may mimic theair-filled cysts of bronchiectasis.The objectives of management for bronchiectasisare the relief of symptoms, the prevention ofcomplications, the control of exacerbations, and areduction in mortality.Antibiotics: Antibiotics are the cornerstonefor the treatment of exacerbations of bronchiectasis.They are used to treat acute exacerbations, toprevent exacerbations, or to reduce the bacterialburden. The bacterial floras include Streptococcuspneumoniae and Haemophilus influenzae, which canbe treated with trimethoprim-sulfamethoxazole,ampicillin-clavulanate acid, or one of the newermacrolide agents. In patients with Pseudomonascolonization, oral therapy requires the use of afluoroquinolone. In some cases, IV administrationof antipseudomonal antibiotics is required.Whether prophylactic antibiotic therapy is necessaryremains an unresolved question. Patientswho experience frequent exacerbations maybenefit from a maintenance regimen, but theevidence for this approach is fairly weak. Strategiesfor prophylaxis with low-dose antibioticsrange from daily to 1 week of each month. Daily,inhaled antibiotic prophylaxis is now recommendedin patients with CF who have been colonizedwith P aeruginosa colonization. There isdebate as to whether P aeruginosa infection has asimilar effect on prognosis in patients with non-CF bronchiectasis as in patients with CF. A fewsmall studies suggest that chronic infection withP aeruginosa can cause an accelerated decrease inlung function that might be altered by the use ofchronic nebulized antibiotic therapy. Large randomizedcontrolled trials, however, are lackingin this population.Bronchodilators: Most patients with bronchiectasishave significant airway hyperresponsiveness,presumably as a result of transmural airwayinflammation. The routine use of bronchodilatorshas the added potential advantage of the stimulationof mucociliary clearance, which is associatedwith the use of β-adrenergic agents. Both aerosolizedβ-agonist therapy and aerosolized anticholinergictherapy should be tried when there isevidence of reversible airway obstruction.Antiinflammatory Agents: Although intenseairway inflammation characterizes bronchiectasis,few studies have looked at the efficacy of<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 7ACC-PULM-09-0302-001.indd 77/10/09 8:00:04 PM


corticosteroids in the treatment of this disorder.Inhaled steroids have been suggested as alternativetherapy and may be useful in some patients,especially those with significant airway hyperreactivity.It has been shown that inhaled corticosteroidscan reduce the levels of inflammatorymediators and improve dyspnea and cough. Inaddition, inhaled corticosteroids appear to reducesputum volume and lead to improvements inquality of life. However, a systematic reviewfound no significant improvement in pulmonaryfunction. Short courses of oral corticosteroid therapyoften are used during acute exacerbations.Nonsteroidal antiinflammatory agents, such asindomethacin (which is not currently approvedin the United States), have been used in Europe,either orally or by inhalation. Leukotriene receptorantagonists may be of benefit in patients withbronchiectasis because they can inhibit neutrophil-mediatedinflammation. However, therehave been no randomized controlled trials publishedconcerning patients in this population.Macrolides suppress inflammation, independentof their antimicrobial action, and have improvedthe clinical status and lung function of patientsin a few small studies of bronchiectasis. Furtherstudy is needed before they can be recommendedroutinely.Airway Clearance Techniques: Postural drainageand chest physiotherapy are useful to enhance thegravity-aided clearance of secretions. It is importantto consider that patients with non-CF bronchiectasisoften have lower-lobe predominatedisease, and thus, postural drainage may play agreater role in these patients. Alternative treatmentincludes the use of a flutter device, a positiveexpiratory pressure mask, chest oscillation,and humidification of inspired air.Mucolytic Agents and Hydration: Adequateoral hydration and the use of nebulized solutionsmay improve airway mucus clearance.Acetylcysteine is beneficial in some patients.To date, there have been no randomized, controlledclinical trials showing mucolytics to beof benefit in the treatment of non-CF bronchiectasis.Recombinant human deoxyribonuclease(rhDNase) breaks down DNA that is releasedfrom degenerating bacteria and neutrophils.DNA has a tendency to form thick viscous gels.DNase improves the clearance of secretions andpulmonary function and reduces the number ofhospitalizations in patients with CF but has notbeen found to be useful in non-CF bronchiectasis.One study has suggested that DNase was ineffectiveand potentially harmful in 300 adult outpatientswith idiopathic bronchiectasis who were instable condition. Therapy with inhaled mannitolmay improve impaired mucociliary clearance byinducing an influx of fluid into the airways andhas shown clinical promise inpatients with non-CF bronchiectasis. This is particularly excitingbecause it is easier to inhale a dry powder than touse a nebulizer.Exercise Training: The role of pulmonary rehabilitationand inspiratory muscle training has onlybeen investigated in one well-designed trial, butit has been suggested that rehabilitation increasesexercise tolerance in patients with bronchiectasis.SurgeryIn patients with localized bronchiectasis, surgicalremoval of the most affected segment or lobemay be considered. The major indications for surgeryinclude the partial obstruction of a segmentor lobe as the result of a tumor or the presence ofa highly resistant organism in the affected area,such as MAC or Aspergillus sp. Patients requiresignificant pulmonary function to withstand surgery.Surgery also may be performed for massivehemoptysis in patients with adequate pulmonaryreserve, although the increased success of bronchialartery embolization for hemoptysis makessurgery less desirable.Lung TransplantationPatients with bronchiectasis and CF initiallywere considered not to be good transplant candidatesbecause of concerns about overwhelminginfection after the use of prolonged immunosuppression.However, double-lung transplantationhas been successful in CF patients, and theSt. Louis International Transplant Registry lists 1,000 CF patients and 200 non-CF bronchiectasispatients who have undergone lungtransplantation, with a 1-year survival rate of72% and a 4-year survival rate of 49% in patientswith CF.8 Unusual Lung Infection, Bronchiectasis, and Cystic Fibrosis (Moores)ACC-PULM-09-0302-001.indd 87/10/09 8:00:04 PM


Cystic FibrosisGeneticsCF is the most common genetic disease in theUnited States, with an incidence of 1 in 3,000 birthsin a white population. The incidence is much lowerin African-Americans, Asians, Hispanics, andNative Americans. CF is an autosomal-recessivedisorder with variable penetrance. Carriers of theCF gene are phenotypically normal. Approximately5% of persons in the white US population are carriersof the CF gene, and approximately 20,000individuals are affected by this disorder. The CFgene, which was sequenced in 1989, is located inthe long arm of chromosome 7 and encodes for theCF transmembrane regulator (CFTR) protein. TheCFTR is located at the cell surface and acts as anion channel that regulates liquid volume on epithelialsurfaces through chloride secretion and theinhibition of sodium absorption.The CFTR protein may also regulate the functionof other epithelial cell proteins. The defectivetransport of ions across the epithelial membraneleads to thick viscous secretions in many organsand to excessive chloride concentration in thesweat of patients with CF. This abnormality is thebasis for the laboratory test that is most frequentlyperformed to diagnose this disorder. The CFTRprotein is expressed in all of the epithelial cellsaffected in patients with CF, including those in thelung, pancreas, sweat glands, and liver. It is alsofound in the large intestine and testes.More than 1,600 mutations of the CFTR genehave been described. The ΔF508 mutation is themost common and accounts for approximately 90%of CF chromosomes in patients of northern Europeandescent but only 60% of CF chromosomesworldwide. The mutation is caused by the deletionof a single phenylalanine residue at position 508.CFTR mutations are categorized into six classes:Class I: absent or defective protein synthesis;Class II: abnormal processing of transport of theprotein to the cell membrane;Class III: abnormal regulation of CFTR function,inhibiting chloride channel activation;Class IV: normal amount of CFTR but reducedfunction (abnormal conductance);Class V: reduced synthesis of fully active CFTR;andClass VI: decreased stability of fully processedand functional CFTR.These multiple mutations lead to varying phenotypicpresentations of the disease. Even patientswith similar genetic mutations may manifest differentclinical manifestations of the disease. Thedegree of organ involvement and perhaps the diseaseseverity correlate with the individual’s sensitivityto the CFTR dysfunction, as well as theamount of functional CFTR, which is influencedby the specific type of mutation. It is now recognizedthat some patients may present with onlyone characteristic feature of CF and a borderlineabnormal sweat test finding in the presence ofeither a known CFTR mutation or an abnormalnasal potential difference (PD), which has led touse of the term atypical or nonclassic CF. NonclassicCF is often associated with class IV or V mutations.These patients typically present later in life, andwill display typical CF symptoms in at least oneorgan, but often have a normal or borderline sweatchloride test. They are typically pancreatic sufficientand have milder pulmonary disease and abetter overall prognosis.PathogenesisCF is initiated by a defect in the gene that isnormally responsible for encoding CFTR protein,which is necessary for the flow of electrolytes andfluid across cell membranes. The resultant abnormalitiesin salt and water transport lead to analteration in the composition of secretions in therespiratory tract, pancreas, GI tract, sweat glands,and other exocrine tissues. In the lung, thesealterations change the properties of the mucuslayer lining the epithelia and the composition ofthe airway surface fluid, ultimately resulting in theclinical features of CF. The net fluid loss on theairway surface leads to the collapse of cilia andimpaired mucociliary clearance, persistent bacterialinfection, excessive host inflammatory response(characterized by the accumulation of leukocytederivedDNA and secretions rich in elastase), andairway obstruction, leading to progressive lungdestruction. Although bacterial infection clearlyleads to an intense inflammatory response, there<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 9ACC-PULM-09-0302-001.indd 97/10/09 8:00:05 PM


is mounting evidence that patients with CF haveincreased basal and inducible airway inflammationindependent of infection. Figure 2 summarizes theinteraction between airway obstruction, infection,inflammation, and destruction.Diagnostic Tests1. Before the mid-1990s, most patients weretreated for CF after presenting with typicalsymptoms. Since that time, newborn screening(NBS) has become much more widespread.Newborns with CF will have increased levelsof serum immunoreactive trypsinogen (IRT),which can be detected via a radioimmunoassayor enzyme-linked immunosorbent assay performedon samples of dried blood. After anabnormal IRT, most NBS programs performDNA testing to identify the known CFTRmutations, although some repeat the IRT after2 weeks. NBS is not a diagnostic test and thusonly identifies newborns at risk for CF. Anincreased IRT must be followed by direct diagnostictesting (typically with the sweat chloridetest). The diagnostic process for screened newbornsis shown in Figure 3.2. Sweat test (pilocarpine iontophoresis): Thesweat chloride test remains the “gold standard”diagnostic test because genetic screeningonly identifies a small number of the mostIncrease in PMNsincreased IL-8Leukotriene B4 and IL-8plugging of airwayelastaserelease of freeradicalssecretion ofmacromoleculesdegradationof elastinstructural damage;bronchiectasiscleavage ofcomplement receptorsand immunoglobulinspersistence of bacteriaFigure 2. Products of polymorphonuclear neutrophil and their effects on inflammation of the airways in patients with CF.Adapted from Ramsey BW. Management of pulmonary disease in patients with CF. N Engl J Med 1996; 335:179–188.Figure 3. The diagnostic process for screened newborns is shown. Adapted from Farrell PM, Rosenstein BJ, White TB, et al:Guidelines for the diagnosis of cystic fibrosis in newborns through older adults: Cystic Fibrosis Foundation consensus report. JPediatr 2008; 153:S4–S14.10 Unusual Lung Infection, Bronchiectasis, and Cystic Fibrosis (Moores)ACC-PULM-09-0302-001.indd 107/10/09 8:00:05 PM


common mutations. The results of this test areabnormal in the large majority of patients withCF. A chloride level of 60 mEq/dL is usuallydiagnostic for CF. However, the test must berepeated at least twice, and an adequate samplecontaining at least 75 mg of sweat must be collectedover a 30-min period. Approximately 1%of patients with CF have normal sweat chloridetest results. Abnormal sweat test results areseen in patients with other disorders (as listedin Table 3). Therefore, a diagnosis should onlybe made if the clinical history is suggestive ofCF. The current diagnostic criteria for CF arelisted in Table 4.3. Molecular diagnosis: Approximately 1,600 mutationshave been identified after sequencing theentire gene in a research laboratory. Genotypingat commercial laboratories usually can identifyonly 20 to 30 of the most common mutations.This accounts for approximately 90% of CFmutations in the general population.4. PD across the respiratory epithelium: This is aresearch tool that may be available at few centersto aid in the diagnosis of CF. The abnormaltransport of chloride leads to a very negativePD across the nasal epithelium. The measurementof this PD can establish a CF diagnosis.Nasal perfusion with amiloride hydrochlorideand with chloride-free solutions leads to characteristicchanges in the PD.Table 3. False-Positive Sweat Test ResultsAdrenal insufficiencyAnorexia nervosaAtopic dermatitisAutonomic dysfunctionCeliac diseaseFamilial cholestasisFucosidosisGlucose-6-phosphate dehydrogenase deficiencyGlycogen storage disease, type IHypogammaglobulinemiaHypoparathyroidismHypothyroidism (untreated)Klinefelter syndromeMucopolysaccharidosis, type IMalnutritionNephrogenic diabetes insipidusNephrosisProstaglandin E1 infusion, long-termPseudohypoaldosteronismTable 4. Criteria for Diagnosis of CFElevated sweat chloride level 60 mmol/L on twooccasionsorIdentification of mutations known to cause CF in both CFTRgenesorIn vivo demonstration of characteristic abnormalities in iontransport across the nasal epitheliumplusOne or more phenotypical features of CFSinopulmonary diseaseCharacteristic GI or nutritional disordersObstructive azoospermiaSalt loss syndromeorSibling with CForPositive NBS resultNonpulmonary Clinical ManifestationsIn the pancreas, dysfunction of the exocrineportion leads to fat malabsorption and malnutrition.Glucose intolerance is present in as many as50 to 70% of older patients with CF; diabetes mellitusoccurs in approximately 15%. Malabsorptionof fat-soluble vitamins such as A, D, E, and K canlead to vitamin deficiency and coagulopathy.The liver is affected by biliary cirrhosis as a resultof thick secretions in the biliary ducts. Thisis manifested by abnormal liver function testresults. Diffuse liver involvement can lead to portalhypertension.Meconium ileus is present in approximately20% of infants with CF and is pathognomonic forthis disorder. A similar syndrome occurs in olderCF patients caused by inspissated mucus in the GItract. This results in the distal intestinal obstructionsyndrome, often requiring surgical treatment.Approximately 95% of male patients with CFare infertile. Although sperm maturation is normal,the Wolffian structures often are not developed.The vas deferens is often completely absent.This absence may indicate a role for the CFTRprotein in the development of the male genitaltract. In fact, in subjects with congenital bilateralabsence of the vas deferens but without any otherabnormalities of CF, mutations in the CFTR gene<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 11ACC-PULM-09-0302-001.indd 117/10/09 8:00:06 PM


have been reported. Women with CF are not infertile,al though they may have difficulty conceivingbecause of thick cervical mucus and/or anovulatorycycles if their nutritional status is poor. Pregnanciescan be successful, and pulmonaryfunction has not been found to deteriorate afterpregnancy.<strong>Pulmonary</strong> ManifestationsThe primary cause of morbidity and mortalityin CF patients is bronchiectasis and obstructivelung disease. <strong>Pulmonary</strong> disease is present in 98%of patients with CF by the time they reach adulthood.Despite the great advances in the managementof this disorder, the majority of the patientssuccumb to respiratory complications. A recurrentcough that becomes persistent is often the firstmanifestation. Airway hyperreactivity and wheezingare common in children. Bronchodilatorresponsiveness tends to decrease with age, perhapsas a result of the destruction of the cartilage.Pansinusitis with opacification of the paranasalsinu ses is a universal finding in patients with CF.Nasal polyps are present in up to 30% of patientswith CF.In patients with CF, the lungs are normal atbirth, but infection occurs early in life and is persistent.The abnormal ionic environment andchronic airway obstruction by thick secretionspromote colonization with pathogenic bacteria,which leads to the accumulation of inflammatorycells. These cells release inflammatory mediatorsthat cause inflammation and damage to the airwaywall, leading to the development of bronchiolitisand subsequently bronchiectasis (Fig 2).H influenzae, Staphylococcus aureus and, later,P aeruginosa, are the pathogens that are most commonlyfound in the airways of patients with CF.The abnormal CFTR may be partly responsible forcolonization with Pseudomonas sp because thenormal CFTR appears to be involved in the clearanceof this organism from the airways. The highsodium content in CF secretions may contribute tochronic infection because low sodium content isrequired for the effective killing of bacteria in airwayepithelia. Colonization with P aeruginosa is notbenign because it has been found to be an independentrisk factor for the accelerated loss of lungfunction and decreased survival. Colonizationwith Burkholderia cepacia denotes an even worseprognosis.The mucoid appearance of Pseudomonas iscaused by the production of alginate. This bacteriumis difficult to eradicate as the result of the poorpenetration of antibiotics into purulent airwaysecretions and to the interference in phagocytickilling by alginate. Bacteria may have native oracquired antibiotic resistance. In addition, thepharmacokinetics differ in patients with CFbecause the volume of distribution of hydrophilicdrugs (eg, penicillins, cephalosporins, and aminoglycosides)is increased as the result of decreasedamounts of adipose tissue. The renal clearance ofaminoglycosides is increased, and therefore, thedosage has to be adjusted, usually at triple thenormal dose. Once-daily administration is notrecommended.Management of <strong>Pulmonary</strong> DiseaseAlthough the focus of this section is on the treatmentof pulmonary disease in patients with CF, itshould be kept in mind that management in generalwill be suboptimal unless all aspects of the diseaseare addressed (ie, exocrine pancreatic supplementation,nutritional support, glucose control, psychosocialissues, and treatment compliance). Progresshas been made in attempts at addressing the underlyingdefect of CF with gene therapy, but currentclinical treatment is aimed at treating the effects ofthe CFTR dysfunction— thickened airway mucus,infection, and inflammation.Clearance of Airway Secretions: The chief nonpharmacologicmeans of enhancing airway secretionclearance has been chest physiotherapy (eithermanual or with use of an oscillatory vest) and posturaldrainage. Other techniques include forcedexpiratory techniques, positive expiratory pressure,and flutter valves. A Cochrane review in early2009 noted that there is no high-level evidence thatany of these techniques are superior to the others.All patients who produce daily sputum should beinstructed in clearance techniques. All of the techniquesrequire a great deal of time, and treatmentcompliance can be an issue. Therefore, severalmethods can be introduced to each patient.Bronchodilator Therapy: As noted previously,many patients with CF will demonstrate a degreeof bronchial hyperactivity and reversibility. The12 Unusual Lung Infection, Bronchiectasis, and Cystic Fibrosis (Moores)ACC-PULM-09-0302-001.indd 127/10/09 8:00:06 PM


obstructive airway disease is typically onlypartially reversible because the underlying causesinclude the chronic infection, inflammation, andresulting structural damage. Bronchodilator therapyshould be considered in patients who haveat least a 10% increase in FEV 1in response to aninhaled bronchodilator. In addition, bronchodilatorsoften are used in all patients with CF beforechest physiotherapy and immediately before theinhalation of hypertonic saline solution, antibiotics,or dornase alfa because these medicationshave the potential to induce nonspecific bronchialconstriction. The role of these medicationsin improving mucociliary clearance has not beenwell defined.Reduction in the Viscosity of Secretions: Theincreased viscosity of the sputum in patientswith CF is caused partly by the presence of manypolymorphonuclear neutrophils and their degradationproducts, including DNA from dyingcells. rhDNase I (or dornase alfa) digests extracellularDNA and can help to reduce the viscosity.A metaanalysis of randomized trials of dornasealfa has concluded that treatment improves lungfunction and is well tolerated. There is some controversyabout when to initiate dornase alfa, butmost clinicians will consider a trial in patientswith documented chronic infection and obstructionin spirometry. A guideline committee for theCF Foundation recommends chronic dornase alfafor all children with CF 6 years old. The recommendeddosage is 2.5 mg/d.Osmotic Therapy: A decreased amount of airwaysurface liquid is an important factor in theunder lying pathophysiology of CF. Therefore,osmotically drawing water onto the airway surfacevia inhalation of a hypertonic substancemight help to clear secretions and restore mucociliarytransport. Inhaled hypertonic saline solutionhas been used for this purpose in patientswith CF, and has been associated with increasedmucus clearance and improvement in lung function,and, in one long-term study, with fewer exacerbationsrequiring antibiotic therapy. Prescribingpatterns vary, but twice-daily treatments (4 mL of7% saline solution) in patients with chronic coughand sputum production should be considered.There have been no randomized controlled trialscomparing dornase alfa with hypertonic salinesolution. Because the mechanism of action differssomewhat between the two, it is reasonable toassume that they maybe complementary.Antibiotic Therapy: Patients with CF havefrequent exacerbations of infection. Acute exacerbationsare manifested by increased coughand sputum volumes, sometimes associatedwith hemoptysis, and are often accompanied bysystemic symptoms such as decreased energy,anorexia, and weight loss. Spirometry demonstratesa decrease in lung function. Parenteralantibiotics are generally administered for 14 to 21days to reduce the burden of bacteria, to decreasesymptoms, and to improve lung function. Themost recent respiratory tract cultures should beused to guide therapy. Cephalothin and nafcillinare used for the treatment of infection withS aureus, and vancomycin is used for patientswith a penicillin allergy. For patients with Pseudomonassp, an antipseudomonal penicillin iscombined with an aminoglycoside. Combinationtherapy reduces the risk of the development ofresistance. Intensified bronchodilator therapyand chest physiotherapy are indicated duringthe treatment of exacerbations. Therapy withsteroids may be used in patients with hyperreactiveairways, but it has not been systemicallystudied.Chronic infection with Pseudomonas sp isassociated with a more rapid decrease in pulmonaryfunction and increased mortality; delayingthe onset of chronic infection is critical. When theorganism is first isolated, a prolonged course ofantibiotics that eradicates the organism has beenshown to delay the onset of chronic infection.Repeated attempts to eradicate the organism mayalso be successful and beneficial. A combinationtherapy consisting of an oral quinolone and aninhaled aminoglycoside is typically used. Somepatients with CF are administered long-termantibiotic therapy in the hopes of suppressingbacterial growth and recurrent infections, althoughthis practice is controversial. The most commoncurrent practice involves the use of nebulizedantipseudomonal antibiotics in patients with documentedchronic infection. The inhaled route isattractive because it allows the delivery of greaterconcentrations to the airways with low systemicabsorption, thus reducing the risk of ototoxicityand nephrotoxicity. A 2009 Cochrane review confirmedthat long-term use is associated with<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 13ACC-PULM-09-0302-001.indd 137/10/09 8:00:06 PM


improved lung function and a decreased numberof hospitalizations.Antiinflammatory Agents: There is a neutrophilpredominantinflammation in the airways ofpatients with CF. These neutrophils perpetuatethe inflammatory cycle through the releaseof cytokines, chemotactants, and proteolyticenzymes (Fig 2). As noted earlier, patients withCF have abnormal basal and inducible airwayinflammation. This inflammation is not entirelyindependent from infectious stimulation. Whenconsidering potential antiinflammatory strategies,several key concepts must be kept in mind:the inflammatory process is primarily endobronchial;it is characterized by persistent neutrophilinflux; intracellular signaling pathways are a keycomponent; the inflammatory response is prolongedand involves heightened oxidative andproteolytic stress.Several early studies evaluated the effects ofsystemic glucocorticoids in patients with CF.Although a beneficial effect on lung function wasobserved in some patients, the adverse effects(glucose intolerance, growth retardation, cataracts)limit widespread use. Therefore, therapy withsystemic glucocorticoids is recommended only forpatients with asthma or ABPA complicating thedisease. There is a lack of good clinical trials ofinhaled corticosteroids in patients with CF, andthese also are only recommended in patients withasthma or ABPA.Ibuprofen also has been studied because of itsability to inhibit the migration and activation ofneutrophils. In high doses, ibuprofen appears toslow the progressive decrease in lung function,particularly in younger patients with a milderform of disease. The 2009 Cochrane review, whichis based on four trials enrolling a total of 287patients, confirms this finding. The beneficialeffects are dependent on achieving adequateserum levels, and thus the drug must be individuallydosed based on measured pharmacokinetics(desired peak plasma concentrations between 50and 100 μg/mL). This therapy has not been usedwidely (only 6% based on current CF patient registrydata) as the result of concerns about thelogistics of dosing and fear of GI and renal toxicity.However, several clinical trials have suggested thatrenal impairment does not occur at an increasedrate, and most GI complaints are related to theunderlying disease. There is a statistically significantincrease in GI bleeding, but the event ratesremain very low. The CF Foundation currentlyrecommends high-dose ibuprofen therapy forchildren 6 to 12 years of age with moderate-toseverelung disease.Macrolide therapy was found to be beneficialin patients with panbronchiolitis, which led toempiric use by some physicians in patients withCF. Several well-conducted clinical trials have nowshown that the long-term use of azithromycin(which appears to act primarily as an antiinflammatoryagent by inhibiting neutrophil migrationand elastase production) is associated withimproved lung function and a reduction in thenumber of exacerbations. The CF Foundation currentlyrecommends azithromycin (500 mg po3 d/wk) for patients with long-term Pseudomonasinfection. Before the initiation of therapy, someexperts recommend that sputum be examined fornontuberculous mycobacteria because macrolideantibiotics are a vital in the treatment of nontuberculousmycobacteria.Other antiinflammatory agents are being investigatedin patients with CF. Increased oxidativestress and reduced levels of glutathione in thelungs of patients with CF have led to trials of theantioxidant N-acetylcysteine, aerosolized glutathione,and oral supplementation with antioxidantvitamins. Evidence suggests that these agentsmight attenuate lung inflammation, but furtherstudy is needed. Pharmacologic therapies thattarget proinflammatory and regulatory cytokineshave been developed for other diseases (rheumatoidarthritis, psoriasis, inflammatory bowel disease).No published data on their use in CF areavailable. In addition, there is some concern thatthese agents might overly suppress the inflammatoryresponse, leading to secondary infectiouscomplications. Finding ways to interrupt intracellularsignaling pathways that lead to increasedinflammation may also be an effective strategy, butmore understanding of the complex roles theseplay in other important cellular mechanisms isneeded. The increased levels of neutrophils in thelung lead to massive amounts of proteases thatoverwhelm native antiproteases. Several trialshave examined the use of inhaled α 1-antitrypsinin patients with CF. Although it appears to be welltolerated, no clear benefit on important outcomes14 Unusual Lung Infection, Bronchiectasis, and Cystic Fibrosis (Moores)ACC-PULM-09-0302-001.indd 147/10/09 8:00:07 PM


have yet been demonstrated. In addition, thistherapy is limited by expense, supply, and the risksof using plasma-derived products. Some of thismay be overcome in the future with recombinantα 1-antitrypsin.New Therapies in Development: Because CFTRmutations lead to defective or absent CFTR protein,the most logical way of treating CF would beto replace the CFTR gene to restore normal production.There are multiple steps to achieving thisgoal: sufficient gene product must be delivered tothe primary target cells and it must be incorporatedso as to allow normal gene expression. Althoughprogress has been made, gene therapy is still someyears away from clinical practice. Other approachesin development include correction of the abnormalprotein folding of the CFTR; improvement in theion channel function; induction of alternative ionchannels, and newer antibiotics.<strong>Pulmonary</strong> ComplicationsHemoptysis: Hemoptysis occurs frequentlyduring bronchitic exacerbations and usually isself-limited. Conservative measures such as bedrest, cough suppression, antibiotics, and correctionof coagulopathy, if present, are adequatetreatment for most patients. Massive hemoptysisis associated with a high mortality rate but mayrespond favorably to bronchial artery embolization.It is successful in as many as 90% of cases,but recurrences can occur in approximately 20%of cases. Repeated procedures can be performed ifnecessary. If this proves unsuccessful, lung resectionof the involved lobe may be the only alternative,but it is often difficult to ascertain withcertainty which lobe or segment is responsible forthe hemorrhage.Pneumothorax: Spontaneous pneumothoraxoccurs in approximately 16% of patients withCF. The vast majority of these patients havesevere pulmonary involvement antecedent tothe pneumothorax, with an FEV 1of 50% predicted.The average recurrence rate is nearly50%, and despite treatment, the mortality rateis high at 30 to 60%. This high mortality raterelates more to the severe underlying parenchymalinvolvement than to the pneumothoraxitself. Chemical pleurodesis may be necessaryfor treating recurrences.Nontuberculous Mycobacterial Infections: Recently, there has been a marked increase in theisolation of nontuberculous Mycobacterium sp(primarily Mycobacterium avium intracellularecomplex and Mycobacterium chelonei) in adultpatients with CF. Several centers have reportedpositive culture findings from up to 18% ofpatients studied. Distinguishing airway colonizationfrom infection can be difficult. HRCT scanningmay be useful in the evaluation. Nodularopacities or a tree-in-bud appearance suggeststhe presence of infection rather than colonization.Transbronchial lung biopsies may be required todemonstrate the presence of infection. If pulmonaryfunction declines and atypical Mycobacteriumsp are found in cultures from at least threesputum samples, treatment with antimycobacterialagents is recommended.ABPA: ABPA develops in a small portionof patients in the United States (1.8% in a 1995report). ABPA should be suspected if there is evidenceof bronchospasm, peripheral eosinophilia,sputum cultures positive for Aspergillus fumigatus,and an immediate skin test response to A fumigatus.Diagnosis is confirmed by total serum IgElevels of 1,000 ng/mL and IgE or IgG specificto A fumigatus.Respiratory Failure and Cor Pulmonale: Respiratoryinsufficiency develops as lung diseaseprogresses, initially with hypoxemia on exercise,then at rest, and eventually with carbon dioxideretention. Right ventricular failure represents theculmination of the pathologic sequence of lungdisease with progressive respiratory failure. Inmost cases, this process heralds the terminal stagein a patient’s course with only limited survivalbeyond a few months.Transplantation: Patients with CF are generallygood transplant candidates because theyare young and otherwise healthy and are used toundergoing complex medical regimens. CF doesnot develop in the transplanted lung. However,transplantation in CF is not without controversy,especially since some prediction modelssuggest that transplantation rarely improvessurvival in CF patients 18 years of age. Othershave found that only those patients with apredicted 5-year survival of 50% and withoutBurkholderia cepacia or CF arthropathy arelikely to have increased survival. These patients<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 15ACC-PULM-09-0302-001.indd 157/10/09 8:00:07 PM


equire bilateral lung transplantation using aclamshell incision.Indications for transplantation include deterioratingrespiratory status despite aggressivemedical therapy and an FEV 1of 30% predictedin compliant patients with good nutritional statusand no other organ impairments. Patients alsocan be considered if they manifest major lifethreateningpulmonary complications (eg, massivehemoptysis), pulmonary hypertension, or increasingantibiotic resistance of bacterial infecting thelungs. The average wait for transplantation isapproximately 2 years. In 1997, the actuarial survivalrate for CF patients was 72% at 1 year and49% at 4 years. Female patients and those 18years of age have a worse prognosis and should beconsidered for earlier listing.Contraindications to transplantation includeother organ failures, noncompliance with therapy,psychosocial instability, profound malnutrition, oractive infection with Aspergillus or atypical Mycobacteriumsp. An increased risk for transplantationis associated with colonization with resistantorganisms (particularly B cepacia), previous thoracicsurgery or pleurodesis, the need for mechanicalventilation, and diabetes mellitus.PrognosisMost patients still die of respiratory complications.The degree of pulmonary impairment ispredictive of survival. In 1992, Kerem et al publishedthe results of a study that demonstrated thatwhen the FEV 1decreased to 30% predicted, the2-year survival rate was 50%. Women appear toexperience a greater deterioration of lung functionwith age.When the first accurate description of CF waspublished by Andersen in 1939, 80% of patientsdied within 1 year of birth. Since then, advances indiagnosis and therapy have been accompanied bya gradual improvement in prognosis and increasedsurvival time. According to data from the CFPatient Registry, the median survival time in theUnited States was approximately 20 years in 1970and had increased to 37.4 years by 2007. Recentestimates have projected that children born withCF in the United States today will survive into theirsixth decade of life.Annotated BibliographyAccurso FJ. Update in CF 2006. Am J Respir Crit CareMed 2007; 175:754–757CF is a complex inherited disease resulting from abnormalitiesin the gene that codes for CFTR. The course of CF is surprisinglyvariable, even when considering CF genotype. Thisupdate concisely reviews the latest research on the geneticsof airway morphology, infections, effects of the nitric oxide,and assessment of inflammation.Barker AF, Couch L, Fiel SB, et al. Tobramycin solutionfor inhalation reduces sputum Pseudomonas aeruginosadensity in bronchiectasis. Am J Respir Crit Care Med2000; 162:481–85More patients treated with tobramycin solution for inhalationthan those treated with placebo reported increasedcough, dyspnea, wheezing, and noncardiac chest pain, butthe symptoms did not limit therapy.Belkin RA, Henig NR, Singer LG, et al. Risk factors fordeath of patients with CF awaiting lung transplantation.Am J Respir Crit Care Med 2006; 173:659–666Precise timing for the listing of CF patients for lung transplantationremains controversial. This retrospective studyof patients listed for lung transplantation at four academicmedical centers identified risk factors for death while awaitingtransplantation. Patients with FEV 1 30% predictedhad a greater risk of death only when their PaCO 2was 50 mm Hg, whereas the increased risk of death with FEV 1 30% predicted was not further influenced by the presenceof hypercapnia.Bilton D. Update on non-cystic fibrosis bronchiectasis.Curr Opin Pulm Med 2008; 14:595–599This review includes a particularly nice discussion of updatesin inherited and noninherited causes of bronchiectasis andinsight into the role of nontuberculous mycobacterial andpseudomonal infections in patients with bronchiectasis.Choi J, Koh WJ, Kim TS, et al. Optimal duration of IVand oral antibiotics in the treatment of thoracic actinomycosis.Chest 2005; 128:2211–2217IV antibiotic therapy for 2 to 6 weeks followed by 6 to 12months of oral antibiotic therapy is usually recommendedfor the treatment of thoracic actinomycosis. However, a retrospectiveof 28 patients with thoracic actinomycosis showedsuccess with varying treatment regimens. Combined surgicaland antibiotic therapy was used in nearly half of thepatients. The duration of IV antibiotic therapy ranged from3 to 17 days, whereas the duration of oral antibiotic therapyranged from 0 to 534 days. Clinical cures were achieved in96% of the patients with no clinical evidence of recurrence.16 Unusual Lung Infection, Bronchiectasis, and Cystic Fibrosis (Moores)ACC-PULM-09-0302-001.indd 167/10/09 8:00:07 PM


Davis JC, Alton E, Bush A. Clinical review: CF. BMJ2007; 335:125551–125259This article includes up-to-date tables with informal evidencegrading of management strategies in CF.Davis PB. Centennial review: CF since 1938. Am JRespir Crit Care Med 2006; 173:475–482This article summarizes the history of CF since it was distinguishedfrom celiac disease in 1938. The standardization of thesweat test in 1959 allowed the identification of milder cases. In1983, chloride transport was identified as the basic CF defect.In 1989, the CF gene was discovered. This article is easy tofollow and explains concepts in an understandable fashion.Elkins MR, Robinson M, Rose BR, et al. A controlledtrial of long-term inhaled hypertonic saline in patientswith CF. N Engl J Med 2006; 354:229–240Seminal article on the long-term use of hypertonic salinesolution in patients with CF.Farrell PM, Rosenstein BJ, White TB, et al. Guidelinesfor the diagnosis of cystic fibrosis in newborns througholder adults: Cystic Fibrosis Foundation consensusreport. J Pediatr 2008; 153:S4–S14Excellent updated review of the diagnostic tests, includingNBS and recommended diagnostic algorithms based uponthese tests.Flotte TR, Laube BL. History and evolution of aerosolizedtherapeutics. Chest 2001; 120(suppl):124S–131STheoretically, CFTR gene replacement during the neonatalperiod can decrease the rates of morbidity and mortalityfrom CF. In vivo gene transfers have been accomplished inpatients with CF. The choice of vector, mode of delivery to theairways, translocation of genetic information, and sufficientexpression level of the normalized CFTR gene are issues thatare currently being addressed.Flume PA, Stange C, Ye X, et al. Pneumothorax in CF.Chest 2005; 128:720–728Pneumothorax is a complication in patients with CF, occurringmore commonly in older patients with more advancedlung disease. Approximately 1 in 167 patients will experiencethis complication each year. The principal risks associatedwith an increased occurrence of pneumothorax includethe presence of P aeruginosa, B cepacia, or Aspergillus spin sputum cultures, FEV 1 30% predicted, enteral feeding,Medicaid insurance, pancreatic insufficiency, ABPA, andmassive hemoptysis. There is an increased 2-year mortalityrate after pneumothorax.Flume PA, Yankaskas JR, Ebeling M, et al. Massivehemoptysis in CF. Chest 2005; 128:729–738Massive hemoptysis is a serious complication in patientswith CF, occurring more commonly in older patients withmore advanced lung disease. The principal risks associatedwith an increased occurrence of massive hemoptysis includedthe presence of S aureus in sputum cultures and diabetes.There was an increased morbidity (eg, increased hospitalizationsand hospital days) and an increased 2-year mortalityrate after massive hemoptysis.Gilljam M, Ellis L, Corey M, et al. Clinical manifestationsof CF among patients with diagnosis in adulthood.Chest 2004; 126:1215–1224Seven percent of CF patients receive diagnoses in adulthood.Patients with CF presenting in adulthood often havepancreatic sufficiency, inconclusive sweat chloride sweattest results, and a high prevalence of mutations that arenot commonly seen in CF diagnosed in childhood. Singleorganmanifestations such as congenital bilateral absence ofthe vas deferens and pancreatitis are seen. The severity oflung disease is variable. Repeated sweat chloride tests andextensive mutation analysis are often required to make adiagnosis.Husain S, McCurry K, Dauber J, et al. Nocardia infectionin lung transplant recipients. J Heart Lung Transplant2002; 21:354–359Nocardia infection tended to involve the native lung in singlelungtransplant recipients. Trimethoprim-sulfamethoxazolefor Pneumocystis carinii prophylaxis at the doses given wasnot protective against nocardiosis in these patients.Kerem E, Reisman J, Corey M, et al. Prediction of mortalityin patients with CF. N Engl J Med 1992; 236:1187–1191In a cohort of 673 patients with CF, FEV 1was predictive ofmortality; when it reached 30% of predicted, the 2-yearmortality rate exceeded 50%.Kim TS, Han J, Koh W, et al. Thoracic actinomycosis:CT features with histopathologic correlation. AJR Am JRoentgenol 2006; 186:225–231Radiologic, pathologic, and histologic images of patientswith pulmonary actinomycosis infection.Kreider M, Kotloff RM. Selection of candidates for lungtransplantation. Proc Am Thorac Soc 2009; 6:20–27Up-to-date review of general consideration in selectingpatients for lung transplantation, as well as specific recommendationsfor patients with CF.Mabeza GF, Macfarlane J. <strong>Pulmonary</strong> actinomycosis.Eur Resp J 2003; 21:545–551Comprehensive clinical review of the pulmonary manifestationsof actinomycosis.Martinez R, Reyes S, Menendez R. <strong>Pulmonary</strong> nocardiosis:risk factors, clinical features, diagnosis and prognosis.Curr Opin Pulm Med 2008; 14:219–227<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 17ACC-PULM-09-0302-001.indd 177/10/09 8:00:07 PM


Most current, comprehensive clinical review of pulmonarynocardiosis (including expanded sections on newer diagnostictechniques and new therapeutic options).McGeinness G, Naidich DP. CT of airways disease andbronchiectasis. Radiol Clin North Am 2002; 40:1–19This article discusses HRCT findings associated withmedium-sized and small airways diseases, focusing on bronchiectasisof both infectious and noninfectious causes.McMullen AH, Pasta DJ, Frederick PD, et al. Impactof pregnancy on women with CF. Chest 2006; 129:706–711Concern exists about the impact of pregnancy on the healthof women with CF. This study suggests that women withCF who become pregnant experience similar respiratoryand health trends compared to nonpregnant women withCF. However, pregnant women used a greater number oftherapies and received more intense monitoring of theirhealth.Moss RB, Rodman D, Spencer LT, et al. Repeatedadeno-associated virus serotype 2 aerosol-mediated CFtrans-membrane regulator gene transfer to the lungs ofpatients with CF: a multicenter, double-blind, placebocontrolledtrial. Chest 2004; 125:509–521This eight-center CF gene therapy study determined thesafety of repeated doses of aerosolized adeno-associated vectorcontaining CFTR complementary DNA. It showed thatthe aerosolized viral vector encoding the complete humanCFTR DNA was safe and well tolerated. There were trendsin improvement in pulmonary function in patients with CFand mild lung disease.Nichols DP, Konstan MW, Chmiel JF. Anti-inflammatorytherapies for cystic fibrosis-related lung disease.Clinc Rev Allerg Immunol 2008; 35:135–153Comprehensive and up-to-date review of the inflammatorymechanisms in CF. Includes a detailed review of severalpharmacologic agents currently use (and those in development)that address these multiple mechanisms.O’Donnell AE. Bronchiectasis. Chest 2008; 134:815–823Bronchiectasis is now being recognized with increased frequencyaround the world. This is a well-written clinicalreview, including a discussion of pathophysiology and naturalhistory, in addition to reviews of current therapeuticapproaches.O’Donnell AE, Barker AF, Ilowite JS, et al. Treatmentof idiopathic bronchiectasis with aerosolized recombinanthuman DNase I. Chest 1998; 113:1329–1334rhDNAse was ineffective and potentially harmful in thisgroup of adult outpatients in stable condition with idiopathicbronchiectasis. This contrasts with previously publishedresults that demonstrated the efficacy of rhDNase in patientswith CF bronchiectasis.Ramsey BW. Drug therapy: management of pulmonarydisease in patients with CF. N Engl J Med 1996;335:179–188Although somewhat out of date, this is an excellent andwell-organized approach to the management of pulmonarycomplications of CF. Includes specific recommendations onantibiotic coverage for acute exacerbations and chronic suppressionof pulmonary infections.Ramsey BW, Pepe MS, Quan JM, et al. Intermittentadministration of inhaled tobramycin in patients withCF. N Engl J Med 1999; 340:23–30This article reports the results of a multicenter trial of intermittentadministration of preservative-free preparation ofinhaled tobramycin in CF patients with P aeruginosa, withresultant improvement in pulmonary function and clinicalcourse.Ratjen F. New pulmonary therapies for CF. Curr Opinin Pulm Med 2007; 13:541–546A very current and concise review of new approaches to themanagement of CF, including gene therapy, CFTR modulation,manipulation of ion channels, and other antiinflammatoryand antiinfective medications.Swensen SJ, Hartman TE, Williams DE. Computedtomographic diagnosis of Mycobacterium avium-intracellularecomplex in patients with bronchiectasis.Chest 1994; 105:49–52A clinical study of 100 patients with bronchiectasis withchest CT scan evidence of MAC infection suggests that thepresence of multiple small lung nodules on CT scans is predictiveof the presence of MAC on culture.ten Hacken NH, Wijkstra PJ, Kersjens HA. Clinicalreview: treatment of bronchiectasis in adults. BMJ2007; 335:1089–1093Succinct, organized review with evidence tables of differingtherapeutic agents.Vlahakis NE, Aksamit TR. Diagnosis and treatment ofallergic bronchopulmonary aspergillosis. Mayo ClinProc 2001; 76:930–938ABPA is an underdiagnosed pulmonary disorder in asthmaticpatients and patients with CF. If ABPA is diagnosedand treated before the development of bronchiectasis andfibrosis, these complications may be prevented. Total serumIgE and IgE-specific A fumigatus antibodies establish thediagnosis. Long-term treatment with corticosteroids is oftenrequired for management. Itraconazole may be effective as acorticosteroid-sparing agent.18 Unusual Lung Infection, Bronchiectasis, and Cystic Fibrosis (Moores)ACC-PULM-09-0302-001.indd 187/10/09 8:00:07 PM


Yankaskas JR, Marshall BC, Sufian B, et al. CF adultcare: consensus conference report. Chest 2004;125(suppl):1S–39SThis supplement is the most up-to-date summary of thediagnosis and management of CF in the adult but is basedon a CF Foundation consensus conference held in 1999.Yildiz O, Doganay M. Actinomycoses and Nocardiapulmonary infections. Curr Opin Pulm Med 2006;12:228–234Well-written, easy-to-understand review and comparison ofpulmonary infections related to Actinomyces and Nocardiaspecies.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 19ACC-PULM-09-0302-001.indd 197/10/09 8:00:08 PM


Notes20 Unusual Lung Infection, Bronchiectasis, and Cystic Fibrosis (Moores)ACC-PULM-09-0302-001.indd 207/10/09 8:00:08 PM


<strong>Pulmonary</strong> Vascular DiseasesLisa K. Moores, MD, FCCPObjectives:• Describe the risk factors and epidemiology of venousthromboembolism (VTE)• Provide approaches to the diagnosis and treatment ofVTE• Discuss approaches to the prevention of VTE• Describe the epidemiology and pathophysiology of pulmonaryhypertension (PH)• Describe the steps in the diagnosis of pulmonary arterialhypertension (PAH)• List the recommended treatment regimens for PHKey words: anticoagulation; hypercoagulable state; pulmonaryarterial hypertension; pulmonary embolism; pulmonaryhypertension; thrombolytic therapy; venous thrombosisVTE<strong>Pulmonary</strong> embolism (PE) and deep venousthrombosis (DVT) constitute the spectrum of onedisease: venous thromboembolism (VTE). Approximately79% of patients who present with PE alsowill have concomitant DVT; conversely, onlyapproximately 50% of patients presenting withDVT will have PE. The estimated annual incidenceof VTE in the United States is one episode per 1,000patients. As many as 3,000,000 people die each yearin the United States from acute PE.Risk FactorsRisk factors associated with VTE often areclassified into acquired and genetic factors, andthese can overlap in an individual patient.Acquired risks include prolonged immobility orparalysis; cancer; major surgery (particularly,operations involving the abdomen, pelvis, andlower extremities); trauma; obesity; high estrogenstates (ie, pregnancy, taking birth control pills, andhormone replacement therapy); varicose veins;congestive heart failure; cor pulmonale; myocardialinfarction; stroke; fractures of the pelvis, hip,or leg; indwelling femoral vein catheters;inflammatory bowel disease; heparin-inducedthrombocytopenia (HIT); and myeloproliferativedisorders, such as polycythemia vera and hyperviscositysyndromes. Patients with a history ofVTE are at increased risk for recurrence. Congenitalor hereditary hypercoagulable statesinclude the following: activated protein C resistance(factor V Leiden mutation); antithrombin IIIdeficiency (can also be acquired with the nephroticsyndrome); protein C deficiency; protein S deficiency;dysfibrinogenemia; disorders of plasminogenand plasminogen activation; the presenceof antiphospholipid antibodies and lupus anticoagulant;and hyperhomocysteinemia. Manypatients may present with multiple risk factors forVTE that can further increase their likelihood ofthis complication developing. Examples includethe elderly patient with congestive heart failurewho sustains a hip fracture and the patient witha myelopro liferative disorder who is undergoingmajor surgery.PathogenesisThe Virchow triad of venous stasis, endothelialvascular injury, and hypercoagulability explainshow various processes can interact to overcomeantithrombotic defenses resulting in VTE. Vascularstasis predisposes the patient to VTE by allowingactivated coagulation factors to remain undilutedand in contact with the vascular endothelium.Vascular trauma or injury, to include the presenceof indwelling catheters, presumably initiatesthrombosis through the release of tissue factorsthat activate coagulation proteins. The term hypercoagulabilityrefers to abnormal fibrinolytic systempathways or acquired and congenital deficienciesor functional abnormalities that predispose thepatient to VTE.The incidence of cancer is slightly increased ina first episode of DVT (11% vs 9%, respectively) orPE (6% vs 5%, respectively) compared with controlsubjects. Cancers of the pancreas, liver, ovary, and<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 21ACC-PULM-09-0302-002.indd 217/10/09 8:03:52 PM


ain, and metastatic disease account for most ofthese. This hypercoagulability relates to the effectsof tumor cells on thrombin generation and fibrinolysis,as well as to tumor-derived activators offactor X. Because these tumors are most commonlyincurable, an exhaustive evaluation for underlyingmalignancy, apart from routinely obtaining amedical history and conducting a physical examination,is not recommended.DVTPE usually arises from the deep venous systemof the body as a complication of DVT. AlthoughDVT usually begins in the lower extremities, occasionalthrombi form in pelvic veins, renal veins,upper-extremity veins, and the right heart. Mostthrombi originate in the soleal veins of the calf,often at sites of decreased blood flow such as thevalve cusps or bifurcations. The majority of calfthrombi resolve spontaneously, and PE is uncommon.Approximately 20 to 30% of DVTs propagateto the popliteal, femoral, or iliac veins. An additional10 to 20% of all DVTs occur in proximal veinswithout previous calf involvement. Iliofemoralthromboses appear to be the source of most clinicallyapparent PEs.The occurrence of DVT can be self-limited, withresolution of the clot in most cases; can embolizeresulting in PE; and/or the postthromboticsyndrome of venous insufficiency and the accompanyingstasis changes can occur. The postthromboticsyndrome can occur in up to one half ofpatients with acute DVT. Its frequency is likelyreduced by the use of compression stockings for 2years after the initial event, but the underlyingpredisposition for the disorder is very poorlyunderstood. Therefore, the prevention of DVT willprevent these complications.Detection of DVTThe clinical diagnosis of DVT of the lowerextremity is insensitive and nonspecific, with DVTrates ranging from 10 to 25% of those suspected ofhaving the disease. Even when the classic signsand symptoms of thrombophlebitis are present,only 45% of patients are found to have DVT byvenography. Therefore, patients presenting withsuspected DVT should first have the pretest probabilityof disease determined. The Wells clinicalprediction rule (Table 1) has been assessed andvalidated in multiple studies, and it can accuratelycategorize patients into low, moderate, or highprobability of having the disease. Obtaining ad-dimer level at this point in the evaluation furthersimplifies the approach and may obviate the needfor further testing. Sensitive d-dimer assays (ie,enzyme-linked immunosorbent assay [ELISA] orsemiquantitative rapid ELISA) have sufficientTable 1. Simplified Clinical Model for Assessment of DVT (Wells Rule)*Clinical VariableScoreActive cancer (treatment ongoing or within previous 6 mo or palliative) 1Paralysis, paresis, or recent plaster immobilization of the lower extremities 1Recently bedridden for 3 d or major surgery within the previous 12 wkrequiring general or regional anesthesia 1Localized tenderness along the distribution of the deep venous system 1Entire leg swelling 1Calf swelling at least 3 cm larger than that on the symptomatic leg (measured 10 cm below the tibial tuberosity) 1Pitting edema confined to the symptomatic leg 1Collateral superficial veins (nonvaricose) 1Previously documented DVT 1Alternative diagnosis at least as likely as DVT 2*High probability if score is 3; moderate if score is 1 to 2; and low if score is 0. Adapted from Wells PS, Owen C, DoucetteS, et al. Does this patient have deep vein thrombosis? JAMA 2006; 265:199−207.22 <strong>Pulmonary</strong> Vascular Diseases (Moores)ACC-PULM-09-0302-002.indd 227/10/09 8:03:53 PM


sensitivity and predictive value to rule out DVT inpatients with a low or moderate clinical probability.The false-negative result rate is too high to beuseful in patients with a high clinical probability.In those patients and in any patients in the low-riskor moderate-risk groups who have an abnormald-dimer level, imaging of the extremity isnecessary.The “gold standard” imaging test for DVT iscompression ultrasound. The diagnosis of DVTusing compression ultrasonography is made byfindings such as abnormal compressibility or lackof compressibility of the vein and abnormal Dopplercolor flow study results. Studies have demonstratedthat a lack of compressibility of a veinis highly sensitive ( 95%) and specific ( 95%)for proximal vein thrombosis. Ultrasound imagingis limited in that it does not detect isolatedcalf vein thrombi, and serial studies may need tobe performed if the initial test result is negativeand the clinical probability is high. A repeat studyfinding that is negative 5 to 7 days later is associatedwith a 1% incidence of subsequent VTEduring the next 3 months. Other tests used in theevaluation of suspected DVT are shown inTable 2. A recent evaluation of the ProspectiveInvestigation of <strong>Pulmonary</strong> Embolism Diagnosis(PIOPED) II data as well as metaanalyses of publishedtrials have shown that CT venography isdiagnostically equivalent to compression sonography.In patients who also have symptoms suggestiveof PE, it may be prudent to perform asingle diagnostic test.PENatural History of PEThe resolution of PE with the reestablishmentof vascular patency begins almost immediatelyafter PE. The National Institutes of Health trialsdemonstrated that with heparin therapy alone, theresolution of PEs occurred during the first severalweeks after the event, with 36% of the vasculardefects resolving by day 5, 52% resolving by day14, 73% resolving by 3 months, and 76% resolvingby 1 year for patients who survive the initial event.Untreated PE has a mortality rate of nearly 30%. Iftreatment is initiated, the mortality rate is 8%.Chronic Thromboembolic <strong>Pulmonary</strong>HypertensionChronic thromboembolic pulmonary hypertension(CTEPH) will develop in approximately 3 to4% of patients with acute PE. It is important todetect this complication because it is the onlypotentially curable form of severe PH without theneed for lung transplantation. A gradual increasein pulmonary vascular resistance occurs in theabsence of documented recurrent PE, and it isthought to be related to an increased resistance toflow through the pulmonary arteries that resultsfrom the initial obstruction and subsequently fromvascular remodeling in small, unobstructed vessels.Risk factors for CTEPH have been identifiedand include ventriculoatrial shunts, infectedTable 2. Diagnostic Tests for DVT of the Lower ExtremitiesTest Invasiveness Complications CommentsContrast venography Most invasive Phlebitis, allergic Reference or “gold standard”reactions, renal failureRadionuclide venography Less invasive Minimal Sensitive for proximal DVT, notcommonly used in clinical practiceImpedance plethysmography Noninvasive Minimal Sensitive/specific for proximal DVT, limitedin congestive heart failure setting andprevious DVTFibrinogen scanning Less invasive Minimal Sensitive for calf vein thrombosis, notcommonly used in clinical practiceReal-time Doppler ultrasound Noninvasive Minimal Sensitive/specific for proximal DVT,operator dependent; clinical “goldstandard”CT venography Noninvasive Allergic reactions Diagnostically equivalent<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 23ACC-PULM-09-0302-002.indd 237/10/09 8:03:53 PM


pacemakers, splenectomy, previous VTE, recurrentVTE, malignancy, thyroid-replacement therapy,and the antiphospholipid antibody syndrome.Patients who experience a major central thromboembolicevent, or have significant hemodynamiccompromise on presentation, a documented thrombophilia,or persistent abnormalities seen on lungperfusion studies on follow-up may need to bemonitored more closely. Ventilation/perfusion(V . /Q . ) scanning is the screening investigation ofchoice because many chest radiologists are notfamiliar with the distinguishing features of CTEPHseen on CT angiography. Patients who are surgicalcandidates should be referred to a specialty centerfor the consideration of pulmonary endarterectomy.All patients should receive life-long anticoagulationtherapy.Diagnosis of PEClinical Suspicion: The clinical suspicion of PEis often arrived at using an unstructured approachbased on a combination of factors, including thepresence or absence of identifiable risk factors (eg,surgery in the past, obesity, or previous VTE);symptoms (eg, dyspnea, pleuritic chest pain, andhemoptysis); physical examination findings (eg,tachypnea); the results of basic laboratory studies(eg, hypoxemia); and the likely presence of alternativediagnoses (eg, asthma, pneumonia, or congestiveheart failure). As with DVT, there are nowprediction rules that can be applied. The two validatedrules are shown in Table 3. It should beemphasized that the tool or approach used is lessimportant than the idea that the clinical pretestprobability of disease must be determined in eachpatient before further testing. As noted in the sectionon “DVT,” moderately to highly sensitive d-dimer assays have a sensitivity of 96 to 98% inpatients with suspected PE, which is sufficient torule out the disease in those patients with anunlikely pretest probability. In patients with a highpretest probability, imaging studies should beperformed instead of the d-dimer assay.Diagnostic Imaging Studies for PEMany studies have been used in diagnosingacute PE. Until the mid-1990s, the first-line imagingstudy was the V . /Q . scan. A normal V . /Q . scanTable 3. Clinical Prediction Rules for Acute PE*Canadian (Wells) Clinical Prediction ScoreVariable and scoreSigns and symptoms of DVT, 3.0PE as or more likely than an alternative diagnosis, 3.0Heart rate 100 beats/min, 1.5Immobilization or surgery in previous 4 wk, 1.5Previous DVT or PE, 1.5Hemoptysis, 1.0Cancer, 1.0Total score 2.0, low pretest probability2.0 to 6.0, moderate pretest probability 6.0, high pretest probabilityDichotomized Wells score 4.0 PE unlikely 4.0 PE likelyRevised Geneva scoreVariable and scoreAge 65 yr, 1.0Previous DVT or PE, 3.0Surgery or lower-limb fracture in previous week, 2.0Active cancer, 2.0Unilateral lower limb pain, 3.0Hemoptysis, 2.0Heart rate75 to 94 beats/min, 3.0 95 beats/min, 5.0Pain on leg palpation or unilateral edema, 4.0Total score0 to 3, low pretest probability4 to 10, moderate pretest probability 11, high pretest probability*Adapted from Tapson VF. Acute pulmonary embolism. NEngl J Med 2008; 358:1037–1052.finding provides compelling evidence against thediagnosis of PE. In one study of 515 consecutivepatients with clinically suspected PE who hadanticoagulation therapy withheld on the basis of anormal perfusion scan finding, only 3 patients hadsymptomatic VTE (PE, 1 patient) during a 3-monthfollow-up period. When the PIOPED criteria areused, a high-probability V . /Q . lung scan findingaccompanied with a high prescan clinical suspicionis associated with confirmed PE in 96% of cases.V . /Q . scan patterns other than normal or high-probabilitypatterns will require additional diagnosticevaluation.Spiral CT pulmonary angiography (CTPA) ofthe pulmonary circulation has emerged as the24 <strong>Pulmonary</strong> Vascular Diseases (Moores)ACC-PULM-09-0302-002.indd 247/10/09 8:03:53 PM


primary diagnostic method for the evaluation ofPE. As a minimally invasive examination, thistechnique is widely available and has replaced theuse of conventional pulmonary angiography inmost centers. However, its use in detecting smallperipheral PEs is unproven (6% of cases in thePIOPED study). Such small clots may not be importantphysiologically, but their detection may beimportant markers for future VTE. Several analyseshave suggested that CT scanning may be a morecost-effective initial test for suspected PE comparedwith the V . /Q . lung scan because of its greater sensitivityfor PE and its ability to identify alternativediagnoses to PE. In addition, the use of CT scanningfor the estimation of clot burden and rightventricular function is emerging as a useful prognosticmeasure in patients with documented acutePE.The use of magnetic resonance angiographyfor the detection of PE is in the initial testingphases. However, the initial results are promising,especially with gadolinium enhancement ofthe vasculature. A major limitation of the study isthe need for an experienced radiologist to interpretthe study. In addition, breath holds of 20 to 30 s arerequired for lung imaging, which limits the efficacyof the test in patients with acute PE. Newer scannersmay reduce the need for breath holds beyond10 to 15 s.<strong>Pulmonary</strong> angiography was considered to bethe reference “gold standard,” especially with thesubselective injection of contrast media and theuse of magnified views. A negative finding for apulmonary angiogram with magnificationappears to exclude clinically relevant PE, basedon data from two follow-up studies. The mortalityrate for patients undergoing the procedure is 0.5%, and the morbidity rate is about 5% (usuallycaused by catheter insertion and contrastreactions). Because of the expense and invasivenessof pulmonary angiography, alternative diagnosticalgorithms have been sought. Despite thelack of an independent “gold standard,” outcomestudies have suggested that spiral CTPA is asaccurate as pulmonary angiography. On the basisof the current evidence, the suggested approachfor the diagnosis of suspected acute PE is shownin Figure 1.Figure 1. Diagnostic algorithm for acute PE. From: Tapson VF. Acute pulmonary embolism. N Engl J Med 2008; 358:1037−1052.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 25ACC-PULM-09-0302-002.indd 257/10/09 8:03:53 PM


Treatment of VTEThe goals of short-term treatment of VTE areto reduce the short-term and long-term complications,such as extension of the thrombus, fatal PE,early or late recurrence, postthrombotic syndrome,and CTEPH. When patients present with acuteVTE, they should be administered parenteral anticoagulationtherapy with IV unfractionated heparin(UFH) or subcutaneous low-molecular-weightheparin (LMWH) or the pentasaccharidefondaparinux. Oral warfarin therapy can be initiatedon the first day of treatment. Treatment usingboth agents should be continued for at least 5 daysand until the international normalized ratio staysin the therapeutic range of 2.0 to 3.0 for 2 consecutivedays (at which time the heparin or fondaparinuxtherapy can be discontinued). Compared withtreatment of warfarin alone, overlap therapy withheparin reduces the risk of recurrence. A validatedweight-based IV nomogram achieves a therapeuticresponse more effectively than empiric heparindosing; thus, it is recommended that such a nomogrambe used if UFH therapy is used. The mostwidely used test for monitoring heparin therapyis the activated partial thromboplastin time, whichis a global anticoagulation test and does not measureheparin levels. Failure to achieve an adequateanticoagulation response with heparin therapy (ie,activated partial thromboplastin time, 1.5 timescontrol) is associated with the increased recurrenceof VTE.The LMWH preparations and pentasaccharidesshare advantages over UFH, as follows: greaterbioavailability, more predictable dosing, and alower risk of HIT. These advantages negate the needfor monitoring in most instances, although measuringthe anti-factor Xa level in morbidly obesepatients or in those with significantly impairedrenal function can be considered. In patients withsuspected or documented HIT, treatment with adirect thrombin inhibitor, such as argatroban orlepirudin, should be considered. Treatment withwarfarin should not be initiated until the plateletcount has returned to normal, especially in thosewith active thrombosis, as there is the potential forworsening thrombotic complications.Bed rest is not recommended for the treatmentof acute DVT unless there is significant pain andswelling. Many patients with DVT can be treatedentirely as outpatients. Although some patientswith PE likely can also be treated as outpatients,the data for this population are less robust.Recently, prognostic scores in patients with acutePE have been developed and validated and mayhelp to accurately identify low-risk patients.Patients with acute VTE require prolongedanticoagulation to prevent extension and recurrence.The current guidelines recommend 3 monthsof therapy for patients in whom thrombosis developsin the setting of documented transient riskfactors. In patients who experience idiopathic(unprovoked) events, treatment should be givenfor at least 3 months. At that point, patients shouldbe evaluated for the risk-benefit ratio of continuingtherapy indefinitely. Studies suggest that d-dimerlevels, persistent thrombus on ultrasound imaging,thrombin levels, and perhaps other markers ofchronic inflammation may help to identify thosepatients who are at the greatest risk of recurrenceafter stopping anticoagulation therapy, but we arenot yet able to use these in a systematic fashionthat allows us to individualize the duration oftherapy. Therefore, in the absence of a high bleedingrisk, most patients with an unprovoked eventshould receive indefinite therapy. There is no evidencethat longer (but finite) initial courses oftherapy (such as 6, 12, or 18 months) are moreeffective than 3 months because the recurrencerates upon discontinuation of therapy are identical.In patients with continuing risk factors (such asactive malignancy) and patients who have had adocumented recurrent event, long-term (perhapslifelong) therapy is recommended.Thrombolytic TherapyThrombolytic therapy is approved for the treatmentof proximal DVT and massive PE. Theseagents dissolve thrombi by activating plasminogeninto plasmin. Plasmin degrades fibrin to solublepeptides in the presence of a thrombus or hemostaticplug. Streptokinase, urokinase, and tissueplasminogen activator are approved for use in theUnited States. Thrombolytic treatment of DVT hasbeen associated with decreased pain, swelling, lossof venous values, and a reduced incidence of postphlebiticsyndrome. For the treatment of PE,thrombolytic agents produce more rapid resolutionof intravascular abnormalities than have been26 <strong>Pulmonary</strong> Vascular Diseases (Moores)ACC-PULM-09-0302-002.indd 267/10/09 8:03:53 PM


demonstrated by either V . /Q . lung scans or pulmonaryangiography. However, no proven effect onmortality has yet been demonstrated. Thrombolytictherapy is currently recommended for patientswith acute PE who show signs of hemodynamicinstability, unless there are major contraindicationsrelated to the bleeding risk.Inferior Vena Cava InterruptionInferior vena cava interruption is most commonlyachieved with the use of an intravascularfilter. A randomized trial found that the use of aninferior vena cava filter reduced the rate of PE, hadno effect on mortality, and was associated with agreater risk of recurrent DVT. These devices arebeing used more frequently, especially now thatmany can be safely retrieved after several months.However, the approved indications remain few, asfollows: absolute contraindication to anticoagulation;life-threatening bleeding while receivinganticoagulation; and documented recurrent VTEwhile receiving therapeutic anticoagulation.Surgical EmbolectomyThis procedure should be reserved for patientswith confirmed massive PEs who are hemodynamicallyunstable and who are not candidates forthrombolysis. The high mortality rates associatedwith this procedure, especially if cardiopulmonaryarrest has occurred, has resulted in a limited enthusiasmfor its use. Interventional catheter extractionor fragmentation may become viable options forthese patients in the near future.Prevention of DVT/PEThe rationale for DVT and PE prophylaxis isbased on the clinically silent nature of the diseaseand the high prevalence of DVT among hospitalizedpatients. There is added importance for prophylaxiswhen one considers that most patientswho die of PE do so within 30 min of the acuteevent. The strategy used for the prevention of DVTand PE should be based on the patient’s level ofrisk, as suggested by the American College of ChestPhysicians Consensus Conference on AntithromboticTherapy. The current approach involvesassessing patient risk by assigning them to a riskgroup based upon their primary reason forhospitalization.Low-Risk PatientsThese patients can be defined as individualswho are undergoing uncomplicated minor surgeryand fully mobile medical patients with no clinicalrisk factors for VTE. Their risk of DVT withoutprophylaxis is 10%, and thus, no specific thromboprophylaxisis recommended.Moderate-Risk PatientsThese patients are most general, open gynecologicor urologic surgery patients or medicalpatients who are at bed rest (including all criticalcare patients). Their approximate risk of DVTwithout prophylaxis is 10 to 40%. These patientsshould receive prophylaxis with LMWH, low-doseUFH, or fondaparinux. Those with a very high riskof bleeding should receive mechanical thromboprophylaxis.High-Risk PatientsThese are patients undergoing hip or kneearthroplasty, hip fracture surgery, or patients whohave had major trauma or spinal cord injury. Therecommended prophylaxis is LMWH, fondaparinux,or an oral vitamin K antagonist (international normalizedratio 2 to 3). As in moderate-risk patients,if the bleeding risk is extremely high, mechanicalmethods of prophylaxis should be used. Pharmacologicprophylaxis should be started as soon asthe bleeding risk decreases.Although VTE is considered to be more commonin surgery or trauma patients, 50 to 70% ofsymptomatic thromboembolic events and 70 to80% of fatal PEs occur in nonsurgical patients.Hospitalization for an acute medical illness isindependently associated with a relative risk forVTE of about 8. Thus, the appropriate prophylaxisof medical inpatients is clearly as important as thatin surgical patients.Several risk factors for VTE have been identifiedin hospitalized medical patients. Major riskfactors include New York Heart Association(NYHA) functional class III and IV heart failure,COPD exacerbations, and sepsis. Additional risk<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 27ACC-PULM-09-0302-002.indd 277/10/09 8:03:54 PM


factors include advanced age, history of VTE, cancer,stroke with lower-extremity weakness, endstagerenal disease, and bed rest. Many medicalpatients have multiple risk factors. Unfortunately,no systematic way of combining these to identifyrisk categories in medical patients has beenaccepted and validated. Therefore, medical patientsare generally classified as low risk if they are fullymobile and have no additional risk factors for VTE.All other medical patients are considered to be atmoderate risk and should receive thromboprophylaxisas noted previously.<strong>Pulmonary</strong> Hypertension (PH)PH is a hemodynamic state that is defined bya mean pulmonary artery pressure (PAP) of 25 mm Hg at rest or 30 mm Hg during exercise,which is shared by many conditions. PH was previouslycategorized into primary or secondary PH,based on the absence or presence of identifiablecauses or associated conditions. A revised classificationsystem that acknowledged the separationof disorders directly affecting the arterial tree fromdisorders affecting the venous circulation or thoseaffecting the circulation by altering respiratoryfunction was proposed in 1998. A revision of thisclassification was proposed at the Third-WorldConference on <strong>Pulmonary</strong> Hypertension, held inVenice, Italy, in 2003 (Table 4). The major changesnoted in these revisions were as follows: (1) abandonmentof the term secondary PH and (2) thereplacement of the term primary PH with idiopathicpulmonary arterial hypertension (IPAH) or, whensupported by genetic testing, familial pulmonaryarterial hypertension (FPAH).PAH is a subset of PH that is defined as a meanPAP of 25 mm Hg at rest or 30 mm Hg duringexercise with a normal pulmonary capillary wedgepressure ( 15 mm Hg) and a lesion localized to thepulmonary arteriole. IPAH is an uncommon disorderaffecting women more commonly than menduring the ages 20 to 40 years. In approximately6% of patients, there is a family history of PAH (ie,FPAH), which has been found to be related to 1 of 70 identified mutations in the gene-encodingbone morphogenetic protein receptor type II(BMPR2), which is localized to chromosome 2q33.Sporadic mutations in this gene are seen in up to25% of patients with IPAH. Asymptomatic carriersTable 4. Revised Nomenclature and Classification of PH (2003)from the Third-World Conference on <strong>Pulmonary</strong> Hypertension*WHO group IPAHIPAHFPAHAssociated withCollagen vascular diseaseCongenital systemic to pulmonary shunts (large, small,repaired, or nonrepaired)Portal hypertensionHIV infection drugs and toxinsOther (glycogen storage disease, Gaucher disease,hereditary hemorrhagic telangiectasia, hemoglobinopathies,myeloproliferative disorders, splenectomy)Associated with significant venous or capillary involvement<strong>Pulmonary</strong> venoocclusive disease<strong>Pulmonary</strong> capillary hemangiomatosisWHO group II<strong>Pulmonary</strong> venous hypertensionLeft-sided atrial or ventricular heart diseaseLeft-sided valvular heart diseaseWHO group III<strong>Pulmonary</strong> hypertension associated with hypoxemiaCOPDInterstitial lung diseaseSleep-disordered breathingAlveolar hypoventilation disordersChronic exposure to high altitudeWHO group IVPH caused by chronic thrombotic and/or embolic diseaseThromboembolic obstruction of proximal pulmonaryarteriesThromboembolic obstruction of distal pulmonary arteries<strong>Pulmonary</strong> embolism (tumor, parasites, foreign material)WHO group VMiscellaneousSarcoidosis, histiocytosis X, lymphangiomatosis,compression of pulmonary vessels (adenopathy,tumor, fibrosing mediastinitis)*Adapted from Rubin LJ. Diagnosis and management ofpulmonary arterial hypertension: <strong>ACCP</strong> evidence-basedclinical practice guidelines. Chest 2004; 126(suppl):7S−10S.of the BMPR2 gene often have an abnormal increasein pulmonary artery (PA) pressures in response toexercise and, given additional environmental orgenetic stimuli, may develop overt disease. Thus,genetic testing and counseling should be offered topatients with either IPAH or FPAH and theirfamilies. The lack of a single genetic mutation, thelow penetrance of overt PAH in BMPR2 mutationcarriers, and the lack of a single environmentalstimulus have led to the development of the multiple-hithypothesis in the etiology of PAH (Fig 2).28 <strong>Pulmonary</strong> Vascular Diseases (Moores)ACC-PULM-09-0302-002.indd 287/10/09 8:03:54 PM


Figure 2. Multiple-hit hypothesis of pathogenesis of PAH: BMPR-II, bone morphogenic protein receptor type 2. From GabbayE, Reed A, Williams TJ. Assessment and treatment of pulmonary arterial hypertension: an Australian perspective in 2006. InternMed J 2006; 37:38−48.PAH related to risk factors or associated conditionsis a heterogeneous group of disorders, includingconnective tissue disease, congenital systemicpulmonary shunt, portal hypertension, HIV infection,drugs and toxin, and those conditions such ashemoglobinopathies and myeloproliferative disorders.The use of appetite-suppressant drugs for 3months has been associated with a 30-fold increasein the risk of development of PH. Hepatosplenicschistosomiasis may account for one of the mostprevalent forms of PAH worldwide. The WorldHealth Organization (WHO) also suggests thatother appetite stimulants such as amphetaminesmay have a very likely causative role in PH.low, the contribution of vasoconstriction relativeto vascular remodeling is minimal.The main cellular changes involve smoothmuscle cells, fibroblasts, endothelial cells, andplatelets. The following three pathways areimportant: prostacyclin (prostaglandin 12),nitric oxide (NO)−cyclic guanosine monophosphate-phosphodiesterase5, and endothelin. Animbalance between the up-regulated vasoconstrictive/proproliferativeendothelin system and thedown- regulated vasodilatory/antiproliferative/antithrombotic NO and prostaglandin 12 systemsPathogenesisThe current understanding of PAH is as a vasoproliferativedisease invoked by mitogenic stimuli.The pathogenesis of PAH is complex and poorlyunderstood and, as mentioned previously, includesboth genetic and environmental factors that altervascular structure and function. The vascularchanges involve the pulmonary arteriole, and theyare characterized by vasoconstriction, vascularremodeling with intimal and medial proliferation,the formation of plexiform lesions, and thrombosis(Fig 3). These changes lead to progressive obstructionof flow, increased pulmonary vascularresistance, and eventual right heart failure anddeath. Because the number of patients who respondsignificantly to acute vasodilator challenge is veryFigure 3. Typical micrograph of a lung from a patient withPAH. The patient has normal alveolar structures but a smallPA, in which the lumen is occluded by concentric fibrosis.The artery also has increased medial smooth muscle and areactive adventitia. Dilated small vessels border the airwayabove the artery. From Newman JH, Phillips JA, Loyd JE. <strong>Pulmonary</strong>hypertension. Ann Intern Med 2008; 148:278−283.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 29ACC-PULM-09-0302-002.indd 297/10/09 8:03:54 PM


is a key contributor. Platelets likely play an importantrole as procoagulants by increasing the plateletrelease of serotonin, vascular endothelial growthfactor, and platelet-derived growth factor.Clinical FeaturesPH frequently presents with nonspecific symptoms(Table 5). The most common symptoms aredyspnea on exertion, fatigue, and syncope, resultingfrom reduced cardiac output during activity.Patients may also occasionally present with anginalikechest pain with normal coronary arteries. Thischest pain may reflect right ventricular ischemia.Hemoptysis as a consequence of pulmonary vascularrupture is rare but may be a catastrophicevent. The Raynaud phenomenon is seen inapproximately 2% of patients with IPAH but is afrequent finding in cases of PH associated withconnective tissue diseases such as scleroderma. Thephysical examination findings are nonspecific butmay give clues to a secondary cause of PH. Withmore advanced disease, the physical examinationreveals signs of right ventricular dysfunction. TheECG most often reveals right atrial or right ventricularhypertrophy and right-axis deviation. Thechest radiograph shows an enlarged cardiac silhouettethat is consistent with right ventricularhypertrophy, with enlarged pulmonary arteries.Diagnostic ApproachThe most important step in the diagnosis ofPAH is to suspect its presence. As noted, the symptomsare vague and nonspecific, and PAH shouldbe considered in any patient with unexplaineddyspnea on exertion (and perhaps in all thoseTable 5. Signs and Symptoms of PHSignsJugular vein distentionProminent right ventricular heaveAccentuated pulmonic valvecomponent (P 2)Right-sided third heart sound (S 3)Tricuspid insufficiency murmurHepatomegalyPeripheral edemaSymptomsDyspnea or exertionFatigueSyncopeAnginal chest painHemoptysisRaynaud phenomenonpatients with an unexplained isolated reduction indiffusing capacity). There are no early symptomsof PAH, and thus, annual screening in high-riskpopulations should be considered. Risk groupsthought to benefit from screening include patientswith known genetic mutations, first-degree relativesof patients with PAH, patients with scleroderma,patients with portal hypertension beforeliver transplantation, and patients with congenitalsystemic to pulmonary shunts. The best screeningtest for the detection of PAH is a transthoracicechocardiogram (TTE). TTE is also a valuable toolto detect features of left-sided heart disease orintracardiac shunts. Increased PA systolic pressureraises the possibility of PH but is clearly not specificfor PAH. If the TTE confirms increased PAP,further testing to identify the underlying cause iswarranted.Spirometry, diffusing capacity, arterial bloodgas measurement, and CT scan of the lung can beperformed to exclude underlying respiratory diseaseand/or hypoxemia. When suggested by history,an overnight polysomnography can beperformed to assess for severe sleep apnea. Thetests for underlying liver disease, HIV, and autoimmunediseases are noninvasive and should beperformed in most patients with unexplained PH.A V . /Q . scan should be performed to evaluate forpotential CTEPH because patients who qualify forthromboendarterectomy can have a dramaticimprovement with this procedure.Right-heart catheterization is the “gold standard”for diagnosis and should be performed inall patients with suspected PH, especially if specifictreatment is planned. Cardiac catheterizationis indicated to establish severity and prognosis,measure wedge pressure, exclude congenital heartdisease, and test the response to selective pulmonaryvasodilators (eg, inhaled NO or IV adenosineor epoprostenol). Most authors recommend heartcatheterization of both the right and left sides.Left-sided heart catheterization can evaluate forcoronary artery disease, as well as measure leftventricular end-diastolic pressure. Right-heartcatheterization measurements should includesequential oxygen saturation for the presence ofan intracardiac shunt, pulmonary angiography forthromboembolic PH, hemodynamic measurementsincluding PAPs, PA occlusion pressure, andcardiac index. A positive response to therapy with30 <strong>Pulmonary</strong> Vascular Diseases (Moores)ACC-PULM-09-0302-002.indd 307/10/09 8:03:54 PM


vasodilators is defined as a decrease in mean PApressure of 10 mm Hg to reach a mean PA pressureof 40 mm Hg, with an increased orunchanged cardiac output. Very few patientsshow a short-term response, but a negative challengeresult will obviate a trial of calcium-channelblockers.Measurement of Exercise CapacityFormal assessment of exercise capacity is anintegral part of the evaluation. A 6-min walk testusually is performed to determine the degree ofexercise limitation because it has proven to bereproducible and to correlate well with other measuresof functional status. This test has been veryuseful in monitoring the response to therapy andhas often been used as a primary end point inclinical trials. Importantly, it has been found to bean independent predictor of survival in patientswith PAH.Treatment of PAHNo treatment for PAH has been found toachieve a cure for this devastating disease. Thereare six main types of treatment for PAH. The firstis prevention. Because some forms of PAH haveclear causal mechanisms, these factors should beeliminated where possible. The second line of treatmentis screening of high-risk patients (as previouslydiscussed) because it is generally believedthat earlier diagnosis and treatment may improveoutcomes. The third is to optimize the therapy forany related diseases, such as heart failure, hypoxemia,sleep disorders, or collagen vascular diseases.The fourth line of therapy is supportive—directedat the consequences of PAH. General measuresinclude pneumococcal and influenza vaccinationsin addition to the avoidance of pregnancy, highaltitudeexposure, tobacco, and medications suchas appetite suppressants, decongestants, and nonsteroidalantiinflammatory agents. In addition, thefollowing supportive treatments are available:1. Anticoagulation: Patients with PAH shouldreceive long-term anticoagulation therapy.Patients with this condition are prone to thromboembolism,in part caused by reduced PAblood flow, right ventricular dilation, venousinsufficiency, and inactivity. The range of anticoagulationthat is recommended is an internationalnormalized ratio of 1.5 to 2.5.2. Oxygen: All patients with PAH and documentedhypoxemia should receive supplementaloxygen to maintain an oxygen saturation of 90%.3. Diuretics: Careful diuresis is indicated inpatients with evidence of right ventricularfailure.4. Inotropic agents: There are no data on thelong-term use of inotropic therapy for PAH.Some authors have suggested the use of digitalisbecause it may improve right ventricularfunction. The use of cardiac glycosides is controversialbecause of the lack of well-designedrandomized controlled trials in this population.The fifth line of treatment is vascular-targetedtherapy aimed at reversing or reducing vasoconstriction,endothelial cell proliferation, and smoothmusclecell proliferation. In patients who exhibitevidence of an acute hemodynamic response tovasodilator challenge testing, long-term treatmentwith orally administered calcium-channel blockersin relatively high doses can occasionally producea sustained hemodynamic response with a reductionin PAP and an increased cardiac output. Thisgroup of responders not only shows improvementin symptoms in exercise tolerance, but they alsohave increased survival. Unfortunately, calciumchannelblockers have no antiproliferative effect,and thus, only a small subset of patients will benefitfrom long-term use. In patients who do notshow a reduction in PAP (with the administrationof vasodilators) during right-heart catheterization,it is not beneficial and is potentially dangerous toinitiate therapy with calcium-channel blockers.Significant adverse effects include systemic hypotension,pulmonary edema, and right ventricularfailure.Current US Food and Drug Administrationapprovedand investigational therapies aimed atreducing endothelial and smooth-muscle cell proliferationtarget the following three major pathwaysinvolved in the pathogenesis of PAH:prostacyclin, NO, and endothelin. Continuous IVepoprostenol, subcutaneous and IV treprostinil,and inhaled iloprost are prostanoids that havebeen shown to improve exercise capacity, functional<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 31ACC-PULM-09-0302-002.indd 317/10/09 8:03:55 PM


capacity, hemodynamics and, in the case of epoprostenol,survival. In the NO pathway, sildenafil(which inhibits phosphodiesterase 5) has demonstratedimprovement in exercise capacity, functionalstatus, and hemodynamics. Bosentan is anoral nonselective endothelin receptor antagonist;sitaxsentan and ambrisentan are selective endothelinA receptor antagonists. These therapies havebeen shown to improve exercise capacity, functionalstatus, and time to clinical worsening. Bosentanhas also been shown to improve hemodynamics.Two studies have shown improved survival withbosentan therapy compared with historical controlsubjects. Although many of the trials that use theseclasses of medications were not powered to detectan improvement in survival, a recent metaanalysissuggests that the rate of mortality is lower inpatients treated with targeted therapies.The approach to the use of these agents in PAHis typically based on their functional class, asdefined by the WHO classification system, whichis a modification of the NYHA functional classsystem for heart failure (Table 6). In general,patients who are in functional class I receive generalcare only, although trials with early use ofspecific therapy are currently underway. Oraltherapies are considered to be first-line agents forpatients who are in functional class II and class III,whereas parental therapies generally are reservedfor patients who are in class IV. There is no clearconsensus as to which agent to start first, nor isthere any current consensus as to the role of combinationtherapy.Few studies have been published that directlycompare individual medications or classes ofmedications. Multidrug treatment is appealingbecause the three classes of medications approvedby the Food and Drug Administration exert theireffects by different mechanisms. The use of two ormore drugs may allow for dosing below the levelsthat cause important side effects. A few small studieshave confirmed this theory, but more data areneeded before firm recommendations can be made.Authors of the American College of Chest Physiciansevidence-based guideline published updatedguidance for medical therapy for PH in June 2007.Their overall recommended approach is shown inFigure 4. If pharmacologic treatment fails, surgical(the sixth line) treatment (lung transplantation oratrial septostomy) should be considered.PrognosisMedian duration of survival of patients diagnosedwith PAH between 1980 and 1985 was2.8 years. Since that time, survival has improved.Patients without evidence of right ventricularfailure may survive 10 years. Responders tocalcium-channel blockers have a 95% 5-year survivalrate. Patients in NYHA classes II and IV whohave been treated with epoprostenol have a 5-yearsurvival rate that is twice that of matched controlpatients.Patients with evidence of right heart failurehave a much lower survival rate. Posttreatmentmarkers of poor prognosis included persistentincreases in right atrial pressure, low cardiac index,low mixed venous oxygen saturation, continuedfunctional class III or IV symptoms, poor exercisecapacity (6-min walk test distance 380 m), andTable 6. WHO Classification of Functional Status of Patients with PH*Class 1Class 2Class 3Class 4Patients with PH but without resulting limitation of physical activity. Ordinary physical activity does not causeundue dyspnea or fatigue, chest pain, or near syncope.Patients with PH resulting in slight limitation of physical activity. These patients are comfortable at rest, but ordinaryphysical activity causes undue dyspnea or fatigue, chest pain, or near syncope.Patients with PH resulting in marked limitation of physical activity. These patients are comfortable at rest, butless than ordinary physical activity causes undue dyspnea or fatigue, chest pain, or near syncope.Patients with PH resulting in inability to perform any physical activity without symptoms. These patients manifestsigns of right-heart failure. Dyspnea and/or fatigue may be present at rest, and discomfort is increased byany physical activity.*Modified from the NYHA classification of patients with cardiac disease. From McGoon, Gutterman D, Steen V, et al. Screening,early detection, and diagnosis of pulmonary arterial hypertension: <strong>ACCP</strong> evidence-based clinical practice guidelines.Chest 2004; 126:18S.32 <strong>Pulmonary</strong> Vascular Diseases (Moores)ACC-PULM-09-0302-002.indd 327/10/09 8:03:55 PM


Figure 4. Treatment algorithm for PAH. The recommended therapies presented in this algorithm have been evaluated mainlyin those patients with IPAH or PAH associated with connective tissue disease or anorexigen use. Extrapolation to other forms ofPAH should be made with caution. Country-specific regulatory agency approval status and functional class indications for PAHmedications vary. From: Badesch DB, Abman SH, Simonneau G, et al. Medical therapy for pulmonary arterial hypertension:updated <strong>ACCP</strong> evidence-based clinical practice guidelines. Chest 2007; 131:1917−1928.increased serum B-type natriuretic peptide levels.Patients with persistent poor functional class orexercise capacity, severely reduced right heart function,a right atrial pressure 15 mm Hg, or cardiacindex 2l/min/m 2 should be considered for lungtransplantation. Either single- or double-lungtransplantation are options, but surgeons in mostcenters prefer double transplantation because ofconcerns about postoperative reperfusion injuryafter a single-lung transplant. Experience hasshown that lung transplant recipients with primaryPH have survival rates of 73% at 1 year, 55% at 3years, and 45% at 5 years. Obliterative bronchiolitisis a common complication in transplant patientswith PAH. The recurrence of PAH after lung transplantationhas not been reported in the transplantedlung.Annotated BibliographyBadesch DB, Abman SH, Ahearn GS, et al. Medicaltherapy for pulmonary arterial hypertension: <strong>ACCP</strong>evidence-based clinical practice guidelines. Chest 2004;126(suppl):35S−62SThis article was included in the July 2004 supplement toCHEST, which was devoted to the diagnosis and managementof PAH. Patients with IPAH should undergo acutevasoreactivity testing with use of a short-acting agentsuch as IV epoprostenol or adenosine or inhaled NO.Patients with IPAH should receive anticoagulation withwarfarin.Badesch DB, Abman SH, Simonneau G, et al. Medicaltherapy for pulmonary arterial hypertension: updated<strong>ACCP</strong> evidence-based clinical practice guidelines.Chest 2007; 131:1917−1928<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 33ACC-PULM-09-0302-002.indd 337/10/09 8:03:55 PM


This article is an update to the one before it, taking intoaccount the rapidly evolving field of therapeutic trials inPAH. The authors give a nice review of the current evidencefor each potential agent and a suggested algorithm for theiruse in patients with PAH.Budhiraja R, Hassoun PM. Portopulmonary hypertension:a tale of two circulations. Chest 2003; 123:562−576<strong>Pulmonary</strong> involvement is common in patients withportal hypertension and can manifest in diverse ways.Changes in pulmonary arterial resistance, manifestingeither as the hepatopulmonary syndrome or portopulmonaryhypertension, have been increasingly recognized.Portopulmonary hypertension is defined as an elevatedPAP in the setting of increased pulmonary vascular resistanceand normal wedge pressure in a patient with portalhypertension.Decousus H, Leizorovicz A, Parent F, et al. A clinicaltrial of vena caval filters in the prevention of pulmonaryembolism in patients with proximal deep-veinthrombosis. N Engl J Med 1998; 338:409−415In high-risk patients with proximal DVT, the initial beneficialeffect of vena caval filters for the prevention of PE wascounterbalanced by an excess of recurrent DVT without anydifference in mortality.DeMonye W, Sanson BJ, MacGillavry MR, et al. Emboluslocation affects the sensitivity of a rapid quantitative d-dimer assay in the diagnosis of pulmonary embolism.Am J Respir Crit Care Med 2002; 165:345−348d-dimer concentration and the accuracy of d-dimer assaysare dependent on embolus location; smaller, subsegmentalemboli may be missed when d-dimer assays are used as thesole test to exclude PE.Douketis JD, Kearon C, Bates S, et al. Risk of fatal pulmonaryembolism in patients with treated venousthromboembolism. JAMA 1998; 279:458−462Among patients with symptomatic PE or DVT who aretreated with anticoagulation therapy for 3 months, the occurrenceof fatal PE is rare during and after anticoagulant therapy.However, patients presenting with PE are more likely todie of recurrent PE than are patients presenting with DVT.Gabbay E, Reed A, Williams TJ. Assessment and treatmentof pulmonary arterial hypertension: an Australianperspective in 2006. Intern Med J 2007; 37:38−48This article, although focusing on the Australian experiencein managing patients with PH, is a very nice overview ofthe current classification, pathogenesis, diagnosis, and treatmentof PH.Galie N, Rubin LJ, eds. <strong>Pulmonary</strong> arterial hypertension:epidemiology, pathobiology, assessment, andtherapy. J Am Coll Cardiol 2004; 43(suppl):1S−90SThis entire supplement is devoted to PH, with articles bynoted authorities.Geerts WH, Bergqvist D, Pineo GF, et al. Prevention ofvenous thromboembolism. American College of ChestPhysicians evidence-based clinical practice guidelines.8th ed. Chest 2008; 133:381S−453SEvidence-based recommendations on prevention of VTEin all surgical and medical populations. Excellent reviewof the literature on risk, risk stratification, and preventivestrategies.Goldhaber SZ, Grodstein F, Stampfer MJ, et al. A prospectivestudy of risk factors for pulmonary embolismin women. JAMA 1997; 277:642−645Obesity, cigarette smoking, and hypertension were associatedwith an increased risk of PE in women. Control of theserisk factors could decrease the risk of PE among women.Hayashino Y, Goto M, Noguchi Y, et al. Ventilation-perfusionscanning and helical CT in suspectedpulmonary embolism: meta-analysis of diagnosticperformance. Radiology 2005; 234:740−748Helical CT scanning has better discriminatory power thanV . /Q . scanning for excluding PE, but the two tests have asimilar discriminatory power in the diagnosis of PE whenthe V . /Q . scan results disclose a “high probability.”Heit JA, Kobbervig CE, James AH, et al. Trends in theincidence of venous thromboembolism during pregnancyor postpartum: a 30-year population-basedstudy. Ann Intern Med 2005; 143:697−706Among pregnant women, the greatest risk for PE and VTEis during the postpartum period. Prophylaxis should beespecially targeted to postpartum women.Highland KB. <strong>Pulmonary</strong> arterial hypertension. Am JMed Sci 2008; 335:40−45This is an excellent succinct, current review of PH by aclinical expert in the field.Humbert M. Update in <strong>Pulmonary</strong> Hypertension 2008.Am J Respir Crit Care Med 2009; 179:650−656.This article summarizes the advances in pulmonary hypertensionas reported in the medical literature over the past24 months.Kearon C, Kahn SR, Agnelli G, et al. Antithrombotictherapy for venous thromboembolic disease. AmericanCollege of Chest Physicians evidence-based clinicalpractice guidelines. 8th ed. Chest 2008; 133:454S−545S.Comprehensive evidence-based literature review and gradedrecommendations on treatment of DVT and PE.Kearon C, Ginsberg JS, Hirsh J. The role of venousultrasonography in the diagnosis of suspected deepvenous thrombosis and pulmonary embolism. AnnIntern Med 1998; 129:1044−104934 <strong>Pulmonary</strong> Vascular Diseases (Moores)ACC-PULM-09-0302-002.indd 347/10/09 8:03:55 PM


Venous ultrasonography is a valuable test for the diagnosisand management of patients with suspected DVT or PE.However, the following factors reduce the positive predictivevalue of venous ultrasonography for detecting acutethromboembolism: asymptomatic for DVT, low clinicalsuspicion for DVT, abnormality confined to a short segmentof a proximal vein, abnormality confined to the calfveins; history of VTE, negative result of another test thatis sensitive for DVT (such as d-dimer), and low prevalenceof DVT in a referral population. (With one or more of theaforementioned factors, venography or other testing shouldbe considered to exclude the possibility of a false-positivetest result.)Mandel J, Mark EJ, Hales C A. <strong>Pulmonary</strong> venoocclusivedisease. Am J Respir Crit Care Med 2000;162:1964−1973This review summarizes the current state of knowledgeregarding pulmonary venoocclusive disease.McGoon M, Gutterman D, Steen V, et al. Screening,early detection, and diagnosis of pulmonary arterialhypertension: <strong>ACCP</strong> evidence-based clinical practiceguidelines. Chest 2004; 126(suppl):14S−34SA comprehensive evidence-based review and recommendationson the diagnostic approach to patients with suspectedPH.McGoon MD, Kane GC. <strong>Pulmonary</strong> hypertension:diagnosis and management. Mayo Clin Proc 2009;84:191−207This is a recent, well-organized clinical review that includesdiscussion of several controversial issues, such as the importanceof mild or exercise-induced PAH, disproportionatePAH, assessment of therapeutic efficacy, combination therapy,and the management of the symptoms other than dyspnea(angina, syncope, and arrhythmias).McNeil K, Dunning J. Chronic thromboembolic pulmonaryhypertension (CTEPH). Heart 2007; 93:1152−1158A description of the natural history, pathophysiology, diagnosis,and treatment of CTEPH are included. The authorsunderscore the fact that this disease is more common thanpreviously thought, and they stress the importance of identifyingit in patients with newly diagnosed PH.Moores LK, Holley AB. Computed tomography pulmonaryangiography and venography: diagnostic andprognostic properties. Semin Respir Crit Care Med2008; 29:3−14An up-to-date review of the diagnostic accuracy and utility,as well as the prognostic features, of spiral CT angiographyand venography.Moores LK, Jackson WL Jr, Shorr AF, et al. Meta-analysis:outcomes in patients with suspected pulmonaryembolism managed with computed tomographicpulmonary angiography. Ann Intern Med 2004;141:866−874The rate of subsequent VTE after negative results of spiralCTPA is similar to that seen after negative results on conventionalpulmonary angiography. It appears safe to withholdanticoagulation therapy after negative CTPA results.Newman JH. <strong>Pulmonary</strong> hypertension. Am J RespirCrit Care Med 2005; 172:1072−1077This “Centennial <strong>Review</strong>” discusses the birth of modern cardiopulmonarymedicine, including pulmonary circulation,primary PH, and hypoxemia associated with PH.Newman JH, Phillips JA, Loyd JE. Narrative review:the enigma of pulmonary arterial hypertension; newinsights from genetic studies. Ann Intern Med 2008;148:278−283A short, easily understandable description of the discoveryand meaning of genetic predispositions to the developmentof PAH.Quinlan DJ, McGuillan A, Eikelboom JW. Lowmolecular-weightheparin compared with intravenousunfractionated heparin for treatment of pulmonaryembolism: a meta-analysis of randomized, controlledtrials. Ann Intern Med 2004; 140:175−183LMWH appears to be similar to UFH in safety and efficacyfor the treatment of either symptomatic or asymptomaticPE.Quiroz R, Kucher N, Zou KH, et al. Clinical validityof a negative computed tomography scan in patientswith suspected pulmonary embolism: a systematicreview. JAMA 2005; 293:2012−2017A negative CT scan finding is as accurate as pulmonaryangiography in ruling out suspected PE. A negative CTscan finding in a low-risk patient rules out PE, whereas anegative CT scan finding in a high-risk patient may requirefurther confirmation.Rubin LJ. American College of Chest Physicians.Diagnosis and management of pulmonary arterialhypertension: <strong>ACCP</strong> evidence-based clinical practiceguidelines. Chest 2004; 126(suppl):75S−105SIn this brief introduction to the guidelines, Dr. Rubin givesan excellent review of the evolution and current classificationof PH.Schulman S, Granquvist S, Holmstrom M, et al. Theduration of oral anticoagulant therapy after a secondepisode of venous thromboembolism. N Engl J Med1997; 336:393−398Prophylactic anticoagulation that was continued for anindefinite period after a second episode of VTE was associatedwith a much lower rate of recurrence during 4 years<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 35ACC-PULM-09-0302-002.indd 357/10/09 8:03:56 PM


of follow-up than treatment for 6 months. However, therewas a trend toward a greater risk of major hemorrhage whenanticoagulation was continued indefinitely.Sorensen HT, Mellemkjaeer L, Steffensen FH, et al. Therisk of a diagnosis of cancer after primary deep venousthrombosis or pulmonary embolism. N Engl J Med1998; 338:1169−1173Among 15,348 patients with DVT and 11,305 patients withPE, there were 1,727 cases of cancer. The expected numberof cancer cases was 1,372 based on national statistics fromDenmark (ratio of observed/expected number of cancer cases,1.3; 95% confidence interval, 1.21 to 1.33). Forty percent ofpatients with a diagnosis of cancer within 1 year of hospitalizationfor VTE had distant metastases at the time of thediagnosis of cancer. The authors conclude that an aggressivesearch for hidden cancer in patients presenting with VTE isnot warranted.Stein PD, Fowler SE, Goodman LR, et al. Multidetectorcomputed tomography for acute pulmonary embolism.N Engl J Med 2006; 354:2317−2327In patients with suspected acute PE, multidetector CT angiography(CTA) in combination with venous phase multidetectorCT venography (CTV) was more sensitive than CTAalone for diagnosing PE. However, it only increased the negativepredicted value from 95 to 97% overall. Therefore, theadded dose of radiation from CTV may not be justified whenadded to the dose from CTA. The multidetector CTA may beaccurate enough to eliminate PE, at least in patients withlow-to-intermediate clinical probability of PE.Stein PD, Hull RD, Patel KC, et al. D-dimer for theexclusion of acute venous thrombosis and pulmonaryembolism: a systematic review. Ann Intern Med 2004;140:589−602For the diagnosis of DVT, the ELISA and quantitativerapid ELISA are more sensitive than latex and wholebloodagglutination assays. For the diagnosis of PE, theELISA and quantitative rapid ELISA are more sensitivethan semiquantitative latex and whole-blood agglutinationassays.Stein PD, Woodard PK, Weg JG, et al. Diagnostic pathwaysin acute pulmonary embolism: recommendationsof the PIOPEDII investigators. Am J Med 2006;119:1048−1055The PIOPED II investigators recommend a stratification ofpatients with suspected PE according to a clinical probabilityassessment. The investigators conclude that the sequenceof diagnostic test for patients with suspected PE will dependon the clinical circumstances.Swensen SJ, Sheedy PF, Ryu JH, et al. Outcomesafter withholding anticoagulation from patients withsuspected acute pulmonary embolism and negativecomputed tomographic findings: a cohort study. MayoClin Proc 2002; 77:130−138Results showed that the incidence of overall DVT, PE, orfatal PE among patients with suspected acute PE, negativeCT scan results, and no other evidence of VTE is low. Withholdinganticoagulation therapy in these patients appears tobe safe.Tapson VF. Acute pulmonary embolism. N Engl J Med2008; 358:1037−1052An updated review of the epidemiology, diagnosis, treatment,and prevention of acute PE from an internationalexpert in the field.Tillie-Leblong I, Marquette C-H, Perez T, et al. <strong>Pulmonary</strong>embolism in patients with unexplained exacerbationof chronic obstructive pulmonary disease:prevalence and risk factors. Ann Intern Med 2006;144:390−396This study showed a 25% prevalence of PE in patients withCOPD hospitalized for severe exacerbation of unknownorigin.Van Belle A, Buller HR, Huisman MV, et al. Effectivenessof managing suspected pulmonary embolismusing an algorithm combining clinical probability, d-dimer testing, and computed tomography. JAMA 2006;295:172−179This study assessed the clinical effectiveness of a simplifiedalgorithm using a clinical decision rule, d-dimer testing, andCT scanning in patients with suspected PE. The use of thisalgorithm was associated with a low risk for subsequent fataland nonfatal VTE.Wells PS, Ginsberg JS, Anderson DR, et al. Use of aclinical model for safe management of patients withsuspected pulmonary embolism. Ann Intern Med 1998;129:997−1005Among inpatients and outpatients with suspected PE, only 3of 665 patients (0.5%; 95% confidence interval, 0.1 to 1.3%)with low or moderate pretest probability and a non—highprobabilityV . /Q . lung scan finding who were considered tobe negative for PE had PE or DVT during a 90-day follow-up.These data suggest that the management of patientswith suspected PE on the basis of pretest probability and theresults of V . /Q . lung scanning is safe.Wells PS, Owen C, Doucette S, et al. Does this patienthave deep vein thrombosis? JAMA 206; 295:199−207This article is part of the rational clinical examinationseries in JAMA and uses a case scenario to walk the readerthrough the approach to the evaluation of a patient with suspectedDVT. In so doing, the authors have also performeda systematic review of the literature regarding clinical36 <strong>Pulmonary</strong> Vascular Diseases (Moores)ACC-PULM-09-0302-002.indd 367/10/09 8:03:56 PM


prediction rules and d-dimer levels and applied these data inthe context of clinical decision making.Wood KE. Major pulmonary embolism: review ofa pathophysiologic approach to the golden hour ofhemodynamically significant pulmonary embolism.Chest 2002; 121:877−905This is an extensive review of a pathophysiologic approach tohemodynamically significant PE. The article also discussesrisk gratification in part based on right ventricular dysfunction.Thrombolytic therapy is acknowledged as a treatmentof choice with embolectomy for those patients in whomthrombolysis is contraindicated.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 37ACC-PULM-09-0302-002.indd 377/10/09 8:03:56 PM


Notes38 <strong>Pulmonary</strong> Vascular Diseases (Moores)ACC-PULM-09-0302-002.indd 387/10/09 8:03:56 PM


Lung CancerW. Michael Alberts, MD, MBA, FCCPObjectives:• Outline the various causes of lung cancer and the types ofclinical and radiographic presentations peculiar to eachcell type• <strong>Review</strong> the paraneoplastic syndromes associated with lungcancer• Place in perspective the appropriate use of laboratorystudies, imaging techniques, and diagnostic approachesto patients with lung cancer• <strong>Review</strong> the results of various treatment modalities for bothsmall cell and non-small cell lung cancersKey words: non-small cell lung cancer; paraneoplastic syndromes;screening; small-cell lung cancer; solitary pulmonarynodule; staging; treatmentThe statistics are no doubt familiar but, nonetheless,staggering: An estimated 215,020 individualsin the United States will receive a diagnosis of lungcancer in 2009. More disconcerting is that 161,830individuals will succumb to the disease within1 year. The numbers from abroad are no morecomforting (and, in many cases, more ominous).It has been estimated that 1 to 2 million peopleworldwide die of this disease each year.Lung cancer is currently the leading cause ofcancer deaths in both men and women in theUnited States. Deaths from lung cancer in womensurpassed those from breast cancer in 1987 and areexpected to account for approximately 26% of allcancer deaths in women in 2009. Thirty-one percentof cancer deaths in men are attributable to lungcancer. Lung cancer causes more deaths than thenext four most common cancers combined (colon,49,962; breast, 40,930; pancreas, 34,290; and prostate,28,660). Fortunately, the death rate from lungcancer in men in the United States began todecrease in 1991, which is reflective of a decreasein smoking that began in the 1950s. The mortalityrate in women has reached a plateau. The anticipateddecrease has lagged behind the decrease inmen, likely to the result of smoking prevalenceamong women. Currently, approximately 20% ofadults in the United States smoke (23.9% of menand 18.1% of women). Lung cancer is now moreprevalent in ex-smokers than active smokers.The status of the treatment of lung cancer is nomore encouraging. The expected 5-year survivalfor the diagnosis of lung cancer is 16% as comparedwith 65% for colon cancer, 89% for breast cancer,and nearly 100% for prostate cancer. Furthermore,progress in treatment has been slow. The currentoverall 5-year survival rate of 16% is only slightlybetter than the 8% survival rate of the early 1960s.Even when a small primary tumor ( 3 cm indiameter) is diagnosed “early” and no apparentmetastatic spread is found (ie, stage IA), theexpected 5-year survival is significantly 100%,with the authors of most studies reporting only65%. The 5-year survival for other potentiallyresectable lesions is even worse (IB 57%, IIA 55%, IIB 39%, and IIIA 23%). When diagnosed“late” with documented extrathoracic spread, thedisease is incurable, with an expected 5-year survivalof 1%. The median survival of patients withuntreated metastatic non-small cell lung canceris 4 to 5 months, with a survival rate at 1 yearof 10%.Before reviewing the subject at hand, it mustbe said that much of the effort evidenced in thisreview might not be necessary but for the realculprit, namely tobacco and tobacco products.Tobacco use is the leading cause of preventabledeath in this country and accounts for one of everyfive deaths. Half of regular smokers die prematurelyof a tobacco-related disease. Not to minimizethe efforts of clinicians and clinical researchers, butthe “biggest bang for the buck” comes in the formof lung cancer prevention. Whether primary, secondary,or tertiary, the prevention of cigarettesmoking has the biggest potential to improve thedismal statistics associated with this cancer.Etiology of Lung CancerTobacco causes 80 to 90% of all lung cancers.There is a clear dose−response relationship betweenthe number of cigarettes smoked per day and the<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 39ACC-PULM-09-0302-003.indd 397/10/09 8:04:21 PM


incidence of lung cancer. Cigarette smoking causesdamage to the DNA in the cells of the bronchialepithelium. There are 20 known carcinogens intobaccos smoke. Nicotine, itself, does not appearto be carcinogenic but leads to smoking addiction.The risk of dying in white men is decreasing as theresult of lower rates of smoking. The increasingrate in women parallels their increasing use ofcigarettes. The risk of lung cancer increases, in cigarand pipe smokers, depending on inhalation practices.The effect of pipe and cigar use on the risk oflung cancer is generally similar to that of lightcigarette smoking. Smokeless tobacco products (eg,chewing tobacco and snuff) are carcinogenic forthe upper aerodigestive tract but not for thelungs.Passive SmokingIt is estimated that up to 25% of lung cancer innonsmokers comes from passive exposure to cigarettesmoke, which translates into an estimate thatpassive smoking causes approximately 1.6 to 5% ofall lung cancers. Passive smoke differs significantlyfrom mainstream smoke inhaled by the activesmoker and may be even more carcinogenic. Severalstudies show that spouses of smokers have atwofold to threefold increased risk of lung cancer.Other Carcinogens for Lung CancerOther carcinogens include asbestos, radondaughters, a variety of polycyclic hydrocarbons,cadmium, chloromethyl ethers (especially for smallcell lung cancer), chromium, nickel, and inorganicarsenic. It has been conjectured that air pollutionmay promote the action of other carcinogens butmay not be carcinogenic alone. As risk factors forlung cancer, most, if not all, of these environmentalfactors either require or are markedly augmentedby concomitant exposure to cigarette smoke.Dietary FactorsDietary factors are thought possibly to decreasethe risk of lung cancer. Vitamin A intake is inverselyassociated with lung cancer risk, especially amongcigarette smokers. The constituents of green and yellowvegetables, such as beta-carotene and selenium,appear to have potential as protective agents againstlung cancer. More than 30 case−control and cohortstudies have suggested that people who eat morevegetables and fruit have a lower risk of lung cancerthan those who eat fewer such foods or who havelower beta-carotene concentrations in their blood.These observations stimulated several large,controlled trials of beta-carotene supplementation.Most trials do not support the observed beneficialassociation or a role for supplemental beta-carotenein the prevention of lung cancer. Instead, thesame studies provide striking evidence of an excessivelung cancer incidence in smokers (an adverseeffect). This finding stresses the importance ofestablishing the efficacy of chemopreventionagents in carefully conducted clinical trials.Genetic Lesions and the Molecular Pathogenesisof Lung CancerIt is becoming apparent through candidategene and genome-wide approaches that clinicallyevident lung cancers have accumulated numerous(perhaps 20) clonal genetic and epigeneticalterations in a multistep process. These alterationsinclude the classical genetic abnormalities of tumorsuppressor gene inactivation and overactivity ofgrowth-promoting oncogenes. These changes leadto the so-called “hallmarks of lung cancer”:(1) abnormalities in self-sufficiency of growthsignals, (2) evading apoptosis, (3) insensitivity toantigrowth signals, (4) limitless replicative potential,(5) sustained angiogenesis, and (6) tissueinvasion and metastases. These findings may haveimportant clinical ramifications.Advances in cell and molecular biology haveincreased our understanding of the multiple eventsthat lead to the development of lung cancer. Thefield cancerization theory suggests that multiplegenetic abnormalities occur throughout the respiratoryepithelium as a result of long-term carcinogenexposure. Mutations may occur during adult life asa result of cigarette smoking, but it is also possiblethat some of them may be acquired during embryonicdevelopment of the bronchial epithelium.Inherited Predisposition to Lung CancerA predisposition to early age of onset of lungcancer may be inherited in a Mendelian codominantfashion. COPD is associated with the development40 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 407/10/09 8:04:22 PM


of lung cancer, and there appears to be a familialcorrelation with the development of such respiratorydisease. Inheritance of an abnormality incarcinogen metabolism is another possibility.Inheriting genes predisposing to malignancy usuallyresults in a high rate of secondary tumors(lung, head and neck, esophagus, and otherorgans). For example, patients cured of laryngealtumors have an increased risk of lung cancer.Lung Cancer in Never SmokersA small proportion of patients with lung cancerdo not have a significant history of cigarette use(defined as 100 cigarettes in a lifetime). In theUnited States, 10 to 15% of lung cancer occursin never smokers (5 to 10% in men and 15 to 25%in women). This rate is greater than that of ovariancancer or Hodgkin disease. Worldwide, 15% ofmen and 53% of women with lung cancer arenever-smokers.The cause of lung cancer in never-smokers isuncertain. Speculation surrounds environmentaltobacco smoke, cooking fumes, indoor air quality,genetic factors, occupational exposures, hormonalfactors, and oncogenic viruses. The study of themolecular basis for oncogenesis has led to speculationthat lung cancer in never smokers and lungcancer in smokers may be different disease entities.Smoking CessationSmokers should be offered counseling andnicotine-replacement therapy (NRT) at every visit.“Treating Tobacco Use and Dependence—2008Update” was released and posted by the US PublicHealth Service Commissioned Corps in May of2008. This document provides the latest informationand techniques.Psychological and behavioral techniques, suchas delivering a strong personalized message,arranging intensive counseling (both individualand group), providing self-help material, referringto a tailored self-help programs, encouraging theuse of telephone “Quit-lines” (telephone therapyis now accessible in every state at 1-800-Quit now),using aversive smoking techniques, and exercise,have proven to be of benefit. Conversely, the useof hypnosis or acupuncture has not proven to bebeneficial.Pharmacotherapy in the form of NRT is beneficial.NRT may be delivered in the form of a gum,a patch, a nasal spray, an inhaler, or as lozenges.Both bupropion, an antidepressant that inhibits thereuptake of dopamine and norepinephrine, andvarenicline, a partial nicotine agonist at a subtypeof the nicotinic acetylcholine receptor, have beenshown to be effective. A recent study demonstratedsmoking abstinence from weeks 9 to 12 of a 3-month study of 44% for users of varenicline, 30%for bupropion, and 18% for placebo. Nausea occursin 30 to 50% of patients on varenicline. In addition,there have been recent alerts from the US Food andDrug Administration on the use of varenicline.There may be an increased risk of neuropsychiatricsymptoms, including agitation, depressed mood,suicidal ideation, and worsening of preexistingpsychiatric disease.Combination therapy may be more effective.For example, the use of a baseline patch plus NRTboost (gum, lozenge, spray) is common. A baselinepatch and oral bupropion plus NRT boost may behelpful for heavy smokers. Pharmacotherapydoubles the cessation success rate at 6 monthswhen compared with placebo. On occasion, longtermtherapy may be necessary.Classification of Lung CancerWorld Health Organization Histologic ClassAn important distinction is to separate smallcell lung cancer (SCLC; 15 to 18% of all lung cancers)from non-small cell bronchogenic carcinomas(82 to 85%). Approximately 10% of SCLCs arecombined with non-small cell lung cancer (NSCLC)components. The management approach to SCLCis significantly different from that for all subtypesof non-small cell bronchogenic carcinomas. Smallcellcarcinoma is generally regarded as a disorderfor which surgery is not indicated.Microscopically, SCLC in made up of small,round cells with large nuclei that stain blue whenhematoxylin and eosin is used. Histologically, theyare characterized by scant cytoplasm, fine chromatin,and nuclear molding. “Crush artifact’ is commonlynoted. SCLC arises from neuroendocrinecells and stains positive for synaptophysin,chromogranin A, and neuron-specific enolase.SCLC is so strongly associated with cigarette<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 41ACC-PULM-09-0302-003.indd 417/10/09 8:04:22 PM


smoking that such a diagnosis in a nonsmokershould be questioned.All other cell types of bronchogenic carcinoma,depending on their stage, can potentially be managedsurgically with the potential for cure. Nonsmallcell carcinomas are split into squamous cell(29% of the total); adenocarcinoma, includingbronchioloalveolar cell carcinoma (32% of thetotal), large-cell carcinoma (9% of the total), andundifferentiated (11% of total). Non-small cell lungcarcinomas are believed to arise from lung epithelialcells. They express cytokeratin in a distinctpattern (negative for CK20 and positive for CK7)and are commonly positive for thyroid transcriptionfactor 1.Adenocarcinomas are the least closely associatedwith cigarette smoking and most commonlyarise in the lung periphery from epithelial cellsinvolved in gland formation. They may grow inacinar, papillary, bronchioloalveolar, or solidgrowth patterns, often in association with the productionof mucus. Up to 50% of patients have neversmoked, and 80% present with metastatic disease.Bronchioloalveolar cell carcinoma (BAC) oftenis characterized as a subtype of adenocarcinoma.BAC is the most common type of lung cancer innever-smokers. BAC tumors grow in lepidic fashion,spreading along the lining of alveolar airspaces, and there is no invasion of the stroma,pleura, or lymphatics. Where such invasion occurs,the disease process is termed adenocarcinoma mixedtype with a predominant BAC pattern. BAC often isvery slow growing and commonly spreads withinthe lung through the airways. BAC is divided intothree subtypes: mucinous, nonmucinous, and amixed type.Squamous cell carcinoma is characterized bykeratin formation, intercellular bridging, and anonglandular appearance. Squamous cell lungcancer typically arises within the major bronchinear the center of the chest. A total of 95% ofpatients with squamous cell are smokers, and 60%present with metastatic disease.Large cell carcinoma lacks the features of eithera squamous cell or an adenocarcinoma. This tumoris a more poorly differentiated form of NSCLC.Large cell tumors that express neuroendocrinetumor markers are less responsive to chemotherapythan SCLC and carry a poor prognosis comparedwith other types of NSCLC.Other cell types constitute a small minority ofall lung neoplasms. For example, carcinoid tumorsmay arise in the lung. These tumors have a neuroendocrinephenotype, like SCLC, but grow slowly,rarely metastasize, and the treatment is often surgeryalone. These tumors are composed of small,round blue cells but can be differentiated from SCLCby their low-grade appearance and lack of mitoticfigures. Carcinoid tumors tend to be highlyresistant to radiation therapy and chemotherapyand are therefore best treated by surgery. Moreaggressive carcinoid tumors with an intermediatenumber of mitotic figures are termed atypical carcinoids.These tumors are faster growing and havethe potential for metastatic spread. Fortunately,atypical carcinoid tumors of the lung are moreresponsive to chemotherapy and radiation therapy.Preinvasive lesions have been identified. Theseinclude squamous dysplasia/carcinoma in situ(leading to squamous cell carcinoma), atypicaladenomatous hyperplasia (leading to adenocarcinoma),and diffuse idiopathic pulmonary neuroendocrinecell hyperplasia (leading to carcinoid).International Staging System for Lung CancerLung cancers are staged by the TNM system(T = 5 primary tumor, N = 5 regional lymph nodes,M = presence or absence of distant metastases). TheTNM system, developed in the late 1940s, is usedas a guide to estimate prognosis, to select treatmentoptions, and to report outcomes. Several changeswere made to the staging system in 1997. Thesechanges were made to more closely align patientclinical presentation to appropriate treatmentoptions and expected outcomes. The seventh editionof the TNM Classification of MalignantTumors is expected to be published in 2009.Stage I Patients Redefined: Only T1N0M0(stage IA) and T2N0M0 (stage IB) patients are nowincluded in stage I. T1N1M0 is now placed in stageII. The prognosis for survival with N1 involvementis significantly worse, especially in thecase of adenocarcinoma, than when there is nolymph node involvement. The prognosis for 5-yearsurvival, with surgical resection, is 70 to 80% forT1N0M0 patients and 50 to 60% for T2N0M0patients.Stage II Patients: The survival pattern forT1N1M0 (stage IIA) patients more closely42 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 427/10/09 8:04:22 PM


esembles the survival patterns for patients withstage II lung cancer than stage I subsets. Stage IIBincludes T2N1M0 and T3N0M0. T3N0M0 wasmoved from stage IIIA to IIB reflecting a prognosismore similar to other stage II tumors.Stage III Patients: Tumors staged in thisgrouping are very heterogeneous. The 10% ofpatients with NSCLC who present with stage IIIAcancer are described by the American College ofChest Physicians (<strong>ACCP</strong>) guidelines as “perhapsthe most therapeutically challenging and controversialsubset” of lung cancer patients, becausethey fit between the generally resectable stage Iand stage II tumors and the generally unresectablestage IIIB and stage IV cases.Although not recognized by the American JointCommittee on Cancer’s Cancer Staging Manual,for treatment purposes, patients with N2 diseasemay be artificially classified into four subgroups:• IIIA 1: incidental nodal metastases found onfinal pathologic examination of the resectionspecimen;• IIIA 2: nodal metastases (single station) recognizedintraoperatively;• IIIA 3: nodal metastases (single or multiple station)recognized by prethoracotomy staging;and• IIIA 4: bulky or fixed multistation N2 disease.Stage IIIA Patients: Patients with involvementof the ipsilateral mediastinal lymph nodes(N2 disease) are categorized as IIIA patients. Abetter outcome is anticipated for these patientsthan when mediastinal node metastasis is moreextensive (N3 disease). Patients with IIIA tumors(IIIA1, IIIA2, and IIIA3 after neoadjuvant) maypotentially undergo complete resection. If nonoperativetherapy is appropriate, randomized controlledtrials have shown that chemotherapy andradiation therapy improve survival over the useof radiation therapy alone. The role of surgicalresection as an adjunct to this combined methodof treatment is still not clear.Stage IIIB Patients: Stage IIIB patients includethose with T4 tumors (T4 is for tumors of any sizethat invade the mediastinum or involve the heart,great vessels, trachea, esophagus, vertebral body,or carina; or for the presence of a malignant pleuraleffusion) and with any N3 metastases (contralateralmediastinal lymph nodes, contralateral hilar lymphnodes, ipsilateral or contralateral scalene, or supraclavicularnodes), but with M0 disease. Selectedstage IIIB patients (good performance status) typicallyare treated with chemoradiation.Stage IV Patients: Stage IV patients includethose with any T designation and any N involvementaccompanied by spread to different lobes ofthe lung than the site of the primary tumor or, byway of the bloodstream, to distant sites within thebody.Exceptions for SCLC: The TNM and stagingsystems (I-IV) are appropriate for SCLC.Many physicians, however, use the old Veteran’sAdministration staging system, the reason beingthat there is less variation in survival for the stagesdescribed by the TNM system when applied toSCLC. Therefore, the more detailed TNM systemoffers no advantage. Both treatment options andprognosis can be adequately communicated bythe use of a two-stage system. In the two-stagesystem developed by the Veteran’s AdministrationLung Cancer study group, small cell lungcancer is staged as “limited disease” or “extensivedisease.” Limited disease for SCLC (30% ofpatients) typically is defined as no detectable diseaseoutside of the hemithorax, with or withoutipsilateral, mediastinal, hilar, or supraclavicularlymph nodes. Patients with unilateral pleuraleffusion also have been classified as having limiteddisease. Extensive disease (70% of patients) isany disease occurring beyond the sites listed forlimited disease.Proposals for the Seventh <strong>Edition</strong> of the TNMClassification and Stage Groupings: The InternationalAssociation for the Study of Lung Cancer(IASLC) recently has issued recommendationsfor the seventh <strong>Edition</strong> of the TNM Classificationand Stage Groupings of Lung Cancer based on anintensive and validated analysis of a database thatincluded 100,000 patients. The recommendedchanges in the T classification are to subclassifyT1 into T1a ( 2 cm) and T1b ( 2− 3 cm) and T2into T2a ( 3− 5 cm) and T2b ( 5− 7 cm) andto reclassify T2 tumors 7 cm into T3. Furthermore,with additional nodules in the same lobeas the primary tumor, T4 would be reclassified asT3. With additional nodules in another ipsilaterallobe, M1 would be reclassified as T4. With pleuraldissemination, T4 would be reclassified as M1.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 43ACC-PULM-09-0302-003.indd 437/10/09 8:04:22 PM


The IASLC recommends that the current Ndescriptors should be maintained and that revisionsto should include grouping cases with malignantpleural effusions and cases with nodules inthe contralateral lung in the M1a category, andcases with distant metastases should be designatedM1b. In addition, cases with nodule(s) in the ipsilateral(nonprimary lobe) currently staged M1should be reclassified as T4M0.Recommendations for revisions to the stagegroupings include additional cutoffs for tumorsize, with tumors 7 cm moving from T2 to T3;reassigning the category given to additional pulmonarynodules in some locations; and reclassifyingpleural effusions as an M descriptor. Inaddition, IASLC suggested that T2bN0M0 cases bemoved from stage IB to stage IIA, T2aN1M0 casesfrom stage IIB to stage IIA, and T4N0-1M0 casesfrom stage IIIB to stage IIIA.Clinical Features of Lung CancerLung cancer symptoms may range from noneto such problems as cough, hemoptysis, dyspnea,fever, and hoarseness. Only 5 to 10% of cases areasymptomatic at discovery, and 15% have extrapulmonarysymptoms as the first clue to the diagnosis.The most common symptoms are listed tofollow: cough (75%), dyspnea (60%), chest pain(45%), hemoptysis (35%), other pain (25%), clubbing(22%), hoarseness (18%), dysphagia (2%), andwheezing (2%). Items that should be included inthe history include weight loss, focal skeletal pain,chest pain, headache, syncope, seizure, extremityweakness, and change in mental status.Resectable lung cancer will seldom be diagnosedbased on the history. Approximately 50%will have demonstrable metastatic lesions or evidenceof unresectability at the time of first diagnosis.Further testing will reveal that another 15% areunresectable. Finally, another 5 to 10% will be foundto be unresectable at surgery. Thus, only 25 to 30%of cases are potentially curable by surgery.Physical examination findings usually parallelthe symptoms. The physical examination willbecome positive only late in the course. If the firstclue to the diagnosis comes from the physicalexamination, it is probably too late to expectany chance for cure. Items that should be includedin the physical include lymphadenopathy,hoarseness, superior vena cava syndrome, bonetenderness, hepatomegaly, focal neurologic signs,papilledema, and soft-tissue mass.Regional Spread of TumorTypical symptoms, when the lung cancer hasspread to the mediastinum, include dysphagia(from esophageal compression or involvement)and the development of effusions (from lymphaticobstruction). Cardiovascular involvement can beassociated with arrhythmias and heart failure(from pericardial involvement). The pericardiumor the myocardium is involved in 15 to 35% ofpatients (autopsy data), but the frequency of clinicalsymptoms is unknown. Malignant pericardialeffusion with tamponade may develop.The superior vena cava syndrome is more commonif the primary tumor is on the right side. InNSCLC (usually squamous) the obstruction developsslowly, allowing development of a collateralvenous system evident at the time of physicalexamination. Edema and suffusion may also bepresent.Regional nervous system involvement includesHorner syndrome (unilateral dilated pupil, enophthalmos,facial dryness, and ptosis) seen withsuperior sulcus (Pancoast) tumors. Symptomsinclude shoulder pain, with radiation to the ulnarnerve distribution of the arm and often with radiographicdestruction of the first and second ribs.Hoarseness is caused by involvement of therecurrent laryngeal nerve. This is more commonon the left side because of the longer course ofthe nerve. Phrenic nerve paralysis produces elevationof the hemidiaphragm and the potential fordyspnea.Malignant Pleural EffusionsThese effusions can compromise lung functionby compressing the lung. Malignant pleural effusionsmay be highly symptomatic, with shortnessof breath and pain. Among NSCLCs, the frequencyis highest with large-cell carcinoma (67%), followedby adenocarcinoma (60%) and squamous cell carcinoma(34%). However, because of the greaterfrequency of adenocarcinoma, this is the cell typemost likely to be associated with a malignant pleuraleffusion.44 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 447/10/09 8:04:23 PM


Metastatic Disease Outside the ThoraxAt autopsy, the frequency of extrathoracicmetastases is 54% for squamous cell carcinoma,82% for adenocarcinoma, and 86% for large-cellcarcinoma. At the time of presentation, most smallcellcarcinoma has already spread outside thethorax, although it may not be clinically evident,even after all applicable staging maneuvers havebeen completed.Bone marrow involvement is encountered in15 to 25% of patients with SCLC at the time ofdiagnosis. Cortical bone involvement occurs inapproximately 22% of patients with SCLC and isthe isolated site of metastatic disease in 10%. TheCNS is commonly involved by metastatic lungcancer. The brain is involved in approximately 10%of cases of SCLC at presentation, but 30% ofpatients will have parenchymal brain metastasessometime during the course of disease.For NSCLC, the most common cell types toinvolve the brain are adenocarcinoma and largecellcarcinoma. Spinal cord involvement usually ispreceded by back pain, followed by symptoms andsigns of compression (bladder/bowel dysfunction,or paraplegia). Carcinomatous meningitis is lesscommon and usually causes death within 4 to 6weeks.Constitutional symptoms and signs includeanorexia and weight loss (31% at time of presentation);weakness; fever (21% of cases, caused bytumor); and clubbing (29%). Cutaneous manifestationsof lung cancer are uncommon but includeskin metastases, acanthosis nigricans, bullouslesions (erythema multiforme), dermatomyositis,scleroderma, and tylosis (hyperkeratosis of thepalms and soles). Vascular and hematologic manifestationsof lung cancer include anemia, thrombophlebitis(especially migratory), disseminatedintravascular coagulopathy, nonbacterial thromboticendocarditis with arterial emboli, granulocytosis,and leukoerythroblastosis.Paraneoplastic SyndromesParaneoplastic syndromes can be found in bothSCLC and NSCLC but are more common in theformer. They may be attributable to secretion ofbiologically active peptides and hormones, or as aresult of tumor-related immune events. Parathyroidhormone secretion with hypercalcemia is a phenomenonassociated with squamous cell lungcancer, whereas most of the other paraneoplasticsyndromes are far more commonly seen withSCLC. It is estimated that approximately 20% ofpatients with SCLC will present with some type ofparaneoplastic syndrome sometime during thecourse of the disease. The reason SCLC is morefrequently accompanied by paraneoplastic syndromesis unclear; it may be related to its putativecell of origin, the APUD cell, which is of neuroendocrineorigin, thus enabling it to elaborate anumber of hormones and other biologically activeproteins.Specific SyndromesParathyroidlike Hormone: Hypercalcemia ismore common from this cause than from skeletalmetastases. Parathyroidlike syndrome is unusualin patients with SCLC; if hypercalcemia is observedin this setting, consider a coexistent non-smallcellhistology, coexisting hyperparathyroidism,or transformation from SCLC to NSCLC. Hypertrophicpulmonary osteoarthropathy is seen moreoften in adenocarcinoma (1 to 10% of cases).Antidiuretic Hormone: Syndrome of inappropriateantidiuretic hormone (SIADH) isthe most common paraneoplastic syndrome inSCLC. Approximately 5 to 10% of patients presentwith this syndrome. An additional 40 to 50%of patients can be shown to have subclinicalabnormalities compatible with SIADH on appropriateclinical testing. The source of the ADH canbe the primary tumor and/or the metastases.Laboratory abnormalities include hyponatremia,increased excretion of sodium in the urine, normalvolume status and adrenal/renal function,and failure to excrete maximally diluted urinewith water challenge.Adrenocorticotropic Hormone: SCLC is themost common tumor associated with ectopic corticotropinproduction. Approximately 3 to 7% ofpatients with SCLC will have Cushing syndrome,but a much greater percentage have a subclinicalform (11 to 72% by radioimmunoassay). Theclinical features of Cushing syndrome tend tobe masked by anorexia and significant weightloss. Severe weakness and the profound mineralocorticoideffects of edema, hypertension, and<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 45ACC-PULM-09-0302-003.indd 457/10/09 8:04:23 PM


hypokalemia are more common. The potassiumlevel is 3.0 mEq/L in 70 to 90% of patients.Hyperpigmentation occurs in approximately 25to 30% of patients. The calcitonin level is elevatedin 38 to 67% of all patients with lung cancer, butSCLC is associated with the greatest frequency.Calcitonin causes an immediate calciuresis, but itdoes not produce symptoms.Paraneoplastic Neurologic Syndromes: Neuromyopathiesare most commonly associated withSCLC. The incidence of neuromyopathies for alllung cancers is 10%. Multiple small brain metastases,carcinomatous meningitis, and spinal cordor peripheral nerve compression by tumor can allmimic neuromyopathies, as can diabetes and useof steroids. Other considerations before diagnosinga neuromyopathy as caused by the lung canceritself are syndrome of inappropriate ADH andinfectious agents (progressive multifocal leukoencephalopathy).The carcinomatous neuromyopathy can occasionallypresent up to 1 year before the clinicaldiagnosis of SCLC, but it usually is apparent at theinitial presentation. Side effects of chemotherapy(eg, Vinca alkaloids and cis-platinum) may be confusedwith paraneoplastic neurologic syndromes.Peripheral neuropathy is the most commonneurologic syndrome in SCLC, occurring in nearly100% of patients sometime during its course. Itis probably related to the high-frequency use ofVinca alkaloids. Perhaps an underlying subclinicalneuropathy exists in most patients with SCLCand the Vinca alkaloids precipitate this to thepoint where it can be clinically detected. The mostcommon symptoms are decreased sensation andparesthesias in the extremities.Dementia is the most common encephalopathyin SCLC. It is characterized by forgetfulness, lossof memory, or confusion. A response to treatmentof SCLC is not necessarily associated with improvementin dementia. There is also a greater incidenceif prophylactic whole-brain irradiation is used.Subacute Cerebellar Degeneration: Bilateralsymmetrical truncal and extremity ataxia arecharacteristic features. Dysarthria and tremorsalso are seen frequently. The clinical course is usuallyrapid, with the patient requiring a wheelchairwithin weeks or months.Eaton-Lambert Syndrome: The clinical pictureis very similar to myasthenia gravis, withproximal muscle weakness and easy fatigability.Symptoms are more pronounced in the lowerextremities, with difficulty in walking, climbingstairs, and getting up from a chair. An electromyogramclearly distinguishes this syndromefrom myasthenia gravis in that there is facilitationof muscular action potentials with repeated stimulation.Unlike the other neuromyopathies, theEaton-Lambert syndrome frequently responds totreatment of the tumor.Early Detection and Screeningfor Lung CancerIn the 1970s, the National Cancer Institute supportedthree mass-screening programs at threemajor institutions (Johns Hopkins, the Mayo Foundation,and Memorial Sloan-Kettering CancerCenter). No mortality difference was observedbetween the screened and the control groups evenwith extended follow-up through 1996. As a resultof these and other studies, no organizations currentlyrecommend screening.The NCI studies screened men who were athigh risk (smokers 45 years of age who smokedone or more packs per day for at least 20 years). Inthe John Hopkins and Memorial Sloan Ketteringprojects, annual chest radiographs in a controlgroup were compared with chest radiographs andcytology in an experimental group. Cytology wasnot associated with a different outcome over chestradiographs alone, but long-term survival in bothstudies was about three times greater than predictedfrom other data.Two randomized trials (the aforementionedMayo study and a Czechoslovakian study) comparedregular and frequent rescreening chestradiographs in an experimental group with sporadicand/or infrequent rescreening in a controlgroup. Both studies demonstrated a strikingadvantage for screening with regard to stage distribution,resectability, survival, and fatality, butmortality was somewhat greater in the screenedgroups. Thus, it can be stated that intensive screening(every 4 months, compared with annual) withchest radiographs and sputum cytology doesdetect lung cancer earlier. The mortality rate fromlung cancer, despite its earlier detection and diagnosis,was not significantly different in the screenedgroup compared with the control group. The better46 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 467/10/09 8:04:23 PM


survival rate may be a function of earlier leadtime.A survival benefit, as opposed to a mortalitybenefit, with screening compared with no screeningmay be anticipated because of several potentialinherent biases. Lead-time bias is defined as apparentimproved survival by earlier diagnosis withscreening even when the outcome, death causedby lung cancer, remains unchanged. Length-timebias has the effect of improving apparent survivalby selecting cancers that, by their slow growth,have a good prognosis compared with cancersfound by symptoms. Overdiagnosis bias improvesthe apparent overall survival by the identificationof cancer that, in the absence of screening, wouldhave gone undetected because the cancer wouldneither cause symptoms nor death because of itsindolent nature.The failure to improve mortality by screeningpatients for lung cancer with plain chest radiographsand sputum cytology led to an indifferenceto screening for 2 decades. There is now arenewed interest in screening, particularly for highriskpopulations (smokers, especially those withairflow obstruction demonstrated by spirometry)because of two new methods to screen for lungcancer: low-dose spiral CT and autofluorescencebronchoscopy.Low-dose CT scanning allows imaging of theentire chest with a single breath hold with lowradiation exposure. An observational study publishedin 2006 reported on 31,567 asymptomaticpersons at risk for lung cancer screened by lowdoseCT. Screening resulted in a diagnosis of lungcancer in 484 participants, and 412 (85%) hadclinical stage I disease. The results confirmed thatannual spiral CT screening can detect lung cancerat a stage when surgical resection is possible. Thisdetection and subsequent treatment results in animprovement in apparent survival but does notconfirm or refute a mortality (“true survival” asopposed to “apparent survival”) benefit.A study published in 2005 reported on lowdoseCT screening coupled with sputum cytologyscreening on 1,520 smokers older than the age of50 years and found noncalcified lung nodules in66% over the course of 1 year. There were 25prevalence cases and incidence cases of lung cancer;CT alone detected 23 of the cases. Twenty-twopatients underwent surgical resection that isthought to be curative; 7 of the patients underwentresection of benign nodules. The authors concludethat the use of spiral CT can detect early-stage lungcancers but caution that this does not necessarilytranslate into decreased mortality. They were particularlyconcerned about the high false-positiverate.Although these studies and others suggestthat lung cancer can be identified and at an earlierstage by low-dose CT screening, many cautionthat this screening method is not yet ready forwidespread application. The high rate of abnormalscreens, most of which are false- positive, createsa necessity for follow-up CT scanning, subspecialtyconsultations, and other testing that createsa high total cost for such screening. The NationalCancer Institute has commissioned a randomizedprospective controlled trial, aiming to enlist 50,000current or former smokers who will be followedup for 8 years after three annual screening lowdoseCTs. Half the subjects will be randomized tobe screened with CT imaging, and the other halfby standard chest radiographs. Until the results ofthis and other longer-term studies are available,there should be caution and reservations aboutspiral CT screening. Screening with low-dosespiral CT should be conducted in the context ofwell-designed clinical trials.A second screening method that holds promisefor the detection of early-stage lung cancers isautofluorescence bronchoscopy, which is alsoundergoing extensive early multicenter trials forsmokers with obstructive lung disease and identifiedcarcinoma cells or severe dysplasia in sputumsamples. Dual screening with both autofluorescencebronchoscopy and low-dose spiral CT maysomeday become the norm for patients at high riskfor lung cancer. However, now is not the time tobegin such dual screening outside the context ofwell-designed clinical trials, pending the outcomeof additional studies to prove efficacy, costefficiency,and a mortality benefit for patients whoundergo such screening.Solitary <strong>Pulmonary</strong> NoduleSolitary pulmonary nodules (SPNs) are seenon a plain chest radiograph, are 3 cm in diameter,and are rounded or slightly ovoid. They are locatedin the lung parenchyma, and there are no other<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 47ACC-PULM-09-0302-003.indd 477/10/09 8:04:23 PM


associated abnormalities on the plain chest radiograph.Granulomas and hamartomas constitute40 to 60% of all SPNs and are the leading cause ofSPNs in persons 35 years of age. In older patients,especially those with a history of cigarette smoking,the key concern is whether the SPN representsa malignancy. Availability of an old chest radiograph/CTthat shows the same lesion, stable ornearly stable in size for at least 2 years, will oftenbe a key arbiter of the proper treatment in such asituation. Malignant lesions have a doubling timeof between 30 and 500 days, whereas benignlesions may double in 30 days or show no growthover the course of 2 years.Although SPNs, especially those 2 cm indiameter, can be definitively diagnosed as malignantwith bronchoscopy ( 65% yield) or transthoracicfine-needle biopsy ( 85% yield), it is the rarepatient who will truly benefit from such anapproach. Generally speaking, when the SPN isstrongly suspected to be malignant, it should beresected both for definitive diagnosis and for cure(assuming it turns out to be malignant). Exceptions,in which a lesser invasive procedure is justifiable,include patients who are poor surgical candidatesor situations in which the surgeon or patientrefuses surgery. In these cases, an invasive diagnosticprocedure less than a thoracotomy is appropriateto define the nature of the lesion and tofacilitate alternative treatment planning (externalbeam radiation therapy).The probability of malignancy varies from 10to 70%, depending on the patient population. Theprobability can be estimated from the smokinghistory, age, size of the lesion, and a history orprevious malignancy. Clinicians should estimatethe pretest probability of malignancy either qualitativelyby using their clinical judgment or quantitativelyby using a validated model. SPNs thathave a lower pretest probability (ie, 5%) formalignancy may be monitored by follow-up serialCT scans to assess stability. A reasonable scheduleis at 3 months for the first follow-up scan, at 6months for the second, a third at 1 year, and afourth at 24 months. For SPNs that have an intermediateprobability of cancer (ie, 5 to 60%), additionaltesting (eg, positron emission tomography[PET] imaging, transthoracic fine-needle aspirationbiopsy, bronchoscopy) is advised. For SPNsthat have a high probability of cancer (ie, 60%),excisional biopsy with an intraoperative frozensection examination is advised.The Role of Diagnostic TestsA complete history and physical examinationremain the key elements in the choice of all laboratorystudies for staging the cancer; particularlyperformance status, a history of weight loss,symptoms and signs suggestive of metastaticdisease, and the function of the vocal cords anddiaphragm.Past and Current Chest Radiographs: Considerwhether the abnormality has been stable during a2-year span; if it is stable, then it is almost alwaysassociated with a benign lesion (less certain if thelesion is a GGO). Certain types of calcificationswithin a lesion indicate that it is benign, eg, concentriclamellated rings. Remember, however, thatnot all calcified lesions are benign.Screening Blood Work and Other “Routine”Studies: Simple blood work is indicated as partof the pretreatment assessment of all patientsknown or strongly suspected of having lung canceron the basis of their clinical and radiographicpresentation. Appropriate studies and the rationalefor doing them include a CBC count (anemiais a poor prognostic indicator); urinalysis(may identify a renal paraneoplastic syndrome);serum liver function tests (alkaline phosphatase,alanine aminotransferase, total bilirubin, aspartateaminotransferase to screen for the presenceof liver metastases—if liver function tests areabnormal, additional investigation is warranted,but liver enzymes are rarely abnormal unlessthere are extensive metastases); serum calciumtest (to screen for parathyroid-like hormone syndromeand bone metastases); serum creatinine(many chemotherapeutic agents are nephrotoxic);and serum albumin (a low value is a poorprognostic factor). No other laboratory tests areroutinely recommended, although a creatinineclearance may be needed if chemotherapy iscontemplated because many chemotherapeuticagents used in the treatment of lung cancer arenephrotoxic.Sputum Cytology: Sputum cytology may bediagnostic in up to 20% of cases (74% of in centrallesions and 5% of peripheral lesions) and is thereforea reasonable first step in selected cases.48 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 487/10/09 8:04:24 PM


Flexible Bronchoscopy: For central (endoscopicallyvisible) cancer, flexible bronchoscopy is 90to 95% sensitive (or greater). The main goal is toestablish a diagnosis and distinguish SCLC fromNSCLC. For peripheral tumors, flexible bronchoscopyhas a reasonable sensitivity (60 to 75%) if thetumor is 2 cm in diameter and fluoroscopy isused. Flexible bronchoscopy is especially importantbefore thoracotomy for NSCLC with curativeintent to judge the proximal extent of the tumorfor transection of the bronchus and to look foran occult central or contralateral second primary(1 to 3% frequency). Flexible bronchoscopy canoften be done by the surgeon at the same anestheticsitting, just prior to thoracotomy (especiallywhen the tumor is peripheral in its location).Transbronchial Needle Aspiration: A transbronchialneedle aspiration (TBNA) biopsy of themediastinal/hilar lymph nodes can stage theextent of the disease at the same time the diagnosisis made. This biopsy is more important whenstage IIIA disease is suspected than in any othersituation. Some centers have a consistently greateryield than others (depending on technique). Agreater yield can be expected if the TBNA is guidedby findings from CT than if done “blindly” or asa routine part of every procedure. The TBNAspecimen should be obtained first, so as to avoidcontamination with possible false-positive results.The need to perform TBNA is predicated, atleast in part, on the institutional philosophy ofwhether stage IIIA NSCLC is a surgically resectabledisease.Endobronchial Ultrasound-Guided Needle Aspiration:Ultrasound-guided transbronchial aspiration(EBUS) of lymph nodes is proving to be avaluable minimally invasive method of samplingsuspicious nodes. In experienced hands, EBUSguidance may allow sampling of level 2, 4R, 7, 10,and 11 lymph nodes.Transesophageal ultrasound-guided needleaspiration (EUS): Ultrasound-guided transesophagealaspiration of mediastinal lymph nodes thatare not accessible by flexible bronchoscopy (orcervical mediastinoscopy) is proving to be a valuableminimally invasive method of sampling suspiciousnodes. A false-negative rate of up to 23% isa potential drawback. In experienced hands, EUSguidance is especially useful for sampling level 4L,7, 8, and 9 lymph nodes. Some have opined thatEBUS plus EUS may allow near-complete minimallyinvasive mediastinal staging in patients withlung cancer.Fine-Needle Aspiration: Percutaneous fine-needleaspiration biopsy has a greater sensitivity (90to 95%) than bronchoscopy for malignant peripheralsolitary nodules, especially if the diameter is 2 cm. A pneumothorax can be expected in 20 to25% of patients; a chest tube will be required in5% of all patients treated this way. Nondiagnosticresults may not obviate the need for thoracotomyif the lesion is likely malignant. Bronchoscopyis still needed before thoracotomy (at the sameanesthetic sitting, however) to exclude a secondprimary.Specific indications for fine-needle aspirationinclude pulmonary masses in a patient unable toundergo a curative thoracotomy (because of compromisedpulmonary function, medical contraindicationsto thoracotomy, or refusal of thoracotomy)who needs a definitive tissue diagnosis; an undiagnosedlocalized or worsening pneumonic infiltratein an immunocompromised patient despitestandard antibiotic therapy; a history of anothermalignancy and an abnormality on chest radiograph;and the evaluation of other masses on achest radiograph or CT scan (eg, mediastinalmasses) that need to be histologically evaluated todevelop a therapeutic plan.Thoracentesis/Pleural Biopsy: In the settingof lung cancer, a pleural effusion usually indicatesthat the cancer has seeded the pleura andthe patient is not a candidate for curative treatment.There are occasional exceptions, however.Some pleural effusions will be parapneumonic innature (benign), when a tumor obstructs a morecentrally located bronchus with a postobstructivepneumonitis. Likewise, there are rare situationsin which a patient’s lymphatics will be obstructedby a tumor that involves the more central lymphnodes, impeding fluid transfer through the pleuralspace with a benign effusion as a result. Therefore,it is important to sample the pleural fluidand to study it cytologically to determine if thetumor has seeded the pleural space, rendering thepatient incurable. Cytologic analysis of a pleuraleffusion ultimately proven to be malignant isassociated with a 50 to 60% true-positive yield. Asecond cytologic analysis of the fluid should bedone for those whose first pleural fluid specimen<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 49ACC-PULM-09-0302-003.indd 497/10/09 8:04:24 PM


is negative for malignant cells. Closed needlebiopsy of the pleura adds only approximately 8%to the overall yield for malignancy, so it is generallynot recommended as part of the approach toa patient with a suspected malignant pleural effusion.Instead, for patients with two negative cytologicstudies of their pleural fluid, a thoracoscopyshould be done, because the true-positive yieldwhen malignancy is present is approximately 98to 99%.Bone Scan: There is general agreement thatradionuclide scans of the bones are not warrantedduring a preoperative evaluation unless symptoms,signs, or abnormal blood test results raise asuspicion of metastatic involvement of the bones.There are many other bone disorders that can giverise to one or more focal areas of increased uptake;thus, the false-positive rate is unacceptably highwhen bone scans are performed routinely. Bonescans should be done when the clinical evaluationreveals bone pain, pathologic fractures, and/or anelevated alkaline phosphatase or serum calciumlevel. Additionally, a bone scan can be obtainedwhen nonspecific findings indicate the presenceof metastatic disease.The increasing availability of whole-body PETscans in the evaluation of the patient with lungcancer may soon affect the frequency with whichbone scans are ordered. A recent study has foundPET scans to be 90% sensitive and 98% specific inthe search for bony metastases whereas comparablefigures for radionuclide bone scanning are90% sensitive and 61% specific.Bone Marrow Aspiration and Biopsy: This testis not performed as part of NSCLC staging. ForSCLC, the incidence of bone marrow involvementis 15 to 25% at the time of diagnosis. Therefore,bilateral bone marrow aspiration and biopsy willshow a greater incidence of involvement. The procedureis not recommended for routine use. Bonemarrow involvement usually occurs in the settingof other metastatic disease. The bone marrow isthe sole site of metastatic disease in 4 to 10% ofpatients at the time of initial presentation.Staging the MediastinumCT assessment is much more sensitive thanradiographs to detect direct extension of the primarytumor and regional lymph node enlargement.In the absence of distant metastases or inadequatecardiopulmonary reserve, chest CT is generallythough to be an integral part of the pretreatmentplanning. Situations in which a chest radiographalone is adequate include obvious metastatic bonelesions or the presence of large, bulky contralateralmediastinal lymph nodes. The sensitivity andspecificity of determining mediastinal lymphnode involvement is a function of the cut pointchosen. With a short axis diameter 1.0 cmdefined as abnormal, the sensitivity and specificityare at best approximately 80%. (Some studieslist 40 to 70%.) Because of their imperfect specificity,however, abnormal mediastinal CT findingsin an otherwise operable patient should neverpreclude thoracotomy unless pathologicallyconfirmed.MRI: MRI is not better than CT for the evaluationof mediastinal metastases. It is more expensivethan CT, but it can be of particular value inselected cases, such as determining whether thechest wall is invaded when the tumor is shownto be contiguous by plain chest radiograph or byCT assessment. MRI may also display invasionof vascular structures in the mediastinum withgreater specificity than is possible with CT.Position Emission Tomography Scanning: Thereare a number of published studies that indicate thatPET scanning is more sensitive and more specificthan CT for staging the mediastinum in lung cancer.PET is a metabolic imaging technique basedon the function of a tissue rather than its anatomy.It is useful to differentiate neoplastic from normaltissue, but it is not infallible because some nonneoplasticdiseases and infections may be positiveon PET imaging. Size limitations are also an issue,with the lower limit of resolution of PET scanningat approximately 1 cm. The sensitivity of PETscanning for mediastinal nodes is approximately85 to 91%, and specificity is approximately 86 to88% from combined studies. Thus, PET has bothgreater sensitivity and specificity for evaluationof mediastinal lymph node staging than CT scanning(sensitivity of 75% and specificity of 66% ina recent study). For metastatic disease to the bone,the PET is up to 90% sensitive and 98% specific,whereas bone scanning is 90% sensitive but only68% specific. Decision analyses demonstrate thatthe use of fluorodeoxyglucose−PET may reduceoverall costs of medical care by identifying50 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 507/10/09 8:04:24 PM


patients with false-negative CT scans in the mediastinumor otherwise-undetected sites of metastases.The advent of fused CT-PET images mayprove to be a significant advance.Although a negative mediastinal PET (NPV of87 to 100%) may obviate the need for mediastinoscopybefore thoracotomy, a positive mediastinalPET (positive predictive value of 74 to 80%) shouldnot be taken as a sign of unresectability becausethere is a possibility of false-positive results on PET.Cost and availability of PET scanning still limit theuse of this promising modality.Cervical Mediastinoscopy: The diagnostic yieldranges from 10 to 75%, depending on (1) the histologictype of tumor, (2) the size and location of theprimary tumor, and (3) the extent of the disease.The value of mediastinoscopy in the preoperativeevaluation is related directly to the therapeuticphilosophy of the surgeon. If involvement ofthe mediastinum indicates that the tumor is notresectable, the surgeon will perform mediastinoscopy(or insist on some other sampling procedureof the mediastinum) routinely. Those who resectmediastinal involvement with curative intentwill use mediastinoscopy less frequently. At thetime of mediastinoscopy, the pretracheal/paratracheallymph nodes can be sampled, as well assome of the subcarinal nodes (right at the bifurcation)and the azygous nodes (primarily levels 2,4, and 7). Mediastinoscopy helps to exclude thoracotomyfor patients with marginal chances forsurvival after thoracotomy: those with poor performancestatus, advanced age, or poor pulmonaryfunction.Anterior Mediastinotomy: Anterior mediastinotomyis performed by resecting the second costalcartilage on either side. There is usually betterexposure and less likelihood of surgical misadventure(uncontrollable bleeding) than withmediastinoscopy. At the time of mediastinotomy,the surgeon can sample the paratracheal, azygous,superior hilar nodes on the right side, andthe nodes in the aortopulmonary window and theanterior mediastinum on the left side. The pretracheal,left paratracheal, and subcarinal lymphnodes cannot be sampled with this approach,however. The morbidity and mortality are essentiallynil with transbronchial/transcranial needleaspiration. The morbidity is 2% with mediastinoscopyor mediastinotomy. The false-negative rateis much greater with needle aspiration, generallyapproximately 8% with mediastinoscopy ormediastinotomy, but it may be as much as 20%in some series.Supraclavicular/Scalene Node Biopsy: This isnot routine in the absence of palpable adenopathy.It is not of value in patients with peripheraltumors. Its value in patients with central tumors,especially adenocarcinoma or those of unknownhistologic type, is still debated. An alternative isneedle aspiration/cytology in palpably enlargedlymph nodes.Other Radionuclide Scans: Brain and liver scanshave been replaced by CT assessment of theseorgans in patients with suspicious clinical findings,abnormal laboratory test results, or intermediateor advanced stages of tumor.Imaging to Detect Occult ExtrathoracicMetastasesPatients undergoing surgical therapy forNSCLC have occult metastatic disease 20% of thetime in the adrenal glands, liver, kidney, and/orbrain.Brain CT Imaging and MRI: Some studies suggestan increased utility of brain imaging, even inthe absence of symptoms or signs of CNS metastases,for patients with stage IB or greater tumors,adenocarcinoma, and/or undifferentiated carcinoma/large-cellcarcinoma. If a comprehensiveclinical evaluation is negative, however, patientswith NSCLC will be found to have brain metastasesby CT assessment only 3% of the time. MRI,particularly when enhanced with gadolinium, ismore sensitive than CT.Brain imaging in patients with NSCLC shouldbe performed for patients with headaches orfor those with nonspecific findings that suggestwidespread disease if metastatic disease has notbeen found elsewhere. A brain CT scan (or preferablyan MRI) usually os warranted for patientswith SCLC.Hepatic and Adrenal (Abdominal) CT Imaging:The liver is one of the most frequent sites of metastaticinvolvement in SCLC, seen in 15 to 28% ofpatients at the time of presentation and 6 to 10%of patients as the sole site of metastatic disease.The use of screening liver function tests helpto determine which patients have metastatic<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 51ACC-PULM-09-0302-003.indd 517/10/09 8:04:24 PM


liver disease, both for SCLC and NSCLC. InNSCLC the absence of hepatomegaly or discreteintrahepatic masses on physical examinationplus normal liver function test results indicatethat hepatic CT imaging should not be doneroutinely. Conversely, routine hepatic imagingis more reasonable for patients with SCLC, asinvolvement is so common and hepatic CT imagingprovides useful information for staging andmonitoring response to therapy. CT imaging ismuch superior to radionuclide imaging, particularlywhen contrast material is injected intravenously.Chest CT, done properly, will extendinto the upper abdomen and provide imagingof the adrenal glands and the upper part of theliver.There are major pitfalls of CT in both organs,however. Cysts and hemangiomas within the liverare quite common. They usually are indistinguishablefrom metastases on CT imaging. When thishappens, an ultrasound study will help to differentiate.Needle biopsy of the liver, or occasionalopen liver biopsy, may be needed before denyingan otherwise resectable patient with NSCLC thechance for a curative thoracotomy.Adrenal adenomas occur in up to 2 to 10% ofthe population and look just like metastases in theadrenal by CT imaging. Thus, CT-guided needleaspiration biopsy to prove they are metastases isessential in NSCLC, provided this is the only siteof extrathoracic disease, before denying an otherwiseresectable patient with NSCLC the opportunityfor curative resection.Tumor Markers: No tumor marker is useful toscreen for lung cancer in the general smoking population.Markers that have been studied includecarcinoembryonic antigen (CEA), total sialic acid,lipid bound sialic acid, subunits of human chorionicgonadotropin, β 2-microglobulin, lipotropin,calcitonin, and parathyroid hormone. The routinemonitoring of any of these substances is not recommendedin the screening, staging, or evaluationof disease progression.The CEA level may have prognostic value inthe follow-up of patients with NSCLC after surgery,because a level 15 ng/mL before surgery isassociated with a reduced possibility of a successfulresection. During follow-up, patients with anelevated CEA level relapse more often than patientswith CEA levels that remain normal.Determination of Resectability and OperabilityIn contemplating potential surgery, one mustask two questions: is the tumor resectable and isthe patient operable. The answer to the first questionshould be apparent after the workup andbasically depends on whether all tumor tissuecan be removed at the time of surgery. It is asearch for metastatic disease, for there is no benefitfrom debulking lung cancer. The answer tothe second question depends most heavily on theoverall health of the patient. In other words, canthe patient withstand the stress of surgery anddo the potential benefits of the surgery outweighthe risks?Operability usually relates to the functionalcapacity of the cardiorespiratory system. Cardiorespiratoryproblems account for at least threefourthsof postoperative morbidity and mortality.Cardiac complications cause approximately 20%of postoperative deaths. A history of cardiac diseasedoubles the risk of major morbidity (18% vs9%). <strong>Pulmonary</strong> complications, however, are morecommon than cardiac problems.The preoperative physiologic assessment of apatient being considered for surgical resection oflung cancer must consider the immediate perioperativerisks from comorbid cardiopulmonarydisease, the long-term risks of pulmonary disability,and the threat to survival due to inadequatelytreated lung cancer. As with any planned majoroperation, especially in a population predisposedto atherosclerotic cardiovascular disease by cigarettesmoking, a cardiovascular evaluation is animportant component in assessing perioperativerisks. Measuring the FEV 1and the diffusing capacityof the lung for carbon monoxide Dlco measurementsshould be viewed as complementaryphysiologic tests for assessing risk related to pulmonaryfunction. If there is evidence of interstitiallung disease on radiographic studies or unduedyspnea on exertion, even though the FEV 1maybe adequate, a Dlco should be obtained.In patients with abnormalities in FEV 1or Dlcoidentified preoperatively, it is essential to estimatethe likely postresection pulmonary reserve. Theamount of lung function lost in lung cancer resectioncan be estimated by using either a perfusionscan or the number of segments removed. A predictedpostoperative FEV 1or Dlco 40% indicates52 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 527/10/09 8:04:25 PM


an increased risk for perioperative complications,including death, from lung cancer resection. Exercisetesting should be performed in these patientsto further define the perioperative risks beforesurgery.Formal cardiopulmonary exercise testing is asophisticated physiologic testing technique thatincludes recording the exercise ECG, heart rateresponse to exercise, minute ventilation, and oxygenuptake per minute, and allows calculation ofmaximal oxygen consumption (V·o 2max). Risk forperioperative complications can generally bestratified by V·o 2max. Patients with preoperativeV·o 2max 20 mL/kg/min are not at increased riskof complications or death; V·o 2max 15 mL/kg/min indicates an increased risk of perioperativecomplications; and patients with V· o 2max 10mL/kg/min have a very high risk for postoperativecomplications. Alternative types of exercisetesting include stair climbing, the shuttle walk, andthe 6-min walk. Although often not performed ina standardized manner, stair climbing can predictV·o 2max. In general terms, patients who can climbfive flights of stairs have V·o 2max 20 mL/kg/min.Conversely, patients who cannot climb one flightof stairs have V·o 2max 10 mL/kg/min. Data onthe shuttle walk and 6-min walk are limited, butpatients who cannot complete 25 shuttles on twooccasions will have V·o 2max O 2max 10 mL/kg/min. Desaturation during an exercise test has beenassociated with an increased risk for perioperativecomplicationsMorbidity and Mortality After Surgery: Theoverall 30-day operative mortality rate is 3.7%.Age is an important variable: for patients 60years old, the mortality rate is 1.3%; between 60and 69 years, 4.1%; and 70 years, 7.1%. Somestudies indicate a lower mortality, even amongoctogenarians, when resection does not requiremore than a lobectomy. The extent of resectionalso influences mortality: pneumonectomy, 6.2%;lobectomy, 2.0%; and lesser resections, 1.4%.Microscopic tumor at the bronchial margin doesnot appear to preclude prolonged survival. Microscopicextramucosal spread to peribronchial tissuesis associated with much poorer outcome.NSCLC with solitary cerebral metastases is aspecial situation because of the potential for curein a few patients with resection of both the metastasis(which should be addressed first) and theprimary tumor. A recent randomized trial suggeststhat surgical resection of the brain metastasis followedby cranial irradiation is associated withbetter survival and much better control of neurologicsymptoms than cranial irradiation alone. It isimportant to recognize that not all pulmonarycancers in previously treated patients with NSCLCrepresent recurrence of the original tumor. Surgicalresection with curative intent is sometimes possiblefor these patients.TreatmentSCLCSCLC, more than other solid tumors, acts as asystemic disease, and therefore, the cornerstone oftherapy is chemotherapy. Combinations of activedrugs are more effective than any single agent.Drugs active as single agents typically are chosenfor combination therapy, such as cisplatin, carboplatin,etoposide, docetaxel, and irinotecan. Theresponse rate to first-line chemotherapy is excellent( 80%), so much so that the superior vena cavasyndrome attributable to SCLC may be treatedwith chemotherapy.First-Line Chemotherapy: Platinum-based che -motherapy is the mainstay of treatment for SCLC.This class of drugs is most commonly combinedwith etoposide. Carboplatin plus etoposideappears to be as effective as cisplatin plus etoposidebut is less toxic (except for increased myelosuppression).Treatment is given for four to sixcycles, which takes about 4 to 5 months to complete.Newer agents are under investigation andmay replace current programs because of greaterefficacy and lesser toxicity. Recent studies havesuggested that irinotecan plus cisplatin is an effectivetreatment. Initial response rates for extensivedisease may be 60 to 85%, with a complete remissionin 20 to 30%. Median survival among thosetreated is 6 to 12 months. Two-year survival is20%, and 5-year survival is 5%.Chemotherapy doses have been escalated upto those requiring bone marrow transplant rescuewithout any benefit. This strategy has failed. Maintenancechemotherapy beyond four to six cyclesonly adds toxicity and expense without benefit.The hypothesis that emergence of drug-resistanttumor cells can be minimized by rapid alteration<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 53ACC-PULM-09-0302-003.indd 537/10/09 8:04:25 PM


of equally effective non--cross-resistant combinationsof drugs has been tested. As yet, there hasbeen no evidence of benefit from the use of thisstrategy.Second-Line Regimens: Despite high initial responserates to chemotherapy (complete responsein 45 to 75% of patients with limited disease and20 to 30% in extensive disease), the responseduration is usually short. In general, the progression-freeperiod is 4 months for extensive diseaseand 12 months for limited disease. It goes withoutsaying, however, that when the relapse appears,the prognosis is grim, with a median survivaltime of 3 to 4 months. Palliation is usually thegoal.Those who relapse 3 months after first-linetherapy may be retreated with the same inductionregimen used initially. The chance of a secondresponse is approximately 50%. Those who relapse 3 months after first-line therapy and are in satisfactoryclinical condition should be offered asecond-line regimen. No single second-line treatmentis standard. A number of single agents andcombinations may be used but a response is foundin 10%.A number of recent studies have looked attopotecan as a second-line therapy. This drug hasnow been approved as second-line therapy.Response to topotecan in patients with a relapse 3 months after therapy is 25 to 35%. Enrollmentin clinical trials is encouraged.Limited-Stage Disease: Combined Modality Treat -ment: Disease confined to one hemithorax (withoutdistant metastasis) and the mediastinum,which can be encompassed within a radiationtherapy treatment field, is treated with concurrentchest radiation and chemotherapy. This producessuperior survival compared with sequential therapyor chemotherapy alone. Concurrent radiationwith the induction phase of chemotherapy isjudged “best” at the present time, as there seemto be greater response rates for local control ofintrathoracic disease. Patients so-treated have amedian survival of 16 to 24 months, a responserate of 65 to 90%, a complete response rate of 45 to70%, and 20% may be alive at 5 years. Hyperfractionatedconcurrent radiation (smaller fractionsdelivered twice daily) results in a slightly superiorsurvival as compared with standard dailyfractionation.Extensive-Stage Disease: chemotherapy: Chemotherapyalone is used. Radiotherapy is notusually administered because patients with metastaticdisease usually relapse in the site of previousmetastases, even if they achieve a completeremission. The cause of death in these patientsis usually widespread metastatic involvement,and not disease in the chest. Treatment is primarilyaimed at the systemic disease, and the patientusually receives chemotherapy alone. Radiationmay palliate symptoms, but it does not prolongsurvival.Surgery: The role of surgery in patients knownto have small cell lung cancer is controversial.Many would say that surgery is never indicated.No randomized trial has shown a survival advantagein patients with SCLC who receive surgicaltreatment. A few highly selected nonrandomizedtrials, however, have reported 5-year survivals inthe 25 to 40% range. A recent report that pooledthe available studies found a 5-year survival of 40to 53%, which is clearly impressive. Taking thisinto account, the <strong>ACCP</strong> Guidelines suggest thatin patients documented to have clinical T1 or T2,NO tumors, surgery is an option in selected casesand should be followed by platinum-based chemotherapy.Prophylactic Cranial Irradiation: The <strong>ACCP</strong>guidelines suggest that prophylactic cranial irradiation(PCI) should be offered to patients witheither limited or extensive stage disease whoachieve a complete response to first-line therapy.The argument for PCI suggests that in those whoachieve complete remission, there is a 60% chancethat CNS metastases will develop within 2 to 3years of starting treatment. The high rate of CNSfailures can be attributed to the fact that most chemotherapyagents do not adequately penetrate theblood−brain barrier. The brain is a sanctuary site.Once overtly present, although often respondingto chemotherapy or radiation therapy (XRT),the patient is rarely if ever cured. The chance ofdeveloping a risk for CNS metastasis can be cut inhalf by delivering whole brain radiation to about2,500 to 3,000 cGy. In addition, one study found amodest improvement in 3-year survival from 15to 21%.The argument against PCI suggests that wholebrainXRT may result in a decrease in neuropsychologicalfunctioning. Several prospective studies,54 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 547/10/09 8:04:25 PM


however, have now shown that patients treatedwith PCI do not have detectably different neuropsychologicalfunctioning than patients not treated,at least in a 2-year follow-up. Any decrease infunctioning appears not to be caused by PCI butby other treatment components or may be a partof the disease.NSCLC• To oversimplify, for stage Ia And Ib, the treatmentis surgery.• For stage IIa and IIb, the treatment is surgery.• For stage IIIA, the treatment is surgery forsome, neoadjuvant therapy followed by surgeryfor some, and chemotherapy/radiationtherapy for some.• For stage IIIb, except for some T4N0M0, thetreatment is chemotherapy and XRT, preferablyconcurrently, if possible.• For stage IV, the treatment is “best supportivecare” or chemotherapy.Surgery: Surgery is the most effective methodfor curing NSCLC. Surgical treatment for cure ispredicated on achieving a complete resection (R0resection). Lobectomy is the most widely usedoperation. Pneumonectomy must be performed toremove all known tumor in 10 to 55% of patients.There is increasing interest in the use of conservativeresections. The minimal operation encompassingall the known areas of disease is used,conserving pulmonary tissue and pulmonaryfunction. Wedge resections, segmental resections,and sleeve lobectomies are increasingly morecommon. Resection via thoracoscopy has beenused, but complete dissection of lymph nodes isnot possible (and is clearly an important part ofoptimal lung cancer surgery). No matter whatthe extent of lung tissue resection, careful intraoperativestaging is needed, including frozen sectionsof biopsies of surgical margins and lymphnodes. More aggressive resections may includethe chest wall, diaphragm, and lower roots ofthe brachial plexus, and occasionally portionsof the atria. Resection of ipsilateral mediastinallymph nodes (N2 disease) is possible.There are a few unique situations such asT3/ NI/MO, which is a IIIA tumor. The guidelinessuggest that those with T3N1 tumors be treated thesame way as patients with stage II disease. Thesetumors are grouped as IIIA by virtue of a 5-yearsurvival more in keeping with other IIIA tumors(29% when staged pathologically and opposed to39 to 55% for stage IIB and IIA, respectively). Additionally,a T4/N0,1,2,3/MO is stage IIIB tumor, butsurgery is an appropriate approach in those veryfew patients who are T4N0 because of either satellitetumor nodules within the primary tumor lobeor carinal involvement.NSCLC of the superior sulcus (Pancoast tumor)is uncommon. For such tumors demonstrated tobe without mediastinal node involvement, preoperativeconcurrent chemotherapy plus radiationtherapy, followed by surgical resection, providesthe best long-term survival.Radiation Therapy: Radiotherapy has a role inearly, medically inoperable, and locally advancedunresectable NSCLC, in the palliation of advancedlung cancer of all types, and in the adjuvant treatmentof limited stage SCLC. Radiation therapyis clearly inferior to surgery in the curative treatmentof NSCLC (15 to 35% cure rate). The limitationsof radiation therapy relate to the toleranceof pulmonary tissues because nearly any tumorcan be eradicated by irradiation given sufficientlyhigh doses. Selected reports indicate a high curerate for early-stage lung cancers, but these studiesare uncontrolled. Radiation therapy leads to a lossof lung function and is a poor choice for patientswho are inoperable because of insufficient pulmonaryreserve. Standard therapy consists ofapproximately 60 Gy, with the dose dividedamong 30 sessions during a period of 6 weeks,although greater doses have been used. Technologicalimprovements in XRT delivery, such asthree-dimensional conformal radiation therapy,intensity modulated radiation therapy, and stereotacticradiosurgery may allow the delivery ofhigher doses of radiation to the tumor.There is no established role for preoperativeradiation therapy in patients with operable lungcancer, although uncontrolled studies suggest abenefit from preoperative radiation therapy forpatients with Pancoast tumors. There is no role forpostoperative radiation therapy for patients withoutnodal involvement.The effect of prophylactic cranial radiation ispronounced in preventing clinical brain metastasesin patients with adenocarcinoma (0% vs 29%) butstill not considered standard therapy. There is no<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 55ACC-PULM-09-0302-003.indd 557/10/09 8:04:25 PM


apparent survival benefit, however, and significantneuropsychological deficits may develop. Palliativeradiation is clearly beneficial in the managementof specific symptoms, such as pain control,hemoptysis, superior vena cava syndrome, andatelectasis.Chemotherapy: Chemotherapy agents availablein the 1980s were relatively ineffective, producingresponse rates of 15%. The combination oftwo active agents produced only marginally betterresponse rates of 20%. Moreover, these slightgains were accompanied by significant toxicityand risk. As a result, many physicians adoptedan attitude of “therapeutic nihilism” (not withoutmerit) and routinely recommended no therapybeyond “best supportive care” for most patients.A 1992 state-of-the-art review of chemotherapyfor patients with lung cancer concluded that theimpact of chemotherapy on survival had notbeen demonstrated and that “it is difficult torecommend incorporating chemotherapy intothe standard care of patients with disseminatedNSCLC.”Several years later, in 1995, a metaanalysis ofclinical trials that compared best supportive carewith cisplatin-based combination chemotherapydemonstrated a slight benefit for the latter, with anincreased median survival of 1.5 months and a 10%improvement in 1-year survival. Later in the 1990s,a number of new agents (taxanes, gemcitabine,vinorelbine, and irinotecan) became available.When combined with platinum compounds, theseagents produced response rates of up to 60%,median survival of 12 months, and 1-year survivalrates of 50% or better in Phase II trials.Unfortunately, as is usually the case, recentlypublished Phase III trials have yielded less optimisticresults. A recent study compared four commonlyused treatment regimens (cisplatin pluspaclitaxel, cisplatin plus gemcitabine, cisplatin plusdocetaxel, carboplatin plus paclitaxel). Theresponse rate ( 50% reduction in measurabledisease) for all 1,155 eligible patients was 19%, witha median survival of 7.9 months, a 1-year survivalof 33%, and a 2-year survival rate of 11%. None ofthe combinations offered a significant advantageover the others.Despite the modest benefit, more recent metaanalysesand randomized trials have suggestedthat chemotherapy should be considered inpatients with stage IV NSCLC who have a goodperformance status (ECG/Zubrod 0, 1, 2, or 70% on the Karnofsky scale). This therapy maymodestly prolong survival, reduce symptoms,and improve quality of life with an acceptablelevel of risk. The cost of chemotherapy compareswith that of other accepted medical interventionsand may be less than best supportive care. It canbe fairly concluded, however, that in 2009, stateof-the-artstandard chemotherapy treatment forthis population provides a median survival ofapproximately 8 to 12 months (a mere 3- to 4-month extension over nontreated) and a 1-yearsurvival rate of 33% (as compared with 10% inuntreated). This modest benefit must be weighedagainst a 3 to 5% risk of chemotherapy-relateddeath, most often the result of infections duringperiods of neutropenia.Preoperative (“Neoadjuvant”) Therapy: This termrefers to the use of nonsurgical therapy as initialtreatment for cases in which surgery is a suboptimalinitial approach. Neoadjuvant therapy mightresult in both tumor shrinkage and early eradicationof systemic micrometastases. Several studies,although not all, have suggested that patientswith stage IIIA NSCLC who are given severalcycles of chemotherapy followed by surgicalresection (postoperative treatment such as radiationtherapy or more chemotherapy has beenvariable in these studies), have a survival advantageover those treated without the addition ofpreoperative chemotherapy. Preoperative chemotherapy(“neoadjuvant chemotherapy”) is bestconsidered an investigational approach requiringfurther validation in controlled trials before it isincorporated as the new standard of practice forsuch patients.Postoperative (“Adjuvant”) Therapy: This refersto the use of radiation or chemotherapy to im -prove survival after a tumor has been completelyresected. Genuine promise exists that systematicadjuvant chemotherapy will increase the curerate in “operable” NSCLC. Because modern combinationchemotherapy produces remission in40 to 50% of good performance status patientswith metastatic NSCLC and because adjuvantchemotherapy increases survival in othercommon malignancies (colorectal and breastcancer), it is logical to suspect that adjuvant chemotherapymay prolong survival in NSCLC. In56 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 567/10/09 8:04:26 PM


theory, chemotherapy with a cytotoxic agent mayeliminate micrometastases, improving survival.The International Adjuvant Lung Cancer Trialevaluated the impact on survival of three to fourcycles of adjuvant cisplatin-based chemotherapyafter complete resection compared with resectionand no chemotherapy. Chemotherapy significantlyimproved survival rates by 3% at 2 years and 5%at 5 years (70% vs 67% at 2 years; 45% vs 40% at5 years; p 0.03) and disease-free survival by 6%at 2 years and 5% at 5 years (61% vs 55% at 2 years;39% vs 34% at 5 years; p 0.003). The authors of ametaanalysis published in 2006 found an overall5.5% survival advantage. The stage II subset demonstrateda 27% reduction in risk. The stage IBsubset trended positive but the benefit was notstatistically significant. Unfortunately, there was anegative impact in the stage IA subset. The <strong>ACCP</strong>Guidelines recommend that adjuvant chemotherapybe considered in completely resected stage IIand stage IIIA patients who can tolerate chemotherapy.On the basis of these and other trials, mostconsider the controversy surrounding adjuvantchemotherapy to be over and state that that adjuvantplatinum-based chemotherapy should berecommended after complete resection of selectedNSCLC in patients with a good performancestatus. It is important to remember that the potentialimprovement in overall survival comes at thecost of possible increased morbidity and evenmortality.The results of adjuvant radiotherapy have beenvariable, with some trials showing a benefit,whereas most showing none. The authors of a largemetaanalysis published in 1998 suggested thatpostoperative radiotherapy was actually detrimentalwith an increase in relative risk of death. Therisk was most marked in stage I patients but notstatistically significant in N2 patients. One mustsay, however, that the data used in the analysis wasvariable and often dated. A recent update of thismetaanalysis in 2005 continues to show decreasedsurvival for N0 and N1 patients. Another significantstudy concluded that adjuvant radiotherapyimproves local control in patients with N2 diseasebut did not improve overall survival. The <strong>ACCP</strong>Guidelines recommend against adjuvant radiationin completely resected stage I and II lung cancerbut recommend that adjuvant XRT be consideredin completely resected stage IIIA disease to reducelocal recurrence. Postoperative radiotherapy is,however, indicated for incompletely resected diseaseand when surgical margins are positive butre-resection is not feasible.Stage IIIA DiseaseThe 10% of patients with NSCLC who presentwith stage IIIA cancer are described by the <strong>ACCP</strong>guidelines as “perhaps the most therapeuticallychallenging and controversial subset” of lung cancerpatients because they fit between the generallyresectable stage I and stage II tumors and the unresectablestage IIIB and stage IV cases. An approachis the combination of induction chemotherapy orchemoradiotherapy followed by surgery, but thishas not been widely used for long enough to providereliable data involving large groups ofpatients.Patients with N2 disease may be divided intofour subgroups for the purpose of treatment:• IIIa 1: incidental nodal metastases found on finalpathologic examination of the resection specimen;• IIIa 2: nodal metastases (single station) recognizedintraoperatively;• IIIa 3: nodal metastases (single or multiple station)recognized by prethoracotomy staging;• IIIa 4: bulky or fixed multistation N2 disease.Patients with incidental N2 disease (stagesIIIA 1and IIIA 2) can undergo resection as plannedif only one nodal station is involved and the nodesand primary tumor are all technically resectable.The guidelines also recommend performing amediastinal lymphadenectomy. Those patientswho are completely resected may benefit fromadjuvant cisplatin-based chemotherapy.In those patients in whom complete resectionis not possible, the <strong>ACCP</strong> Guidelines recommendaborting the planned surgery but conducting systematicmediastinal lymph node sampling orcomplete mediastinal lymph node dissection.Adjuvant radiation therapy may reduce the risk oflocal recurrence in selected patients.Potentially resectable N2 disease (stage IIIA 3)should be treated with two or more modalities,according to the guidelines because surgery alonehas a poor prognosis. Multidisciplinary evaluation<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 57ACC-PULM-09-0302-003.indd 577/10/09 8:04:26 PM


of each patient will assist in determining the besttreatment plan.Bulky N2 disease (stage IIIA 4) traditionally wastreated with radiation alone until the advent ofmore effective platinum-based chemotherapyregimens made the combination of both therapiesfeasible. The guidelines recommend using bothmodalities in patients with good performancestatus.Stage IIIB DiseaseSignificant advances in the management of“unresectable” locally advanced NSCLC (stage IIIBand many IIIA patients) have been made. The oldertraditional standard treatment of radiation alonehas been replaced by chemoradiotherapy. Althoughconsensus exists that chemoradiotherapy is superior,major issues that remain unresolved are:sequential vs concurrent modality treatment;induction chemotherapy vs concurrent treatment;and the specific choice of dose and schedule ofchemotherapeutic agents to be used with radiation.One common regimen is cisplatin plus etoposideplus radiation concurrently. Results of the RTOG9410 trial showed that concurrent chemoradiotherapywas associated with better survival ratesthan sequential therapy. The improved outcomecomes at a cost. Patients treated concurrentlyhave a greater rate of acute side effects, and severeesophagitis is common.Stage IV DiseaseSingle-agent chemotherapy is of little value.The most active traditional drugs include cisplatinand vinblastine. In recent years, several newdrugs have demonstrated activity, such asvinorelbine, paclitaxel, docetaxel, carboplatin,topotecan, and gemcitabine. Combination chemotherapyoffers a small survival advantage comparedwith best supportive care. Two of the bestregimens are carboplatin plus paclitaxel and cisplatinplus gemcitabine. These therapies are costeffective compared with best supportive care.They should be offered to ambulatory patients.However, nonambulatory (poor performancestatus; Eastern Cooperative Oncology Groupstatus 2 or greater) patients have greater toxicityand lower response rates. Consequently, theymay be managed with single-agent therapy orbest supportive care.Patients with stage IV disease should be limitedto three or four (perhaps up to six) cycles of twoagentchemotherapy. Because virtually all treatmentfor stage IV disease fails, second-line therapyoften will be necessary. Studies have suggested thatsingle-agent docetaxel or pemetrexed (and perhapsother agents) may offer some survival benefit.Erlotinib is approved for the treatment of patientswith locally advanced or metastatic NSCLC afterfailure of first-line or second-line chemotherapy.The use of this drug conveys a small survival benefit(6.7 vs 4.7 months with placebo).A recent study has shown that the addition ofbevacizumab to carboplatin and paclitaxel conveysa 2-month survival benefit (10.2 vs 12.5 months)over carboplatin and paclitaxel alone. Bevacizumabis a humanized monoclonal antibody that targetsthe vascular endothelial growth factor (VEGF). Bypreventing the growth factor from interacting withendothelial cell, tumor-induced angiogenesis istheoretically hampered. Hemoptysis has beennoted and is more commonly seen in patients withsquamous cell histology. As a result, patients withsquamous cell histology have been excluded fromrecent trialsVirtually all patients with stage IV disease willexperience recurrence. In appropriate patients, asecond- or third-line chemotherapy regimen maybe used. In this setting, docetaxel has an 8.8%response rate, and the median survival is improvedby 2.2 months. Similar results have been shown forpemetrexed and erlotinib in this setting.Targeted Therapy: A promising area of researchemanates from the study of tumor cell biology andthe molecular mechanisms of oncogenesis. Therecognition that alterations in proteins and genesinvolved in cell signaling, the cell cycle, and thecontrol of programmed cell death has suggestedthat therapies targeted directly at the molecularalterations inherent in neoplastic cells may leadto reversal of the malignant phenotype. Such “targetedtherapy” is distinguished by inhibition ofspecific pathways and processes in the cancer cellrather than a generalized attack on cell proliferationcharacterized by standard cytotoxic chemotherapy.A number of such “targeted therapies”are under active investigation. As an example,gefitinib and erlotinib affect signal transduction58 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 587/10/09 8:04:26 PM


initiated by the binding of epidermal growth factorto cell surface receptors.The initial trials of epidermal growth factorreceptor (EGFR)-blocking agents revealed that theseagents were not uniformly effective. A closer look atthe data indicated that patients with certain clinicalcharacteristics (never smokers, women, Asian ethnicity,and adenocarcinoma) may exhibit dramaticresults. This important clinical observation led to asearch for predictors of responsiveness to theseagents. Research focused on the detection of EGFRmutations. Mutations have been demonstrated inthe first four exons (18 to 21) of the EGFR-TKdomain. The highest frequency of EGFR mutationshas been demonstrated in those patients with clinicalcharacteristics that predict TKI response.Multiple prospective trials have demonstratedthat patients whose tumors harbor these geneticchanges have a collective response rate of nearly75% to EGFR-TK inhibitors. In addition, it has beenshown that increased EGFR gene copy number (asdetermined by fluorescent in situ hybridization) isa positive predictor of EGFR-TKI response. Interestingly,a mutated K-ras gene (KRAS) is a strongpredictor of nonresponse to EGFR inhibitors.Tumors rarely have mutations in both EGFR andKRAS. K-ras is a critical downstream effector of theEGFR pathway that has been found to be mutatedin 15 to 30% of adenocarcinomas and in associatedwith tobacco smoking.Another target is neoangiogenesis. An increasedblood supply is necessary for tumor growthbeyond a certain volume. Tumor-derived and stromal-derivedVEGF binds to receptors on endothelialcells triggering intracellular pathways that leadto endothelial cell proliferation and increased permeability.Countering this process, as an example,bevacizumab (a humanized anti-VEGF monoclonalantibody) binds to and neutralizes VEGF, thuspreventing the initiation of the signaling pathwaythat leads to angiogenesis. Many other drugs targetingmany other molecular level processes areunder study.Tailoring therapy based on genetic or proteinexpression profiles of individual tumors and utilizingtargeted therapy based on this information hassignificant theoretical promise and may prove tobe a significant advance.Personalized Therapy: Recent studies have suggestedthat tailoring chemotherapy based on theindividual’s tumor profile may be efficacious.One such study based therapy on whether thetumor overexpressed ribonucleotide reductasesubunit 1 (RRM1) and excision repair crosscomplementinggroup 1 (ERCC1). RRM catalyzesthe biosynthesis of deoxyribonucleotidesfrom corresponding ribonucleotides. RRM1, theregulatory subunit of RRM, plays a role in suppressingtumor cell migration and metastasis formation.It is the molecular target of gemcitabine.High expression of RRM1 correlates negativelywith gemcitabine efficacy (ie, tumors with lowRRM1 expression responded better to treatmentcompared with tumors with high levels of expression).ERCC1 is a DNA damage repair gene thatencodes the 5 endonuclease of the nucleotide excisionrepair complex. Increased ERCC1 predictsresistance to cis-platin. In this study, selection ofdouble-agent chemotherapy based on RRM1 andERCC1 expression proved feasible and also effectivewhen compared with historical controls in aPhase II study (eg, survival at 1 year was 59% vsa historical survival of 33%). Phase III studies areunderway.PrognosisSCLCWith changes in the treatment during the past15 to 20 years, the median survival time hasincreased from 3 months to approximately 1 year.For patients with limited disease, the median survivalranges from 1 to 2 years, and approximately10 to 20% remain alive at 2 years. There is a potentialfor long-term cure. Thirty percent of patientswith small-cell carcinoma die of local tumor complications,and 70% die of carcinomatosis. Fiftypercent of patients with small-cell carcinoma havebrain metastases at autopsy.NSCLCThe most important determinant of survivalduration is stage of the disease. Seventy-five percentof patients with squamous cell carcinoma dieof complications of the thoracic tumor, as only 25%have extrathoracic dissemination at autopsy. Fortypercent of patients with adenocarcinoma and largecellcarcinoma die of intrathoracic complications,<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 59ACC-PULM-09-0302-003.indd 597/10/09 8:04:26 PM


whereas 55% have distant metastases. Fifty percentof patients with large-cell carcinoma and adenocarcinomahave brain metastases at autopsy. CNSdeaths are seen in 8 to 9% of patients with large-cellcarcinoma and adenocarcinoma. Two percent ofCNS deaths are found in patients with squamouscell carcinoma. Thus, there is a clear distinctionbetween squamous cell carcinoma and other formsof lung cancer in terms of the mode of death.Local/regional disease dominates squamous cellcarcinoma. Carcinomatosis dominates the lattergroups of patients. The median survival withmetastatic NSCLC varies from 6 weeks to morethan l year, depending on the initial Karnofskyperformance status (the most important variable),the extent of disease, and the presence or absenceof weight loss in the 6 months before diagnosis.Biological prognostic factors, including mutationof the tumor suppressor gene (p53), activationof K-ras oncogenes, and other biological markers,may have significant value in predicting a poorprognosisLung Cancer as a Second PrimaryThis disease may occur after a previous diagnosisof nonpulmonary primary cancer (which isthe case in up to 9.7% of all lung cancer patients).Lung cancer may occur as a second or even a thirdprimary tumor after successful treatment of a previousprimary lung cancer. Subsequent lung cancersmay be any cell type, but the second tumor isusually NSCLC. This is a particular problem inlong-term survivors of combined modality treatmentfor small-cell carcinoma (up to 4.4 per personper year). This underscores the need for patientswith a lung cancer to stop smoking cigarettes. The<strong>ACCP</strong> Guidelines recommend that, after successfulsurgery, patients be examined and undergo animaging study every 6 months for 2 years thenyearly thereafter.ConclusionAlthough a 5-year survival of 16% is meagerand still dismal, the near doubling of the 5-yearsurvival rate (when compared with the 1950s) hasprovided some room for optimism and has begunto shift the nihilism associated with lung cancertreatment into a guarded optimism. In addition, anumber of promising new drugs have been incorporatedinto clinical trials, and many more are inthe pipeline. Specifically targeted biological therapiesare particularly promising. New diagnosticmodalities, such as PET, are finding widespreaduse and may alter our diagnostic and therapeuticalgorithms. New surgical procedures and techniqueshave been developed and perfected. Saferand more effective methods of delivering radiationare coming into clinical use and the medical communityis hopeful that lung cancer screening willprove to convey a mortality benefit and be costeffective. Even with all this, it must be said oncemore that all the research, all the clinical care, andeven this review would not be necessary, if we couldfind a way to eliminate the problem of tobacco.Annotated Bibliography<strong>Review</strong>sHerbst RS, Heymach JV, Lippman SM. Molecularorigins of cancer: lung cancer. N Engl J Med 2008;359:1367−1380This review focuses on the major advances in the molecularstudy of the origins and biology of lung cancer.Molina JR, Yang P, Cassivi SD, et al. Non-small celllung cancer: epidemiology, risk factor, treatment andsurvivorship. Mayo Clin Proc 2008; 83:584−594Excellent short review.Spira A, Ettinger DS. Multidisciplinary managementof lung cancer. N Engl J Med 2004; 350:379−392Concise review of radiotherapy, chemotherapy, and the combinationin the treatment of lung cancer.Subramanian J, Govidan R. Lung cancer in never smokers:a review. J Clin Oncol 2007; 25:561−570Although up to 90 % of lung cancer is “smoking assisted,”lung cancer in never smokers is being recognized more frequently.ScreeningBach PB. Lung cancer screening. J Natl Compr CancNetw 2008; 6:271−275Bach PB, Jett JR, Pastorino U, et al. Impact of computedtomography screening on lung cancer outcomes. JAMA2007; 297:1−9Henschke CI, Yankelevitz DF, Libby DM, et al. Survivalof patients with stage I lung cancer detected onCT screening. N Engl J Med 2006; 355:1763−177160 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 607/10/09 8:04:26 PM


CT screening can identify lung cancers when they aresmall and predominantly stage I. It is hoped that the randomizedcontrolled trials currently underway will provideevidence relating to the important issue of mortalitybenefit.Midthun DE, Swensen SJ, Hartman TE, et al. Lung cancerscreening results: easily misunderstood. Mayo ClinProc 2007; 82:14−15Excellent short reviews of the arguments for and againstscreening for lung cancer with a discussion of future methods.StagingMountain CF. Revisions in the international system forstaging lung cancer. Chest 1997; 111:1710−1717Mountain CF. Regional lymph node classification forlung cancer staging. Chest 1997; 111:1718−1723Classic articles that describe important modifications in theTNM system for staging patients with lung cancer.Rami-Porta R, Ball D, Crowley J, et al. The IASLC lungcancer staging project: proposals for the revision of theT descriptors in the forthcoming (seventh) edition ofthe TNM classification for lung cancer. J Thorac Onc2007; 2:593−602Rami-Porta R, Crowley JJ, Goldstraw P. The revisedTNM staging system for lung cancer. Ann Thorac CardiovascSurg 2009; 15:4−9Proposals for changes in the descriptors and staging of lungcancer.Clinical Practice GuidelinesAlberts WM, American College of Chest Physicians.Diagnosis and management of lung cancer: <strong>ACCP</strong> evidence-basedguidelines. Chest 2007; 132:1S−422SClearly and simply, the “gold standard.” The second editionof the Guidelines was published in 2007 as a supplement toCHEST.British Thoracic Society guidelines: guidelines on theselection of patients with lung cancer for surgery. Thorax2001; 56:89−108This guideline addresses the work-up necessary to answerthe questions of resectability and operabilityEttinger DS, Akerley W, Bepler G, et al. Non-small celllung cancer. Clinical practice guidelines in oncology.J Natl Compr Canc Netw 2008; 6:228−269Clinical Practice Guidelines from a well-respected organizationpresented in an algorithm type format.Pfister DG, Johnson DH, Azzoli CG, et al. AmericanSociety of Clinical Oncology Treatment of unresectablenon-small-cell lung cancer guideline: update 2003J Clin Onc 2004; 22:330−353Update of the consensus based ASCO Guidelines firstpublished in 1997. Excellent discussion of the recent literatureon the diagnosis and management of unresectabledisease.SmokingEbbert JO, Sood A, Hays JT, et al. Treating tobaccodependence: review of the best and latest treatmentoptions. J Thorac Oncol 2007; 2:249−256Wilson JF. In the clinic. Smoking cessation. Ann InternMed 2007; 146:ITC2-1−ICT2-16Quick reviews of the latest treatment options.TreatmentAlberts WM. Follow-up and surveillance of the patientwith lung cancer: what do you do after surgery? Respirology2007;12:16−21Discussion of the appropriate follow-up for patients afterdefinitive surgical treatment monitoring for recurrence ofthe original tumor and for the potential development of asecond primaryMolina JR, Adjei AA, Jett JR. Advances in chemotherapyof non-small cell lung cancer. Chest 2006;130:1211−1219Concise discussion of the history and current state-of-the-artof chemotherapy in the management of lung cancer.Price A. State of the art radiotherapy for lung cancer.Thorax 2003; 58:447−452Concise discussion of current indications for radiation therapyand the methods of delivery.Schiller JH, Harrington D, Belani C, et al. Comparisonof four chemotherapy regimens for advanced nonsmall-celllung cancer. N Engl J Med 2002; 346:92−98This study compared four commonly used chemotherapydoublets and found that no combination provided superiorresults.Silvestri GA, Rivera MP. Targeted therapy for the treatmentof advanced non-small cell lung cancer. Chest2005; 128:3975−3984Cytotoxic chemotherapy has likely reached a plateau in efficacy.Fortunately, the study of molecular oncogenesis hassuggested targets for future drug development.Visbal AL, Leight NB, Feld R, et al. Adjuvant chemotherapyfor early-stage non-small cell lung cancer.Chest 2005; 128:2933−2943New addition to treatment guidelines.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 61ACC-PULM-09-0302-003.indd 617/10/09 8:04:27 PM


Targeted and Personalized TherapyBepler G, Begum M, Simon GR. Molecular analysisbasedtreatment strategies for non-small cell lung cancer.Cancer Control 2008; 15:130−139Hodkinson PS, MacKinnon A, Sethi T. Targetinggrowth factors in lung cancer. Chest 2008; 133:1209−1216Personalized treatment regimens and the use of targetedagents may prove to be a major advance in the treatment oflung cancer.Small Cell Lung CancerKalemkerian GP, Akerley W, Downey RJ, et al. Smallcell lung cancer. Clinical practice guidelines in oncology.J Natl Compr Canc Netw 2008; 6:294−314Clinical Practice Guideline from a well-respected organizationpresented in an algorithm type format.Sher T, Dy GK, Adjei AA. Small cell lung cancer. MayoClin Proc 2008; 83:355−367Discussion of the currently recommended treatment for limitedand extensive small cell lung cancer.62 Lung Cancer (Alberts)ACC-PULM-09-0302-003.indd 627/10/09 8:04:27 PM


Hypersensitivity PneumonitisW. Michael Alberts, MD, MBA, FCCPObjectives:• Discuss the features of dusts and particulate matter thatdictate possible clinical manifestations• Identify the most common causes and the usual clinicalpresentation of hypersensitivity pneumonitis• Describe those clinical syndromes that are similar tohypersensitivity pneumonitis but are clinically distinctentities• <strong>Review</strong> the most likely pathogenetic mechanisms and thecontinuing gaps in our understanding• Outline an approach to the diagnosis of the diseaseKey words: allergic alveolitis; farmer’s lung; inhalation fever;organic dust toxic syndrome; pigeon breeder’s diseaseInspired air may contain a wide variety of potentiallyharmful substances. These substances maybe in the form of aerosols, gases, vapors, fumes,or particulate matter. Inhaled particles may causerespiratory system dysfunction as the resultof an irritant, toxic, or hypersensitivity effect.Hypersensitivity pneumonitis (HP) is one type ofhypersensitivity reaction caused by the inhalationof particulate matter. The type of hypersensitivityreaction that develops in response to inhaled particlesis dependent on a number of factors, suchas the size of the particle, the concentration of theparticles, the duration of exposure, the frequencyof exposure, the nature of the substance, and individualsusceptibility.A major factor is the size of the particle. Inhaledparticles of 10 m in diameter are intercepted bythe nose and never reach the tracheobronchialtree. In the nose, they are cleared by the usualmechanisms or cause such problems as allergicrhinitis. Particles between 5 and 10 m impact onthe tracheobronchial mucosa. Here, they may becleared by the ciliary transport mechanism or causeextrinsic asthma. Particles between 0.3 and 5 mmay escape deposition and reach the alveolarspaces. Here, they may be cleared by pulmonarymacrophages or cause a hypersensitivity reaction.The hypersensitivity reaction that develops in thealveoli and terminal bronchioles is HP. Particles 0.3 m will normally act as vapors or fumes andare exhaled in the expired gas.DefinitionHP is an immunologic-induced, non-IgE-mediated inflammatory lung disease resultingfrom the sensitization and subsequent recurrentexposure to any of a wide variety of inhaledorganic dusts. The disease is a diffuse, predominantlymononuclear inflammation of the lungparenchyma, particularly the terminal bronchioles,interstitium, and alveoli. There is little, if any,involvement of the larger airways. The lymphocyticinflammation and monocyte accumulationoften organize into granulomas and may progressto fibrosis. The disease is caused by the inhalationof an organic dust in a sensitized individual.Recently, the definition has been broadened toinclude some diseases caused by inorganic chemicalsthat are capable of serving as haptens. Thesereactive chemicals join native airway proteins toform the complete antigen. HP is also known asextrinsic allergic alveolitis. Previously considereda rare disease, HP may be more common thanpreviously estimated.EtiologySince the original description of farmer’s lungby Campbell in 1932, 300 different occupationaland environmental sources of antigen exposurehave been identified. HP was once thought to beprimarily an occupational hazard. The disorder isnow known to be a recreational, avocational, andgeneral environmental hazard as well. The list ofoffending agents seems to be ever-expanding. Theassociated diseases often have been given colorfuland descriptive names. For example, bagassosis iscaused by exposure to moldy sugar cane, maplebark stripper’s disease is caused by exposure to afungus found under the bark (Cryptostroma corticale),paprika slicer’s lung is caused by exposureto Mucor stolinfer, and duck fever is caused by<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 63ACC-PULM-09-0302-004.indd 637/10/09 8:04:52 PM


exposure to duck feathers. The main etiologicagents are microbial spores, animal products (eg,avian-related antigens), and saprophytic fungi.Farmer’s lung is the best known of the HPsyndromes. Farmer’s lung is caused by the inhalationof one of the thermophilic actinomycetes, suchas Faenia rectivirgula (also known as Micropolysporafaeni or Saccharopolyspora rectivirgula) and Thermoactinomycesvulgaris. These ubiquitous unicellularorganisms usually are thought of as fungi but, inreality, are bacteria. These organisms grow exuberantlyin hay that has not been properly dried orstored. Farm workers are exposed when the storedhay is subsequently raked or turned over. Thermophilicactinomycetes also have been isolated fromthe soil, manure, and grain compost.The thermophilic actinomycetes also grow inair conditioners, home humidifiers, hot air furnaces,swimming pools, hot tubs, and fireplaceflues. When HP is acquired by these routes,the disease may be termed forced air disease ormore commonly humidifier lung. Common to allis the presence of warm stagnant water.Another common mode of acquiring HP is theinhalation of animal products (eg, serum proteins,urine proteins, dried excrement, feathers, eggwhites, and bloom). The best-known example ispigeon breeder’s disease (or bird fancier’s lung),but also gerbil keeper’s lung, turkey handler’slung, and pituitary snuff taker’s lung.HP may be caused by exposure to true fungi.Examples include malt worker’s lung caused byAspergillus clavatus; cheese worker’s lung causedby Penicillium caseii; and suberosis caused byPenicillium frequetans growing on moldy cork. Aunique form of disease recently has been describedin Japan, occurring in summer. Summer-type HPis associated with contamination of homes withTrichosporon asahii or Trichosporon mucoides.Highly reactive chemicals that are capable offorming hapten-protein complexes with airwayproteins have been shown to cause HP. HP associatedwith exposure to toluene diisocyanate, diphenylmethane diisocyanate, hexamethylenediisocyanate, and trimellitic anhydride have beendescribed.A number of ingested drugs have also beenshown to cause an “HP-like illness.” The best characterizedis probably amiodarone, but also gold,minocycline, procarbazine, hydrochlorothiazide,among others. In general, these reactions arenot considered HP as the inciting agent is administeredsystemically and the pathogenesis is likelydifferent.During the past several years, 100 cases ofHP have been reported in workers using waterbasedmetalworking fluids in the automotiveindustry. An HP-like syndrome may develop afterexposure to aerosolized Mycobacterium avium-intracellulare.Many exposures come from the use ofrecreational hot tubs, thus the common term hot tublung. Separating actual infection from an immunologicresponse to the organism often is difficult.EpidemiologyThe incidence of HP in exposed populationsvaries widely. A survey of 1,000 farmers in Wisconsinrevealed precipitins in 8.5%. Clinical manifestationsof the disease occurred in nearly one halfof the precipitin-positive group. In another study,the prevalence of farmer’s lung in three agriculturalareas of Scotland ranged from 2.3 to 8.6%.Studies of pigeon breeder clubs have found precipitinsin 40 to 50% of members and diseasemanifestations in from 6 to 21%. Reports of outbreaksof HP resulting from microbially contaminatedoffice buildings and industrial sites havereported attack rates of 15% and up to 70% of thoseexposed.Because the disease is present in only a smallminority of identically exposed individuals, hostfactors must be important. The reason for individualsensitivity is unknown. A number of studieshave confirmed that HP occurs more frequently innonsmokers than in smokers. The mechanism forthis effect is not known. Atopic patients are not atincreased risk. Concomitant viral infections withexposure may play a roleA related unanswered question is the frequencywith which undiagnosed HP presents as end-stageidiopathic pulmonary fibrosis (IPF). It has beenestimated that 10% of patients referred to NationalJewish in Denver with IPF had histologic featuressuggestive of HP.Related DisordersSeveral disorders often are confused with HP.Silo filler’s disease (not to be confused with64 Hypersensitivity Pneumonitis (Alberts)ACC-PULM-09-0302-004.indd 647/10/09 8:04:53 PM


farmer’s lung) is an acute toxic lung injury causedby the inhalation of nitrogen dioxide. Nitrogendioxide develops as a result of an interaction ofnitric oxide and freshly stored silage. This chemical,when inhaled, will affect the terminal airways ofall who are sufficiently exposed, regardless of previousexposure or sensitization. This is, in essence,a chemical burn manifesting as bronchitis, bronchiolitis,or acute respiratory distress syndrome.Organic dust toxic syndrome (ODTS) is aninflammatory pneumonitis that develops when amassive dose of organic dust is inhaled. This disorderwas originally termed pulmonary mycotoxicosisby Emanuel in 1975 to describe fever and coughin Wisconsin dairy farmers. The term ODTS hasbeen described as “clumsy and ugly.” As a result,some prefer to use the term inhalation fever todescribe this and related illnesses. Clinically, ODTSresembles HP with symptoms that are flu-like anddominated by fever and chills. ODTS, however,does not require previous sensitization, and mostof the time this disorder will develop in subjectswho are sufficiently exposed. The attack rate canbe 70% or greater, given sufficient exposure. Nolong-term effects have been documented, andcomplete recovery may be expected. In fact, theafflicted worker should be able to work with the dustagain, if the levels are kept low. This may be veryimportant in rendering employment advice.Bacterial endotoxins, mycotoxins, or sporesthat reach the terminal airspaces may provoke anintense inflammatory reaction. This reaction maybe caused by the direct activation of the alternatecomplement pathway and/or stimulation of theproduction of proinflammatory cytokines. ODTSis most commonly recognized in an agriculturalsetting. Exposure to dusts in swine-confinementand poultry-raising buildings and exposure tograin dust, resulting in grain fever, are commonforms of the disorder. The National Institute ofOccupational Safety and Health has estimated thatup to 30 to 40% of all heavily exposed workers,predominantly in the agricultural sector, mightexperience ODTS after organic dust inhalation.ODTS is distinct from HP and far more common.ODTS, or inhalation fever, also has beenreported with exposure to contaminated humidifierwater as found in industrial humidifiers, airconditioning units, central heating and humidificationsystems, and cool mist home vaporizers.Although HP (termed humidifi er lung) has beenacquired by this route, sufficient evidence hasshown that inhalation fever (termed humidifierfever) is a separate and distinct entity. Humidifierfever is an ill-defined feeling of malaise, fever,cough, and myalgia. These flu-like symptomsresolve within 24 h. Symptoms tend to be worse atthe beginning of the work week (much like byssinosis).There are no radiographic changes orlong-term effects. One theory of pathogenesis isthat recirculated water becomes contaminated byGram-negative organisms that produce endotoxin.Subsequent inhalation of the endotoxin triggersthe clinical syndrome.Metal fume fever is a nonspecific, self-limitedacute illness that resembles an attack of influenza.The syndrome is chiefly caused by exposure to zincoxide. The most common modern industrial activityassociated with zinc oxide fume inhalation iselectric arc welding on galvanized steel. Althoughthe data are sparse, metal fume fever also has beenreported to occur after exposure to copper, magnesium,antimony, and iron. The pathogenesis isbelieved to be a direct toxic effect rather than ahypersensitivity reaction. Polymer fume fever is asimilar syndrome resulting from exposure to combustionproducts of polymers and reactive chemicalsused in polymer production.Ornithosis (not to be confused with pigeonbreeder’s disease) is an infectious disease acquiredby inhaling contaminated droppings. The illnessis usually caused by Chlamydia psittaci. The diseaseusually resembles a flu-like illness but may be afulminant toxic syndrome.Clinical PresentationThe clinical picture of HP is very similar regardlessof the inciting agent. Although there is considerableoverlap and an individual patient may bedifficult to categorize, it is conceptually expedientto describe three forms of the disease, namelyacute, subacute, and chronic. Some authors list twoforms by combining the acute with the subacuteform. The chronic form, in turn, appears to be thefinal common pathway for either multiple episodesof acute HP or prolonged subacute disease.Acute HP is commonly caused by a briefbut-intermittentintense exposure. Abrupt onset ofsymptoms occurs 4 to 8 h after exposure. The<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 65ACC-PULM-09-0302-004.indd 657/10/09 8:04:53 PM


illness resembles the “24-h flu” with cough, dyspnea,tachypnea, diffuse end inspiratory rales,chills, fever, malaise, diaphoresis, headache, andmyalgias. The clinical picture is often mistaken fora viral or Mycoplasma pneumonia. Symptoms lastfor 12 to 24 h, with a peak at 8 to 12 h after exposure.Between attacks, the patient is usually asymptomatic.In most instances, the acute febrile episodeoccurs after each contact with the responsibleantigen.Subacute HP is likely attributable to a lessintense but more prolonged exposure. The diseaseis insidious, and symptoms resemble progressivechronic bronchitis with chronic cough, exertionaldyspnea, malaise, anorexia, fatigue, and weightloss. Symptoms are much less specific and do notreadily associate themselves with exposure toinhaled substances. This form of the disease presentsa diagnostic challenge to even the most astuteclinician. Nevertheless, subacute disease must berecognized because it may progress to the chronicform if avoidance of the responsible agent is notpracticed.Chronic HP may be the final result of inadequatelytreated or unrecognized acute or subacutedisease. Signs and symptoms of chronic HP arethose of any chronic interstitial pulmonary fibrosisin which irreversible lung damage has occurred.At this stage, the entire picture is indistinguishablefrom end-stage IPF of any cause.PathogenesisThe pathogenesis of HP is complex andincompletely understood. This is an area ofintense controversy and ongoing research. Thefinal story is not in at this time, but there is muchspeculation. In fact, several mechanisms may beinvolved. We do know that HP occurs in the settingof exposure to dust and particles of the correctsize and composition. Particles must beapproximately between 0.5 and 5 m. Particlesmust be capable of providing an antigenic stimulusfor not all particles of this size will cause thesyndrome. Particles must be relatively resistantto degradation so that the antigen can persist longenough in lung tissue to allow sensitization. Thepatient must have been sensitized during a previousexposure. The period of sensitization is variableand may be as short as several months ormay require a number of years of exposure. Notall of those exposed will become sensitized, andthe disease may not occur, upon being re-exposed,in all who do. Individual susceptibility is importantbut not well characterized.Current speculation suggests that the pathogenesismay involve both immune complexmediateddisease and cell-mediated immunity,perhaps in a sequential fashion. Immune complexmechanisms may initiate and mediate the acutesyndrome. There is little to suggest that type I ortype II immune mechanisms are involved.BAL studies in humans have provided someinsights into possible pathogenetic mechanisms.Fluid analysis 48 h after the last exposureis characterized by neutrophilia. Fluid analysis, 2 to 5 days after last exposure, may reveal T-celldifferential counts of as high as 70 to 80%. Thereappears to be a preferential expansion of CD8+cells (suppressor/cytotoxic T cells). HP has thereforebeen termed a T-suppressor cell alveolitis.Recent studies have suggested that proinflammatorycytokines and chemokines are involved inpathogenesis. These substances activate alveolarmacrophages, cause an influx of CD8+ lymphocytes,facilitate granuloma formation, and promotethe development of fibrosis. Interferon gamma,interleukin-10, tumor necrosis factor-, and transforminggrowth factor- have been implicated.In summary, it is clear that a type III or a typeIV reaction alone cannot explain all the observedphenomena, and there is little evidence for a typeI or type II reaction. As a result, a possible scenariois that an immune complex-mediated hypersensitivityreaction initiates acute lung injury andinduces the syndrome. This is followed byT-cell-mediated hypersensitivity mechanisms thatperpetuate the acute injury and induce the inflammation,granuloma formation, and interstitialfibrosis seen in subacute and chronic disease. Onceagain, this is an active area of ongoing research.The pathogenetic mechanisms discussed may ormay not prove to be correct.DiagnosisNo single clinical feature or laboratory testresult is diagnostic of the disease. The diagnosisshould therefore be made from a combination ofcharacteristic symptoms, physical findings,66 Hypersensitivity Pneumonitis (Alberts)ACC-PULM-09-0302-004.indd 667/10/09 8:04:53 PM


adiographic changes, pulmonary function testresults, and immunologic test results. Of primeimportance is a high index of suspicion. In patientswith recurrent bouts of an influenza-like illness orin patients with active interstitial lung disease, acomplete occupational, avocational, and environmentalhistory is vital.Laboratory studies are neither sensitive norspecific. In the setting of acute disease, a mild leukocytosis(25,000 mL) with a left shift is usuallyseen. Eosinophilia is variable. Increased erythrocytesedimentation rate, rheumatoid factor, andantinuclear antibody levels are common. A nonspecificpolyclonal increase in gamma globulinsoften is noted. IgE levels are normal.The chest radiograph is variable, and normalradiographs do not rule out the disease. In onestudy, 4% of acute cases of farmer’s lung had normalradiographs and another 40 to 45% had minimalchanges that might have otherwise beenoverlooked. In acute or subacute disease, a fine softpoorly defined reticulonodular pattern (oftendescribed as “alveolar mottling”) is noted. Themottling is less distinct than miliary tuberculosisand is widespread or predominantly located in thelower lung fields. Another common picture is scatteredpatchy interstitial infiltrates that maycoalesce. Segmental or lobar consolidation isunusual. Pleural effusions, thickening, and hilaradenopathy are rare. The degree of radiographicabnormalities correlates poorly with symptoms.The symptoms often are worse than the chestradiograph as opposed to Mycoplasma pneumoniain which the chest radiograph may look “worse”than the patient. The chest radiograph usuallyreturns to normal in several days to weeks after anacute episode.High-resolution CT (HRCT) may be helpful,although no single pathognomonic feature identifiesthe syndrome. Some believe that HRCT can beused to distinguish IPF from HP in most, but notall cases. Authors of one studied noted that desquamativeinterstitial pneumonitis looked likeacute or subacute HP. Chronic HP looked likeIPF. These investigators concluded that a lungbiopsy specimen was still the “gold standard” fordiagnosis.Characteristic findings in subacute HP onHRCT included the following: poorly defined centrilobularmicronodules, widespread ground-glassopacities, a mosaic attenuation on inspiratoryimages, and air trapping on expiratory CT images.A predominance of disease in upper and middlelung zones often is mentioned. A pattern of centrilobularground-glass nodules is fairly specific forthe diagnosis of HP with the appropriate clinicalhistory.In chronic stages, the chest radiograph is indistinguishablefrom diffuse interstitial fibrosis of anycause. A reticular pattern throughout the lungfields and/or honeycombing is common. Diffusevolume loss is evident, and features of right ventricularenlargement and pulmonary hypertensionmay dominate the film. In some cases, upper lungzones contract with upward retraction of the pulmonaryarteries. The final appearance may beindistinguishable from tuberculosis, sarcoidosis,ankylosing spondylitis, and eosinophilic granuloma,all of which are in the differential for upperlung zone fibrosis. The HRCT picture of chronicHP is that of reticulation superimposed on thefindings of subacute HP.<strong>Pulmonary</strong> function testing most commonlysuggests a restrictive ventilatory impairment,although obstructive and mixed patterns occasionallyare noted. In acute disease, the forced vitalcapacity is commonly approximately 80% of predicted.Decreased flow rates in concert with volumes(ie, a normal ratio) are usually noted. Thediffusing capacity is reduced (commonly about60% of predicted). Decreased lung compliance anda decreased Po 2with an increased alveolar-arterialoxygen pressure difference gradient are noted.<strong>Pulmonary</strong> function usually returns to normalbetween episodes of acute disease but may requireseveral weeks to completely recover. Progressiveand irreversible restrictive changes are found inthe chronic form of the disease.Skin testing is neither appropriate nor practicalin most patients. Few antigens are available and,where available, they often are nonspecificallyirritating, not well standardized, and/or impure.Where good antigens are available, an immediatewheal-and-flare reaction is observed in 80%. Thisreaction is not IgE mediated and is either nonspecificor IgG4 mediated. A classic Arthus skin reactionmay be noted 6 to 8 h after placement butdelayed skin reactions are rarely, if ever, seenPrecipitating antibodies were once believed tobe the sine qua non of the disease. They are now<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 67ACC-PULM-09-0302-004.indd 677/10/09 8:04:53 PM


more useful in establishing evidence of exposureto specific antigens. Serum precipitins are antigenspecificcomplement-fixing antibodies of the IgGclass, although IgM and IgA have also beendetected. Precipitating antibody to the offendingagent is found in 90 to 100% of patients with clinicaldisease. False-negative results are likely to becaused by methodologic problems or an inabilityto define the relevant antigen. Serum precipitinsmay disappear over time. Antibodies persist for atleast a year, and many are present at 3 years aftercessation of exposure. There are a number of commerciallyavailable farmer’s lung and bird fancierbatteries. These batteries are not for screeningpurposes nor are they useful in the absence ofa clinical suspicion of HP. Reference laboratoriesmay be able to prepare antigens for testing fromdust extracts or fungal cultures.Although the composition of BAL fluid varieswith the stage of the disease, certain findings havebeen suggested as being of clinical use. There is anincrease in lymphocytes in HP vs non-HP patients(65% vs 6%). Moreover, HP appears to be mainlycharacterized by a low CD4 + /CD8 + ratio. In contrast,a high CD4 + /CD8 + ratio frequently is foundin sarcoidosis. The presence of plasma cells in BALfluid samples is highly suggestive of HP. BAL hasalso been touted as useful in separating IPF fromHP. BAL fluid analysis in patients with IPF commonlyreveals a neutrophil predominance asopposed to the lymphocyte predominance in HP.Unfortunately, BAL findings do not distinguishdisease from exposure, CD4 + /CD8 + ratios varywidely (and are not stable over time), do not correlatewith disease activity, and do not serve asprognostic indicators. These latter findings wouldsuggest that BAL analysis may not fulfill its earlydiagnostic promise.A biopsy may not be necessary but often isuseful in establishing a diagnosis. Unfortunately,histopathology varies depending on the stage ofthe illness and is distinctive but not diagnostic.Lung tissue usually is obtained by open biopsyrather than transbronchial lung biopsy. Very fewbiopsies have been performed during the acuteform of the disease. Where available, they haveshown alveolar and interstitial accumulation ofpolymorphonuclear leukocytes, fluid, and macrophages.One of the most characteristic features isa centrilobular distribution of changes.More often, biopsies are conducted in the subacuteor chronic phases. At this stage of the disease,biopsy specimens show a predominant lymphocyte,plasma cell, and macrophage infiltration inthe alveolar wall and interstitium. Seventy percentshow noncaseating granulomas described as lesscompact and less well defined than those seen insarcoidosis. Sixty-five percent show varyingdegrees of fibrosis. Foam cells are seen in 65%.Foam cells are large foamy histiocytes representingactivated macrophages. Fifty percent of specimensshow bronchiolitis obliterans. Vasculitis andeosinophils are distinctly unusual. The typicalbiopsy specimen is best described by the findingof cellular bronchiolitis, a bronchiolocentric lymphocyticinterstitial pneumonitis, and poorlyformed noncaseating granulomas.Inhalation challenge tests are rarely performedbut may be available in specialized centers. Thesechallenge tests are not as well standardized, as isthe case for asthma. Each test must be individualized.The test is not to be taken lightly, as the reactionmay be quite severe. Once a reaction isestablished, bronchodilators are of little benefit.Corticosteroids must be used. Outside of researchsettings, it would be difficult to recommend testing.In general, after a simulated exposure, the patientis observed for 12 to 24 h. Reproduction of theclinical syndrome is a positive test. Some or all ofthe following parameters are monitored: temperature,WBC count, lung volumes, physical examination,carbon monoxide diffusing capacity, arterialblood gases, and spirometry.In summary, the diagnosis of HP may be challenging.A high degree of suspicion bolstered by acompatible history may permit a presumptivediagnosis. Removal of the suspected antigen sourceor avoidance by the patient followed by resolutionof symptoms is very strong circumstantial evidence.The authors of several recent articles haveemphasized the difficulties encountered in makingthe diagnosis.In 1997, Schuyler and Cormier suggested sixmajor and three minor criteria for the diagnosis. Adiagnosis of HP may be confirmed if four of themajor and two of the minor criteria are present.Major criteria are as follows: (1) symptoms compatiblewith HP, (2) evidence of exposure to appropriateantigen by history or detection of antibody inserum and/or BAL fluid, (3) findings compatible68 Hypersensitivity Pneumonitis (Alberts)ACC-PULM-09-0302-004.indd 687/10/09 8:04:54 PM


with HP on chest radiograph or HRCT, (4) BALfluid lymphocytosis, (5) histologic changes compatiblewith HP, and (6) positive “natural challenge”(ie, reproduction of symptoms and laboratoryabnormalities after exposure to the suspectedenvironment). Minor criteria are (1) bibasilar rales,(2) decreased diffusing capacity, and (3) arterialhypoxemia.In 2003, the Hypersensitivity PneumonitisStudy Group attempted to develop a clinical predictionrule for this diagnosis. Six significant predictorswere identified: (1) exposure to a knownoffending antigen, (2) positive precipitating antibodiesto the offending antigen, (3) recurrent episodesof symptoms, (4) inspiratory crackles onphysical examination, (5) symptoms occurring 4 to8 h after exposure, and (6) weight loss.TreatmentThe key is prevention and avoidance of reexposureis the top priority. This may require a jobchange, discarding a hobby, or moving to a newhome. These changes may not be well received bythe patient. In one study, 75% of pigeon breederswith disease were still breeding (pigeons, that is)at a 10-year follow-up. Changes in industrial procedureshave proven beneficial. For example, maplebark stripper’s disease has all but vanished becauseof the adoption of alternate work procedures. Theact of spraying bagasse with dilute propionic acidhas decreased the incidence of bagassosis. Improvementsin ventilation, air filtering systems, or masksmay help in some circumstances.For symptomatic relief of the inflammatorypneumonitis, antipyretics and supplemental oxygenmay be needed. Bronchodilators generally arenot beneficial. Corticosteroids, such as prednisoneat 0.5 to 1 mg/kg/d for 1 to 4 weeks, may decreasetoxicity of the acute form but probably offer nolong-term advantage. Recurrence of acute farmer’slung, however, may be more common in patientstreated with corticosteroids suggesting that thistreatment may also suppress the counterregulatoryaspects of the immune response in these patients.Longer periods of treatment may be necessary forthe subacute form (up to 3 to 6 months), but onceagain, these are of variable and questionable benefit.Once the chronic phase has developed, anytreatment is of marginal benefit.PrognosisPrognosis is incompletely understood. In theacute form, recovery is usually complete. Symptomshave usually disappeared in 12 to 48 h, butfatigue, lassitude, and exertional dyspnea may lastfor several weeks. The patient usually is asymptomaticbetween attacks. In the subacute form,recognition is crucial. With avoidance, signs andsymptoms may completely disappear. In one study,30 to 60% of patients with farmer’s lung who continuedto be exposed were disabled in 5 years and10 to 15% were dead. However, this also meansthat 40 to 70% did not suffer clinical deteriorationdespite continued exposure. A study in pigeonbreeders in Mexico demonstrated a 5-year mortalityof 30%. Some patients with HP, for example,pigeon breeders, may experience progression ofthe disease despite avoidance of repeated exposure.In the chronic form, progressive functionaldeterioration to respiratory insufficiency is likely,but progression to end-stage disease is not a uniformfinding.Building-Related Lung DisordersAmericans generally spend 90% of their timeindoors, and more than half of the workforce nowworks primarily in offices or commercial buildings.Therefore, it is not surprising that the quality ofindoor air has become an important public concern.A number of well-publicized outbreaks of workrelatedillness acquired from the indoor workenvironment have heightened the public awarenessof this issue.Unfortunately, the popular media has begunto lump all signs and symptoms related to theindoor working environment under the catchy butsomewhat inflammatory term, sick building syndrome.There are, however, a number of clinicalsyndromes associated with indoor working environmentthat are distinctly different in terms ofdiagnosis, prognosis, and management. The sickbuilding syndrome is but one of these syndromes.A more inclusive term, building-related illness, betterdescribes the entire spectrum of disorders acquiredin the indoor working environment.The building-related illnesses may be furtherclassified or into the following: (1) specificdisorders and (2) nonspecific disorders (ie, sick<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 69ACC-PULM-09-0302-004.indd 697/10/09 8:04:54 PM


uilding syndrome). Specific disorders are characterizedby well-defined signs, symptoms, andlaboratory findings. A specific cause can be establishedand a long recovery time is the norm.Examples would include HP, humidifier fever,asthma, and various respiratory tract infections.Conversely, nonspecific building-related disordersare characterized by nonspecific symptoms andvariable signs. There are no characteristic laboratoryfindings, and no single cause can be identified.Symptoms appear soon after arriving at work,worsen during the day, and dramatically disappearon leaving the building.Worldwide, it has been estimated that in 25%of the investigations of apparent outbreaks ofbuilding-related illness, a specific cause can beidentified, such as microbial contamination ofhumidification systems or the accumulationof motor vehicle exhausts. The remaining 75% ofoutbreaks are unexplained and are considered tobe due to the sick building syndrome. In a reviewof 356 cases of building-related illness in the UnitedStates, the National Institute for OccupationalSafety and Health found that 39% involved identifiablecontaminants and 50% were associated withinadequate provision of fresh air with no identifiablecontaminants. In 11%, no cause was found.Building-Related Illness: Specific DisordersThese disorders are conditions with a uniformclinical picture attributable to a specific identifiablecause. The symptoms of specific building-relateddisorders are due to identifiable allergens, toxins,or infectious agents that may be identified byappropriate laboratory tests or identification of thesource in the building. In turn, the treatment ofthese disorders hinges on removing the sourcerather than merely altering building ventilation.The etiology of the specific disorders may becategorized into those caused by the following:(1) antigens and organic dusts, (2) infectiousagents, and (3) toxins and irritants.Building-Related Illness: Sick BuildingSyndromeIn the 1970s, a remarkably consistent pattern ofcomplaints from office workers began to surface: dryeyes, dry skin, stuffy nose, fatigue, and headache.This was the time of the energy crisis, and newlyconstructed or remodeled buildings were beingdesigned with energy efficiency in mind. Most ofthese buildings had no windows that opened, andoutside air was provided only through a recirculatingair-conditioning system. Moreover, ventilationstandards had been lowered from 20 cubic feet ofoutdoor air per minute per occupant to 5 cubic feetper minute per occupant. This lower standard isbelow the circulation rate of 15 cubic feet per minuteper occupant, which is necessary to ensure a CO 2level of 1,000 parts per million. A CO 2level of 1,000 parts per million ppm generally is believedto indicate an adequate fresh air supply. As a result,the building-related symptoms were assumed to beattributable to lower rates of ventilation that resultedin inadequate dilution of irritants. The eye, nose, andthroat irritation, headache, and difficulty concentratingthat occurred in workers in these buildings cameto be called the sick building syndrome.In the 1990s, similar complaints were heardfrom occupants of many buildings, regardless ofdesign, age, or geography. The World Health Organizationestimated that 30% of newly constructedor remodeled buildings are associated with healthand discomfort problems and that between 10 and30% of the occupants of these buildings maydevelop symptoms of sick building syndrome.The World Health Organization has definedthe sick building syndrome as follows: an excessof work-related irritations of the skin and mucousmembranes and other symptoms, including headache,fatigue, and difficulty concentrating, reportedby workers in modern office buildings. Becausemost of these symptoms are commonly experiencedby the general population, an outbreak ofsick building syndrome is defined by an excessivereporting of one or more symptoms by the occupantsof a building. Excessive is defined as inexcess of what would ordinarily be expected as the“background” level of such complaints. It has beenestimated that the background rate of similar vaguesymptoms in any population of people is 10 to 20%.In addition, symptoms must be work-related.Symptoms must develop after coming to work,worsen during work, and disappear on leavingwork. Finally, there should be no evidence ofapparent nonoccupational cause (such as apreexisting medical condition) nor an obviousexposure to occupational toxic materials.70 Hypersensitivity Pneumonitis (Alberts)ACC-PULM-09-0302-004.indd 707/10/09 8:04:54 PM


Core symptoms of sick building syndrome areremarkably consistent and consist of lethargy;mucous membrane irritation (dry throat, stuffynose); headache; eye symptoms (itchy, irritated,watery); and dry skin. These symptoms are not theresult of tissue damage detectable by physicalexamination or laboratory tests. The most commonlyreported symptoms are mucous membraneirritation (46%), headache (43%), and lethargy(57%). These symptoms, although not threatening,can be very unpleasant, disruptive to work andhome life, cause lost work time and decreasedproductivity, and create concern about more serioushealth problems.Annotated Bibliography<strong>Review</strong>sIsmail T, McSherry C, Boyd G. Extrinsic allergic alveolitis.Respirology 2006; 11:262–268Kurup VP, Zacharisen, Fink JN. Hypersensitiviy pneumonitis.Ind J Chest Dis Allied Sci 2006; 48:115–128Madison JM. Hypersensitivity pneumonitis. Clinicalperspectives. Arch Pathol Lab Med 2008; 312:195–198Short but comprehensive review.Salvaggio JE. Hypersensitivity pneumonitis. J AllergyClin Immunol 1987; 79:558–571“Classic” review of the subject.DiagnosisGlazer CS, Rose CS, Lynch DA. Clinical and radiologicmanifestations of hypersensitivity pneumonitis.,J Thorac Imaging 2002; 17:261–272Krasnick J, Meuwissen HJ, Nakao MA, et al. Hypersensitivitypneumonitis: problems in diagnosis. J AllergyClin Immunol 1996; 97:1027–1030Through a discussion of a challenging case, the authorsdescribe the difficulties encountered in the diagnosis of HP.Lacasse Y, Selman M, Costabel U, et al. Clinical diagnosisof hypersensitivity pneumonitis. Am J Respir CritCare Med 2003; 168:952–958Members of the Hypersensitivity Pneumonitis StudyGroup attempt to develop a clinical prediction rule fordiagnosis.Schuyler M, Cormier Y. The diagnosis of hypersensitivitypneumonitis. Chest 1997; 111:534–536Since the clinical and laboratory findings of HP overlap withthose of many other pulmonary disease, the diagnosis canbe difficult. The authors of the editorial suggest major andminor criteria that should be present in a patient with thisdisorder.Silva CI, Churg A, Muller NL. Hypersensitivity pneumonitis:spectrum of high-resolution CT and pathologicfindings. Am J Roentgenol 2007; 188:334–344There are several high-resolution CT findings that are typicalof HP.PathologyBarrios RJ. Hypersensitivity pneumonitis. Histopathology.Arch Pathol Lab Med 2008; 132:199–203Churg A, Muller NL, Flint J, et al. Chronic hypersensitivitypneumonitis. Am J Surg Pathol 2006; 30:201–208Takemura T, Akashi T, Ohtani, et al. Pathology ofhypersensitivity pneumonitis. Curr Opin Pulm Med2008; 14:440–454.Discussion of the pathologic features of this disease.MechanismMcSharry C, Anderson K, Bourke SJ. Takes your breathaway—the immunology of allergic alveolitis. Clin ExpImmunol 2002; 128:3–9Useful discussion of the current thinking on pathogenesis.Woda BA. Hypersensitivity pneumonitis. An immunopathologyreview. Arch Pathol Lab Med 2008;132:204–205Recent thoughts on the still incompletely understood pathogenesis.Inhalation Fever (ODTS)Linaker C, Smedley J. Respiratory illness in agriculturalworker. Occup Med 2002; 52:451–459Inhalation fever (ODTS) may be more common than HP.These reviews highlight the distinct nature of these disorders.Seifert SA, Von Essen AS, Jacobitz K, et al. Organic dusttoxic syndrome: a review. J Toxicol Clin Toxicol 2003;41:185–193Building-Related IllnessAlberts WM. Building-related illness. J Respir Dis 1994;15:899–910Brooks SM, Spaul W, McCluskey JD. The spectrumof building-related airways disorders. Chest 2005;128:1720–1727<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 71ACC-PULM-09-0302-004.indd 717/10/09 8:04:54 PM


Burge PS. Sick building syndrome. Occup EnvironMed 2004; 61:185–190Jones TF, Craig AS, Hoy D, et al. Mass psychogenicillness attributed to toxic exposure at a high school.N Engl J Med 2000;342:96–100 (editorial: Wessely S.Responding to mass psychogenic illness. N Engl J Med2000; 342:129–130)Report and discussion of a nonspecific building-related disorder.Excellent discussion of physiologic symptoms in apsychogenic disorder.Menzies D, Bourbeau J. Building-related illnesses.N Engl J Med 1997; 337:1524–1530All four are short reviews of the entire spectrum of buildingrelatedillnesses, including the sick building syndrome. Thefirst is a classic.72 Hypersensitivity Pneumonitis (Alberts)ACC-PULM-09-0302-004.indd 727/10/09 8:04:54 PM


Eosinophilic Lung DiseasesW. Michael Alberts, MD, MBA, FCCPObjectives:• Highlight the heterogeneous nature of these disorders andsuggest a usable classification system• Discuss important aspects of the eosinophil• <strong>Review</strong> diseases that involve the eosinophil and theairways• Discuss known parenchymal lung disorders that are associatedwith peripheral and/or tissue eosinophilia• Highlight the idiopathic eosinophilic lung diseasesKey words: acute eosinophilic pneumonia; allergic bronchopulmonaryaspergillosis; chronic eosinophilic pneumonia;eosinophil; Loeffler pneumoniaThe eosinophilic lung diseases are a heterogeneousgroup of clinical entities in which there isan increased number of eosinophils in the airwaysand/or lung parenchyma. These disorders mayor may not be accompanied by peripheral eosinophilia.The presence of eosinophils does not establisha cause-and-effect relationship. In some cases,the eosinophil is merely a part of the inflammatoryprocess and may even be present to protect hosttissues. In other cases, the eosinophil appears to bedirectly responsible for the tissue damage.Classification of the EosinophilicLung SyndromesThere have been many attempts to create aclinically useful classification system. In the 1950s,the first classification system was based on Crofton’sfive syndromes, marked by peripheral bloodeosinophilia and pulmonary infiltrates. The fivecategories were as follows: (1) simple pulmonaryeosinophilia, (2) prolonged pulmonary eosinophilia,(3) pulmonary eosinophilia associated withasthma, (4) tropical eosinophilia, and (5) pulmonaryeosinophilia with polyarteritis nodosa. Theseare the so-called PIE syndromes (pulmonary infiltrateswith eosinophilia) as coined by Reeder andGoodrich.In the 1960s, a new group of diseaseswas included. These disorders had increasedparenchymal eosinophils but did not necessarilyhave peripheral blood eosinophilia. A lung biopsyspecimen was needed to identify these disorders.Chronic eosinophilic pneumonia, as reported byCarrington, is an example. In the 1980s, additionaldiseases were added to this heterogeneous collectionof syndromes by virtue of their associationwith an increased percentage of eosinophils inBAL specimens; “acute eosinophilic pneumonia”is an example.Currently, there is no universally accepted oroptimal way to classify these disorders. In theirstate-of-the-art review, Allen and Davis state thatan improved understanding of the role of cytokinesand the other factors that control eosinophil trafficin the lung may ultimately permit a scientificallyplausible classification of these disorders. One suggestedclassification scheme is listed in Table 1 andwill be used in this review.Table 1. Classification SchemeAirway disordersAsthmaEosinophilic bronchitisAllergic bronchopulmonary aspergillosisBronchocentric granulomatosisInterstitial disordersSecondary (associated with known underlying diseaseprocesses)Bacterial infections (eg, brucellosis, mycobacterial)Fungal infections (eg, Coccidiomycosis and Aspergillus)Interstitial lung diseasesIdiopathic pulmonary fibrosisSarcoidosisSystemic lupus erythematosusEosinophilic granulomaHypereosinophilic syndromeAIDS-associated Pneumocystis carinii pneumoniaParasitic infections<strong>Pulmonary</strong> vasculitisHodgkin diseaseDrug reactionsLung cancerOthersPrimary (idiopathic eosinophilic pneumonia)Simple pulmonary eosinophiliaChronic eosinophilic pneumoniaAcute pulmonary pneumonia<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 73ACC-PULM-09-0302-005.indd 737/10/09 8:05:17 PM


The EosinophilThe eosinophil is a polymorphonuclear leukocytethat is produced in the bone marrow. Theindividual cell is 12 to 15 m in diameter. Thenucleus has two lobes. There is abundant endoplasmicreticulum and an active Golgi apparatus.Eosinophils contain three different types of granules:two smaller granule populations and onelarger eosin-specific granule population. There areapproximately 200 of these larger granules per cell.The larger granules are responsible for the characteristicred staining. The eosin-specific granule hasa very characteristic electron microscopic appearance.The internal core of the eosin-specific granuleis comprised primarily of major basic protein. Majorbasic protein is an arginine-rich, highly cationicpolypeptide that is highly toxic to parasites, tumorcells, and respiratory epithelial cells. The cell containsother cationic proteins, such as eosinophilcationic protein, eosinophil peroxidase, eosinophilderivedneurotoxin, and several collagenases.Another unique constituent of this cell is the membraneassociated Charcot-Leyden crystal protein.Differentiation of the eosinophil in the bonemarrow is under the control of interleukin (IL)-3,IL-5, and granulocyte macrophage colonystimulatingfactor. There are two populations ofcirculating eosinophils that differ in their biochemicalmakeup and their functional activities.They are termed normodense cells and hypodensecells. An increased proportion of hypodense cellscan be found both in the tissues and blood ineosinophilic disorders. These cells are more metabolicallyactive and may represent “activated” cells.The eosinophil is primarily a tissue cell. Aftera brief time in the peripheral circulation (usually13 to 18 h), the cell moves into tissues and organsthroughout the body. There are 100 to 400 times asmany eosinophils in the tissues as compared withthe circulation at any one time. The eosinophiltends to settle in submucosal areas of organsexposed to the environment such as the lungs, theGI tract, and the genitourinary tract. The movementof eosinophils into the tissues is controlledby many factors. Known chemotactic factorsinclude complement components, histamine,eosinophil chemotactic factor of anaphylaxis, platelet-activatingfactor, leukotrienes, lymphokines,tumor-associated factors, and IL-5.There are several teleological theories of eosinophilfunction: host defense against parasites,modulator of inflammation, and tissue-destructivecell. The host defense theory builds on the wellestablishedclinical observation that the tissueinvasive stages of parasitic infection are associatedwith a striking peripheral eosinophilia. In vitrostudies have shown that eosinophils can degranulateinto and kill parasites. The modulator ofinflammation theory holds that the eosinophilprotects host tissues by dampening the damagingeffects of inflammation. Eosinophils contain a varietyof substances that have been shown to dampenor modulate the effect of various inflammatorymediators. Eosinophils tend to encircle areas ofacute and chronic inflammation, perhaps representingan attempt to contain inflammation. The tissuedestructivecell theory holds that the eosinophil isactivated at sites of inflammation and is directlyresponsible for the damage of host tissues.Eosinophilic Lung DiseasesAirway DisordersAsthma: Peripheral eosinophilia is presentin many asthmatics and is independent of thetype of asthma (intrinsic vs extrinsic). Eosinophilcounts may in fact be used to follow the courseof the disease in a given patient. In many cases,there is a quantitative linear relationship betweenthe eosinophil count and the degree of expiratoryairflow limitation.Eosinophilic Bronchitis (Without Asthma):Eosinophilic bronchitis with a high percentageof eosinophils (approximately 40%) in sputumis a well-recognized cause of chronic cough.This entity is responsive to corticosteroid treatment.In contrast to patients with asthma, thosewith nonasthmatic eosinophilic bronchitis donot have variable airflow limitation or bronchialhyperresponsiveness.Fungus-Induced Asthmatic Reactions: Inhalationof fungal spores by the asthmatic may result inseveral types of reactions: (1) IgE-mediated allergicrhinitis and asthma and (2) allergic bronchopulmonarymycosis. Fungal IgE-mediatedasthma is a noninfectious disease resulting fromthe immune response of the atopic host. Theresponse to inhaled fungal spores or mycelial74 Eosinophilic Lung Diseases (Alberts)ACC-PULM-09-0302-005.indd 747/10/09 8:05:18 PM


antigens resembles the response to other inhalantallergens, such as house dust mite. Allergicbronchopulmonary mycosis is an “infectiousdisease” characterized by periods of persistentfungal growth or colonization of the respiratorytract. Both types of disorders can cause longstandingasthmatic manifestations in susceptibleindividuals.Allergic Bronchopulmonary Aspergillosis: Allergicbronchopulmonary aspergillosis (ABPA)is the most common type of allergic bronchopulmonarymycosis. ABPA is a complication ofallergic asthma in which the ubiquitous fungus,Aspergillus fumigatus, colonizes the lowerrespiratory tract. Patients with cystic fibrosis areparticularly vulnerable. ABPA may be found inup to 15% of patients with cystic fibrosis and 2to 28% of asthmatics. An ABPA-like disordermay be caused by other fungi such as Candidaalbicans, Helminthosporium species, and Curvularialunata. Each of these organisms is capable ofhyphal growth at body temperature. The hyphaepersist in the airways and continually releaseantigens. The antigen combines with specific IgGand IgE to cause actual tissue damage. This tissuedamage may result in permanent damage,as evidenced by proximal bronchiectasis andirreversible airways obstruction.Bronchocentric Granulomatosis: Bronchocentricgranulomatosis (BG) is a rare granulomatous disordercharacterized by granuloma formation andnecrosis centered on and limited to bronchi andbronchioles. Vasculitis is not a major component.BG may be a nonspecific response to a number ofsubstances. A surgical biopsy specimen is requiredfor diagnosis. The disease is defined by morphologiccriteria and is not a clearly defined clinicalsyndrome. It has been debated whether BG isa separate disease entity or purely a pathologicdescription of one of the limited ways in whichbronchi and bronchioles respond to injury.Parenchymal Disorders (Associated With OtherKnown Disease Entities)Interstitial Lung Diseases: Increased BAL eosinophils( 5%) may be present in approximately10 to 20% of interstitial lung diseases. Commoninterstitial diseases associated with increasedBAL eosinophils include idiopathic pulmonaryfibrosis, sarcoidosis, systemic lupus erythematosus,and eosinophilic granuloma. The pathogeneticimportance of eosinophils in these disordersis unknown. The presence of BAL eosinophiliain idiopathic pulmonary fibrosis may correlatewith clinical deterioration and may predict a poorresponse to therapy.Drug Reactions: A number of drugs have beenassociated with the PIE syndrome. Most havebeen reported as isolated cases, such as nitrofurantoin,sulfasalazine, phenytoin, bleomycin, andtetracycline. The reactions are generally mild and,therefore, little is known about the pathologiccondition of the pulmonary infiltrates or evenwhether the reaction involves eosinophilic lunginflammation. Most patients with drug-inducedeosinophilic lung disease will improve by simplydiscontinuing the medication. In severe or persistentcases, corticosteroids have been used withsome success.<strong>Pulmonary</strong> Vasculitis: Eosinophils may be associatedwith lung lesions that accompany pulmonaryvasculitis syndromes. <strong>Pulmonary</strong> vascularinflammation is found most frequently as a manifestationof primary systemic vasculitis but alsooccurs in association with a number of conditions,including rheumatologic disorders (eg, systemiclupus erythematosus and polymyositis); chronicinfection; lymphoma; sarcoidosis; and extrinsicallergic alveolitis. Primary vasculitic processesaffecting the lung include giant cell arteritis, pulmonarycapillaritis, Takayasu arteritis, and thoseassociated with circulating antibodies to neutrophilcytoplasmic enzymes (eg, Churg-Strausssyndrome, Wegener granulomatosis, and microscopicpolyangiitis).The Churg-Strauss syndrome, also known asallergic granulomatosis and angiitis, is a rare butdistinctive disorder. Patients with the Churg-Strauss syndrome generally have long-standingestablished asthma and a dramatic peripheraleosinophilia. Manifestations of the disease appearto be attributable to a granulomatous inflammatoryresponse that results in vascular necrosis, primarilyinvolving the lungs. An open lung biopsy specimenusually is necessary for diagnosis. In the past, thepathophysiology of this disorder was presumed tobe caused by immune complex deposition in thewalls of small- and medium-sized arteries andveins. More recent speculation involves the role of<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 75ACC-PULM-09-0302-005.indd 757/10/09 8:05:18 PM


anti-neutrophilic cytoplasmic antibody (ANCA) andthe characteristic vasculitis. The association withasthma and eosinophilia is unexplained.Clinical features include asthma, a history ofatopic disease, fever, malaise, and weight loss.Systemic manifestations of the illness may includeupper airway involvement (sinusitis, rhinitis, nasalpolyps); skin changes (nodules, purpura, urticaria);arthralgias; myalgias, mononeuritis multiplex;abdominal symptoms (pain, diarrhea,bleeding); cardiac findings (heart failure, pericarditis,hypertension); and microscopic hematuria.Chest radiographs commonly reveal patchy andtransient infiltrates, but large and small nodulesalso have been reported. Thin-section CT findingsinclude bilateral subpleural consolidation withlobular distribution, centrilobular nodules (especiallywithin the ground-glass opacity), or multiplenodules, especially in association with bronchialwall thickening. Pleural effusions may be noted inone-third of cases. The effusions are exudative andmay contain a significant number of eosinophils.Common laboratory abnormalities include leukocytosiswith marked eosinophilia, a high percentageof BAL eosinophils, prolonged erythrocytesedimentation rate, and anemia. The IgE level isoften markedly increased and appears to correlatewith disease activity. In Churg-Strauss syndrome,the prevalence of a positive serum ANCA resultranges from 44 to 66%. The p-ANCA pattern isoften present as opposed to the c-ANCA observedin Wegener granulomatosis.Clinically, there are three distinct phases: (1) aprodromal phase that may persist for many years,consisting of asthma, often preceded by allergicrhinitis; (2) a second phase of marked peripheralblood eosinophilia and eosinophilic tissue infiltratesresembling Loeffler syndrome, or chroniceosinophilic pneumonia, which may recur duringa period of years; and (3) a third, life-threateningvasculitic phase.Survival is dramatically enhanced with treatment.In the past, without treatment, 50% ofpatients died within 3 months of onset. Myocardialinvolvement was the most frequent cause of death.In patients treated with corticosteroids, a meansurvival of 9 years has been reported. Corticosteroidsalone usually are effective in the treatmentof Churg-Strauss syndrome. The addition of oralcyclophosphamide may reduce the rate of relapsebut has not been shown to improve survival. Inpatients who fail to respond to corticosteroidtherapy, however, “pulse” methylprednisolone,azathioprine, or cyclophosphamide may be effective.It is important to separate the Churg-Strausssyndrome from other necrotizing vasculitides, suchas Wegener granulomatosis and polyarteritisnodosa, which may require treatment withcytotoxic agents.The association of leukotriene antagonists withthe Churg-Strauss syndrome originally describedin patients taking zafirlukast has now been recognizedto occur with montelukast as well and istherefore most likely an uncommon but documentedclass association with an incidence ofapproximately 1 in 20,000 patients. The pathophysiologyis unknown, but these patients mayhave had a primary eosinophilic infiltrative disorderthat had been clinically recognized as asthma,was quelled by steroid treatment, and wasunmasked after corticosteroid withdrawalfacilitated by the initiation of the leukotrieneinhibitor.Parasitic Disease: PIE syndrome may developin conjunction with a number of parasitic infections.In the United States, the most commoninfections are caused by Strongyloides, Ascaris,Toxocara, and Ancylostoma. The syndrome mayarise because of the presence of the parasite in thelung at certain stages of its life cycle. At this point,eosinophils may be recruited to the lung to killthe parasite. Eosinophils have been shown to bepresent in the lung, however, when parasites arenot demonstrable, which suggests that immunologicmechanisms may be involved. GI symptomsusually dominate the clinical picture of parasiticinfestation. Respiratory symptoms of cough andwheezing are usually mild. The lung diseasecommonly resolves with therapy directed at thespecific parasite. Corticosteroids usually are notneeded.The most serious and best-characterized parasiticeosinophilic lung disease is tropical pulmonaryeosinophilia. This disorder is caused by the filarialworms Wuchereria bancrofti and Brugia malayi.These organisms are found primarily in India,Africa, South America, and Southeast Asia. Microfilariaereleased from adult worms cause an intenseinflammatory reaction in the lung. This reactioncommonly causes nocturnal cough, dyspnea,76 Eosinophilic Lung Diseases (Alberts)ACC-PULM-09-0302-005.indd 767/10/09 8:05:18 PM


wheezing, fever, weight loss, and malaise. A historyof residence in a filarial endemic region and afinding of peripheral eosinophilia 3,000/mm 3should initiate a consideration of this disease.Cases of tropical pulmonary eosinophilia typicallyhave been reported to masquerade as acute orrefractory asthma. The recommended treatment isdiethylcarbamazine. If left untreated or treated late,the disease may lead to long-term sequelae of pulmonaryfibrosis or chronic bronchitis with chronicrespiratory failure.Idiopathic Hypereosinophilic Syndrome: Idiopathichypereosinophilic syndrome is a rareillness of unknown etiology. This disorder affectsmultiple organ systems, primarily as the result ofthe infiltration of mature eosinophils. This disorderis likely related to unchecked T-cell secretionof IL-3, IL-5, or granulocyte macrophage colonystimulatingfactor. Specific diagnostic criteria forthe hypereosinophilic syndrome have been established:(1) peripheral eosinophilia ( 1,500 cellsper microliter) for 6 months; (2) involvement ofvarious organ systems with evidence of end organdamage; and (3) no evidence of parasitic, allergic,vasculitic, or other known causes of eosinophilia.The illness may be mild or fatal. The majorcause of morbidity and mortality is cardiac diseasewhere endocardial fibrosis; restrictive cardiomyopathy;valvular damage (supportivestructures around the valves, especially the mitralvalve, are prone to fibrosis); and mural thrombusformation occur. Lung involvement occurs in 40%of cases. Symptoms are nonproductive cough anddyspnea. The chest radiograph may show pulmonaryedema and pleural effusions associated withcardiac dysfunction but may also reveal interstitialinfiltrates presumably due to perivasculareosinophilic infiltration or fibrosis. Patients usuallyare treated with oral corticosteroids, but onlyabout 50% will have a good clinical response.Other drugs, such as busulfan, interferon alfa,and hydroxyurea, may be used in steroid-unresponsivepatients. Advances in molecular diagnosticshave enabled the identification of subtypesof the hypereosinophilic syndrome that mayrespond to imatinib, a small molecule tyrosinekinase inhibitor. Additionally, studies havereported success in treatment with the use ofmepolizumab, which is a monoclonal antibodyagainst IL-5.Miscellaneous: Bronchogenic carcinoma occasionallyis associated with lung and peripheraleosinophilia. Eosinophils have been shown toinvade the tumor, which suggests that eosinophilsmay be involved in host defense againsttumors. Hodgkin disease may be associated withperipheral, BAL, and lung eosinophilia. Fungaldisease may be associated with peripheral eosinophilia.Peripheral blood eosinophilia is noted inthe majority of cases of primary coccidiomycosis.<strong>Pulmonary</strong> eosinophilic infiltrates may be notedon biopsy specimen or BAL. Administration ofcorticosteroids to patients early in their infectioncan result in an acceleration of the infection withpossible fatal dissemination. In AIDS-associatedPneumocystis carinii pneumonia, 15% of patientshad BAL eosinophils 5%. A number of otherdiseases have been reported to be associatedwith pulmonary infiltrates and blood or alveolareosinophilia. They include bronchiolitis obliteransorganizing pneumonia, ulcerative colitis,mycobacterial infection, Sjögren syndrome, andpostradiation fibrosis.Parenchymal Disorders (Idiopathic)Simple <strong>Pulmonary</strong> Eosinophilia (Loeffler pneumonia):Simple pulmonary eosinophilia was originallydescribed by Loeffler in 1932. This disorderis characterized by migratory pulmonary infiltratesaccompanied by peripheral eosinophilia.Respiratory symptoms are minimal or absent.Malaise, fever, and cough may be noted. At times,the chest radiographic pattern may be almostdiagnostic, with transitory and migratory illdefinedperipheral, nonsegmental, and relativelyhomogeneous densities. By definition, the diseaseresolves within 4 weeks. Afflicted patients havean excellent prognosis. Complete resolution withor without treatment is the rule. In the originaldescription of this disorder, most of the patientslikely had a parasitic infection or a drug reaction.Currently, it is estimated that up to one-third ofcases do not have a clinically identifiable cause.This latter group is included in the idiopathiceosinophilic pneumonias. Apparent simple pulmonaryeosinophilia, however, should be viewedas a sign of possible underlying disease. A carefulsearch for parasitic infection or drug reactionshould be pursued.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 77ACC-PULM-09-0302-005.indd 777/10/09 8:05:19 PM


Chronic Eosinophilic Pneumonia: Althoughuncommon, this entity is the best characterizedof the idiopathic eosinophilic pneumonia syndromes.Both histologic and BAL studies stronglyimplicate the eosinophil in the pathogenesis ofthe disorder. A recent study found strikinglyincreased levels of IL-5, IL-6, and IL-10 in BALfluid recovered from involved lung segments. Nosuch increase was noted in serum or uninvolvedlung segments. Furthermore, the presence of circulatingimmune complexes, increased levels ofIgE, and a frequently positive rheumatoid factorsuggest an immunopathogenic mechanism.Chronic eosinophilic pneumonia (CEP) is aserious disease that requires specific treatment. Thedisease usually affects middle-aged atopic women,but it has been reported in both sexes and all ages.Onset of the disorder is insidious with progressiverespiratory and constitutional symptoms. In onestudy, symptoms have been present for an averageof 7.7 months before the correct diagnosis is made.The most common symptoms are cough, dyspnea,fever, night sweats, malaise, and weight loss.Asthma is present in 50 to 60% of patients and isusually of recent onset. Laboratory evaluation mayreveal increased IgE levels, and the levels may correspondto the clinical activity. Peripheral eosinophiliaoccurs in up to 88%.Diffuse peripherally based infiltrates in theouter two thirds of the lung fields are found in 63%of patients. On occasion, the chest radiograph maybe so characteristic as to be diagnostic. Bilateralperipheral or pleural-based dense infiltrates withoutsegmental or lobar distribution may be noted.These infiltrates may form a pattern described as“the photographic negative of pulmonary edema.”Unfortunately, 50% of patients demonstrate thisclassic plain radiographic picture. A CT scan willreveal peripheral infiltrates in all afflicted. Biopsyspecimens show interstitial and alveolar eosinophilsthat are degranulated. Common findingsinclude eosinophilic microabscesses (which areaggregates of necrotic eosinophils surrounded bya rim of palisading histiocytes), low-grade vasculitis,and interstitial fibrosis. BAL specimens usuallyshow eosinophilia that average 44%.Fewer than 10% of patients will have spontaneousresolution, and deaths from CEP have beenreported. The disease responds quickly anddramatically to corticosteroid therapy. The patientmay become asymptomatic in hours or a few days.The prognosis is excellent, but treatment for prolongedperiods is usually necessary, at least 6months. Steroid therapy must be tapered slowlyas the disease tends to relapse. If a relapse occurs,treatment must be continued for a prolongedperiod before trying to taper again.Acute Eosinophilic Pneumonia: In 1989, severalwell-documented cases of acute respiratoryfailure associated with increased BAL or tissueeosinophils were reported. The cause is unknownbut may be a unique hypersensitivity reaction toan inhaled antigen. Various drug toxicities havebeen reported to produce this syndrome. The followingdiagnostic criteria have been suggested:(1) acute febrile illness of 5 days in duration,(2) hypoxemic respiratory failure, (3) diffusemixed alveolar and interstitial chest radiographicinfiltrates, (4) BAL eosinophilia ( 25%), (5) noapparent infectious etiology, (6) rapid and completeresponse to corticosteroid therapy, and(7) no relapse after discontinuing corticosteroidtherapy. Recent studies have suggested that newcigarette smokers may be at increased risk.Patients typically present with an acute febrileillness accompanied by myalgias, pleuritic chestpain, and hypoxemic respiratory failure, oftenrequiring mechanical ventilation. Clinically, thereis little to distinguish acute eosinophilic pneumoniafrom an acute infectious process or ARDS. It isthe BAL that provides the clue to the diagnosis.Peripheral blood eosinophil percentage is usuallynormal, but a very high percentage of BAL eosinophilsis characteristic, with an average of 42% inone series. Small-to-moderate pleural effusions arefrequent. Fluid analysis may reveal a high percentageof eosinophils.Patients usually respond rapidly to high dosesof corticosteroids, usually within 24 to 48 h. Thedose is then tapered, but treatment is usually continuedfor 2 to 4 weeks. Most patients survive andrecover normal lung function. Acute eosinophilicpneumonia is a diagnosis of exclusion. An infectiousetiology should be pursued even after corticosteroidtherapy is begun. It is especially importantto look closely for disseminated fungal disease inall patients and P carinii in the HIV-positivepatient.78 Eosinophilic Lung Diseases (Alberts)ACC-PULM-09-0302-005.indd 787/10/09 8:05:19 PM


ConclusionIn these disorders, teleologically, the eosinophilmay be: (1) a “good guy,” (2) a “bad guy,” or (3) an“innocent bystander.” The eosinophil may be agood guy. The “host defense” theory builds on theclinical observation that the tissue invasive stagesof parasitic infections are associated with a strikingeosinophilia. In vitro studies have shown thateosinophils can kill parasites. Conversely, theeosinophil may be a bad guy. The eosinophil isactivated at sites of inflammation and thus may beresponsible for the actual tissue damage. Alternatively,the eosinophil may be an innocent bystander.The appearance of the cell may be part of thebody’s attempt to dampen or contain the effects ofinflammation—the so-called modulator of inflammationtheory.Whatever the function of the eosinophil, it isimportant to remember that the disease processeslumped together as the eosinophilic lung diseasesare a heterogeneous group of diseases. In anattempt to categorize these disorders, we may haveeither appropriately or artificially connected themby their association with the eosinophil.Annotated Bibliography“State-of-the-Art” <strong>Review</strong>Allen JN, Davis WB. Eosinophilic lung diseases. Am JRespir Crit Care Med 1994; 150:1423–1438Most authoritative and complete, yet readable, review of theeosinophilic lung diseases. This article should be in yourreprint collection as a ready reference.Jeong YJ, Kim KI, Seo IJ, et al. Eosinophilic lungdiseases: a clinical, radiologic, and pathologic overview.Radiographic 2007; 27:617–637More recent comprehensive review highlighting the radiologicmanifestations.The EosinophilAfshar K, Vucinic V, Sharma. Eosinophil cell: pray tellus what you do! Curr Opin Pulm Med 2007; 13:414–421This review concentrates on the cell rather than the clinicalsyndromes.Short <strong>Review</strong>sAlberts WM. Eosinophilic interstitial lung disease.Curr Opin Pulm Med 2004; 10:419–424Katz U, Shoenfeld Y. <strong>Pulmonary</strong> eosinophilia. Clin RevAllerg Immunol 2008; 34:367–371Wechsler ME. <strong>Pulmonary</strong> eosinophilic syndromes.Immunol Allergy Clin North Am 2007; 27:477–492Shorter reviews with emphasis on the recent literature.Acute Eosinophilic PneumoniaAllen J. Acute eosinophilic pneumonia. Semin RespirCrit Care Med 2006; 27:142–147Short recent review of this entity.Allen JN, Pacht ER, Gadek JE, et al. Acute eosinophilicpneumonia as a reversible cause of noninfectious respiratoryfailure. N Engl J Med 1989; 321:569–574Badesch DB, King TE, Schwarz MI. Acute eosinophilicpneumonia: a hypersensitivity phenomenon? Am RevRespir Dis 1989; 139:249–252Original descriptions of the“newest” idiopathic eosinophilicpneumonia.Chronic Eosinophilic PneumoniaAlam M, Burki NK. Chronic eosinophilic pneumonia.N Engl J Med 1969; 280:787–798Original description of the most common idiopathic eosinophilicpneumonia syndrome.Carrington CB, Addington WW, Goff AM, et al.Chronic eosinophilic pneumonia: a review. South MedJ 2007; 100:49–53Short recent review.Bronchocentric GranulomatosisOrtiz-Saracho J, Vidal R, Delgado E, et al. Bronchocentricgranulomatosis in a non-asthmatic patient without etiologicagent. Respiration 1996; 63:129–132In a report of a nonasthmatic patient with bronchogenicgranulomatosis, the authors discuss different proposedpathogenetic mechanisms.Ward S, Heyneman LE, Flint JDA, et al. Bronchocentricgranulomatosis: computed tomographic findings infive patients. Clin Radiol 2000; 55:296–300Discussion of the radiographic findings in this rarelydiagnosed disorder.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 79ACC-PULM-09-0302-005.indd 797/10/09 8:05:19 PM


Allergic Bronchopulmonary AspergillosisLazarus AA, Thilagar B, McKay SA. Diagnosis andtreatment of allergic bronchopulmonary aspergillosis.Mayo Clin Proc 2001; 76:930–938More comprehensive review of the topic.Vlahakis NE, Aksamit T. Allergic bronchopulmonaryaspergillosis. Dis Mon 2008; 54:547–564Short, but complete, update of this entity. Suitable for aquick review.Classics in Eosinophilic PneumoniaCrofton JW, Livingstone JL, Oswald NC, et al.<strong>Pulmonary</strong> eosinophilia. Thorax 1952; 7:1–35First attempt to categorize a group of disorders that sharedthe features of pulmonary infiltrates and peripheral eosinophilia.Reeder WH, Goodrich BE. <strong>Pulmonary</strong> infiltration witheosinophilia (PIE syndrome). Ann Intern Med 1952;36:1217–1240Reeder and Goodrich coined the term PIE syndrome in thisarticle.Churg-Strauss SyndromeAlberts WM. <strong>Pulmonary</strong> manifestations of the Chrug-Strauss syndrome and related idiopathic small vessel vasculitissyndromes. Curr Opin Pulm Med 207; 12:445–450Excellent short reviews of the pertinent literature with anemphasis on the thoracic manifestations.Choi YH, Im JG, Han BK, et al. Thoracic manifestationsof Churg-Strauss syndrome: radiologic and clinicalfindings. Chest 2000; 117:117–124Noth I, Strek ME, Leff AR. Churg-Strauss syndrome.Lancet 2003; 361:587–594Brief, readable review of this rare disorder.Parasitic and Tropical <strong>Pulmonary</strong> EosinophiliaChitkara RK, Krishna G. Tropical pulmonary eosinophilia:pathogenesis, diagnosis, and management.Curr Opin Pulm Med 2007; 13:428–432<strong>Review</strong> of eosinophilia and pulmonary eosinophilia associatedwith parasites.Vijayan VK. Parasitic pulmonary eosinophilia. SeminRespir Crit Care Med 2006; 27:171–184Excellent discussion of a rare, but becoming more common,cause of eosinophilia.Nonasthmatic Eosinophilic BronchitisGonlugur U, Gonluger TE. Chronic cough due tononasthamatic eosinophilic bronchitis. Chest 2006;129:116S–121SAn underrecognized cause of chronic cough.Brightling CE. Eosinophilic bronchitis without asthma.Int Arch Allergy Immunol 2008; 147:1–580 Eosinophilic Lung Diseases (Alberts)ACC-PULM-09-0302-005.indd 807/10/09 8:05:19 PM


AsthmaSidney S. Braman, MD, FCCPObjectives:• Understand the epidemiology of asthma and the contributionsof genetic predisposition and environmentalfactors• Learn the pathology and pathophysiology of asthma andthe importance of inflammatory mechanisms of thisdisease• Appreciate the risk factors for asthma and the causes ofexacerbation of symptoms• Be able to objectively and accurately diagnose thisdisease• Learn the strategies of treating asthma based on the NationalInstitutes of Health NAEPP (National Asthma Education andPrevention Program) guidelinesKey words: asthma; atopy; diagnosis; epidemiology; pathophysiology;treatmentDefinition of AsthmaA universally accepted definition of asthmawas elusive for most of the 20th century. Many ofthe clinical and physiologic features appeared tooverlap with other diseases of the airways.Although the inflammatory nature of asthma wasknown in the latter part of the 19th century andwas even reported in Sir William Osler’s Textbookof <strong>Medicine</strong> during that time period, this importantdistinctive aspect of asthma was put aside for mostof the 20th century.The definition instead emphasized the clinicaland physiologic consequences of reversible bronchospasmand bronchial hyperresponsiveness(BHR). This emphasis resulted in controversybecause these two major features of asthma arefound in other common pulmonary conditionssuch as COPD and acute viral tracheobronchitis.During the past decade or so, the emphasis inasthma research and patient care has again shiftedto inflammation and the recognition that this distinctfeature allows for a more precise definition ofthis disease.An expert panel of the National Institutes ofHealth has provided us with the most widelyaccepted definition of asthma. The first report ofthis group was published in 1991. It is oftenreferred to as the NAEPP Asthma Guidelines. It wasupdated as the “Expert Panel Report 3” in 2007.The guidelines define asthma asa chronic inflammatory disease of the airwaysin which many cells play a role, in particular,mast cells, eosinophils, and T lymphocytes.In susceptible individuals this inflammationcauses recurrent episodes of wheezing,breathlessness, chest tightness, and cough, particularlyat night and/or in the early morning.These symptoms are usually associated withwidespread but variable airflow limitation thatis at least partially reversible either spontaneouslyor with treatment. This inflammationalso causes an associated increase in airwayhyperresponsiveness to a variety of stimuli.PathophysiologyAsthma is caused by a complex interaction ofcells, mediators, and cytokines that results ininflammation. This interaction causes the followingimportant pathophysiologic changes in asthmapatients: airway smooth-muscle contraction,hypertrophy and hyperplasia, microvascular leakage,activation of airway neurons, stimulation ofmucus-secreting cells, disruption of the ciliatedepithelium, and the laying down of collagen in thelamina reticularis of the basement membrane layer.The characteristic cellular changes involve the following:(1) constitutive cells such as epithelial cells,mucous glands, endothelial cells, and myofibroblasts;(2) resident cells such as bone marrowderivedmast cells and macrophages; and (3)infiltrating cells such as eosinophils, CD4 (helpercell)T lymphocytes, neutrophils, basophils, andplatelets. These inflammatory cells, includinghistamine, platelet-activating factor, and a numberof derivatives of the arachidonic cascade suchas prostaglandin (PG) D 2and the cysteinyl leukotriene(LT) C 4, LTD 4, and LTE 4, are capable ofgenerating a wide variety of mediators that can<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 81ACC-PULM-09-0302-006.indd 817/10/09 8:05:41 PM


induce bronchoconstriction. A number of typicalhistopathologic findings can be found in theairways of patients with asthma.Histopathologic FindingsInfiltration of the airways by inflammatorycells such as mast cells, eosinophils, activated Tlymphocytes, and neutrophils can be demonstratedby bronchial biopsies and inferred by demonstratingincreased numbers of these cells in BAL fluid.The extensive bronchial wall infiltration by eosinophilsmay be accompanied in severe cases byeosinophilic infiltration of the alveolar septae andadjacent arteries. The number of activated lymphocytesfound in bronchial biopsy specimens hasbeen correlated with the number of local activatedeosinophils and the severity of asthma.Specific cytokines, most of which are productsof lymphocytes and macrophages, appear to directthe movement of cells to the site of airway inflammation.They also activate the cells, causing themto release their mediators. For example, interleukin(IL)-3, IL-5, and granulocyte-macrophage colonystimulatingfactor direct eosinophils to release anumber of mediators from their preformed granules.These mediators are major basic protein,eosinophil cationic protein, eosinophil-derivedneurotoxin, and eosinophil peroxidase. All haveinflammatory effects that contribute to the pathogenesisof asthma.Mast cells, usually as a result of IgE-mediatedstimulation, also release preformed mediators,such as histamine and proteases (eg, tryptase, chymotryptase,carboxypeptidase, and kininogenase)and further act as a regulator of inflammation byproducing cytokines that promote eosinophil infiltrationand activation. Several mast-cell products(histamine, LTC 4, and PGD 2) are smooth-muscleconstrictors and, hence, induce bronchoconstriction.LTs also cause increased mucus secretion.Increased vascular permeability is caused byhistamine, LTC 4, platelet-activating factor, andbradykinin.Denudation of the airway epithelium can leadto airway edema and the loss of substances in themucosa that protect the airway. Epithelial damagepromotes BHR because access by irritating substancesto sensory nerve endings is increased.Similarly, aeroallergens can more readily penetratethe airways. It is believed that eosinophil-derivedbasic proteins, together with partial reductiveproducts of oxygen and proteases, contribute toepithelial fragility. Regeneration first appears assimple or stratified squamous epithelium beforedifferentiation and maturation to new ciliated andmucus (goblet) cells.One of the characteristic findings in patientswith severe asthma is the presence of tenaciousmucus plugs in the airways. There are approximately20 mammalian mucin genes. There areidentified with the letters MUC followed by anumber. MUC5AC is the principal mucin up-regulatedby asthma inflammation. Strategies to suppressmucin secretion may be possible in the future.Death from asthma usually occurs from blockageof the airways by diffuse mucus plugging. In fatalasthma, evidence shows 98% of the airways areoccluded to some extent with mucus. The presenceof mucus is associated with hyperplasia and metaplasiaof goblet cells; it may cause lung hyperinflationcaused by the trapping of air or, when moresevere and located in central airways, can causeatelectasis.Edema of the airway mucosa is caused byincreased capillary permeability with leakage ofserum proteins into the interstitium. A number ofcell-derived mediators are capable of inducingedema formation, including histamine, PGE, LTC 4,LTD 4, LTE 4, platelet-activating factor, and bradykinin.The intense mucosal thickening in asthmacontributes to the airway wall thickness and, therefore,airway wall narrowing.Growing evidence points to asthma as a diseaseof airway wall restructuring that involves the activationof the epithelial mesenchymal trophic unit(EMTU). Inflammatory cells and mediators interactwith the EMTU to cause the changes we refer to asairway remodeling. The airway epithelium becomesan important source of mediators, growth factors,and chemokines that perpetuate the inflammatoryand repair response. Biopsy studies in patientswith chronic asthma have shown a number ofpathologic changes that support the hypothesis ofincreased release of fibroproliferative and fibrogenicgrowth factors in in vivo models of asthma.Growth factors that are thought to be important inasthma include fibroblast growth factor-2, insulinlikegrowth factor-1, platelet-derived growth factor,endothelin-1, and transforming growth factor-.82 Asthma (Braman)ACC-PULM-09-0302-006.indd 827/10/09 8:05:41 PM


Thickening of the reticular basement membrane,the laminar reticularis, is observed withlight microscopy, is a constant feature of asthma,and is diagnostic of the disease. In other chronicrespiratory diseases such as COPD, bronchiectasis,and tuberculosis, basement membrane thickeningappears to be more focal and variable.There is evidence of increased bronchialsmooth-muscle mass that contributes considerablyto the thickness of the airway wall. The smoothmuscle of asthmatic patients does not behaveabnormally after isolation; there is no correlationbetween airway hyperresponsiveness in vivo andincreased airway muscle sensitivity measures invitro. New evidence suggests that the smoothmusclecell may secrete cytokines and chemokinesand express cellular adhesion molecules. Therefore,the smooth muscle that traditionally has beenthought of only as a passive effector cell of asthma,responsible for bronchomotor tone, may also contributeto the inflammation of asthma. This immunomodulatoryfunction of the smooth muscle issimilar to that found in epithelial cells. There is alsoevidence that smooth-muscle cells in the airwaysare capable of producing growth factors that canthemselves promote proliferation in an autocrinemanner. The airway architecture is changed by thedeposition of type III and V collagen and fibronectinbeneath the basement membrane. This depositionhas been referred to as subepithelial fibrosis.Increased numbers of myofibroblasts are also seenin the reticular layer in biopsy material obtainedfrom asthmatic patients. There is a microvascularcomponent to airway remodeling in asthma, withevidence of angiogenesis in biopsy material andincreased levels of vascular endothelial growthfactor in BAL fluid and sputum.The structural changes that occur in the airwaysof asthmatic patients are likely to be detrimentaland contribute to fixed airway narrowing.This is particularly true in patients with severeasthma. Biopsy studies have shown the changes ofairway remodeling in such patients despite treatmentwith high doses of inhaled and oral corticosteroids.Thickening the inner airway wall mayalso amplify the degree of luminal narrowing fora given degree of smooth-muscle shortening,which might result in an exaggerated narrowingof the airway after a bronchoconstricting stimulus.It has been argued that some aspects of remodelinghave beneficial effects, as follows: stiffening theairway may result in decreased compressibilityallowing the airways to better resist dynamic compression;and extra connective tissue surroundingthe smooth-muscle cells may provide a radial constraintto maximal shortening.Physiologic MechanismsNeurogenic Influences: There is growing evidencethat the neural control of the airways isabnormal in patients with asthma and that neurogenicmechanisms may augment or modulate theinflammatory response. The autonomic nervoussystem regulates many aspects of airway function,such as airway tone, airway secretions, blood flow,microvascular permeability, and the release ofinflammatory cells. A primary defect in autonomiccontrol, the -adrenergic receptor theory, has beenpostulated for asthma. It is more likely, however,that autonomic dysfunction is a secondary defectcaused by inflammation or by the effects of treatment.For instance, inflammatory mediators canmodulate the release of neurotransmitters fromairway nerves such as irritant receptors and C-fiberendings. They can also directly act on autonomicreceptors such as those that cause reflex bronchoconstrictionfrom gastroesophageal reflux.There is also evidence that the nonadrenergic,noncholinergic nervous system is important in thepathogenesis of asthma. As the airway epitheliumbecomes denuded, sensory nerve endings maybecome exposed, causing a release of potent neuropeptidessuch as substance P, neurokinin A, andcalcitonin gene-related protein. These neuromediatorscan lead to bronchoconstriction, microvascularleakage, and mucus hypersecretion.This neurogenic inflammation of the airways,triggered by sensitized sensory nerve endings, hasbeen postulated to cause airway “hyperalgesia”and result in symptoms such as cough and chesttightness. Nitric oxide (NO) also appears to be oneof the neurotransmitters of the nonadrenergic,noncholinergic system and is an important brakingmechanism that can initiate bronchodilation.BHR: Airway inflammation is thought to bea key factor in producing BHR, which is a cardinalfeature of asthma. This inflammation can bedescribed as an exaggerated bronchoconstrictiveresponse by the airways to a variety of stimuli such<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 83ACC-PULM-09-0302-006.indd 837/10/09 8:05:41 PM


as aeroallergens, histamine, methacholine, cold air,and environmental irritants. It is not clear whetherbronchial hyperreactivity is acquired or is presentat birth and genetically determined to appear withthe appropriate stimulus. It is thought that airwayinflammation is the stimulus for BHR because itmay be induced by a number of inciting events.These events include viral respiratory infections, anIgE-mediated allergic reaction, and the inhalationof noxious agents such as ozone or sulfur dioxide.Antiinflammatory agents such as inhaled corticosteroids(ICS) are effective in reducing BHR.The degree of BHR can be determined in thepulmonary function laboratory by standard inhalationchallenge testing. The methacholine andhistamine inhalation challenges are the most frequentlyused clinical tools to determine the presenceand degree of BHR. Cold air and hypotonicsaline solution challenges also are performed.Although the end result of this testing, airway narrowing,is the same for these provocative agents,the mechanisms that cause the airways to constrictvary. Methacholine and histamine act directly onairway smooth muscle. Exercise and hyperosmolaror hypoosmolar solutions act indirectly by releasingpharmacologically active substances frommediator-secreting cells such as mast cells. Sulfurdioxide and bradykinin act by directly stimulatingairway sensory nerve endings.The inhalation of hypertonic saline solution ormannitol and adenosine 5’-monophosphate havebeen advocated as producing more diagnosticprovocation for asthma than histamine or methacholinebut have not attained widespread useclinically. The degree of BHR usually correlateswith the clinical severity of asthma and the medicationneeds of the patient. In addition, fluctuationsof diurnal peak flow measurements correlate withthe degree of BHR.Airway Obstruction: Airway obstruction isanother cardinal feature of asthma. The causesof airflow limitation in patients with asthma arelisted in Table 1. During an acute asthma attack,lung hyperinflation occurs. This results in anincrease in lung elastic recoil forces, which act onthe airways to prevent further narrowing. Thisoccurs because of the interdependence of lungvolume and airway caliber because parenchymalattachments cause greater tethering of the airwaysat greater lung volumes.Table 1. Causes of Airflow Limitation in AsthmaAcute bronchoconstrictionMucus plugging of airwaysBronchial wall edemaInflammatory cell infiltrationAirway wall remodeling (fibrosis)Smooth-muscle hypertrophyUncoupling of elastic recoil forcesInflammation of the bronchial wall, however,may uncouple the mechanical linkage between theparenchyma and the airway, which may contributeto airway narrowing and also to BHR. During anacute attack of asthma, airway resistance increases,and all measures of airflow (eg, peak expiratoryflow rate [PEFR], FEV 1, FEV 1/FVC ratio, and specificconductance) are abnormal. During remission,the results of these tests may be normal, and yetconsiderable dysfunction may be present in peripheralairways. Studies in which the authors used awedged bronchoscope technique have shown thatperipheral resistance can be 10-fold greater thanthat in healthy subjects. There is evidence that thisresistance is linked to BHR. In many asthmaticpatients, particularly children and younger adultswith milder disease, airflow obstruction is completelyreversible. In most elderly asthmaticpatients and those of any age group with moresevere and persistent symptoms, airflow obstructionis only partially reversible despite continuousintense antiinflammatory and bronchodilatortherapy, including large doses of oral corticosteroids.The reasons for these permanent physiologicchanges remain obscure, and it is suspected thatairway remodeling is, at least in part, responsible.Atopic AsthmaAtopy may be defined as the largely geneticsusceptibility for developing IgE directed to epitopesexpressed on common environmental allergenssuch as dust mites, animal proteins, pollens,and fungi. Atopic (allergic) reactions in the upperairways (eg, nose and sinuses) and lower airwaysare both important in the pathogenesis of asthma.The term extrinsic asthma has been used to describeasthma that is triggered by exposure to inhaledaeroallergens. Cellular responses may occur withthe first exposure to a specific antigen in suchgenetically predisposed individuals. The antigen84 Asthma (Braman)ACC-PULM-09-0302-006.indd 847/10/09 8:05:42 PM


is trapped in airway mucus and is exposed tounderlying epithelial cells and resident dendritic(Langerhans) cells. As the antigen penetratesbeneath the mucosa, it is likely exposed to granulocytesand tissue macrophages and eventuallyenters the lymphatic system after enzymaticdegradation.In performing these functions, antigenpresentingcells (APCs) are setting the stage forsubsequent lymphocyte activation to occur. It isnot definitely known which cells are the APCs.There is a key role for naive CD4 T thymocytes(called TH0 cells). As they come in contact with theantigen on the APCs, these lymphocytes becomedifferentiated and activated (CD4 /CD25 state).The specific subclass of TH0 cells that become differentiatedduring this allergic reaction are calledTH2 cells and are to be distinguished from the TH1subclass that is associated with the delayed hypersensitivityreaction. The TH2 subclass is associatedwith the release of specific ILs and other cytokinesthat are responsible for episodic and chronic allergicreactions. For example, in the sensitizationstage, the release of IL-4 enhances the synthesis ofIgEs by antigen stimulated B lymphocytes (plasmacells). IgE infiltrates the airways and becomes fixedto mast cells, basophils, and dendritic cells throughhigh-affinity cell surface receptors. This step setsthe stage for the acute allergic response with theinhalation of more antigens.This reaction can be divided into an early (minutesafter exposure) bronchospastic response anda late (hours after exposure) inflammatoryresponse. First, the antigen bridges to at least twoIgE antibodies. Then, mast-cell degranulation andsynthesis of proinflammatory molecules occurs.The synthesized mediators are IL-3, IL-4, IL-5,tumor necrosis factor (TNF)-α, granulocyte-macrophagecolony-stimulating factor, and arachidonicacid derivatives LTs, PGD 2, and thromboxane A 2.The late-phase response is initiated by mast-cellmediators as follows: IL-5 is a chemoattractant foreosinophils, IL-3 and IL-5 are chemoattractants forbasophils, LTB 4appears to attract neutrophils, andTNF-α increases the number of adhesion moleculeson endothelial cells that enhance the adhesion ofleukocytes.The formation of antigen-specific IgE antibodyusually does not occur until the second or thirdyear of life. The prevalence of allergic asthma is,therefore, less common during infancy. Allergenexposure during infancy to cat antigen has beenshown to be a predictor of skin test sensitivity andasthma at age 6 years. The prevalence of atopyincreases throughout childhood and adolescenceand peaks in the second decade of life. Sensitizationto indoor aeroallergens (eg, dust mites, cockroaches,molds, and animals) and outdooraeroallergens (eg, pollens and molds) may be animportant cause of acute and chronic recurrentsymptoms in patients in these age groups. Oftenallergic asthma has a well-defined seasonal variation;avoidance of the offending antigen may resultin a dramatic improvement of symptoms and inpulmonary function and BHR.Nonatopic AsthmaThe term intrinsic asthma is used to describepatients who have none of the typical features ofatopy, including a positive family history of allergyand asthma, positive immediate hypersensitivityreactions to skin-prick tests after exposure to avariety of aeroallergens, and an increased serumIgE level. Usually, such patients are older thanatopic asthmatic patients and have a later onset ofasthma. Bronchial biopsy studies in patients withintrinsic asthma have been compared with a groupof patients with extrinsic asthma with a comparableseverity of symptoms. There is a more intenseinflammatory cell infiltrate in the bronchial mucosaof the intrinsic asthmatic patients with leukocytes,macrophages, and CD3 and CD4 cells. Patientswith intrinsic asthma have an exaggerated T-cellresponse to maintain the same degree of symptomsand BHR. This may mean that intrinsic asthma mayinvolve activation by an as-yet unidentified antigen.Putative nonallergic antigens that may causesuch reactions include viral antigens or inappropriatelyrecognized “self-antigens.”Risk Factors for AsthmaGenetics of AsthmaAlthough it is generally agreed that there is amajor hereditary contribution to the etiology ofasthma, the inheritance patterns are complex, andasthma cannot be simply classified as havingan autosomal-dominant, autosomal-recessive, or<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 85ACC-PULM-09-0302-006.indd 857/10/09 8:05:42 PM


sex-linked mode of inheritance. This may be inlarge part caused by the marked heterogeneity ofthe asthma phenotype. For example, the role ofcontributing factors such as atopy, viruses, aspirinsensitivity, exercise, and occupational exposurediffers greatly among patients with asthma. Variabilityin clinical symptoms, severity of disease,degree of BHR, and response to therapy are otherexamples of disease heterogeneity. However,genetic studies have discovered multiple chromosomalregions that may contain genes that contributeto asthma. High serum IgE levels have beenlinked to chromosomes 5q, 11q, and 12q. Clinically,there is a strong correlation between BHR andincreased IgE levels, and evidence shows there iscoinheritance of genes for atopy and BHR foundon these same chromosomes.Population studies that have conductedgenome-wide screens have contributed to ourunderstanding of the inheritance of asthma. Theyhave shown the linkage of asthma, BHR, total IgElevels, and elevated eosinophil counts to multiplechromosomal regions. Chromosomes 2q, 3p, 5q,6p, 12q, 13q, 19q, and 21q are likely to containgenes for allergy and asthma. These genomescreens also have found numerous loci that containpotential candidate genes that can regulate theimmune response of asthma. For example, genesthat regulate IgE production, 2-adrenergic andglucocorticoid receptor function, and the inflammatorymediator response have been found. Theymay be important in determining genetic susceptibility,responses to environmental stimuli, andresponses to treatment.After the identification and characterization ofthe -receptor gene, variants of the gene wereevaluated to see whether they could be responsiblefor the expression of asthma. One mutation (thesubstitution of glycine for arginine at position 16)was associated with more severe asthma and especiallywith more severe nocturnal symptoms. Thismutation causes an increase in the degree of agonist-promoteddown-regulation of receptor expression.Studies by researchers in China have shownthat the increased risk for the development ofasthma in homozygotes for this allele is profoundlyaffected by cigarette smoking. Another mutant atposition 27 renders the receptor more resistant to-agonist down-regulation and has been associatedwith less intense BHR in patients with asthma.Genes determining the specificity of the immuneresponse also may be important to the pathogenesisof asthma. Genes located on the human leukocyteantigen complex may govern the response toaeroallergens in some individuals. Genes on chromosomes11, 12, and 13 may direct the control ofproinflammatory cytokines. Chromosome 12, forexample, contains genes that encode for interferon-7, mast-cell growth factor, insulin-like growth factor,and NO synthase. Polymorphisms of the IL4RAgene have been associated with severe exacerbationsand poor lung function, and the ADAM33gene, which is expressed on smooth-muscle cells,fibroblasts, and myofibroblasts and plays a role incell signaling, appears to be a susceptibility genefor asthma and is associated with an accelerateddecrease in lung function. It should be noted thatdespite many encouraging reports, most studieson the genetics of asthma that show an associationbetween asthma-related phenotypes and alleleswithin a specific gene have generally lacked statisticalpower.Sex and RaceChildhood asthma is more prevalent in boys,but this prevalence is reversed in puberty andadulthood. The overall prevalence, therefore, isgreater in female individuals. Black race/ethnicityis associated with a greater risk of asthma death,independent of socioeconomic status and education.The greater incidence of asthma that has beenobserved with urbanization suggests that environmentalfactors maybe as important as genetic andracial factors. In fact, subjects of different racesacquire the risk of the population to which theymove.Environmental FactorsAllergens and occupational factors are consideredto be the most important causes of asthma.Several epidemiologic studies have shown a correlationbetween allergen exposure and the prevalenceof asthma and the improvement of asthmawhen allergen exposure ceases. Important indoorallergens include domestic (house dust) mites;animal allergens (eg, cats, dogs, and rodents); cockroachallergen; and fungi (Alternaria, Aspergillus,Cladosporium, and Candida sp). House dust is86 Asthma (Braman)ACC-PULM-09-0302-006.indd 867/10/09 8:05:42 PM


composed of several organic and inorganic compounds,including insects and insect feces, moldspores, mammalian dander, pollen grains, fibers,mites, and mite feces. Outdoor allergens includepollens (mainly from trees, weeds, and grasses);fungi; molds; and yeasts. Occupational asthma isdiscussed in a later section.Air PollutionBoth outdoor and indoor pollutants contributeto worsening asthma symptoms by triggeringbronchoconstriction, increasing BHR, and enhancingresponses to inhaled aeroallergens. The twomain outdoor pollutants are industrial smog (sulfurdioxide particulate complex) and photochemicalsmog (ozone and nitrogen oxides). Whetherlong-term exposure is further injurious is not clear.Modern construction techniques have been suspectedof causing greater indoor air pollution. Inenergy-efficient buildings, there is 50% less turnoverof fresh air. Indoor pollutants include cookingand heating fuel exhausts as well as insulatingproducts, paints, and varnishes that contain formaldehydeand isocyanates. Long-term passivecigarette smoke exposure has been linked to newonsetasthma in children and adults as well asworsening asthma symptoms in those with preexistingasthma. The effect of cleaning solutions andsprays in the home a number of times a week hasbeen linked to an increased risk of asthma developing.Health-care professionals also have increasedodds of asthma developing when exposed to aerosolizedirritants, cleaning solutions, nebulizedmedications, and powdered latex gloves. Nursesshow the greatest risk; there is a correlation withthe length of time on the job.Respiratory InfectionsThe role of respiratory infections in the pathogenesisof asthma has been under intense investigation.It has been established that viral respiratoryinfections are common precipitators of asthma.Whether they can initiate the disease de novo is suspectedbut has not been strongly suggested. Thereare strong data to suggest that some of the atypicalbacteria, such as Chlamydia pneumoniae and Mycoplasmapneumoniae, also may be involved. Theseorganisms are temporally related to exacerbationsof asthma and can be found by the use of polymerasechain reaction techniques to be present inbronchial biopsy specimens obtained from patientswith chronic stable asthma. There have also beenreports of asthma appearing for the first time afteran active infection with both of these organisms, aswell as significant improvement in lung functionand symptoms with antimicrobial therapy directedto these organisms. In one prospective study, Mpneumoniae or C pneumoniae was found in the airwaysof 60% of a group of stable patients withchronic asthma. Therapy with the macrolide antibioticclarithromycin improved lung function anddecreased tissue expression of IL-5 in those patientswho were polymerase chain reaction positive tothese organisms. This finding suggests that antibioticuse in some patients with persistent asthmamay help to control the disease.Allergic RhinitisNasal and sinus diseases are common comorbiditiesin patients with asthma. Allergic rhinitis(AR) and asthma commonly occur together; 40%of children with AR have asthma. Furthermore, thepresence of AR in a child 7 years of age predictsfuture asthma; smokers with AR have a threefoldincreased risk for developing asthma. Whether thisrisk is because AR and asthma are the same diseaseprocess or are two different diseases that affect thesame susceptible population is not known. The factthat the eosinophilic and lymphocytic inflammationin the upper and lower airways are the same,and also that the severity of both diseases occursin parallel, support the former theory. It has beendiscovered that patients with AR, despite the paucityof pulmonary symptoms, have abnormal lungfunction and, when tested with methacholine orhistamine, demonstrate BHR. Exhaled NO also hasshown to be increased in such patients, and bronchialbiopsies have demonstrated eosinophils andmast cells. There has been considerable interest inwhether treatment of the upper airway improvesasthma control. Retrospective studies suggest it,but prospective studies have been disappointing.Other FactorsA number of interesting epidemiologic associationshave come to light concerning the risk of<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 87ACC-PULM-09-0302-006.indd 877/10/09 8:05:42 PM


the development of asthma and atopy. For instance,being underweight and being overweight are bothassociated with an increased risk of the developmentof asthma. There is some evidence thatweight loss in obese asthmatic patients is associatedwith improved symptoms and lung function,especially PEFR variability. Growing up on a farmdecreases the risk of atopy and allergic rhinitis inadulthood, suggesting that environmental factorsmay have a lifelong protective effect against thedevelopment of allergy. In developing countries,the move to cities is associated with a switch frombiomass fuels such as wood, charcoal, and animalwastes to gas and electricity. The use of modernfuels has been associated with an increased rate ofallergic sensitization and symptoms. Studies indicatea reverse relationship between family size andasthma. This finding and others suggest that theexposure of young children to older children athome or to children at day-care centers protectsagainst the development of asthma. There has beensome suggestion that the use of dietary antioxidantsmay protect against the development ofasthma. In one study, for example, the severity ofasthma was negatively associated with the consumptionof red wine. This is another bit of evidenceto suggest that flavonoids and other dietaryantioxidants may protect against asthma.The Natural History of AsthmaThe most reliable information on the naturalhistory of asthma comes from large communitysurveys. Comparisons from country to country andeven within the same country are difficult to makebecause of the varying criteria used to make thediagnosis and the inclusion of cigarette smokerswho might also have underlying COPD. The peakprevalence of asthma in all studies is in childhoodand is approximately 10% of the population. Thisrate decreases to approximately 5 to 6% in adolescenceand early adulthood, when remission ratesare quite high. The prevalence increases again duringlater adulthood to 7 to 9%.The relationship of atopy and asthma has beencarefully studied. In children and adults, the presenceof bronchial hyperreactivity correlates withthe presence and number of positive immediatehypersensitivity skin test results to inhalant allergens.The presence of positive skin test results ininfants of allergic parents correlates with the onsetof asthma in later childhood, and the number ofpositive skin test results shows a correlation withthe severity of asthma in childhood and earlyadulthood. Studies correlating skin test reactivityto house dust mite exposure and asthma alsostrongly implicate atopy in the pathogenesis ofasthma. One alternative theory is that there is onecommon mechanism involved in the developmentof both IgE-mediated hypersensitivity and bronchialhyperreactivity or that the two may be relatedby genetic linkage. This would suggest that thereis not a causal relation between atopy and asthmaand would explain why certain patients (intrinsicasthmatic patients) have no allergic basis for theirasthma.It has been stated that most patients withasthma during childhood will outgrow it by earlyadulthood. Statistics show that this is true in only30 to 50% of children with asthma and is morelikely in male patients than in female patients.Lung growth appears to be relatively normal inchildren with asthma but can be reduced in somewith more severe and persistent symptoms. Evenin those who become asymptomatic at puberty,lung function frequently remains abnormal, andcough and BHR persist. The following three featuresof childhood asthma may predict those whoare more likely to have persistent symptoms intoadulthood: severe atopy determined by skin testing;a marked degree of bronchial reactivity; anddifficult-to-control asthma. This is particularlyevident in young girls. Elevations in serum IgElevels, eosinophilia, and skin test reactivity toaeroallergens are seen in children with persistentwheezing from early childhood to later childhoodbut not in those who wheeze in the first 6 years oflife but outgrow their wheezing by age 6 years (ie,transient wheezers).In adulthood, there is a steady incidence ofnew-onset asthma through patients of all ages, evenin elderly patients. Many patients begin with recurrentwheezing after respiratory viral infections.This pattern may gradually or abruptly developinto persistent wheezing and often severe, poorlyresponsive disease. At other times, asthma developsexplosively, with no previous respiratorysymptoms, immediately after the onset of a typicalviral respiratory infection. Longitudinal studies ofasthmatic patient populations have shown that88 Asthma (Braman)ACC-PULM-09-0302-006.indd 887/10/09 8:05:43 PM


although remission from asthma is common in thesecond decade, it is much less common in older agegroups, although it still may be as high as 20 to30%. Asthmatic subjects who have severe symptoms,reduced ventilatory function, and a concomitantdiagnosis of COPD are much less likelyto experience remission. Not infrequently, heavycigarette smokers in whom COPD has developedmay also manifest acute bronchospasm that isresponsive to therapy with inhaled bronchodilators.This pattern is similar to that seen withasthma. Because of the underlying emphysema andchronic bronchitis, these patients have an elementof irreversible airflow obstruction, and they oftenare labeled as having asthmatic bronchitis to distinguishthem from those with pure asthma and toidentify the irreversible component of their disease.Unlike some adult asthmatic patients, thosepatients with asthmatic bronchitis will not everhave a complete remission of their disease. Asthmamay also result from exposure to occupationalhazards. Remission usually occurs when the patientis removed from the offending environment, butunfortunately, this is not always the case.Clinical Features of AsthmaSymptoms of AsthmaThe typical triad of symptoms of asthma iswheezing, shortness of breath, and cough with orwithout sputum production. These symptoms arenot specific for asthma and can be seen in otheracute and chronic airway diseases. For example,an acute viral tracheobronchitis associated with orafter typical upper respiratory infection symptomscan cause the asthma triad of symptoms and canbe associated with BHR for up to 6 weeks. Unlikeasthma, these symptoms usually resolve completelyover time. COPD caused by emphysemaand chronic bronchitis can also cause typicalasthma symptoms, and the distinction betweenthese conditions can become quite difficult, especiallywhen COPD is complicated by acute viral orirritant-induced bronchospasm.In general, the symptoms of asthma are considerablymore episodic and of more sudden onsetthan those of COPD, and periods of prolongedremission are typical. Attacks of asthma are likelyto be provoked by known aeroallergens, especiallyin the younger atopic population. It is not unusualfor asthmatic patients to have a history of wheezingand shortness of breath after exposure to householdpets such as cats and dogs. Sensitivity to rats,guinea pigs, and other small animals may similarlydevelop in animal handlers and laboratory workers.Asthma may occur seasonally, such as duringragweed season in the fall or during flower andtree blooming in the spring. It is believed thatexposure to household mites that live in bedding,in floor rugs, and on other fabrics found in thehouse also is a major cause of asthma symptoms,especially in warmer climates that favor theirgrowth. Similarly, in areas of the inner city, especiallywhere poverty is found, cockroach exposureis an important inciting agent of atopic asthma.The immediate hypersensitivity reaction(atopic reaction) has two phases, an immediatereaction that causes symptoms within minutesafter exposure and a delayed reaction that occursup to 6 h after exposure. Sometimes exposure toan aeroallergen results in a predominantly delayedallergic reaction. The patient may not recognize theoffending agent since the exposure has occurredmany hours previously.Typically, the triad of symptoms of asthmapresents simultaneously, but this is not always thecase. There is evidence that some patients withasthma perceive their symptoms poorly. In studiesof acutely ill asthmatic patients, up to 10% of themhave no shortness of breath and complain only ofwheezing and cough. The reasons for the lack ofdyspnea remain obscure, but the following severalobservations are relevant:• Asthmatic subjects have a greater threshold fortolerating resistive loads than normal subjects.• Asthmatic subjects with greater resting baselineairflow obstruction are less likely to perceiveworsening lung function after a cholinergicinhalation challenge.• Symptoms caused by a precipitous decrease inlung function caused by the immediate hypersensitivityreaction are better perceived thanthose that may occur as a result of a slowerequal decline in lung function caused by thelate-onset reaction.• Abnormalities in the perception of asthmasymptoms have important therapeutic implications,eg, an asthmatic patient who is not able to<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 89ACC-PULM-09-0302-006.indd 897/10/09 8:05:43 PM


detect increased airway resistance will not reachfor appropriate medication when necessary.• Finally, it has been postulated that poor perceptionof asthma in some patients may be responsiblefor fatal and near-fatal attacks, especiallyin the elderly because impaired perception ofbronchospasm is also a feature of aging. Objectivemonitoring of airway function by means ofspirometry and peak flowmeters is essential forasthma prevention.Other variant manifestations of asthma alsooccur. Patients with asthma may present with onlycough or dyspnea as isolated symptoms. A nonproductivecough may be present for years as asign of asthma before the full triad of symptomsbegins. Cough is probably caused by the stimulationof airway sensory nerves by inflammatorymediators. The asthmatic cough is often provokedby respiratory irritants such as cigarette smoke andby cold air, laughter, and cough itself. Many times,a single cough will begin a series of violent coughingparoxysms that may last for many minutes andlead to exhaustion. Cough may be provoked alsoby deep inhalation and by forced exhalation. Thismechanism may be useful as a bedside test becausea coughing paroxysm induced by a deep-breathingmaneuver suggests hyperreactive asthmaticairways.Because patients with isolated cough or dyspneamay have normal findings on pulmonaryfunction studies, reversible airflow obstructionmay not be possible to document. In such cases,bronchoprovocation testing with histamine ormethacholine may be useful in making the properdiagnosis. Other symptoms of asthma are chesttightness, substernal pressure, chest pain, andnocturnal awakenings. These symptoms are morelikely to be confused with cardiac disease, especiallyischemic heart disease, in the older individual.Gastroesophageal reflux also causes a feelingof chest tightness in addition to heartburn. Refluxcan exacerbate asthma, especially in children, andsometimes the treatment of reflux can result inimprovement of asthma. In patients with symptomaticreflux, upper respiratory symptoms arevery common, and they can be confused with thesymptoms of asthma. For example, cough may bethe sole presenting manifestation of gastroesophagealreflux. Symptoms of nasal and laryngealirritation are common to both gastroesophagealreflux and asthma, and objective measures of airwaydisease can be helpful in distinguishing thetwo.Physical Signs of AsthmaThe physical findings associated with acutebronchospasm of asthma are (1) the direct result ofdiffuse airway narrowing and hypersecretion ofmucus, and (2) the indirect result of reflex influencesfrom an increase in the work of breathing,increased metabolic demands on the body, anddiffuse sympathetic nervous discharge. Tachypneaand tachycardia are universal features of acuteasthma. The average respiratory rate is between 25and 28 breaths/min, and the average pulse rate is100 beats/min. Respiratory rates of 30 breaths/min and heart rates of 120 beats/min are notuncommon and may be seen in as many as 25 to30% of patients.Sinus tachycardia is also a known complicationof asthma therapy that occurs with the use of drugssuch as sympathomimetics and theophylline. Usually,however, because airway function improveswith sympathomimetic treatment, sinus tachycardiaactually improves rather than worsens. Inpatients with acute asthma, premature ventricularcontractions occur occasionally, whereas atrialarrhythmias are extremely uncommon. Other ECGabnormalities seen in acute severe asthma includeP-pulmonale, right axis shift, right bundle-branchblock, and right ventricular strain.Diffuse musical wheezes are characteristic ofasthma, but their presence or intensity does notreliably predict the severity of asthma. By the timewheezing can be detected by stethoscope, peakflow rates may be decreased by as much as 25%.In general, wheezing during inspiration and expiration,loud wheezing, and high-pitched wheezingare associated with greater airway obstruction. Invery severe cases, wheezing may be absent. Thissuggests very poor air movement and impendingrespiratory failure.A prolonged phase of exhalation is typicallyseen, as is chest hyperinflation. These are theresult of airflow obstruction and trapped air,respectively. Accessory muscle use, pulsus paradoxus,and diaphoresis are associated with severeairflow obstruction, although their absence does90 Asthma (Braman)ACC-PULM-09-0302-006.indd 907/10/09 8:05:43 PM


not rule out a severe attack. Cyanosis and signs ofacute hypercarbic acidosis, such as mental obtundation,are absent in all but extreme cases. Accessorymuscle use and pulsus paradoxus are causedby large negative changes in intrapleural pressure,and they may not be manifest in the patient withrapid, shallow breathing. They are reported to bepresent in 30 to 40% of patients with acuteasthma.Objective Measures of Asthmaand Asthma SeverityThe use of objective measurements of pulmonaryfunction is important in patients with asthmabecause the perception of asthma symptoms oftenis poor and the physician’s findings on physicalexamination may either overestimate or underestimatethe severity of the airflow obstruction. Thefollowing four tests of pulmonary function areextremely useful for establishing the diagnosis ofasthma and/or for following the clinical course ofthe patient once the diagnosis has been made: (1)spirometry, (2) PEFR measurements with a peakflowmeter, (3) arterial blood gas analysis, and (4)bronchoprovocation testing.During an acute attack of asthma, on averagethe FEV 1is reduced to approximately 30 to 35% ofpredicted, and the FVC to approximately 50% ofpredicted. The FEV 1in absolute terms is reducedto approximately 1 L. The PEFR is reduced toapproximately 150 L/min. Reversible airflowobstruction is characteristic of asthma and can bedemonstrated by spirometry: a 15% improvementin the FEV 1(and an absolute value 200 mL)5 to 10 min after treatment with a short-actinginhaled -agonist generally is accepted as a significantchange. Alternatively, after outpatienttreatment, a 15% improvement in the FEV 1between office visits proves the presence of reversibleairway disease and supports a diagnosis ofasthma.Short-term home monitoring with the PEFRmeter can be useful in the diagnosis of asthma andidentification of environmental triggers of asthmaand also can be used to detect early signs of deteriorationwhen symptoms change. Long-termmonitoring is useful for those with severe brittleasthma and for those who have a poor perceptionof asthma symptoms. Peak flow measurementsideally should be performed early in the morning(when measurements tend to be lowest) and in theevening (when they should be highest), 5 to 10 minafter the patient inhales a -agonist. Each patientshould establish a personal best PEFR after aperiod of maximal therapy. The severity of asthmais reflected not only by the level of baseline obstructionbut also its variability during a 24-h period.A zone system has been provided by the NAEPPguidelines for patient ease, as follows: green is 80to 100% of the personal best and shows good control;yellow is 50 to 80% of the personal best andsignals caution that asthma is not under sufficientcontrol, and red is 50% of personal best andsignifies danger and the need for immediate physicianintervention.A diurnal variation in PEFR of 20% is diagnosticof asthma. The magnitude of peak flow variabilityis in general proportional to the severity ofthe disease. A high degree of variability signalsunstable asthma that demands increased medication.Virtually all asthmatic patients have hypoxemiaduring an acute exacerbation; the more severethe attack, the lower the arterial oxygen tension.Arterial blood gas analysis is helpful during asevere attack. During an acute attack, an assessmentof oxygen saturation by transcutaneousoximetry is helpful to ensure adequate oxygenation.The mechanism of arterial hypoxemia isventilation/perfusion mismatch. Areas with lowventilation/perfusion ratios cause hypoxemia andare caused by bronchospasm, mucus plugging, andmucosal swelling from inflammation.Hypocarbia and respiratory alkalosis are presentin approximately 75% of patients with acuteasthma. When the obstruction worsens and theFEV 1approaches about 15 to 20% of predicted, thePco 2normalizes. Carbon dioxide retention occurswhen the FEV 1reaches 15% of predicted and theabsolute FEV 1is very severely reduced to 0.5 L.Acute respiratory acidosis results. This occurs inapproximately 10% of asthmatic patients who seekemergency care. The mechanism of carbon dioxideretention in patients with severe asthma is severeventilation-perfusion mismatch. High ventilation/perfusion ratios result in dead space or wastedventilation. As the work of breathing increases andcarbon dioxide production rises, the lungs becomeincapable of removing the carbon dioxide that isproduced, and respiratory acidosis occurs.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 91ACC-PULM-09-0302-006.indd 917/10/09 8:05:43 PM


It is believed that respiratory muscle fatiguealso contributes to respiratory failure and, inextreme cases, sudden respiratory muscle failurecan result in acute cardiopulmonary arrest. Superimposedmetabolic acidosis is observed on bloodgas analysis in severe cases. This may be the resultof lactic acidosis caused by vigorous muscle contractionor to inadequate cardiac output and maypossibly be seen as a complication of excessivesymptomatic use. Inhalation challenge testing isuseful in some patients when the diagnosis ofasthma is not secure. This occurs when the patienthas normal peak flow rates and spirometry findingsyet typical asthma symptoms are present. In suchcases, an FEV 1response to an inhaled bronchodilatorcannot be used as a criterion for asthma. If thefluctuation in diurnal PEFR is also not useful,methacholine or histamine bronchoprovocationtesting can be helpful. Low concentrations of theagonist are inhaled after baseline spirometryensures normal or nearly normal airway function.Gradually increasing doses are administered,either by tidal volume breathing or by single deepbreaths from a dosimeter, and the forced spirogramis repeated after each dose. A calculation of theFEV 1at the administration of each dose producesa dose-response curve. If the FEV 1decreases 20%below baseline at any standard dose, the test resultis positive. An inhaled -agonist is administered,which promptly returns lung function to baseline.A positive methacholine challenge result is typicalof asthma, but it is not specific because it can beseen in a number of other inflammatory airwaydiseases and in patients who have allergic rhinitiswithout asthma. A normal response to methacholineor histamine is incompatible with asthma, andan alternative diagnosis should be considered.Exercise testing is a form of bronchoprovocationtesting that is especially useful in children.Using a standard 6-min protocol, a 15% decreasein FEV 1or a 20% decrease in PEFR from baseline 5to 15 min after exercise is considered to be diagnostic.Although the presence and quantificationof various inflammatory cells and mediators insputum and body fluids have been used to reflectthe activity of inflammation in asthma, there isunfortunately there are no biomarkers of diseaseactivity that can be used clinically. NO is a reactivegas that is formed from arginine through the actionof NO synthase. In asthma, there is some evidencethat this enzyme is up-regulated, and elevatedlevels of NO have been detected in the exhaled airof asthmatic subjects. Exhaled NO is well establishedas a marker of eosinophilic inflammationand has been shown to correlate with sputumeosinophil counts, BHR, serum IgE levels asthmasymptoms and lung function tests. It has beenshown to be a fairly effective screening tool with asensitivity of 88% when there are levels of 20parts per billion in patients with asthma symptoms.Unfortunately, exhaled NO levels do notcorrelate with disease severity, but serial measurementsmay aid in monitoring disease activity. Theyhave shown promise, for example, in identifyingpending exacerbations of asthma. A standardizedmethod of measurement must be carefully followedto prevent spurious results. The availabilityof a low-cost, hand-held analyzer may lead to moreuse, especially when adequate reimbursement byinsurance companies is available.The Treatment of AsthmaThere are both short-term and long-term therapeuticobjectives for every asthmatic patient. Theshort-term objectives are the control of immediatesymptoms and the response to decreasing peakflow rate measurements. Long-term objectives arethose directed at disease prevention because thereare now well-proven strategies to avoid seriousexacerbations of acute bronchospasm that oftenlead to ED visits or hospitalization. To meet thesetherapeutic objectives, four components of asthmacare should be addressed. First, the optimal treatmentof asthma depends on a careful assessment ofthe patient’s symptoms as well as objective monitoringby office spirometry and home PEFR measurements.Second, the treatment of asthma withbronchodilator and antiinflammatory medicationsis tailored to the patient’s needs and relies on astaging system that is based on the symptoms andobjective measures of lung function. Third, measuresshould be taken to avoid exposure to respiratoryallergens and irritants that can cause theworsening of symptoms, including the avoidanceof outdoor allergens such as ragweed, grass, pollens,and molds and indoor allergens such as animaldander, house-dust mites, and cockroach antigen.When exposure cannot be avoided, allergen immunotherapyshould be considered, although this92 Asthma (Braman)ACC-PULM-09-0302-006.indd 927/10/09 8:05:44 PM


modality is somewhat controversial. Indoor irritantssuch as smoke from cigarettes and woodburningstoves, strong odors, and cleaning solutionsare particularly troublesome. Fourth, patient educationcan be a powerful tool in asthma control. Familymembers also can be helpful, especially withchildren and elderly adults. Active participation bya patient in monitoring lung function, the avoidanceof provocative agents, and decisions regardingmedications provide asthma management skillsthat give that patient the confidence to control hisor her own disease.Medications for AsthmaAntiinflammatory Agents: Antiinflammatoryagents are capable of reducing airway inflammationand thus improving lung function, decreasingbronchial hyperreactivity, reducing symptoms, andimproving the overall quality of life. Corticosteroidsare the most useful antiinflammatory agents.They act by preventing the migration and activationof inflammatory cells, interfering with the productionof PGs and LTs, reducing microvascularleakage, and enhancing the action of 2-adrenergicreceptors on airway smooth muscle.Corticosteroids are available for oral, parenteral,and inhaled use. Oral preparations such asprednisone are useful for the treatment of acuteexacerbations of asthma that are unresponsive tobronchodilator therapy. Doses of 40 to 60 mg/d areadministered until the patient responds, and thenthe dosage can be slowly tapered down. Often,poorly controlled asthma requires daily or everyother-daymaintenance therapy with prednisonein dosages of 10 to 15 mg. IV corticosteroids, usuallyadministered as methylprednisolone, 60 to 80mg every 6 to 8 h for 1 or 2 days, are effective within4 to 6 h of administration in preventing furtherprogression of the severe asthma exacerbation thatrequires hospitalization.ICS are safe and effective treatment for moderate-to-severeasthma and have been in use for 20years. Formulations of beclomethasone, triamcinolone,flunisolide, fluticasone, budesonide, andciclesonide can reduce airway inflammation withone to several months of treatment. Several studieshave shown that the therapeutic effects of ICS areimparted in active cigarette smokers. The longtermuse of ICS has been associated with a goodsafety profile. High doses of inhaled steroids (eg, 1,000 μg/d) are capable of causing hypophyseal−pituitary−adrenal axis suppression. Local adverseeffects such as hoarseness, dysphonia, cough, andoral candidiasis do occur but can usually beavoided by the use of a spacer or holding chamberand by rinsing the mouth after each use. Oral andparenteral corticosteroids are associated with manyside effects such as the risk of osteoporosis, cataracts,diabetes mellitus and, rarely, depression ofimmunity to infection. Attempts to reduce dependenceon oral corticosteroids should be made,especially by the use of inhaled agents.Cromolyn sodium and nedocromil sodium aretwo antiinflammatory agents that are available ininhaled form and have an extremely good safetyprofile. A 4- to 6-week trial may be useful to determinetheir effectiveness in the prevention of asthmasymptoms. These agents are capable of preventingallergen-induced bronchospasm. LT pathwaymodifiers (LPMs) are also medications that areconsidered to be asthma controllers. There are twosubclasses, as follows: the 5-lipoxygenase inhibitors,which inhibit the cysteinyl LTs and also LTB 4,and the LTD 4receptor antagonists (LTRAs) of thecysteinyl LTs LTC 4, LTD 4, and LTE 4. These agentshave been shown to be effective in preventingallergen-induced asthma, exercise-induced asthma(EIA), and aspirin-induced bronchospasm andallergic rhinitis. The use of antihistamines with theLPMs provides a more complete protection toallergen challenge. Studies comparing the LPMswith low-dose ICS have favored the latter withrespect to FEV 1; however, symptom scores, daily 2-agonist use, and patient preferences are similar.The use of LPMs also may reduce asthma exacerbationrates and the need for steroid bursts.The authors of several studies examined therole of LTRAs as steroid-sparing agents. In patientswhose symptoms were well controlled on an ICS/LABA combination, switching to a regimen ofLTRA/LABA resulted in greater numbers of treatmentfailure (26% vs 9%), showing clear inferiorityof the LTRA/LABA combination. In a large randomizedtrial of asthmatic patients with poorlycontrolled symptoms, an LTRA (or theophylline)did not add any benefit when added to existingtreatment regimens. There is some evidence thatthe use of LTRAs in active smokers may improveoutcomes compared with the use of an ICS.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 93ACC-PULM-09-0302-006.indd 937/10/09 8:05:44 PM


The LPMs are generally very safe, which has ledto more use in younger asthmatics, especially 4years of age. The rare cases of Churg-Strauss vasculitishave occurred in patients with severe steroid-dependentasthma who have had a recentsteroid taper. Because this form of vasculitis hasbeen seen in patients in whom inhaled steroidshave been substituted when oral steroids havebeen tapered, the reaction may not be specific tothe LT-modifying agents.Omalizumab, a humanized murine monoclonalantibody that inhibits the binding of IgE to mastcells by forming complexes with circulating freeIgE, has been shown to be effective in the treatmentof patients with atopic asthma. The reduction offree IgE in the circulation leads to a down-regulationof basophil and mast-cell receptors. This furtherreduces the potential for mast cell/basophilactivation and the subsequent release of inflammatorymediators. Randomized, controlled trials havebeen conducted in patients with moderate-tosevereasthma. The agent has been shown to reducethe asthma exacerbation rate and the use of ICS.Many patients are able to completely eliminate theneed for corticosteroid therapy. Symptom scoresand quality-of-life indexes also improve. Themedication must be administered by injectiontwice a month, and many patients can be taughtself-administration. The drug is well tolerated,with serious adverse events rarely reported. TheUS Food and Drug Administration (FDA) issueda warning regarding omalizumab in 2007 after itwas notified of case reports of anaphylaxis afterinjection of this agent. The reaction may occur atrandom time points and not just after the first injection.Physicians are advised to have epinephrineavailable for such an adverse reaction after treatment,and patients should carry an epinephrine kitafter leaving the office.Bronchodilators: Inhaled short-acting -adrenergicagonists (SABAs) are the treatment of choicefor the acute exacerbation of asthma symptoms.Inhaled agents can be delivered by metered-doseinhaler, dry-powder capsules, and compressordrivennebulizers. Long-acting -agonists(LABAs) may be helpful for long-term maintenancetherapy and also used to control nocturnalsymptoms. Studies of inadequately controlledasthmatic patients have shown a benefit to addinga LABA to a moderate-dose ICS regimen ratherthan doubling the dose of the ICS, adding theophylline,or adding an LPM. However, becauseof potential safety concerns with LABA therapy,the most recent NAEPP guidelines (Expert PanelReport 3) retracted previous recommendations togive a low-dose ICS/LABA combination in preferenceover doubling the dose of an ICS. The GlobalInitiative for Asthma (GINA) guidelines have notmade this change because of the robust literaturesupporting combination therapy superiority.The safety of LABA therapy was questionedmainly as a result of a study called the SMART trial,a 28-week randomized observational study inwhich the authors found a small but statisticallysignificant increase in respiratory-related andasthma deaths in patients receiving salmeterol. Asubgroup analysis suggested that the risk wasgreater in African Americans. A post hoc analysissuggested that deaths were mainly found in patientswho were not receiving effective controller medication.It has been argued that the bronchodilator effectof the LABA offered symptomatic relief, therebymasking inflammation and worsening the disease.The need for regularly scheduled doses ofSABAs should alert the physician to the need formore intense antiinflammatory medication. Regularlyscheduled doses of SABA agents have beenassociated with diminished control of asthma andheightened bronchial reactivity. In addition, epidemiologicevidence has linked excessive -agonistuse to increased mortality. Ideally, SABA therapyshould be prescribed for acute symptom relief onan as-needed basis.With the discovery and sequencing of the-agonist receptor gene (ADRB2), interest hasturned to the role of -adrenergic polymorphismsin the control of asthma. The variant at amino acidposition 16 (Gly16Arg) has received the mostclinical attention. Most studies have shown noinfluence on bronchodilator responses. Prospectivestudies such as the -Adrenergic Response byGenotype Study study have shown that patientshomozygous for Arg16 have worse asthma controlwhen treated with regularly scheduled short acting-antagonist, albuterol, than those asthmaticshomozygous for Gly 16. The effect of continuouslong acting -adrenergic therapy on asthma controlis less clear. Numerous large retrospective pharmacogeneticanalyses have failed to show an effectin all racial groups and in children and adults.94 Asthma (Braman)ACC-PULM-09-0302-006.indd 947/10/09 8:05:44 PM


Theophylline is an effective bronchodilator andhas antiinflammatory properties. It is available asa sustained-release preparation and can be takenonce or twice daily. The monitoring of theophyllineblood levels is important to avoid toxicity, especiallyin the elderly, who are more prone to adverseeffects. GI side effects are seen with mild toxicityand blood levels in the range of 20 to 30 μg/mL.Serious cardiac arrhythmias and seizures mayoccur with blood levels in excess of this range.A blood level of 8 to 15 μg/mL is generally consideredto be therapeutic.Inhaled anticholinergic agents such as ipratropiumproduce bronchodilatation by reducing vagaltone and are widely used in patients with COPD.They may be useful in combination with -agonistsfor a severe acute exacerbation of asthma. Their rolein long-term maintenance therapy for asthma hasnot been established. They may be tried as a substitutebronchodilator when side effects precludethe use of -agonists, although it must be rememberedthat they have a slower onset of action of 30to 60 min until the maximal effect is produced.Treatment Protocols for AsthmaA global strategy for asthma management andprevention has been offered in the NAEPP guidelines.The most recent 2007 NAEPP guidelines(Expert Panel Report 3) have offered two new treatmentplans. Treatment protocols use step-carepharmacologic therapy based on the intensity ofasthma symptoms and future risk as measures ofcontrol and the clinical response to these interventions.As symptoms and lung function worsen,step-up or add-on therapy is administered. Assymptoms improve, therapy can be steppeddown.Asthma Severity: The severity classification andrecommended treatment protocols are listed inTable 2.Asthma Control: The classification of asthmaby severity has proven useful when decisions arebeing made about management at the time of theinitial assessment of a patient. Asthma severity,however, involves both the severity of the underlyingdisease and its responsiveness to treatment.This classification of severity is no longer recommendedas the basis for ongoing treatment decisionsin the new NAEPP guidelines. Instead, theconcept of asthma control has been suggested asa basis for treatment decisions. The term controlrefers to control of the manifestations of disease,including symptoms, effects on the quality ofdaily living, and use of rescue medication. A periodicassessment of asthma control is useful andstrongly recommended. The new classification ofasthma control is listed in Table 3.Table 2. Asthma Treatment by Severity of DiseaseMild intermittent asthmaSeverity: symptoms 2 times/wk nocturnal symptoms 2 times/mo FEV 1and PEFR 80% of predicted; PEFR variability 20%Treatment: NEAPP guidelines call this step 1 therapy: inhaled SABA as neededMild persistent asthmaSeverity: symptoms 2 times/wk and 1 time/d; nighttime symptoms 2 times/mo; FEV 1and PEFR 80% of predicted;PEFR variability of 20 to 30%.Treatment: step 2 therapy: Begin antiinflammatory therapy; low-dose ICS are preferred, or consider a LPM or cromolyn.Use short-acting -agonist as needed for quick relief; however, increased use means need for additional controller therapy.Moderate persistent asthmaSeverity: daily symptoms and daily use of rescue -agonist; 2 exacerbations/wk and 1 nighttime exacerbation/mo;FEV 1and PEFR of 60 to 80% of predicted; PEFR variability 30%.Treatment: Step therapy is advised in the 2007 NAEPP guidelines. Adding an LABA is given equal weight to increasing tomedium-dose ICS; if symptoms still persist, may increase dose of ICS or add LABA, whichever applies, or medium-doseICS with LPM or consider theophylline; continue a SABA for short-term relief. Step 5 is high-dose ICS plus LABA andconsider omalizumab.Severe persistent asthmaSeverity: continuous symptoms; limited physical activity; frequent exacerbations; frequent nighttime symptoms; FEV 1andPEFR 60% of predicted; PEFR variability 30%.Treatment: high-dose ICS, LABA, possibly theophylline, oral corticosteroids, and consider omalizumab.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 95ACC-PULM-09-0302-006.indd 957/10/09 8:05:44 PM


Table 3. Classification of Asthma Control in Youths 12 Years of Age and AdultsCharacteristicsControlled (All ofthe Following)Not Well Controlled (AnyMeasure Present in Any Week)Very PoorlyControlledDaytime symptoms None ( 2/wk) 2 times/wk Throughout dayInterference with normal activities None Any Extremely limitedNocturnal symptoms/awakening 2 times/ mo 1 to 3 times/wk 4 times/wkNeed for SABA (rescue treatment) None ( 2/wk) 2 times/wk Several times a dayLung function (PEFR or FEV 1) 80% of predicted or 60 to 80% of predicted or 60% of predictedValidated questionnaire(eg, Asthma Control Test score)personal bestpersonal bestor personal best 20 16 to 19 15Step-down Care. There are no hard and fast ruleson stepping down asthma care when asthma controlis improved. Studies conducted by the AsthmaClinical Research Centers have shown that patientswith mild persistent asthma controlled with a lowdose of ICS do better with continuation of the corticosteroidthan switching to a LTRA but do as wellwhen switched to a once-a-day regimen of ICS/LABA. In another study with similar patients, asneededuse of a ICS/SABA combination did as wellas daily ICS twice daily with less total inhaledcorticosteroid use, but patients randomized to asneededSABA did worse. The use of an ICS/LABAcombination concurrently as a scheduled maintenancecontroller and as an as needed controllerreliever is another new approach (not approved bythe FDA in the United States). The use of abudesonide/formoterol combination product inthis manner has led to fewer exacerbations andbetter lung function.Asthma SyndromesEIANearly all asthmatic patients experienceincreased bronchospasm after vigorous exercise atsome point in their history (ie, EIA). If they aretested in the pulmonary function laboratory whilenot receiving medication, 70 to 80% of patients willshow a characteristic decrease in peak flow or FEV 1immediately after a 6- to 8-min exercise challenge.During exercise, patients seem to be protected asthe airways actually dilate, possibly attributable tothe presence of circulating catecholamines. Symptomsof cough and shortness of breath parallel thedecrease in lung function that usually reaches itspeak response 5 to 10 min after exercise. Oftenpatients hear no wheezing during the attack, andin some patients, atypical presentations of chesttightness or chest pain alone will occur. A cardiacworkup may often precede pulmonary testing insuch patients. The symptoms usually abate withouttreatment, although the administration of a shortacting-agonist immediately returns lung functionto normal. Occasionally, the attack after exercisecan be very severe and may require emergencytreatment.The mechanisms that cause exercise-inducedbronchospasm in susceptible patients (eg, asthmaticpatients, some patients with allergic rhinitis,and patients with cystic fibrosis) have beenintensely studied. As ventilation increases withexercise, cooler, dryer air is drawn into the airwaysin greater quantities. This action causes heat andwater loss from the airways into the airstream.Airway cooling and mucosal drying are thoughtto trigger mast cells to release their mediators. Thegreater the heat and water loss, the greater thedegree of bronchospasm. A number of factors mayenhance or reduce the risk of exercise-inducedbronchospasm, including the following:• Patients with poorly controlled asthma are morelikely to have EIA.• Exercise in cold, dry air is more likely to causeEIA.• Running, especially outdoors, is more likelyto cause EIA than other activities such asswimming.• The greater the intensity of the exercise, themore likely the bronchospasm.• Bronchoconstriction becomes less intense duringa second challenge if performed an hour or96 Asthma (Braman)ACC-PULM-09-0302-006.indd 967/10/09 8:05:44 PM


so after the first. An intense warm-up period30 to 60 min before competitive sports can behelpful.• Pretreatment can effectively block exerciseinducedbronchospasm in 90% of patients.-agonists are the first choice for treatment,and cromolyn is a good alternative. They canbe used together if needed.Nocturnal AsthmaExacerbations of wheezing and shortness ofbreath are a common cause of sleep deprivation inasthma patients and tend to be underreported bypatients and overlooked by physicians. Evenpatients who have mild disease that is apparentlycontrolled by medication may awake at least onetime during the night because of their asthma.Deaths from asthma have been increasing in manycountries during the past few decades, and most ofthese deaths from asthma occur at night betweenmidnight and 8:00 am. The mechanisms of nocturnalasthma are not completely understood. Bronchialbiopsy specimens have not shown increases inT cells, eosinophils, or mast cells at 4:00 am in asthmaticpatients with nocturnal asthma. Transbronchialbiopsy specimens have shown accumulationsof eosinophils and macrophages in alveolar andperibronchial tissue in these patients, leading to theconcept that adventitial inflammation in the peripheralairways leads to excessive airway narrowing.This leads to an altered airway-parenchymal interdependenceand the loss of the normal elastic recoilforces that normally keeps the airways open.It has been observed that lung function has adefinite circadian rhythm and that PEFRs are highestat 4:00 pm and lowest at 4:00 am. Healthy individualsshow a fluctuation of only about 8%,whereas many asthmatic patients can, during periodsof unstable symptoms, show as much as a 50%variation in peak expiratory rates in a 24-h period.This variation can result in frequent nocturnal awakeningsand the need for frequent rescue therapy.Sustained-release theophylline has proven to be avaluable tool in the control of nocturnal asthma, andLABAs such as oral extended-release albuterol andinhaled salmeterol are also helpful. Because researchwith transbronchial biopsy specimens has shownthat the levels of inflammatory cells are greatest at4:00 am, the key to controlling nocturnal asthma isthe control of the inflammatory response with ICSor, if necessary, oral corticosteroids.Aspirin-Induced AsthmaAcute idiosyncratic reactions to aspirin mayresult in either acute bronchospasm or urticaria/angioedema, rarely both together. Aspirin-inducedasthma is a syndrome that predominantly affectsadults rather than children and nonatopic individuals.One third of patients, however, maymanifest positive immediate skin-test hypersensitivityto inhaled aeroallergens. The disease usuallyevolves during the course decades and first beginsas chronic nonallergic perennial rhinitis, oftencomplicated by nasal polyposis and recurrent bacterialsinusitis. Asthma then appears, is oftensevere and unremitting, and requires treatmentwith systemic steroids.The symptoms of rhinosinusitis, however, maycontinue to be more troubling to the patient thanthe asthma. Although aspirin may have previouslybeen tolerated, suddenly acute bronchospasmdevelops in the patient 30 min to several hoursafter ingesting a standard dose of oral aspirin. Theuse of aspirin also may be associated with flushingof the face and ocular and nasal congestion. Crosssensitivityoccurs with the use of other nonsteroidalantiinflammatory drugs such as indomethacin,naproxen, ibuprofen, sulindac, and piroxicam, allof which are inhibitors of PG synthesis from thecyclooxygenase pathway of arachidonic acidmetabolism. Noninhibitors such as sodium salicylate,salicylamide, propoxyphene, and acetaminophendo not cause the reaction.It is believed that, by blocking the cyclooxygenasepathway, arachidonic acid metabolism ispreferentially shifted to the 5-lipoxygenase pathwayto produce LTs, including the cysteinyl LTsLTC 4, LTD 4, and LTE 4. They are potent inflammatorymediators that can induce bronchoconstriction,mucous secretion, bronchial wall edema, andswelling of the nasal mucosa. They are also potentchemokines for eosinophils. There is a fourfoldincrease in urinary LTE 4levels after aspirin provocationin the aspirin-sensitive patient. It has alsobeen shown that aspirin-sensitive patients haveincreased expression of the cysteinyl LT receptorCysLT1 on leukocytes from nasal mucosalbiopsyspecimens compared with those who do not have<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 97ACC-PULM-09-0302-006.indd 977/10/09 8:05:45 PM


this syndrome. The precise mechanism for thissyndrome is not known. The overexpression of theenzyme LTC 4synthase has been reported, andsome studies have described polymorphisms ofthe LTC 4synthase promoter region.The prevalence of aspirin-induced bronchospasmin the population of asthmatic patientsdepends on the population studied. Approximately3 to 5% of hospitalized asthmatic patients report ahistory of a reaction to aspirin. This number isgreater at highly specialized referral centers atwhich patients with the most severe asthma arecared for. The treatment is, of course, the avoidanceof aspirin and nonsteroidal antiinflammatory drugswith known cross-reactivity. For the occasionalpatient who needs these drugs, oral desensitizationprotocols using gradually increasing doses of aspirinare effective. With the use of aspirin desensitization,approximately two thirds of the patients reporta significant improvement in their rhinosinusitis,and approximately one half have improvement intheir asthma. The addition of LT inhibitors (eg,zileuton) and receptor antagonists (eg, montelukast)as therapeutic options for asthma is importantbecause the agents have been shown to block aspirin-inducedbronchospasm in sensitive patients.Occupational AsthmaAsthma that is precipitated by a particularoccupational environment and not to stimuli outsideof the workplace is called occupational asthma.The following two types have been described: (1)asthma that follows a latent period of exposure toeither a high-molecular-weight or a low-molecularweight“sensitizing” antigen and (2) asthma thatfollows exposure to workplace irritants. One formof irritant-induced asthma is called reactive airwaysdysfunction syndrome, a condition that usuallyresults from the sudden inhalation of a large doseof a highly irritating substance. Occupationalasthma associated with a latency period causesasthma by immunologic events that affect only asmall proportion of exposed subjects.Although some compounds induce asthmathrough the production of specific IgE antibodies,the immunologic mechanisms responsible for otheragents have not been identified. The reactionbetween specific IgE antibodies and antigens maycause an isolated early asthmatic reaction thatoccurs within a few minutes after the occupationalexposure. It typically reaches a maximum within30 min and subsides within 1 to 1.5 h. Also, a biphasicreaction may occur, and the spontaneous recoverymay be followed by a late bronchospasticreaction 4 to 6 h after exposure. This reaches maximalintensity within 8 to 10 h and subsides after24 to 48 h. When an isolated late reaction occurs,the affected individual may be away from theworkplace, making the diagnosis less obvious. Thepresence of sensitization to occupational agents canbe detected by skin tests, radioallergosorbent tests,or enzyme-linked immunosorbent assay, which canstrongly support the diagnosis. Unfortunately,there are very few standardized commerciallyavailable materials for skin tests or for radioallergosorbenttests in patients with this disorder.The following six major categories of agentshave been described in the pathogenesis of occupationalasthma:1. Exposure to animals, shellfish, fish, and arthropodscan cause asthma in farmers, laboratoryworkers, grain handlers, and poultry workers.2. Exposure to wood, plants, and vegetables cancause asthma in woodworkers, carpenters,grain handlers, bakers, and tobacco workers.3. Enzymes and pharmaceuticals are known tocause asthma in pharmaceutical workers, pharmacists,and detergent industry workers.4. Low-molecular-weight chemicals cause asthmain solderers, spray painters, chemical manufacturers,polyurethane manufacturers, and electronicsworkers.5. Metals and metal salts such as aluminum,chromium, cobalt, nickel, and platinum fumescan cause asthma in electroplaters, hard metalworkers, polishers, and solderers.6. Dusts, fumes, and gases may cause asthma viaa sudden, high-concentration exposure, usuallyas the result of an accident in the workplace orvia less severe repetitive exposures.Estimates of the prevalence of occupationalasthma vary, but it is estimated that 2 to 15% of allcases of adult-onset asthma occur from workplaceexposure. Seeking the occupational causes ofasthma in the patient’s history can be very rewardingand can lead to primary prevention by theavoidance of the offending environment.98 Asthma (Braman)ACC-PULM-09-0302-006.indd 987/10/09 8:05:45 PM


Cough-Variant AsthmaAt times, cough may be the sole presentingmanifestation of asthma, and the characteristicfindings of variable airflow obstruction may notbe present. Such patients represent a subgroup ofpatients with asthma, rather than being asthmaticpatients who cough. As a result, cough-variantasthma (CVA) is significantly underdiagnosed.Prospective studies have shown that asthma isamong the most common causes of chronic cough(24 to 29%) in adult nonsmokers.There is evidence that cough receptors in theairways are separate from bronchoconstrictivereceptors, and this may explain why such patientsdo not wheeze. Cough receptors are more abundantin the central airways. Patients with CVAhave demonstrated heightened sensitivity oftheir cough reflexes, whereas those with thetypical form of asthma do not differ from healthyvolunteers when cough reflex sensitivity isexperimentally measured. Methacholine sensitivity,on the other hand, occurs less often in patientswith CVA when compared with those with typicalasthma. The cough receptors are presumablystimulated by inflammatory cell mediators thatare released by airway inflammation. Subepitheliallayer thickening, which is a pathologic featureof airway remodeling that is thought to becaused by chronic inflammation, is present inCVA.The diagnosis of CVA should be suspected ina patient who has a chronic cough, usually presentfor months and even years, and no obvious causeseen on the chest radiograph. The history is oftensuggestive in that the patient coughs vigorouslyafter exposure to fumes, cigarette smoke, and otherirritants and often after laughter and exercise. Thediagnosis is supported by a positive response tobronchial inhalation challenge with methacholine,histamine or exercise, and is confirmed by a clinicalresponse to bronchodilators and/or antiinflammatoryagents. Most patients will respond totreatment with the use of inhaled bronchodilatorsand ICS. A subgroup of patients will require theaddition of LT receptor antagonists and/or a shortcourse of oral corticosteroids. A negative methacholineinhalation challenge test result essentiallyexcludes asthma from the differential diagnosis ofchronic cough.Allergic Bronchopulmonary AspergillosisAllergic bronchopulmonary aspergillosis(ABPA) is a syndrome most commonly found inpatients with asthma and cystic fibrosis. Patientspresent with wheezing, fleeting, pulmonary infiltrates,and expectoration of brown mucus plugs.ABPA is suspected on clinical grounds, andthe diagnosis is confirmed by radiologic and serologictesting. In the United States, the acceptedcriteria for diagnosis are as follows: (1) asthma, (2)immediate cutaneous reactivity to Aspergillusfumigatus, (3) precipitating (IgG) antibodies toA fumigatus, (4) elevated total serum IgE levels( 1,000 ng/mL), (5) elevated serum IgE antibodiesto A fumigatus, (6) proximal bronchiectasis, (7) infiltratesseen on the chest radiograph, and (8) peripheraleosinophilia with radiographic infiltrates.A radiologic classification of ABPA has beenproposed: ABPA-S (seropositive) and ABPA-CB(central bronchiectasis). Patients with ABPA-Srepresent the earliest form of disease, and theclinical outcome is more favorable. The presenceof central bronchiectasis, often accompanied bycentral mucoid impaction of the bronchus, is anindependent predictor of failure to achieve longtermremission.It is wise to diagnose ABPA early and begin thetreatment of patients to prevent the developmentof bronchiectasis. The mainstay of treatment forABPA is oral corticosteroids to control the asthmasymptoms as well as the immunologic response tothe Aspergillus antigens. Usually, prednisone, 40to 60 mg, is required for complete clearing. The useof low-dose corticosteroids has been associatedwith relapse and corticosteroid dependence, andICS are ineffective. On effective therapy the IgEconcentration should decrease in 1 to 2 months,and a subsequent increase suggests recurrence ofthe disease. The goal is to decrease the IgE level by30 to 50%. Azoles such as itraconazole have alsoshown effectiveness in the treatment of this conditionand usually are reserved for relapses or corticosteroiddependency.Gastroesophageal Reflux and AsthmaThe prevalence of gastroesophageal reflux inasthma is quite high and has been discovered in asmany as 30 to 90% of patients. In patients with<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 99ACC-PULM-09-0302-006.indd 997/10/09 8:05:45 PM


eflux symptoms, 24-h pH monitoring has provenreflux to be present in 80% of patients. Evenwhen reflux symptoms are not present, reflux canbe demonstrated with this study nearly 30% of thetime. Asthmatic patients have been shown to havedecreased lower esophageal sphincter pressureswhen compared with healthy individuals. Medicationssuch as theophylline and -agonists canlower the lower esophageal sphincter as can certainfoods, such as those with a high fat content, chocolate,caffeinated beverages, alcohol, and peppermint.In addition, the administration of prednisoneorally has been shown to increase esophageal acidcontact time.A large study of veterans showed that patientswith significant esophageal disease, such as erosiveesophagitis and/or esophageal stricture, had a 1.5odds ratio of having asthma compared withhealthy subjects. In patients with difficult-tocontrolasthma, reflux should be considered as apossible triggering agent. A favorable asthmaresponse to antireflux medical therapy has beenshown to occur, especially in patients with refluxsymptoms, but also in as many as 24% of patientswith no reflux symptoms. Unfortunately, suchstudies have not shown consistent results. A trialof therapy with a proton-pump inhibitor used inhigh doses (eg, omeprazole, 20 mg bid) may benecessary and has been shown to be more effectivethan the use of histamine type 2 blocker therapy.Consensus has not been reached on the role ofsurgery when medical therapy fails. The followingtwo possible mechanisms of gastroesophagealreflux-induced asthma exacerbation have beenproposed: (1) a vagally mediated reflex caused byacid in the esophagus, which stimulates sensorymucosal receptors, and (2) microaspiration into theupper airway, which stimulates other vagallymediated reflexes. With either mechanism, heightenedcholinergic tone could result in significantbronchoconstriction.Corticosteroid-Resistant AsthmaPatients with severe unremitting asthma symptomsdespite the use of high doses of corticosteroidshave been termed corticosteroid-resistantasthmatic patients when no confounding factorshave been discovered. These factors include pooradherence to recommended therapy, uncontrolledgastroesophageal reflux, unrecognized food orenvironmental antigen exposure, aspirin-sensitiveasthma, ABPA, systemic vasculitis, and fixed airwaydisease secondary to COPD. By definition,such patients have an FEV 1 75% of predicted,with a failure to improve by 15% after an adequatedose and duration of glucocorticoid therapy(eg, 40 mg/d of prednisone for 1 to 2 weeks); areclinically resistant to corticosteroid therapy; andhave a longer duration of asthma, lower morninglung function, and a greater degree of bronchialreactivity than corticosteroid-sensitive asthmaticpatients.Although some abnormalities have been demonstratedin other tissues, such as the cutaneousvasoconstrictor responses to beclomethasonedipropionate, metabolic effects suggesting glucocorticoidresistance in other tissues such as bonehave not been seen, and the hypothalamic pituitaryadrenal axis response in such patients has beenshown to be normal. Glucocorticoid resistanceprobably is produced by a number of heterogeneousmechanisms. It is thought that the majormechanisms for resistance occur distal to thenuclear translocation step. The management ofpatients with glucocorticoid-resistant asthma presentsa difficult challenge as there are no effectiveand well-tolerated alternatives to therapy withsteroids. Methotrexate and cyclosporine have beenused.Fatal and Near-Fatal AsthmaDespite our better understanding of the pathogenesisof asthma and more effective treatmentprograms, the rates of morbidity and mortalityassociated with this condition still remain a problemin most countries in the industrialized world.Many factors have been proposed to explain thisdilemma. These factors include poor compliancewith medication and asthma monitoring; an underestimationby the patient and physician alike of theseverity of the asthma; poor long-term access tomedical care, especially for the poor; and overrelianceon inhaled -agonist therapy. Near-fatal andfatal attacks of asthma are the result of acute hypercarbicrespiratory acidosis; fatalities rarely arecaused by serious cardiac arrhythmias.Despite severe degrees of acidosis, even to pHlevels of 7.0, recovery can be rapid and complete.100 Asthma (Braman)ACC-PULM-09-0302-006.indd 1007/10/09 8:05:45 PM


In fact, there is evidence that when the asthmaattack is sudden and respiratory failure occurswithin a few hours, recovery too is rapid. Improvementoccurs more rapidly than when the asthmaattack has been slowly progressive over days. Theterm sudden asphyxic asthma has been given to suchexplosive attacks. There is evidence that suchpatients are immunohistologically different in thattheir airways have a relative paucity of eosinophilsand a larger number of neutrophils compared withthe classic asthma picture. Typically, in patientswho die of asthma, the lungs are overinflated, andboth large and small airways are filled with plugsconsisting of a mixture of mucus, serum proteins,inflammatory cells, and cell debris. Microscopically,there is infiltration of the airway lumen andwall with eosinophils and mononuclear cells. Thisis accompanied by vasodilation, evidence of microvascularleakage, and epithelial cell sloughing.The most commonly identified causes of nearfatalasthma are listed in Table 4. Risk factors forfatal or near-fatal asthma are as follows: (1) highmedication use (three or more medications), (2)overuse of inhaled -agonist agents, (3) a historyof recurrent hospitalizations, (4) previous occurrencesof life-threatening attacks, and (5) markedfluctuations in morning and evening peak flow ratemeasurements. In addition, investigations haveshown that patients with near-fatal attacks haveabnormal respiratory control mechanisms such asa blunted perception of dyspnea and a reducedhypoxic ventilatory response to hypoxia. They arealso more likely to have a drinking problem, tosmoke cigarettes, and to have family problems.Asthma management plans are becomingpopular, especially with high-risk patients. Studieshave shown that death from asthma can be reducedby the use of a peak flowmeter, a written actionplan, and oral glucocorticoids prior to a severeattack. The risk of death is enhanced by the use ofTable 4. Identified Causes of a Near-Fatal Asthma AttackEnvironmental allergen exposureUpper respiratory viral infectionAspirin and nonsteroidal antiinflammatory drug ingestionProfound emotional upsets-Blocker drugsExposure to indoor air pollutantsAir pollution after a thermal inversiona nebulizer for symptomatic relief in the monthbefore death. Because blood levels of albuterol are2.5 times greater in patients dying of asthma comparedwith control subjects, closer supervision of-agonist use should reduce asthma mortality.Structured educational plans to teach managementmay also offer improved mortality statistics inasthma patients.Annotated BibliographyAbramson MJ, Bailey MJ, Couper FJ, et al. Are asthmamedications and management related to deaths fromasthma? Am J Respir Crit Care Med 2001; 163:12–18A questionnaire was administered to the next-of-kin ofpatients who had died of asthma. Smoking, drinking, andfamily problems were more likely among patients dying ofasthma than in a group of control subjects. The risk of deathwas reduced by the use of a peak flowmeter in the previousyear. The use of oral glucocorticoids also appeared to reducethe risk of death. The risk of death was increased by the useof a nebulizer for symptomatic relief in the previous month.Blood albuterol levels were considerably greater in patientsdying of asthma, suggesting that albuterol overuse is amarker for asthma mortality.Adams RJ, Fuhlbrigge A, Guilbert T, et al. Inadequateuse of asthma medication in the United States: resultsof the asthma in America National Population survey.I Allergy Clin Immunol 2002; 110:58–64This study is based on a cross-sectional national telephonesurvey. It reveals that appropriate therapy for asthma remainsinadequate. Antiinflammatory therapy is still greatly underused,especially in socioeconomically disadvantaged groups.Agarwal R, Gupta D, Aggarwal AN, et al. Allergic bronchopulmonaryaspergillosis: lessons from 126 patientsattending a chest clinic in north India. Chest 2006; 130:442–448This article describes the clinical findings in 126 cases ofABPA. A total of 27% of the patients had ABPA-S (serologicpositivity alone) and the remainder had ABPA-CB (with centralbronchiectasis); many of the latter had other radiologicfindings such as patchy infiltrates and atelectasis. This studyshowed no significant difference between the stages of ABPAand the duration of illness, the severity of asthma, and theserologic findings (ie, absolute eosinophil count, total IgElevels, and IgE levels for A fumigatus). All patients wentinto “remission” at 6 weeks on corticosteroids. Twenty-fivepatients had a “relapse” during the course of their treatment.One hundred nine patients had “complete remission,” 17patients were classified as having “glucocorticoid-dependent<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 101ACC-PULM-09-0302-006.indd 1017/10/09 8:05:46 PM


ABPA,” and 7 patients were classified as having “end-stageABPA.”Barnes PJ. The cytokine network in asthma and chronicobstructive pulmonary disease. J Clin Invest 2008;118:3546–3556This article discusses how cytokines orchestrate the chronicinflammation and structural changes of the respiratory tractin both asthma and COPD and have become important targetsfor the development of new therapeutic strategies inthese diseases.Barnes PJ, Adcock IM. How do corticosteroids work inasthma? Ann Intern Med 2003; 139:359–370Corticosteroids suppress a number of pathways of chronicinflammation in asthma. Inflammatory genes are regulated byproinflammatory transcription factors such as nuclear factorκBand activator protein-1. These factors can switch on genetranscription. Corticosteroids act to suppress the action ofthese factors. This review discusses the mechanisms involved.Bateman ED, Boushey HA, Bousquet J, et al. Can guideline-definedasthma control be achieved? Am J RespirCrit Care Med 2004; 170:836–844This is the Gaining Optimal Asthma Control study, whichexamined the use of inhaled ICS/LABAs such as fluticasoneand salmeterol for the treatment of asthma. This study specificallylooked at the guideline-defined approach, using doseescalation as necessary to achieve improved asthma care. Thestudy looked at increasing the dose of ICS alone or increasingthe dose of ICS and coupling it with a LABA for asthmacontrol. Total control was achieved across all levels of treatmentin 41% vs 28% of patients, respectively, at 1 year forICS-LABA therapy vs ICS therapy alone. Asthma was wellcontrolled (ie, nearly totally controlled) in 71% vs 59%of patients, respectively. This study confirmed that guideline-derivedasthma control can be achieved in a majority ofpatients; however, the asthma of a significant percentage ofpatients cannot even be well controlled with this combinationtherapy.Berry MA, Hargadon B, Shelley M, et al. Evidenceof a role of tumor necrosis factor alpha in refractoryasthma. N Engl J Med 2006; 354:697–708A double-blind, placebo-controlled study was conducted inpatients with severe asthma to determine the effect of blockingTNF-α with etanercept. When etanercept was added tousual care, significant improvements in FEV 1and BHR werefound. Further studies are needed because the study groupwas quite small (ie, 10 patients). Blocking TNF-α may proveto be an effective disease-modifying strategy for asthma.Blaiss MS. Management of rhinitis and asthma in pregnancy.Ann Allergy Asthma Immunol 2003; 90(suppl):16–22This report from the FDA, the American College of Allergy,Asthma, and Immunology, and the American College ofObstetricians and Gynecologists underscores the importantdifferences in pregnant and nonpregnant patients. For pregnantwomen with asthma, inhaled cromolyn should be thefirst-line therapy, followed by inhaled budesonide if symptomsworsen. Other agents, such as salmeterol, LT modifiers,and the newer ICS, may be considered if the patient had agood response to these drugs before pregnancy.Blitz M, Blitz S, Beasely R, et al. Inhaled magnesiumsulfate in the treatment of acute asthma. CochraneDatabase Syst Rev. Issue 4, 2005This Cochrane Database review concludes that nebulizedinhaled magnesium sulfate in addition to a -agonist for thetreatment of an acute asthma exacerbation appears to havebenefits with respect to improved pulmonary function. Inpatients with severe asthma, there is a trend toward benefitin terms of hospital admissions.Borish L. Allergic rhinitis: systemic inflammation andimplications for management. J Allergy Clin Immunol2003; 112:1021–1031This review of the inflammatory mechanisms of allergic rhinitisreminds us that some inflammatory cells and mediatorsmay pass into the bloodstream and create a systemic reaction.It offers justification for the use of LT modifiers to treatthe systemic antiinflammatory effects of rhinitis.Boulet LP, Lemière C, Archambault F, et al. Smokingand asthma: clinical and radiologic features, lung function,and airway inflammation. Chest 2006; 129:661–668This study shows that compared with nonsmoking asthmapatients, smoking asthma patients have features similar towhat could be found in patients with early stages of COPD.Braman SS, Hanania NA. Asthma in older adults. ClinChest Med 2007; 28:685–702This review underscores the differences between asthma inthe young and elderly. Airflow obstruction in the elderlyis often difficult to reverse, and many elderly patients havefixed airway disease, even with maximum therapy. Pitfallsin asthma therapy in treating the elderly are discussed aswell as differences in pathogenesis.Chalmers GW, Macleod KJ, Little SA, et al. Influence ofcigarette smoking on inhaled corticosteroid treatmentin mild asthma. Thorax 2002; 57:226–230This study showed that active cigarette smoking impairs theefficacy of short-term ICS treatment in patients with mildasthma. This study obviously has important implications forpatients with mild asthma who smoke.Chung KF. Individual cytokines contributing to asthmapathophysiology: valid targets for asthma therapy?Curr Opin Invest Drugs 2003; 4:1320–1326102 Asthma (Braman)ACC-PULM-09-0302-006.indd 1027/10/09 8:05:46 PM


This article reviews the evidence for cytokine-targeted therapy.IL- 4, IL-5, and IL-13 contribute to the inflammation ofasthma; the inhibition of these cytokines might be useful inthe future.Cochrane MG, Bala MV, Downs KE, et al. Inhaled corticosteroidsfor asthma therapy: patient compliance,devices, and inhalation technique. Chest 2000; 117:542–550This article presents a systematic literature review concerningcompliance, inhalation techniques, and ICS devices forthe treatment asthma.Compalati E, Penagos M, Tarantini F, et al. Specificimmunotherapy for respiratory allergy: state of theart according to current meta-analyses. Ann AllergyAsthma Immunol 2009; 102:22–28A meta analyses of studies that met the inclusion criteria,five evaluating sublingual immunotherapy and two evaluatingsubcutaneous immunotherapy, reported a reduction insymptom and medication scores. The authors conclude thataccording to evidence-based criteria, specific immunotherapycan be recommended for the treatment of respiratory allergybecause of its efficacy in reducing asthma and rhinitis symptomsand that future methodologic approaches that considersafety and costs should corroborate this positive evaluation.Cote J, Bowie DM, Robichaud P, et al. Evaluation of twodifferent educational interventions for adult patientsconsulting with an acute asthma exacerbation. Am JRespir Crit Care Med 2001; 163:1415–1419These authors show that a structured educational planemphasizing self-capacity to manage asthma exacerbationand the use of peak flow rate reduce the morbidity of patientswith asthma.Creticos PS, Reed CE, Norman PS, et al. Ragweedimmunotherapy in adult asthma. N Engl J Med 1996;334:501–506This article examined the efficacy of immunotherapy forasthma in ragweed-sensitive patients. Patients were followedup for 2 years of treatment. Although immunotherapyfor adults in this study had positive effects and producedobjective measures of improvement of asthma and allergyduring the first year of therapy, the improvement was notsustained during the second year.Currie GP, Devereux GS, Lee DK, et al. Recent developmentsin asthma management. BMJ 2005; 330:585–589This is an evidence-based review of developments from thepast few years in asthma management. It is a comprehensiveliterature search up to 2005 on a number of approaches toasthma therapy including allergen avoidance, dietary manipulation,asthma action plans, and pharmacologic management.It is an up-to-date review with excellent references.Dicpinigaitis PV. Cough: 4: cough in asthma and eosinophilicbronchitis. Thorax 2004; 59:71–72This author discusses the difference between CVA and eosinophilicbronchitis. The main distinction is that the former isassociated with BHR. Eosinophilic bronchitis patients havea negative methacholine challenge result. Evidence suggeststhat LT receptor antagonists might be useful in treatingCVA. Eosinophilic bronchitis responds to therapy with ICS.Dicpinigaitis PV. Chronic cough due to asthma: <strong>ACCP</strong>evidence-based clinical practice guidelines. Chest 2006;129(suppl):75S–79SAsthma should be considered a potential etiology in anypatient with chronic cough. A subgroup of asthmaticpatients presents with CVA. This group of patients has onlycough with no other asthma symptoms. The cough shouldrespond to standard antiasthma therapy. This article is takenfrom the American College of Chest Physicians Evidence-Based Practice Guidelines on Cough, which was publishedin 2006.Donohue JF. The expanding role of long-acting -agonists.Chest 2000; 118:283–285The author summarizes the use of salmeterol, an LABA, forhospital inpatients. He first provides evidence that salmeterolusers are at no greater risk for asthma-related hospitalizations,emergency department visits, or ICU stays. Thetreatment of an acute asthma attack is not compromised bymaintenance therapy with this long-acting agent. In addition,salmeterol can be safely used for the treatment of acuteexacerbation of COPD along with albuterol. Minimal cardiaceffects are seen in most patients; however, there are veryfew data on this subject in patients with preexisting heartdisease. Careful monitoring is urged if both agents are used.Drazen J. Asthma and the human genome project: summaryof the 45th annual Thomas L. Petty Aspen LungConference. Chest 2003; 123(suppl):447S–449SThis article discusses more recent thinking on the geneticsof asthma.Forbes L. Do exogenous estrogens and progesteroneinfluence asthma? Thorax 1999; 54:265–267This interesting article discusses the influence of estrogensand progesterone on asthma. The author believes that therewas no convincing evidence to suggest that hormonal contraceptiveseither exacerbate or improve asthmatic symptoms.Their use in patients with severe asthma or those withpremenstrual exacerbations of asthma will require futureclinical trials. The author concludes that, “[a]t the present[time,] there is not enough evidence to suggest any change tocurrent prescribing practice.”Gern JE. Viral respiratory infection and the link toasthma. Pediatr Infect Dis J 2004; 23(suppl):S78–S86<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 103ACC-PULM-09-0302-006.indd 1037/10/09 8:05:46 PM


Acute asthma symptoms have been correlated with a varietyof viral pathogens. Most commonly, it is respiratory syncytialvirus in infancy and rhinovirus in older children. Thisarticle enhances our knowledge of how viruses can adverselyaffect lung or immune development in asthma patients.Grau RG. Churg-Strauss syndrome: 2005–2008 update.Curr Rheumatol Rep. 2008; 10:453–458This article provides a recent update on this disease.Greenstone IR, Ni Chroinin MN, Masse V, et al. Combinationof inhaled long-acting 2-agonists and inhaledsteroids versus higher dose of inhaled steroids in childrenand adults with persistent asthma. CochraneDatabase Issue 4, 2005This Cochrane Database review presents data on the use ofhigh-dose ICS vs ICS plus LABA. It concludes that the combinationof LAB A and ICS leads to greater improvement inlung function, symptoms, and the use 2-agonists for rescue.There are also fewer withdrawals from poor asthma controlwhen the combination therapy is used. However, in headto-headcomparison, the combination of LABA and ICS vsgreater-dose ICS showed no significant differences for the preventionof exacerbations requiring systemic corticosteroids.Harding SM. Recent clinical investigations examiningthe association of asthma and gastroesophageal reflux.Am J Med 2003; 115(suppl):39S–44SThis article reviews the association between gastroesophagealreflux disease and asthma. Esophageal acid may alter BHR.Medical and surgical therapy for gastroesophageal refluxdisease may improve asthma outcomes in selected patients.Holgate ST, Chuchalin AG, Herbert J, et al. Efficacyand safety of a recombinant anti-immunoglobulin Eantibody (omalizumab) in severe allergic asthma. ClinExp Allergy 2004; 34:632–638This study investigated the role of omalizumab, the IgGmonoclonal antibody, in patients with severe allergic asthmathat is refractory to corticosteroid therapy. In this 16-weekstudy, the primary end point was corticosteroid withdrawal.Anti-IgE therapy led to greater reductions in ICS dosagecompared with placebo. It also led to significantly improvedsymptoms, lung function, quality of life, and reduced controlor medication use.Holgate ST, Davies DE, Puddicombe S, et al. Mechanismsof airway epithelial damage: epithelial-mesenchymalinteractions in the pathogenesis of asthma. EurRespir J 2003; 44(suppl):24S–29SThe authors provide evidence that asthma is a disease of airwaywall restructuring that engages activation of the EMTU.Activation of this unit provides a stimulus for altered airwaywall structure and function and may explain the decline inlung function observed over time with asthma. In asthmapatients, the epithelium shows considerable evidence ofinjury and activation as well as goblet-cell metaplasia. Thealtered epithelium becomes an important source of mediators,chemokines, and growth factors that sustain ongoinginflammation. Injury to the epithelial cells is thought toresult in increased release of fibroproliferative and fibrogenicgrowth factors. This results in the deposition of interstitialcollagens in the lamina reticularis, the deposition of collagenin the submucosa, smooth-muscle hyperplasia, and microvascularneuronal proliferation.Israel E, Deykin A, Mitra N, et al. Use of regularlyscheduled albuterol treatment in asthma: genotypestratified,randomised, placebo-controlled cross-overtrial. Lancet 2004; 364:1505–1512Genetic polymorphisms of the -adrenergic receptor havebeen described. Homozygosity for arginine (Arg/Arg)rather than glycine (Gly/Gly) predicts adverse events withasthma. This is a prospective clinical trial in patients withmild asthma that show that when albuterol was kept to aminimum, patients with Arg/Arg genotype had an increasein morning peak flow rate, whereas the change in patientswith Gly/Gly genotype was not significant. Patients withGly/Gly genotype using regularly scheduled albuterol hadan improvement in PEFR. By contract, the Arg/Arg genotypegroup had lower morning peak flow rates during thetreatment with albuterol. This suggests that bronchodilatortreatments avoiding albuterol may be appropriate forpatients with Arg/Arg genotype.Jones SL, Kittelson J, Cowan JO, et al. The predictivevalue of exhaled nitric oxide measurements in assessingchanges in asthma control. Am J Respir Crit CareMed 2001; 164:738–743ENO has been confirmed as a marker of airway inflammation.It is present in greater concentrations in steroid-naiveasthmatic patients compared with healthy control individuals.This group of investigators measured exhaled NO in75 asthmatic subjects with mild-to-moderate asthma. ICStherapy was withdrawn and loss of asthma control ensued.There were highly significant correlations between exhaledNO and symptoms, sputum eosinophils, and BHR. ExhaledNO had an 80 to 90% predictive value for the loss of asthmacontrol. It has also been previously confirmed that exhaledNO is closely related to several other markers of asthmacontrol such as asthma symptoms, dyspnea score, and dailyuse of rescue medication. This technique is noninvasive andcan be performed repeatedly. It appears that a change in theexhaled NO levels measured over time could be useful inmonitoring asthma control.Källén B. The safety of asthma medications duringpregnancy. Exp Opin Drug Saf 2007; 6:15–26104 Asthma (Braman)ACC-PULM-09-0302-006.indd 1047/10/09 8:05:46 PM


This article reviews the literature on asthma or use of antiasthmaticdrugs during pregnancy.Lee-Wong M, Dayrit FM, Kohli AR, et al. Comparisonof high-dose inhaled flunisolide to systemic corticosteroidsin severe adult asthma. Chest 2002; 122:1208–1213High-dose ICS are as effective as systemic corticosteroids inthe 7 days after hospital admission for severe asthma.Le Gall C, Pham S, Vignes S, et al. Inhaled corticosteroidsand Churg-Strauss syndrome: a report of fivecases. Eur Respir J 2000; 15:978–981The association of Churg-Strauss syndrome with LPMshas been well documented. This group presents five cases ofChurg-Strauss syndrome in severe steroid-dependent asthmaticsubjects in which the use of ICS allowed systemic corticosteroidwithdrawal. It gives further evidence for systemicsteroid withdrawal as a cause of Churg-Strauss syndromesince these patients were not receiving LPMs.Leone FT, Fish JE, Szefler SJ, et al. Systematic reviewof the evidence regarding potential complications ofinhaled corticosteroid use in asthma: collaborationof American College of Chest Physicians, AmericanAcademy of Allergy, Asthma, and Immunology, andAmerican College of Allergy, Asthma, and Immunology.Chest 2003; 124:2329–2340This multisociety report discusses the complications of ICSuse in asthma patients. ICS use is not associated with areduction in bone density in children with asthma. This canbe said in adults but less conclusively. The risk of cataractsmay be negligible in young asthmatic patients but it maybe increased in older patients. The risk of glaucoma may besmall, and further study is warranted. Therapy with ICS inchildren is associated with a decrease in short-term growthrates in children, but the overall effect is small and may not besustained with long-term therapy. The risk of skin thinningand easy bruising is elevated in patients receiving ICS.Lima JJ, Blake KV, Tantisira KG, et al. Pharmacogeneticsof asthma. Curr Opin Pulm Med 2009; 15:57–62This article discusses genetic variations in patient responsesto drugs, including topics of -agonist polymorphisms andresponses to corticosteroids and LT antagonists.Malo JL, Lemiere C, Gautrin D, et al. Occupationalasthma. Curr Opin Pulm Med 2004; 10:57–61This article reviews some of the more recent informationregarding the pathophysiology, diagnosis, and preventionof occupational asthma. High-molecular-weight inhaledantigens work through an IgE mechanism. The second categoryof agents that induce occupational asthma are lowmolecular-weightagents. They are generally chemicals,and the pathophysiology is not as well understood. There islittle evidence for an IgE-mediated mechanism. The effectsof exposure to isocyanates, the chemical most frequentlyresponsible for occupational asthma, may work through theoverproduction of matrix metal-loproteinase-9, which maycause symptoms and BHR. There is a predominant sputumneutrophilia after exposure to isocyanates. Specific inhalationchallenges represent the “gold standard” for the diagnosisof occupational asthma. When not available, the serialrecording of PEFRs has been advocated. The risk for men isgreatest among bakers, laundry workers, shoemakers, animalskin and hide workers, and metal plating and coatingworkers. For women, the greatest risks are for shoemakers,railway station personnel, jewelry engravers, engine roomcrews, foundry molders, round-timber workers, and bakers.Results indicate that respiratory symptoms and airwayresponsiveness to methacholine persist in subjects who areremoved from exposure to the isocyanates for 10 years.Matricardi PM, Rosmini F, Panetta V, et al. Hay feverand asthma in relation to markers of infection in theUnited States. J Allergy Clin Immunol 2002; 110:381–387This study shows that hay fever and asthma were less commonin participants who were seropositive for hepatitis A,Toxoplasma gondii, and herpes simplex virus 1 comparedwith seronegative participants. This finding suggests thatthe acquisition of certain infections, primarily food-borneand orofecal infections, is linked with a lower probabilityof having hay fever and asthma. These data are from theNational Health and Nutrition Examination Study III.McFadden ER Jr. Acute severe asthma. Am J RespirCrit Care Med 2003; 168:740–759This review discusses the current therapeutic options for theacute exacerbation of asthma. Seventy to 80% of patients inemergency departments clear within 2 h with standard care.The relapse rate is between 7% and 15%. The 20 to 30% ofpatients who are resistant to -agonist therapy in the emergencydepartment slowly reverse their condition over 36 to48 h but require intensive treatment with corticosteroids.McFadden ER Jr., Zawadski DK. Vocal cord dysfunctionmasquerading as exercise-induced asthma. Am JRespir Crit Care Med 1996; 153:942–947This article discusses an interesting group of patients andextends the syndrome of vocal cord dysfunction to those whohave only exercise-induced symptoms. Although the combinationof exertion and wheezing suggests EIA, when patientspresented in this article underwent clinical and physiologicevaluations, including bronchoprovocation testing, asthmawas ruled out. Laryngoscopy was helpful in diagnosingvocal cord dysfunction.McParland BE, Macklem PT, Pare PD. Airway remodeling:friend or foe? J Appl Physiol 2003; 95:426–434<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 105ACC-PULM-09-0302-006.indd 1057/10/09 8:05:46 PM


In asthmatic patients, chronic inflammation of the airwaywall results in an abnormal repair process. Changes thatoccur include subepithelial fibrosis, smooth-muscle hyperplasiaand hypertrophy myofibroblast hyperplasia, epithelialhypertrophy, and mucus gland and goblet-cell hyperplasia.Collectively, these changes are called airway remodeling.These structural changes are thought to be detrimental asthey contribute to fixed airway narrowing. Thickening ofthe inner airway wall may also amplify the degree of luminalnarrowing for a given degree of smooth-muscle shortening.This might result in an exaggerated narrowing ofthe airway after a bronchoconstricting stimulus. It hasbeen argued that some aspects of remodeling have beneficialeffects: stiffening the airway may result in decreased compressibilityallowing the airways to better resist dynamiccompression; and extra connective tissue, surroundingthe smooth-muscle cells may provide a radial constraint tomaximal shortening.Mitsunobu F, Tanizaki Y. The use of computed tomographyto assess asthma severity. Curr Opin AllergyClin Immunol 2005; 5:85–90Chronic inflammation in asthma can lead to airway remodelingand airway narrowing. This can be quantified by theuse of CT scanning. High-resolution CT scanning is a usefultool for imaging the airways of asthmatic patients. Asthmaticpatients have thicker airways compared with controlsubjects. The thickness is related to disease severity, airflowobstruction, and airway reactivity. This is a new tool forassessing asthma severity.Moore WC. Update in asthma 2007. Am J Respir CritCare Med 2008; 177:1068–1073This review discussed the articles published in 2007 in thisrespiratory journal.Morgan WJ, Crain EF, Gruchalla RS, et al. Results ofa home-based environmental intervention amongurban children with asthma. N Engl J Med 2004; 351:1068–1080This group looked at a cohort of children with atopic asthma.A controlled trial of environmental intervention to reduceallergies and environmental smoke that included the reductionof exposure to indoor allergens, including cockroach anddust mite allergens, resulted in reduced asthma-associatedmorbidity.Morgan WJ, Stern DA, Sherrill DL, et al. Outcome ofasthma and wheezing in the first 6 years of life: followupthrough adolescence. Am J Respir Crit Care Med2005; 172:1253–1258The authors found that in children who start having asthmalikesymptoms before the preschool years, the prevalence ofwheezing and the levels of lung function are established byage 6 years and do not appear to change significantly by age16 years. Transient early wheezers (ie, children who wheezeduring early life but who were not wheezing at age 6 years)were no more likely to wheeze after age 6 years than healthychildren. These children have lower lung function before thetransient wheezing and continue to have lower levels of lungfunction through adolescence. It is known from other studiesthat COPD is more likely to develop during later adult yearsin individuals who enter adult life with lung function deficits.Whether early wheezing predicts COPD is not known.The study also showed that deficits in lung function that areobserved in children with asthma are not the consequenceof ongoing disease, but rather are due to changes before theage of 6 years. Although the transient wheezers have lowerlung function that is present as early as the first 3 monthsof life, they do not have the elevations in serum IgE levels,eosinophilia, and skin test reactivity to aeroallergens that areseen in those with persistent wheezing from early childhoodto later childhood.Murphy VE, Clifton VL, Gibson PG. Asthma exacerbationsduring pregnancy: incidence and associationwith adverse pregnancy outcomes. Thorax 2006;61:169–176This article reminds us that exacerbations during pregnancyoccur primarily in the late second trimester. The major triggersare viral infection and nonadherence to ICS therapy.Women who have an exacerbation during pregnancy are ata significantly increased risk of having a low-birth-weightbaby compared to women without asthma.Murugan A, Prys-Picard C, Calhoun WJ. Biomarkersin asthma. Curr Opin Pulm Med. 2009; 15:12–18This article reviews the current literature about the use ofbiomarkers for the diagnosis and monitoring of asthma. Todate, evidence suggests that exhaled NO may become a clinicaltest for asthma control. The article discusses the limitationof this measurement and recent clinical trials showingsome promise.Nathan RA, Sorkness CA, Kosinski M, et al. Developmentof the asthma control test: a survey for assessingasthma control. J Allergy Clin Immunol 2004; 113:59–65The NAEPP guidelines have concentrated on a classificationof asthma that is static and does not really discuss asthmacontrol and response to therapy. This asthma control testpresents a scoring system for asthma control that includesfive items for gauging the status of current asthma. Theseinclude asthma symptoms, the use of rescue medications,and questions regarding the impact of asthma on everydayfunctioning. This is a brief, easy-to-administer, patientbasedindex of asthma control.106 Asthma (Braman)ACC-PULM-09-0302-006.indd 1067/10/09 8:05:47 PM


National Asthma Education and Prevention Program.Expert panel report 3: guidelines for the diagnosis andmanagement of asthma. Bethesda, MD: National Institutesof Health, August 2007This guideline should be on the bookshelf of every pulmonologist.This particular report is the most recent update of thefirst report published in 1991. This updated report is postedon the Internet (http://www.nhlbi.NAEPP.gov/guidelines/asthma/index.htm). Key differences in the report include thefollowing comments: (1) The critical role of inflammationhas been further substantiated, but evidence is emerging forconsiderable variability in the pattern of inflammation, thusindicating phenotypic differences that may influence treatmentresponses. (2) Gene-by-environmental interactions areimportant to the development and expression of asthma. Ofthe environmental factors, allergic reactions remain important.Evidence also suggests a key and expanding role forviral respiratory infections in these processes. (3) The onset ofasthma for most patients begins early in life, with the patternof disease persistence determined by early, recognizable riskfactors, including atopic disease, recurrent wheezing, and aparental history of asthma. (4) Current asthma treatmentwith antiinflammatory therapy does not appear to preventprogression of the underlying disease severity. (5) The keyelements of assessment and monitoring are refined to includethe separate, but related, concepts of severity, control, andresponsiveness to treatment. Classifying severity is emphasizedfor initiating therapy; assessing control is emphasizedfor monitoring and adjusting therapy. Asthma severity andcontrol are defined in terms of two domains: impairment andrisk. (6) The distinction between the domains of impairmentand risk for assessing asthma severity and control emphasizesthe need to consider separately asthma’s effects on qualityof life and functional capacity on an ongoing basis (ie,in the present) and the risks it presents for adverse eventsin the future, such as exacerbations and progressive loss ofpulmonary. (7) Emphasis on the many potential points ofcare and sites available in which to provide asthma education,including review of new evidence regarding the efficacyof asthma self-management and education outside the usualoffice setting. (8) Greater emphasis on the two aspects of thewritten asthma action plan—(a) daily management and (b)how to recognize and handle worsening asthma. Use of theterminology “written asthma action plan” encompassesboth aspects. This change addresses confusion over the previousguidelines’ use of different terms. One term is nowused for the written asthma action plan, although in somestudies cited, investigators may have used a variation of thisterm. (9) New sections on the impact of cultural and ethnicfactors and health literacy that affect delivery of asthmaself-management education. (10) Information about asthmamedications has been updated based on review of evidencepublished since 1997. This updated report (Expert PanelReport 3) continues to emphasize that the most effectivemedications for long-term therapy are those shown to haveantiinflammatory effects. (11) New medications—immunomodulators—areavailable for long-term control of asthma.(12) New data on the safety of LABAs are discussed, andthe position of LABA in therapy has been revised). The mostsignificant difference is that for youths 12 years of age andadults who have moderate persistent asthma or asthma inadequatelycontrolled on low-dose ICS, the option of increasingthe dose of medium-dose ICS should be given equal weightto the option of adding LABA to low-dose ICS. (13) Theestimated clinical comparability of different ICS preparationshas been updated. The significant role of ICS in asthmatherapy continues to be supported.Nelson HS, Weiss SC, Bleecker ER, et al. The SalmeterolMulticenter Asthma Research Trial: a comparisonof usual pharmacotherapy for asthma or usual pharmacotherapyplus salmeterol. Chest 2006; 129:15–26The safety of salmeterol vs placebo was studied in a 28-weekrandomized observational study. In a telephone follow-up, itwas found that there were small but statistically significantiincreases in respiratory-related and asthma deaths in thepopulation of patients receiving placebo. Subgroup analysissuggested that the risk was greater in African Americans.A post hoc analysis was conducted to explore the effect ofICS use on the results of this trial, often called the SMARTtrial. The number of deaths that occurred in patients whowere not receiving ICS at baseline was greater in the salmeterolgroup, suggesting a relationship. The study wasnot designed to evaluate the effects of ICS therapy; therefore,adequate conclusions could not be drawn. The studyraised the alert that therapy with LABAs may have adverseeffects in some populations. Further studies need to clarifythe veracity of these findings. In the meantime, the US Foodand Drug Administration has placed a warning on the packageinsert because of this study.Newman KB, Mason UG III, Schmaling KB. Clinicalfeatures of vocal cord dysfunction. Am J Respir CritCare Med 1995; 152:1382–1386This article reviews vocal cord dysfunction in asthmatic andnonasthmatic patients. All patients had laryngoscopic evidenceof paradoxical vocal cord motion with inspiratory and/or early expiratory vocal cord adduction. The patients werepredominantly women and had a misdiagnosis of asthma foran average of almost 5 years. Many were steroid dependent,and medical utilization was enormous. Twenty-eight percentof the patients had been intubated. This study helps to<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 107ACC-PULM-09-0302-006.indd 1077/10/09 8:05:47 PM


define the historical and clinical features of vocal cord dysfunction.Ngoc LP, Gold DR, Tzianabos AO, et al. Cytokines,allergy, and asthma. Curr Opin Allergy Clin Immunol2005; 5:161–166This article is a review of more recent articles on the relationship of cytokines to allergy and asthma. It discusses theimmune mechanisms involved in the phenotypic expressionof allergic diseases, including the allergen-specific T-helpertype 2 responses that release IL-4, IL-13, and IL-5.O’Bryne PM, Bisgaard H, Godard PP, et al. Budesonide/formoterol combination therapy as both maintenanceand reliever medication in asthma. Am J Respir CritCare Med 2005; 171:129–136This group hypothesized that the conditions of patientsreceiving low-dose ICS/LABA therapy whose asthma wasnot under good control would be improved with as-neededICS/LABA (ie, budesonide/formoterol) therapy rather thanshort-acting 2-agonist reliever medication. In a doubleblindrandomized study of 2,760 patients with asthma,the budesonide/formoterol maintenance therapy plus reliefmedication prolonged the time to first exacerbations andresulted in a 45% lower exacerbation risk vs therapy withbudesonide/formoterol plus a short-acting 2-agonist. Thetherapy also improved symptoms, nighttime awakenings,and lung function.O’Byrne PM, Inman MD. Airway hyperresponsiveness.Chest 2003; 123(suppl):411S–416SThis is a review of the features of airway hyperresponsiveness,a feature that is constant in asthma. Measurements ofairway responsiveness are useful in making a diagnosis ofasthma, especially when symptoms are present and there isno airflow obstruction.O’Byrne PM, Pedersen S. Measuring efficacy andsafety of different inhaled corticosteroid preparations.J Allergy Clin Immunol 1998; 102:879–886This is a review of ICS therapy in patients with asthma.An excellent section on potential systemic effects is alsoincluded. Differences in pharmacologic properties amongthe ICS are discussed.Panettieri RA Jr. Airway smooth muscle: an immunomodulatorycell. J Allergy Clin Immunol 2002; 110:S269–S274The airway smooth muscle is a major effector cell of asthmathat is responsible for bronchomotor tone. Some evidencesuggests that it also may be active in secreting cytokinesand chemokines and also may express cellular adhesion molecules.Growth factors may be formed from airway smoothmusclecells, and hence, an autocrine-like proliferativeresponse may occur.Pearlman DS. Pathophysiology of the inflammatoryresponse. J Allergy Clin Immunol 1999; 104:S132–S137This article provides a review of the pathophysiology of theatopic inflammatory response. An in-depth discussion ofantigen sensitization and the subsequent cytokine responseis provided.Phipps P, Garrard CS. The pulmonary physician incritical care: acute severe asthma in the intensive careunit. Thorax 2003; 58:81–88This review discusses the approach to status asthmaticusin the ICU, including a good discussion on modes of ventilation,ventilator setting, and the use of extrinsic positiveend-expiratory pressure. Other modalities, including theuse of helium, magnesium sulfate, and inhalation anestheticagents, also are discussed.Restrepo RD, Peters J. Near-fatal asthma: recognitionand management. Curr Opin Pulm Med 2008; 14:13–23This review discusses recent advances in our understandingof the pathophysiology, diagnosis, and treatment of nearfatalasthmaRingdal N, Chuchalin A, Chovan L, et al. Evaluationof different inhaled combination therapies (EDICT):a randomised, double-blind comparison of Seretide(50/250 microg bd Diskus vs formoterol (12 microgbd) and budesonide (800 microg bd) given concurrently(both via Turbuhaler) in patients with moderate-to-severeasthma. Respir Med 2002; 96:851–861This study was the first comparison of LABA/ICS combinationtherapies. Salmeterol plus fluticasone was at least aseffective as formoterol plus budesonide in improving pulmonaryfunction despite a lower corticosteroid dose. The formercombination also significantly reduced exacerbation ratesand nocturnal symptoms. More studies are needed to confirmthese results.Rodrigo GJ, Rodrigo C, Pollack CV, et al. Use of heliumoxygenmixtures in the treatment of acute asthma: asystematic review. Chest 2003; 123:891–896This study discusses the use of helium in the treatment ofacute asthma and concludes that evidence does not providesupport for heliox mixtures in patients with acute severeasthma. Conclusions, however, are based on small studies.Salvi SS, Krishna MT, Sampson, et al. The anti-inflammatoryeffects of leukotriene-modifying drugs andtheir use in asthma. Chest 2001; 119:1533–1546This article reviews the beneficial effects of LT-modifyingdrugs in the management of all grades of asthma severity.It concludes that certain patient groups, such as those withEIA or aspirin-induced asthma, may be particularly suitablefor such therapy.108 Asthma (Braman)ACC-PULM-09-0302-006.indd 1087/10/09 8:05:47 PM


Schaub B, von Mutius E. Obesity and asthma, whatare the links? Curr Opin Allergy Clin Immunol 2005;5:185–193This article reviews the relationship between obesity andasthma. A number of prospective studies have shown thatweight gain can antedate the development of asthma. Severalhypotheses have been proposed to explain the epidemiologicassociations and are discussed in this article, including airwaymechanics, influences on immune responses, and hormonalinfluences. Particularly, the diminished tidal lungexpansion in overweight individuals may partially accountfor the findings. Studies in animal models have indicatedthat elevations of IL-6 levels may contribute to and up-regulateinflammation in the airways. The levels of the hormoneleptin, a member of the IL-6 family, are increased in obesepersons. Leptin may have an effect on inflammation by promotingthe release of IL-6 from macrophages in lymphocytes.Other potential factors also are discussed.Sciurba FC. Physiologic similarities and differencesbetween COPD and asthma. Chest 2004; 126(suppl):117S–124SThere are significant differences in the physiologic consequencesof COPD in asthma, as described in this article.However, it also takes a closer inspection of the literatureand reveals significant overlaps between the two conditions.It reminds us that there is a subgroup of COPD patients thathas features indistinguishable from asthma.Sears MR, Greene JM, Willan AR, et al. A longitudinal,population-based, cohort study of childhood asthma followedto adulthood. N Engl J Med 2003; 349:1414–1422The outcome of childhood asthma is reported in this populationbasedstudy. The risk factors for the persistence and relapse ofasthma into adulthood are described. In this study, more thanone in four children had wheezing that persisted from childhoodto adulthood or that relapsed after remission. The factors predictingpersistence or relapse were sensitization to house dustmites, airway hyperresponsiveness, female sex, smoking, andearly age at onset. These findings, together with persistentlylow lung function, suggest that outcomes in adult patientswith asthma may be determined primarily in early childhood.Shore SA, Drazen JM. -Agonists and asthma: toomuch of a good thing? J Clin Invest 2003; 112:495–497This interesting article discusses a theory of why a paradoxicalresponse sometimes develops in asthmatic patients whoreceive long-term treatment with short-acting -agonists.Sin DD, Sutherland ER. Obesity and the lung: 4.Obesity and asthma. Thorax 2008; 63:1018–1023This review examines the clinical and epidemiological relationshipbetween obesity and asthma and the purportedmechanisms that may link these two processes together.Smith AD, Cowan J, Brassett K, et al. Exhaled nitricoxide: a predictor of steroid response. Am J Respir CritCare Med 2005; 172:453–459Patients with asthma symptoms who had not previouslyreceived a diagnosis of asthma were studied to determinewhether exhaled NO measurements would predict steroidresponsivephenotypes. Levels of exhaled NO proved to bemore sensitive and specific for identifying steroid respondersthan was FEV 1percentage of predicted, peak flow variability,or tests of BHR.Smith AD, Cowan J, Brassett K, et al. Use of inhalednitric oxide measurements to guide treatment ofchronic asthma. N Engl J Med 2005; 352:2163–2173These authors determined whether measurements of thefraction of exhaled NO constitute a noninvasive marker ofinflammation that may be a useful alternative for the adjustmentof ICS treatment. A threshold of 15 parts per billionof exhaled NO was used to increase the dose of ICS. Theyfound that with fraction of exhaled NO measurements maintenancedoses of ICS could be significantly reduced withoutcompromising asthma control.Smith AD, Taylor DR. Is exhaled nitric oxide measurementa useful clinical test in asthma? Curr Opin AllergyClin Immunol 2005; 5:49–56This is a good review on the use of exhaled NO measurementsin the management of asthma. Epidemiologic dataconfirm that exhaled NO measurements reflect the presenceand severity of airway inflammation in asthma patients.Reference values and thresholds for an abnormal test need tobe agreed on internationally.Sont JK, Willems LNA, Bel EH, et al. Clinical controland histopathologic outcome of asthma when usingairway hyperresponsiveness as an additional guide tolong-term treatment. Am J Respir Crit Care Med 1999;159:1043–1051This group of investigators explored whether a treatmentstrategy aimed at reducing airway hyperresponsivenessadded to the strategies used in the asthma guidelines wouldprovide more effective control of asthma and greater improvementin chronic airway inflammation. They did show thatreducing airway hyperresponsiveness led to more effectivecontrol of asthma. This implies that the monitoring of airwayhyperresponsiveness (by repeated bronchial inhalation challenge)may improve the long-term management of asthma.Suissa S, Ernst P, Benayoun S, et al. Low-dose inhaledcorticosteroids and the prevention of death fromasthma. N Engl J Med 2000; 343:332–336This important article gives evidence that the regular useof low-dose ICS is associated with a decreased risk of deathfrom asthma.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 109ACC-PULM-09-0302-006.indd 1097/10/09 8:05:47 PM


Sur S, Crotty TB, Kephart GM, et al. Sudden-onset fatalasthma: a distinct entity with few eosinophils and relativelymore neutrophils in the airway submucosa? AmRev Respir Dis 1993; 148:713–719This study determined the histologic differences in the airwaysof patients who died of sudden-onset asthma ( 1 h)compared with those who had the more commonly seen slowonsetasthma. Patients with slow-onset asthma had moreeosinophils and fewer neutrophils than patients with suddenonsetasthma. The authors concluded that sudden-onsetasthma is immunohistologically distinct from the slow-onsettype because of these differences in eosinophilic and neutrophilicairway infiltration. They raise the possibility that themechanisms involved in these two distinct forms of asthmaare different.Sutherland ER, Martin RJ. Asthma and atypical bacterialinfection Chest. 2007; 132:1962–1966This article reviews the basic and clinical science that implicatesthe atypical bacterial pathogens M pneumoniae andChlamydophila (formerly Chlamydia) pneumoniaeas potentially important factors in asthma. Although theirexact contribution to asthma development and/or persistenceremains to be determined, evidence links them to new-onsetasthma and asthma exacerbations.Szczeklik A, Stevenson DD. Aspirin-induced asthma:advances in pathogenesis, diagnosis, and management.J Allergy Clin Immunol 2003; 111:913–921This article discusses aspirin-induced asthma, including theclinical presentation and the molecular biology. It concludesthat aspirin-induced asthma runs a protracted course even iftherapy with cyclooxygenase-1 inhibitors is avoided. Aspirindesensitization followed by daily aspirin treatment is avaluable therapeutic option.Tan WC. Viruses in asthma exacerbations. Curr OpinPulm Med 2005; 11:21–26This review discusses the role of viruses as triggers foracute exacerbations of asthma. The application of moleculardiagnostic methods has shown that rhinovirus is the mostcommon cause but coinfection is frequent. Viruses provokeasthma attacks by additive or synergistic interactions withallergen exposure or air pollution. Respiratory viruses causeasthma exacerbations by triggering the recruitment of T-helper type 2 cells into the lung.Tarlo SM, Balmes J, Balkissoon R, et al. Diagnosis andmanagement of work-related asthma: American Collegeof Chest Physicians Consensus Statement. Chest2008; 134(3 suppl):1S–41SThis review replaces a previous American College of ChestPhysicians Consensus Statement on asthma in the workplacewas published in 1995. This current Consensus Statementwas written by a panel of experts, including allergists, pulmonologists,and occupational medicine physicians, The ConsensusDocument defined work-related asthma to includeoccupational asthma (ie, asthma induced by sensitizer orirritant work exposures) and work-exacerbated asthma (ie,preexisting or concurrent asthma worsened by work factors).The Consensus Document focuses on the diagnosis and management(including diagnostic tests, and work and compensationissues), as well as preventive measures.Turner WS, Palmer LJ, Rye PJ, et al. The relationshipbetween infant airway function, childhood airwayresponsiveness and asthma. Am J Respir Crit CareMed 2004; 169:921–927This study sought to determine whether there are early lifefactors that can predict the development of asthma later inchildhood. A total of 243 Australian children were prospectivelystudied during the course of 11 years. The presence ofwheezing at age 11 years was associated with lower lung functionduring infancy and was independent of increased airwayresponsiveness and atopy at a younger age. This suggeststhat intrinsic disturbances in lung function, possibly relatedto lung development, maternal factors, and/or environmentalfactors close to the time of birth, have a role in the later developmentof asthma. Ongoing atopy and BHR predicted wheezingthat persisted from ages 4 to 6 years to age 11.Van Hove CL, Maes T, Joos GF, Chronic inflammationin asthma: a contest of persistence vs resolution.Allergy 2008; 63:1095–109Recent investigations have highlighted that endogenousantiinflammatory mediators and immune-regulating mechanismsare important for the resolution of inflammatory processes.A disruption of these mechanisms can be causallyrelated not only to the initiation of unnecessary inflammation,but also to the persistence of several chronic inflammatorydiseases. This article discusses these potential mechanismswith asthma.Vonk JM, Jongepier H, Panhusen CIM, et al. Risk factorsassociated with the presence of irreversible airflowlimitation and reduced transfer coefficient in patientswith asthma after 26 years of follow up. Thorax 2003;58:322–327Fixed airflow obstruction develops in some asthmaticpatients. In this study, after 20 to 30 years of follow-up,adults with a history of asthma were reexamined to assessthe risk factors for the development of irreversible airwayobstruction and low diffusing capacity. In this study, irreversibleairflow obstruction developed in 16% of patients.This was associated with a lower FEV 1(percentage of predicted),a lower level of BHR, less reversibility at initial testing,and with less use of corticosteroids at follow-up.110 Asthma (Braman)ACC-PULM-09-0302-006.indd 1107/10/09 8:05:48 PM


Walker S, Monteil M, Phelan K, et al. Anti-IgE forchronic asthma. Cochrane Database Syst Rev (databaseonline). Issue 3, 2004This Cochrane Database report discusses the use of recombinanthumanized monoclonal antibody directed against IgE.This therapy (called omalizumab) is significantly more effectivethan placebo at increasing the number of patients whoare able to reduce or withdraw from therapy with ICS. It alsoreduces the number of asthma exacerbations.Wark PA, Gibson PG, Wilson AJ. Azoles for allergicbronchopulmonary aspergillosis associated withasthma. Cochrane Database Syst Rev (database online).Issue 34, 2004This review concludes that itraconazole modifies the immunologicaction associated with ABPA and improves clinicaloutcomes.Wenzel S. Severe/fatal asthma. Chest 2003; 123(suppl):405S–410SThis is a discussion of severe fatal asthma, including thegenetic and environmental factors that are responsible. Itdiscusses the pathologic findings, including eosinophilicinflammation, structural changes, and distal disease.Wenzel S. Severe asthma in adults. Am J Respir CritCare Med 2005; 172:149–160This is a pulmonary perspective on severe asthma, whichdisproportionately consumes health-care resources relatedto this disease. Early-onset severe asthma is a more allergicassociated disease. Severe asthma with persistent eosinophilia(either early onset or late onset) is more symptomaticand has more near-fatal events. At least 50% of patients withsevere asthma have little identifiable inflammation. Steroidresistance in severe asthma is also discussed in this article.Wenzel SE, Szefler SJ, Leung DY, et al. Bronchoscopicevaluation of severe asthma. Am J Respir Crit CareMed 1997; 156:737–743This study was designed to evaluate the type of airwayinflammation in patients with severe asthma who arereceiving high-dose oral glucocorticoids. Eosinophils werenot found in healthy control individuals or patients withsevere asthma but were seen in patients with moderateasthma. In contrast, patients with severe asthma demonstrateda greater concentration of neutrophils in BAL fluid.Findings suggest that inflammation remains in symptomaticpatients with severe asthma despite glucocorticoidtreatment.West PM, Fernandez C. Safety of COX-2 inhibitors inasthma patients with aspirin hypersensitivity. AnnPharmacother 2003; 37:1497–1501This study concludes that cyclooxygenase-2 inhibitorsprovide a potentially safe alternative for the treatment ofinflammatory conditions in patients with aspirin-inducedasthma.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 111ACC-PULM-09-0302-006.indd 1117/10/09 8:05:48 PM


Notes112 Asthma (Braman)ACC-PULM-09-0302-006.indd 1127/10/09 8:05:48 PM


Cardiopulmonary Exercise TestingDarcy D. Marciniuk, MD, FCCPObjectives:• Provide a brief overview of normal exercise physiologyand responses• Outline the indications, conduct, and interpretation ofcardiopulmonary exercise testing• Highlight characteristic responses commonly demonstratedby patients with various disorders frequently assessed bythe pulmonologistKey words: cardiopulmonary exercise testing; exercise;interpretation; pulmonary function laboratory; pulmonaryfunction testingExercise in the normal human involves the effectiveintegration of respiratory, cardiovascular, neuromuscular,and metabolic functions. The organsinvolved in these varied and important roles havea sizeable reserve, with the consequence that clinicalmanifestations of a disease state or abnormalitymay not become readily apparent until thefunctional capacity of the organ(s) is markedlyimpaired. When this manifestation occurs, patientsoften experience the distressing and disablingsymptoms of shortness of breath with activityand exercise limitation. Cardiopulmonary exercisetesting (CPET) has secured an essential role inour practice of clinical medicine by allowing theclinician to objectively evaluate these importantfunctions and symptoms.Objective assessment and measurement ofvarious parameters during exercise, which placesan increased physiologic demand on the functionalreserve capacity of these organs, can alsoprovide a sensitive method for the early detectionof abnormal function and response(s). The resultsfrom exercise testing parallel functional capacityand quality of life more closely than measurementsobtained only at rest, and they have beenshown to accurately predict important outcomes,such as the rate of mortality, in a variety ofpatients and clinical circumstances. Although theuse of CPET has been previously viewed asbeing merely interesting in the hands of a fewindividuals, CPET is now cemented into mainstreamclinical practice. In view of these meaningfulbenefits, the conduct and interpretationof CPET is now an essential competency for practicingpulmonologists.Normal Exercise PhysiologyKnowledge and a thorough understanding ofnormal exercise physiology are essential for theappropriate interpretation of exercise responses indisease. Although relevant principles of normalexercise physiology will be briefly summarized, acomprehensive review of this subject topic isbeyond the intent of this course and syllabus. Thereader is encouraged to consult more detailedappropriate source literature and documents onthis topic (see the references section).To meet the increased metabolic demands ofexercise, the respiratory and cardiovascular systemsmust be able to augment oxygen delivery tothe working skeletal muscles. Oxygen transport inthe body depends on a series of linked mechanismsthat can be schematically expressed as follows(adapted from Jones; 1). Potential factors that mayaffect these mechanisms are shown in parentheses,but these are not inclusive:Inspired O 2(altitude)↓Respiratory ventilation(alveolar ventilation, distribution)↓Respiratory gas exchange(diffusion, ventilation/perfusion)↓Arterial O 2(oxygen capacity, saturation)↓Cardiac output(SV, cardiac frequency)↓<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 113ACC-PULM-09-0302-007.indd 1137/10/09 8:06:20 PM


Circulation and muscle blood flow(peripheral vascular resistance)↓O 2extraction(capillary-tissue diffusion)↓Muscle O 2utilization → O 2stores(respiratory enzymes)↓Venous O 2Similarly, the removal of carbon dioxide (CO 2),which is a byproduct of metabolism, can beexpressed with the following scheme (adaptedfrom Jones; 1):CO 2production (aerobic, anaerobic)↓ → CO stores2Muscle blood flow and circulation↓Venous CO 2↓Venous return(cardiac output)↓Respiratory gas exchange(ventilation/perfusion)↓Respiratory ventilation(alveolar ventilation, distribution)↓Meanwhile, when the inspired CO 2concentrationis negligible, the respiratory response to exercisecan be represented by the followingequation:(3) Va k ⋅ V • co 2Paco 2or(4) V • e k ⋅ V • co 2Paco 2(1 Vd/Vt)where V • e is minute ventilation, the sum of alveolarventilation (Va) and dead space ventilation (Vd).Vd/Vt is the ratio of physiologic dead space totidal volume. These relationships are used tounderstand the ventilatory response during exercisebecause V . e and Va are associated more closelyto V • co 2than to V • o 2. Graphical representations ofhow the respiratory and cardiovascular systemsrespond to meet the demands of exercise in thenormal human are shown in Figures 1 and 2.These relationships and equations also serveto illustrate and emphasize the integrative aspectsof exercise, which result in the total response beinggreater than could be supported by any individualcontribution alone. For instance, oxygen supply ina fit athlete may increase to 20 times the restingoxygen consumption (ie, from 0.25 L/min at restto 5.0 L/min at peak exercise). This net increaseExpired CO 2Understanding the cardiorespiratory and metabolicresponses to exercise is further facilitated by closerexamination of the direct physiologic determinantsof oxygen uptake (V • o 2) and carbon dioxide output(V • co 2) as depicted in the following equations. Thecardiovascular responses are represented by rearrangementof the Fick equation:(1) V • o 2 Qt (Cao 2 Cvo 2)or(2) V • o 2 (HR ⋅ SV) (Cao 2 Cvo 2)where Qt is cardiac output, the product of strokevolume (SV) and heart rate (HR), and Cao 2 Cvo 2is the oxygen content difference between systemicarterial and mixed venous blood.Figure 1. Respiratory system responses during exercise.114 Cardiopulmonary Exercise Testing (Marciniuk)ACC-PULM-09-0302-007.indd 1147/10/09 8:06:20 PM


Figure 2. Cardiovascular system responses during exercise.is not brought about by a 20-fold increase in anyone mechanism but rather is shared betweenmechanisms. For example, the increase in V • e duringexercise occurs not only because of the typicalhyperventilation (attributable to metabolic acidosis)at the end of exercise but also because of asignificant decrease in Vd/Vt during exercise.The fractional changes in various componentsof the cardiovascular and respiratory systems duringexercise are summarized in Figure 3, whichshows representative data from a population ofnormal middle-aged men. These data illustrate themajor demands placed on cardiorespiratory functionduring exercise, and they exemplify the significantreserve that exists to meet the increaseddemands of exercise. These demands are met byincreases in ventilation (breathing frequency andVt) and cardiac output (HR and SV) as well as byredistribution of blood to the exercising muscles.In addition, Vd decreases as Va becomes moreefficient. The V • o 2increases with increasing exerciseand, in some fit and motivated individuals, reachesa plateau near the end of exercise. Meanwhile, asCO 2production continues, further fueled byanaerobic respiration and progressive increases inlactate production later in exercise, the respiratoryexchange ratio (RER) increases in the normalhuman. The peak V • o 2is usually determined by thecapacity of the cardiovascular system to deliveroxygen to the working muscles. Except in welltrainedathletes and perhaps the fit elderly, gasexchange remains well preserved during exercise,although the hyperventilation induced by lacticacid production typically results in a slight decreasein Paco 2and an increase in Pao 2.Clinical Indications for CPETThe indications for CPET are varied anddepend on the clinical setting and question(s) tobe addressed. Shortness of breath with exercise andlimitation of activity are cardinal and commonsymptoms of dysfunction and are therefore someVOCaO CvO VCOVA VD VT VEFigure 3. Relationship between resting and maximal exercise values in normal humans. Adapted from Gallagher. 2<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 115ACC-PULM-09-0302-007.indd 1157/10/09 8:06:21 PM


of the most frequent reasons for testing in theclinical laboratory. CPET also has significant utilityin distinguishing normal from abnormal responsesand in establishing cardiovascular from respiratorycauses for activity limitation.Our understanding of the clinical utility ofCPET is also increasing as further work is publisheddemonstrating the prognostic value of exercisetesting in various clinical settings. 1,3,4,5 Results fromexercise testing have been shown to correlate withimportant clinical outcomes, including rate of mortality,in COPD, interstitial lung disease (ILD),cystic fibrosis, primary pulmonary hypertension,and chronic heart failure. Generally acceptedclinical indications for CPET are listed in Table 1.Equipment, Conduct, andMeasurementsExercise testing is most commonly performedwith either a stationary cycle ergometer or a motorizedtreadmill. Both cycle ergometer and treadmilltesting are appropriate, although the cycle ergometeris used more frequently. The cycle ergometerallows for a direct measurement of work rate, hasless potential for the creation of artifacts, and seemsto be better tolerated by patients, particularly thosewith significant lower-limb joint problems. Alternatively,the treadmill yields greater values for thepeak V • o 2(approximately 8 to 12%) and is betteraccepted by fit normal subjects. 5Specialized equipment is required for CPET,necessitating both increased cost and expertise.Although reliable and user-friendly equipment isnow available as the result of advances in technology,these advances also have ushered in a multitudeof confusing numbers and graphs that thisTable 1. Clinical Indications for CPET*Objective assessment of symptomsEvaluation of severity of impairmentAppraisal of contributors to exercise limitationEarly detection of disease or impairmentAssessment of response to therapyDisability assessmentAssessment/titration of supplemental oxygen therapyIdentification of exercise-associated bronchoconstrictionPreoperative risk and transplantation assessment*Adapted from Palange et al 4 and Weisman et al. 5new equipment now produces. Similar to pulmonaryfunction testing equipment, exercise testingequipment systems, whether mixing chamber orbreath-by-breath ones, require meticulous calibrationprocedures to ensure that measurements areand remain accurate and precise. In addition todaily calibration, routine physiologic calibrationshould be performed. This calibration is bestundertaken by having a healthy staff memberexercise at several constant work rates for a specifiedduration while V • e, V • o 2, and V • co 2are measured.This procedure might entail measurementsafter 4 to 6 min each of rest, 50 W and 100 W. Valuesshould remain consistent ( 5% variation) withmeasurements collected previously under identicalcircumstances. If not, the cause of any discrepancymust be investigated and corrected. The ventilatorymeasurements unique to CPET are listed in Figure4, in which measured variables are denoted bysolid lines, whereas derived variables are denotedby dashed lines.In addition to the aforementioned measurements,the Pao 2(arterial sampling) or arterialoxygen saturation (Spo 2; ie, oxygen saturation bypulse oximetry) work rate (cycle ergometer), BP(cuff sphygmomanometer), heart rate/rhythm(ECG), and symptoms of shortness of breath andleg fatigue (modified Borg scale or visual analogscale) are measured. The reason(s) for stoppingexercise should always be noted.Together, these measurements allow for a comprehensiveevaluation of behaviors during exerciseand for the determination of various derived variablesthat provide additional information in theinterpretation process. However, as noted, it isimportant to be focused on the overriding principles,values, and relationships of both normal andabnormal responses rather than become confusedand misled by the often-unnecessary plethora ofconfusing numbers and graphs.Cycle ergometers and treadmills shouldundergo periodic calibration, although this calibrationis typically necessary only every few years orif the equipment has been moved. The manufacturershould be contacted to facilitate these procedures,which are more detailed and require morespecialized expertise and techniques.The patient should wear comfortable clothesand shoes, and he or she should refrain from heavyactivity or meals for 2 h before testing. All testing116 Cardiopulmonary Exercise Testing (Marciniuk)ACC-PULM-09-0302-007.indd 1167/10/09 8:06:21 PM


Figure 4. Measured and derived ventilatory measurements.procedures should be explained, and informedconsent should be obtained. A brief review of theclinical history and a physical examination shouldbe performed. Baseline spirometry is measured,and the patient should be familiarized with use ofthe equipment before the test. If a cycle will beused, the seat and handlebars should be adjustedfor comfort. In the case of a treadmill, the patientshould be shown how to get on and off the movingbelt comfortably and safely.The choice of which test to perform dependsin large part on the clinical question beingaddressed. Maximal symptom-limited incrementalprotocols are most commonly performed. Enduranceexercise protocols often are conducted inclinical trials to determine a treatment effect, andthey may have clinical utility in the individualpatient when the clinician is assessing a responseto an intervention. They are often completed at aconstant work rate (approximately 75% of themaximal work rate is typical) and continued untilthe patient is no longer able to exercise. Exercisechallenge testing frequently is performed to objectivelyassess for the presence of exercise-inducedbronchoconstriction (EIB). The test is performed toillicit approximately 6 min of very intense exerciseassociated with high levels of V • e. Spirometry isthen performed after exercise at frequent intervals(1, 5, 10, 15, 20, and 30 min are recommended), witha decrease in the FEV 1of 10% indicating possibleand 15% probable EIB. 6 Readers are asked to consultreferenced guideline statements for furtherinformation and detail regarding the conduct ofthis specialized testing. 5,6The performance and conduct of the 6-minwalk distance test will not be addressed in thischapter, but it is discussed in more detail elsewhere.7 However, timed walk tests (typically6 min or 12 min) are the standard “simple” test forthe assessment of activity limitation. They are safeand practical, and they tend to mimic activities ofdaily living. Their utility has been confirmed toimprove with standardization, although they aresusceptible to a training effect (similar to othertests). Their major limitation is they providerestricted information regarding physiologic contributorsand mechanisms of exercise limitation.Nonetheless, they have become embedded in ourfield because of the meaningful information theyyield. 8,9Shuttle walk tests are emerging as anotheroption in this setting, and reports 10,11 have validatedtheir usefulness in detecting a treatmenteffect. The obstacle to more widespread use at thistime relates to a lack of familiarity. However, asfurther reports validating their utility are published,our understanding of the role of both incrementaland endurance shuttle walk test protocolswill mature.In short, CPET is the current “gold standard”for assessing exercise performance. It allows determinationof mechanistic insights, recognition ofcoexistent and multiple exercise-limiting factors,and provides a thorough evaluation of respiratory<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 117ACC-PULM-09-0302-007.indd 1177/10/09 8:06:21 PM


esponses and constraint. Both incremental andendurance protocols may be used, and the resultsfrom endurance exercise testing have been foundto be more responsive to a treatment effect thaneither incremental exercise or 6-min walk testing.If a thorough evaluation of exercise performanceis required, particularly in patients with multiplecomorbidities, CEPT remains the testing methodof choice.CPET is safe, with the risk of death for patientsapproximating 2 to 5 per 100,000 exercise testsperformed. 5 These risks can be minimized by anumber of practical and common sense precautions,including the following:• direct physician supervision (the importanceof this cannot be overemphasized) and a thoroughunderstanding of the contraindications toundergoing testing, as well as the indicationsfor terminating exercise testing;• appropriate cardiac and BP monitoring;• resuscitation equipment and expertise; and• accurate Sao 2monitoring and availability ofsupplemental oxygen.Specific contraindications for testing are listedin Table 2. Indications for prematurely terminatingCPET are listed in Table 3.Specific Variables Assessed DuringExerciseAlthough a discussion of individual measurementsis presented, it is important to recognize thatCPET involves their collective integration for interpretation.In many instances, specific individualmeasurements are of lesser importance and areTable 2. Contraindications to CPETUnstable angina or recent acute coronary syndromeUncontrolled arrhythmias causing symptoms or hemodynamiccompromiseActive endocarditis, myocarditis, or pericarditisSymptomatic, severe aortic stenosisPoorly or uncontrolled heart failureAcute pulmonary embolism or infarctionThrombosis of the lower limbsSuspected dissecting aortic aneurysmUncontrolled asthma<strong>Pulmonary</strong> edemaSignificant hypoxemiaTable 3. Indications for Terminating CPET in the ClinicalLaboratory*Unstable angina or chest pain suggestive of myocardialischemiaECG changes suggestive of ischemia, complex ectopy, orsecond- or third-degree heart blockSystolic BP decrease 20 mm Hg from highest valueSystolic BP 250 mm Hg, diastolic BP 120 mm HgDesaturation to 80%Sudden pallor or dizzinessMental confusionSigns of respiratory failure*Adapted from Weisman et al. 5dictated by the clinical question being addressedand, in part, by responses demonstrated by thepatient during testing.V • o 2V • o 2is determined by cellular oxygen demandand maximal oxygen transport (see the section“Normal Exercise Physiology”). It is typically presentedtogether with work rate, and it increasesnearly linearly as external work increases. Theslope of this relationship reflects the efficiency ofmetabolic conversion of energy to work and themechanical efficiency of the musculoskeletal system.Values of 8.5 to 11.0 mL/min/W are normalfor the ΔV • o 2/Δ work rate relationship and are unaffectedby age, height, or sex. Although obese individualsmay have an increased ratio, the slope ofthat relationship remains normal.As exercise progresses, V • o 2increases until oneof more of its determinants approach limitation(HR, SV, or tissue extraction), at which time theV • o 2/work rate may begin to plateau. This plateauhas been used as the best evidence of maximumV • o 2, which is the “gold standard” used for assessingcardiorespiratory fitness. However, in theclinical setting, a plateau may not be reached, andthe peak V • o 2is often used for this purpose. Thepeak V • o 2is often normalized for body size bydividing it by weight in kilograms. Unfortunately,normalization by body weight in obese individualsmay provide a falsely low value. Although useof the lean body mass or height may be moredesirable, there is no consensus on how best toaccount for body size.118 Cardiopulmonary Exercise Testing (Marciniuk)ACC-PULM-09-0302-007.indd 1187/10/09 8:06:22 PM


A reduced peak V • o 2is the starting point forinterpretation, and underlying causes responsiblefor reduced exercise capacity are determined byinspecting the pattern of responses in other variables.The peak V • o 2should be expressed as bothan absolute value and also as a percentage of thepredicted value.V • co 2V • co 2is determined by factors similar to V • o 2,but because CO 2is more soluble, V • co 2is moreclosely related to V • e than to V • o 2. Importantly, thebody uses CO 2to compensate for acute metabolicacidosis, which contributes to the V • co 2vs workintensity relationship above the point of anaerobicmetabolism. An accurate measurement of V • co 2isimportant because it serves in the calculation ofseveral meaningful derived variables. Moreover,because V • e is so closely related to V • co 2, it is helpfulto analyze V • e in relation to V • co 2.RERThe RER is the ratio of V • co 2/V • o 2, which understeady-state conditions approximates the respiratoryquotient. An RER of 1.0 indicates metabolismby primarily carbohydrates, 0.7 by primarily fat,and 0.8 by primarily protein. Values 1.0 mayindicate carbohydrate metabolism but also CO 2derived from lactic acidosis or, importantly, hyperventilation.The RER should be reported as a functionof the V • o 2.Anaerobic ThresholdAnaerobic threshold (AT), also often referred toas the lactate threshold or gas exchange threshold, isconsidered an indicator of the onset of metabolicacido sis caused predominantly by increased arteriallactate during exercise. The AT should be expressedas a percentage of the peak V • o 2. In normal individuals,the AT occurs at 50 to 60% of peak V • o 2,with the range of normal being 40 to 80%. The ATis affected by the type of exercise (lower for arm vsleg exercise) and method of testing (lower for cyclevs treadmill testing). From a practical point of view,the AT denotes the upper limit of exercise intensitythat can be accomplished aerobically, and exerciseabove the AT is associated with a progressivedecrease in exercise tolerance.Invasive methods for measuring the AT includearterial blood sampling of lactate or bicarbonate.There are various proposed methods for estimatingthe AT noninvasively (Fig 5), including the ventilatoryequivalents method (V • e/V • o 2, V • e/V • co 2, endtidalPo 2[Peto 2], and end-tidal Pco 2[Petco 2]) andthe V-slope method (V • co 2vs V • o 2). Confirmatoryevidence is provided by noting the change in slopeof the V • E vs V • o 2relationship, and when the RERapproximates 1.Like the peak V • o 2, a reduced AT is nonspecificand requires inspection of other variables to determinethe underlying etiology of the reduction.Values 40% may be witnessed with a wide varietyof cardiovascular, respiratory, and musculoskeletalconditions. In some patients with severe respiratorylimitation (for example, severe COPD), the ATcannot be determined noninvasively.HR and HR/V • o 2In normal healthy individuals, HR increaseslinearly with increasing exercise and V • o 2. Predictedmaximal HR is often estimated by 220 –age, but this equation may underestimatemaximal HR in the elderly. The differencebetween predicted maximal HR and the observedmaximal HR is called the HR reserve. In normalindividuals, there is little or no HR reserve at theend of exercise, which suggests a maximal ornear-maximal patient effort. Peak HR may bereduced in a variety of cardiovascular conditions(including with pharmacologic agents) or in asubmaximal study, but if a patient achieves predictedmaximal HR, it suggests that cardiovascularfunction contributed to exercise limitation.In this instance, examination of other variables(ie, V • o 2, AT) will assist in the determinationof whether this occurrence was normal orabnormal.The slope of the HR-V • o 2relationship is a functionof SV; the greater the SV, the lower the HR(Figs 2 and 6). In patients with significant lungdisease, the opposite may be seen, often reflectingdeconditioning, mechanical ventilatory limitation,or potentially the hemodynamic consequences ofdynamic hyperinflation.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 119ACC-PULM-09-0302-007.indd 1197/10/09 8:06:22 PM


Figure 5. Noninvasive estimations of the AT.Figure 6. Cardiovascular responses during exercise in different clinical situations.120 Cardiopulmonary Exercise Testing (Marciniuk)ACC-PULM-09-0302-007.indd 1207/10/09 8:06:22 PM


Oxygen PulseThe ratio of V • o 2to HR is referred to as the oxygenpulse and is often used as a correlate of SVduring exercise. Although under ideal conditionsthis relationship generally holds true (see equations1 and 2), in many settings the variable mustbe used with caution. The assumptions in the equationsare not valid in the presence of significantdesaturation or with impaired skeletal muscleoxygen extraction. A low oxygen pulse may thereforenot only reflect cardiovascular disease but alsodeconditioning, early exercise termination causedby respiratory factors, symptoms or submaximaleffort, and/or the presence of arterial oxygendesaturation.BPSystolic BP typically increases progressivelyduring exercise as V • o 2increases, whereas diastolicBP typically increases only slightly or remainsrelatively unchanged. Abnormal patterns ofresponse during exercise include an exaggeratedincrease, a reduced increase, or a decrease. Exaggeratedincreases are found with resting hypertension,but in those without a known diagnosis, theymay also predict the future onset of resting hypertension.A reduced increase may suggest underlyingcardiovascular or sympathetic controlabnormalities. A decreasing BP during exercise isvery serious, and it is an indication for immediatetermination of the test. If a decrease occurs, seriousefforts to exclude heart failure, ischemia, or outflowtract obstruction should be undertaken.Breathing Pattern and VentilationThe increase in V • e with exercise is accompaniedby increases in both the depth and frequencyof breathing (Fig 1). Increases in Vt are primarilyresponsible for the increase in V • e at low levels ofexercise, but as exercise progresses, both Vt andrespiratory rate increase. At 70 to 80% of peakexercise, increases in V • e are achieved primarily byincreases in respiratory rate.There are various methods used to estimatepeak V • e, including actual measurement of themaximal voluntary ventilation. Predicted maximalV • e may also be estimated with various equations,most commonly the FEV 1× 37 to 40. Unfortunately,none of these methods are perfect, and a pressingneed exists for enhancements in this area. The useof the maximal voluntary ventilation is limited byconcerns regarding patient effort and repeatabilityand with the unique breathing strategy adoptedduring the maneuver that does not parallel thestrategy used during exercise. Equations are thereforemost commonly used, but they may be lesssuited for patients with neuromuscular disordersor respiratory muscle weakness. Despite theselimitations and regardless of the method used,estimating the peak V • e has significant clinical utilityand has withstood the test of time.The terms breathing reserve or ventilatory reserveare used to denote the relationship betweenpredicted peak V • e and the actual measured peakV • e, displayed both as an absolute value (in liters)and as a percentage of the predicted peak V • e.Patients with respiratory disease characteristicallyhave reduced ventilatory capacity and increasedventilatory demand, resulting in reduced ventilatoryreserve. Ventilatory demand is usually increasedboth at rest and during exercise in patients withCOPD, ILD, and pulmonary vascular disease, forexample, as the result of ventilation/perfusioninequality and increased Vd/Vt , hypoxemia, and/or increased stimulation of lung receptors (whichserves to increase ventilation). It is also dependenton other factors such as metabolic requirements,lactic acidosis, behavioral factors, deconditioning,body weight, and mode of testing. In most healthyadults, peak V • e at the end of exercise approaches70% of the maximal V • e, although this percentagemay be greater with increased fitness and with aging(Fig 7). In patients with significant respiratory diseaseand mechanical abnormalities, the patient’send-of-exercise V • e may reach or even exceed thepredicted maximal peak V • e. A plot of V • e vs eitherV • co 2or V • o 2is acceptable for the graphical representationof ventilatory data.Ventilatory Equivalents for V • o 2and V • co 2The ratio of V • e to V • o 2is called the ventilatoryequivalent for oxygen and the ratio of V • e to V • co 2iscalled the ventilatory equivalent for CO 2. They areboth related to Vd/Vt and are greater as Vd/Vtincreases. They also both increase with hyperventilation.Normal responses for these variables are<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 121ACC-PULM-09-0302-007.indd 1217/10/09 8:06:23 PM


Figure 7. Respiratory responses during exercise in different clinical situations.shown in Figure 5 (bottom left). The initial increasein the V • e/V • o 2(while the V • e/V • co 2has still notincreased) that typically occurs in concert withmetabolic acidosis is different than the patterndemonstrated with hyperventilation (attributableto anxiety, pain, or hypoxemia), whereby both theV • e/V • o 2and the V • e/V • co 2increase together. Thesubsequent increase in the V • e/V • co 2representsthe respiratory compensation associated with adecrease in the Paco 2and the Petco 2.The V • e/V • co 2is usually 34 at the AT and usually 37 to 40 at the end of exercise. Increasedvalues reflect either an increased Vd/Vt or a lowPaco 2. A lack of an increase with exercise reflectseither insensitivity to the stimulus of metabolicacidosis, or an inability of the respiratory systemto respond to that stimulus (ie, significantCOPD).PETO 2and PETCO 2The characteristic response of these variablesduring exercise is shown in Figure 5 (bottom right).The period of increasing Peto 2with relativelystable Petco 2has been termed isocapnic buffering.A high V • e/V • co 2without a corresponding decreasein Petco 2suggests increased Vd/Vt , whereas adecrease in Petco 2when the V • e/V • co 2is high suggestshyperventilation.Flow-Volume CurvesAlthough first reported in 1961, flow-volumecurves during exercise were not adopted nor wellstudied until the 1990s. Since that time, our insightof their usefulness to better understand respiratoryresponses and symptoms during exercise hasgrown (12). This greater understanding, coupledwith advances in technology, has led to the routineanalysis of flow-volume curves during exercise.The added value garnered from their use relates totheir ability to provide information about the overallbreathing strategy adopted by patients duringexercise. They also enable an appreciation of thebehavior of operational lung volumes, includingthe end-expiratory lung volume (derived from thetotal lung capacity and the inspiratory capacity)and the end-inspiratory lung volume. Additionally,analysis of flow-volume curves provides an objectiveassessment of the presence and degree of flowlimitation during exercise.In patients with significant disease, there are anumber of characteristic patterns of response inlung volumes that differ from normal patients122 Cardiopulmonary Exercise Testing (Marciniuk)ACC-PULM-09-0302-007.indd 1227/10/09 8:06:23 PM


Figure 8. Behavior of operational lung volumes during exercise.(Fig 8: volume vs V • e; Fig 9: flow vs volume). Inaddition to COPD and ILD, patients with significantobesity and central airway obstruction alsodemonstrate distinguishing responses. 13 Analysisof flow-volume curves during exercise is also especiallyuseful in helping to assess a therapeuticresponse. 14,15In these examples, the end-expiratory lungvolume, the inspiratory reserve volume, and thepresence/absence of flow limitation serve to distinguishthe various disease states from normal,these changes being unique and characteristic. Thevalue of flow-volume curves during exercise isbeing further studied, and it is likely that additionalindications and applications such as usingthem to better estimate maximal ventilatory capacitymay be adopted.Pao 2, Sao 2, and Alveolar-Arterial Po 2Normal exercise responses are enabled by efficientgas exchange. Arterial hypoxemia is uncommonduring exercise in normal humans, but it mayoccur in some elite or aging athletes during highintensityexercise. Inefficient gas exchange is demonstratedby the alveolar-arterial oxygen pressuredifference, or P(A-a)o 2, gradient and by the Pao 2.A significant widening of the gradient and decreasein the Pao 2are abnormal, and they are most characteristicof ILD and significant right-to-left shuntsas well as some patients with COPD and pulmonaryvascular disease (PVD). A reduced Pao 2witha normal P(A-a)o 2may be seen in processes associatedwith abnormal respiratory control and in ahypoxic environment (ie, altitude), although theFigure 9. Maximal and tidal flow-volume curves at rest and during exercise.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 123ACC-PULM-09-0302-007.indd 1237/10/09 8:06:23 PM


latter is rarely reproduced in the clinical laboratory.A reduced Pao 2with an abnormally widenedP(A-a)o 2would suggest worsening ventilation/perfusion inequalities, right-to-left shunt, and/ordiffusion limitation, potentially accompanied by adecrease in the mixed venous Po 2.The P(A-a)o 2is normally 10 mm Hg at restbut may normally increase to 20 mm Hg duringexercise. Values 35 mm Hg are abnormal andindicate possible gas exchange abnormalities.Occasionally, a less-than-optimal substitute, Spo 2,is used as an alternative to arterial blood gas sampling.In the normal individual, both the Pao 2andthe Spo 2are not appreciably different during exercisewhen compared with rest.VD/VTVd/Vt is another index of gas exchange efficiency.An increase in Vd/Vt represents anincreased inefficiency of ventilation, which requiresan increase in V • e to maintain Paco 2. Vd/Vt ishighly dependent on the breathing pattern becausea rapid shallow breathing pattern increases Vd/Vtwithout any other abnormalities. The Vd/Vt iscalculated from the Paco 2and the Peco 2using thefollowing equation:(5) Vd/Vt (Paco 2 PECO 2)Paco 2where the Peco 2is the mixed expired CO 2value ofalveolar and dead space gas. Vd/Vt is obtaineddirectly by collecting expired gas and measuring itsCO 2concentration. This test can easily be performedwith a mixing chamber or, for breath-by-breath systems,after determination of the V • e/V • co 2ratio.Approaches that substitute end-tidal Pco 2for Paco 2yield unreliable results, particularly in disease, becausepulmonary gas exchange abnormalities in themselvesaffect the difference between Paco 2and Petco 2.At rest, the Vd/Vt may normally be approximately0.30 and decrease to approximately 0.10 to0.15 at the end of exercise. Patients with respiratorydisease may have at rest either normal or increasedvalues that fail to decrease normally during exercise.In some instances, the Vd/Vt will increase duringexercise in patients with significant disease. However,the Vd/Vt is neither sensitive nor specific forlung disease and, thus, an isolated abnormalityshould be interpreted with caution. This findingalso emphasizes the importance of evaluating thepatterns of response from a collection of variablesrather than reacting to just a single measurement.Symptoms During ExercisePatients often report that breathlessness and/or leg fatigue limits their exercise, although otherreasons such as musculoskeletal complaints andexhaustion are also reported. Objective evaluationof these end points is valuable, and the use of eithera visual analog scale 16 or a modified Borg scale 17 isrecommended. These scales have shown to berepeatable and responsive and are commonly usedin clinical laboratories. The reproducibility ofresults obtained from these scales can be enhancedby providing a consistent set of written instructionsto the patient prior to testing. As noted, thereason(s) for discontinuing exercise should berecorded. The value of these ratings is furtherenhanced when relationships such as V • e or workrate vs dyspnea or leg fatigue are examined, particularlyin serial studies or after interventions.Reference ValuesThe selection of appropriate reference valuesfor use in interpreting CPET is essential. The1, 5, 18reader is advised to consult source referencesfor guidance in selecting which reference valuesmay be most appropriate for their specific clinicallaboratory.Reproducibility of CPET ResultsWhenever serial testing is undertaken, forexample, to assess the response to therapy, thevariability of the measurements during CEPT mustbe considered before a beneficial, detrimental, orno effect can be concluded. The coefficient ofvariation (ratio of the SD to the mean, expressedas a percentage) reported for serial measurementsduring maximal testing in the same subject are asfollows (range of reported values listed):V • o 23.0–8.4% V • co 25.0–9.6%HR 1.4–8.6% V • e 5.0–12.3%AT 9.2–13.0% Systolic B/P 2.2–6.7%Sao 22.5% Duration(work rate)3.6–13.8124 Cardiopulmonary Exercise Testing (Marciniuk)ACC-PULM-09-0302-007.indd 1247/10/09 8:06:24 PM


It should be understood that the variabilitydepends on a number of factors, such as the populationstudied, the type of testing performed, andthe specific variable examined. However, it appearsreasonable to assume that differences in most measurementsobtained during CEPT should exceedapproximately 12 to 20% to be considered clinicallysignificant (ie, twice the coefficient of variation)and not caused by inherent variability alone.Interpretation and Reporting of CPETInterpretation of CPET begins with the requisitionform. Requisition forms should be designedto encourage the requesting physician to provideas much clinical information as is reasonable toenable the interpretation to be valuable. This informationwill enhance the ability of the reportingphysician to provide the most meaningful interpretationpossible from the available measureddata. There are a number of important fundamentalquestions that the reporting physician mustaddress when interpreting CPET (19), includingthe following:1. Are the results normal or abnormal?2. How limited is the patient?3. What factors are responsible for the limitation?4. What abnormal patterns of response are demonstrated?5. What clinical disorders may result in these patternsof response?Determining whether a test is physiologicallymaximal is necessary in excluding potential underlyingdiseases or abnormalities. Useful indicatorsmay include the following:1. A plateau of peak V • o 2, or peak V • o 2is achieved.2. Maximum work rate is achieved.3. HR or V • e reach predicted maximum.4. RER 1.15.5. Blood lactate level 4 (mmol/L).6. Patient exhaustion.The clinician must also decide on contributors toexercise limitation and whether they are normal (ie,appropriate) or abnormal. Potential factors that maycontribute to limit exercise include the following:1. Cardiovascular function.2. Respiratory mechanics (and/or respiratorymuscle function).3. Arterial hypoxemia.4. Dyspnea.5. Unfitness/deconditioning.6. Musculoskeletal disorders, peripheral vasculardisease.7. Other factors, ie, motivation, secondary gain,and technical factors.Although responses during exercise may varyin normal individuals, and the range of normal isquite broad, an understanding of “normalcy” andnormal exercise physiology is essential for therecognition of disease. This understanding is essentialfor the appropriate interpretation of CPET.Suggested normal values for various measurementsobtained during CPET are listed in Table 4.In interpreting CPET, it is important to focus onwhat is most important; patterns of responses duringexercise and the reason(s) for testing. This focuswill ensure that a correct and meaningful interpretationwill result. The multitude of both graphicaland numerical results and an overreliance andoverdependence on complicated algorithms hascontributed to confusion. They distract and deterus from cardinal measurements and relationships,the fundamental importance of patient’s symptomsduring exercise, and the basic physiologic principlesthat would help us to better understand and appreciatethe role and benefit (as well as the limitations)of exercise testing in clinical practice. They alsodetract from the practical reality that, unlike intextbooks, guideline statements, and/or the lecturehall, CPET is never ordered nor interpreted in isolation.It should be requested and its results translatedwithin the context of all available clinical informationfrom the patient, with the specific goal ofaddressing the question(s) being asked. CPET doesnot replace common sense, starting with the clinicalassessment of the patient. For example, there aremore appropriate methods than CPET to differentiateCOPD from ILD. To aid in achieving this goal,the characteristic patterns of response during exercisedemonstrated in various disease states arenoted in Table 5. These values highlight the importanceof understanding that no single measurementis diagnostic of any specific disease entity.SummaryCPET typically involves the measurement ofrespiratory gas exchange (V • o 2, V • co 2, V • e, and other<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 125ACC-PULM-09-0302-007.indd 1257/10/09 8:06:24 PM


Table 4. Suggested Normal Values for Selected CPET Variables*VariablePeak V • o 2Normal Value 85% predictedAT 40% predicted peak V • o 2Peak HR 90% predicted peak HRHR reserve 15 beats/minBP 200/90 mm HgOxygen pulse (V • o 2/HR) 80% predictedPeak V • e 85% (peak V • e/predicted peak V • e)Ventilatory reserve 11 L (predicted peak V • e [minus] peak V • e)RR 60 breaths/minV • e/Vco 2 34 at AT; 37–40 at end of exerciseVd/Vt 0.30Pao 2 80 mm HgSao 2desaturation 5%P(A-a)o 2 35 mm HgHR 1.4 to 8.6%AT 9.2 to 13.0%V • co 25.0 to 9.6%V • e 5.0 to 12.3%Systolic BP 2.2 to 6.7%Duration (work rate) 3.6 to 13.8%*Adapted from Weisman et al. 5variables) while the ECG, BP, Spo 2, and perceivedexertion (Borg scale) is monitored during amaximal symptom-limited incremental exercisetest on a cycle ergometer or on a treadmill. In somecircumstances, a constant workload exercise test(based on maximal test results) may be performed.Measurement of arterial blood gases provides moredetailed information on pulmonary gas exchange.Resting and exercise tidal flow-volume loopsshould also be monitored to accurately assess/understand the degree of ventilatory constraint.CPET provides a global assessment of the integrativeexercise responses that are not adequatelyreflected by measurement of individual organ systemfunction at rest: Resting values cannot reliablypredict exercise performance and functional capacity.CPET is safe, comorbidities can be identified,and enhanced understanding and insight intoTable 5. Characteristic Patterns of Response During Exercise Demonstrated in Various Disease States*VariableCongestiveHeart FailureCOPDILD<strong>Pulmonary</strong>Vascular DiseaseDeconditionedPeak V • o 2↓ ↓ ↓ ↓ ↓AT ↓ v or indeterminate ↓ ↓ ↔ or ↓Peak HR v ↔ or ↓ ↓ ↔ or ↓ ↔ or ↓Oxygen pulse ↓ ↔ or ↓ ↔ or ↓ ↓ ↓Peak V • e/maximalvoluntary ventilation↔ or ↓ ↑ ↔ or ↑ ↔ ↔V • e/V • co 2↑ ↑ ↑ ↑ ↔Vd/Vt ↑ ↑ ↑ ↑ ↔Pao 2↔ v ↓ ↓ ↔P(A-a)o 2↔ v ↑ ↑ ↔*Adapted from several sources. 4,5,18,20,21 ↓ decreased; ↔ unchanged from normal; ↑ increased; v variable.126 Cardiopulmonary Exercise Testing (Marciniuk)ACC-PULM-09-0302-007.indd 1267/10/09 8:06:24 PM


various responses, including exercise limitingfactors, is possible. Importantly, CPET promotes anintegrative approach to assessing metabolic, ventilatory,and cardiac function and reserve. Althoughtechnically more demanding than simpler tests, theuse of CPET provides the clinician with a measurementof the peak V • o 2(and relationships involvingthe peak V • o 2), which remains the “gold standard”for assessing aerobic exercise capacity.AcknowledgmentThe author is indebted to Mr. R. Clemens forhis expert assistance in the preparation and reviewof the figures and the manuscript.References1. Jones NL. Clinical exercise testing. 4th ed. Philadelphia,PA: Saunders, 19972. Gallagher CG. Exercise and chronic obstructivepulmonary disease. Med Clin N Am 1990;74:619−6413. Palange P. The prognostic value of exercise testing.Breathe 2009; 5:229−2344. Palange P, Ward SA, Carlsen K-H, et al. Recommendationson the use of exercise testing in clinicalpractice—ERS Task Force. Eur Respir J 2007;29:185−2095. Weisman IM, Beck K, Casaburi R, et al. AmericanThoracic Society/American College of ChestPhysicians joint statement on cardiopulmonaryexercise testing. Am J Respir Crit Care Med 2003;167:211−2776. American Thoracic Society. ATS statement: guidelinesfor methacholine and exercise challengetesting—1999. Am J Respir Crit Care Med 2000;161:309−3297. American Thoracic Society. ATS statement: guidelinesfor the six-minute walk test. Am J Respir CritCare Med 2002; 166:111−1178. Marciniuk DD, Butcher SJ, Reid JK, et al. The effectsof helium-hyperoxia on 6-min walking distance inCOPD: a randomized, controlled trial. Chest 2007;131:1659−16659. Celli BR, Cote CG, Marin JM, et al. The body-massindex, airflow obstruction, dyspnea and exercisecapacity index in chronic obstructive pulmonarydisease. N Engl J Med 2004; 350:1005−101210. Pepin V, Brodeur J, Lacasse Y, et al. 6-minute walkingversus shuttle walking: responsiveness tobronchodilation in chronic obstructive pulmonarydisease. Thorax 2007; 62:291−29811. Brouillard D, Pepin V, Milot J, et al. Enduranceshuttle waling test: responsiveness to salmeterol inCOPD. Eur Respir J 2008; 31:579−58412. Johnson BD, Weisman IM, Zeballos RJ, et al.Emerging concepts in the evaluation of ventilatorylimitation during exercise: the exercise tidal flowvolumeloop. Chest 1999; 116:488−50313. Babb TG. Mechanical ventilatory constraints inagain, lung disease, and obesity: perspectivesand brief review. Med Sci Sports Exerc 1999; 31:S12−S2214. Maltais F, Marciniuk DD, Hernandez P, et al.Improvements in symptom-limited exercise performanceover 8 h with once-daily tiotropium inpatients with COPD. Chest 2005; 128:1168−117815. O’Donnell DE, Sciurba F, Celli B, et al. Effect offluticasone propionate/salmeterol on lung hyperinflationand exercise endurance in COPD. Chest2006; 130:647−656,16. Gift AG. Visual analogue scales: measurement ofsubjective phenomena. Nurs Res 1989; 38:286−28817. Borg GA. Psychophysical bases of perceived exertion.Med Sci Sports Exerc 1982; 14:377−38118. Wasserman K, Hansen JE, Sue DY, et al. Principlesof exercise testing and interpretation. 4th ed. Philadelphia,PA: Lippincott Williams and Wilkins, 200519. Younes M. Interpretation of clinical exercise testingin respiratory disease. Clin Chest Med 1984;5:189−20620. Marciniuk DD, Gallagher CG. Clinical exercisetesting in interstitial lung disease. Clin Chest Med1994; 15:287−30321. Marciniuk DD, Gallagher CG. Clinical exercisetesting in chronic airflow limitation. Med Clin NAm 1996; 80:565−587<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 127ACC-PULM-09-0302-007.indd 1277/10/09 8:06:25 PM


Notes128 Cardiopulmonary Exercise Testing (Marciniuk)ACC-PULM-09-0302-007.indd 1287/10/09 8:06:25 PM


Hypercapnic Respiratory FailureMark J. Rosen, MD, FCCPObjectives:• Understand the determinants of Paco 2and the types ofphysiologic derangements that lead to hypercapnia• <strong>Review</strong> specific neurologic and muscular disorders thatcause respiratory failure• Discuss respiratory muscle function and the assessment ofrespiratory muscle strength• Outline the types of situations in which an imbalance betweenthe work of breathing and energy supply to respiratorymuscles leads to fatigue• Discuss how hypercapnia may occur when excessive concentrationsof inspired oxygen are administered to susceptiblepersonsKey words: hypercapnia; neuromuscular diseases; respiratoryfailure; respiratory musclesComponents of the RespiratorySystemThe respiratory system can be considered astwo distinct parts: the lungs and a ventilatorypump. Disorders of the lungs interfere with gasexchange, manifested mainly by hypoxemia. Disordersof the pump impair ventilation, manifestedby hypercapnia. Arterial carbon dioxide is relatedto carbon dioxide output (V . co 2) by metabolism andis removed by alveolar ventilation (V . a). V . a is thetotal minute ventilation (V . e) minus dead spaceventilation (V . d). Paco 2is determined by the following:Paco 2 K(V . co 2/V . a), where K 0.863 mmHg; (V . a V . e V . d) or, rearranged, Paco 2K(V . co 2/V . e (1 V . d/tidal volume [Vt]).Increased V . CO 2Hypercapnia may occur when V . co 2in creaseswithout a corresponding increase in V . a.V . co 2increases in proportion to the metabolic rate,as with fever and exercise. Carbohydrate utilizationalso increases the respiratory quotient (orV . co 2/oxygen consumption) compared with fatand protein metabolism, producing more carbondioxide for each calorie expended. In patients withlimited ventilatory reserve, an increase in V . co 2mayculminate in an increase in Paco 2.Reduced V . AAny disorder causing a reduction in V . a willincrease Paco 2. This occurs with three types ofpathophysiologic derangements:• V . e (the product of respiratory rate and Vt) isreduced.• Vt is reduced. Even if the overall V . e is normalor high, the portion of each breath that is distributedto anatomic or physiologic dead space(dead space fraction, or V . d/Vt) increases, andV . a is reduced. If V . e does not increase proportionally,Paco 2increases. This occurs commonlyin patients with rapid, shallow breathing patterns.• V . d/Vt increases, with normal or high V . e andVt. An extreme example would be a patientwith fixed V . e who has a massive pulmonaryembolism. Here, although V . e and Vt remain thesame, the increase in V . d/Vt that results fromthe newly underperfused lung units reducesV . a and increases Paco 2.Disorders of VentilationCentral ControlThe medullary central controller sends signalsthat set the respiratory rate and Vt. This centerreceives inputs from higher CNS centers, peripheralchemoreceptors, and receptors in the lungs,chest wall, and airways. The carbon dioxide sensor(chemoreceptor) is located in the medulla andresponds to changes in pH of extracellular fluid.Oxygen sensors in the carotid body and aortic bodysense changes in Po 2. Disorders of central respiratorycontrol include the following: (1) congenitalcentral hypoventilation syndrome (CCHS, or<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 129ACC-PULM-09-0302-008.indd 1297/11/09 12:02:19 AM


Ondine’s curse) and inborn failure of autonomiccontrol of breathing, linked to mutations in thePHOX2B gene; (2) central sleep apnea; (3) overdoseswith narcotics or sedatives; (4) diseases ofthe medulla (infarct, tumor); (5) hypothyroidism;(6) metabolic alkalosis, in which the Paco 2usuallyincreases 0.7 to 0.8 mm Hg for every 1.0 mEq/Lincrease in plasma bicarbonate; and (7) rabies.Rabies is a viral infection transmitted in thesaliva of infected mammals that causes an encephalomyelitisthat is almost always fatal. In NorthAmerica, bats are implicated in transmission in 90% of cases. The virus may involve the entirenervous system in infected persons, and respiratoryfailure may occur from either central respiratorydepression or ascending spastic paralysis,which occurs in 20% of cases. Postexposure prophylaxiswith rabies immune globulin and rabiesvaccine is recommended for the unvaccinatedperson who is exposed to bat saliva. Because theconsequences of rabies are so devastating, postexposureprophylaxis should be considered even ifa bite cannot be reasonably excluded.Motor NeuronsInterruption of transmission impulses from therespiratory centers to the respiratory muscles alsoleads to hypoventilation. Motor neuron diseasesinclude the following: (1) spinal cord injury, especiallyat the C3 level; (2) tetanus; and (3) diseasesinvolving the anterior horn cells, including amyotrophiclateral sclerosis (ALS) and poliomyelitis. Intetanus, the anaerobe Clostridium tetani may inoculatea disruption of the skin barrier and elaboratethe exotoxin tetanospasmin, which reaches thespinal cord by retrograde axonal transport andblocks the release of inhibitory transmitters. Trismusis often the first symptom, which progressesto spastic paralysis, especially of the laryngeal andrespiratory muscles, causing respiratory failure thatmay persist for several months. Subcutaneous injectionof heroin, or “skin popping,” accounted for theincreasing incidence of tetanus and would botulismin California since 1997. In surgical wound infection,contraction of the muscles surrounding thewound could be the first sign. The disease is preventablethrough vaccination; patients with tetanusshould be treated with wound care, tetanusimmune globulin to neutralize remaining exotoxin,support of ventilation and hemodyna mics, andmedications to suppress muscle rigidity.ALS is characterized by the presence of dysarthria,tongue atrophy and fasciculations, amyotrophy(muscle atrophy), extremity fasciculations,weakness, and hyperreflexia. The diagnosis of ALSis confirmed with electrophysiologic studies.Although new cases of poliomyelitis are almostunheard of in developed nations, late manifestationsmay occur many years after the primaryinfection. The “postpolio syndrome” may occur inup to half of the survivors of paralytic poliomyelitis,with the onset of new weakness or abnormalmuscle fatigue, muscle atrophy, or generalizedfatigue. Muscle weakness may even lead to respiratoryfailure in a few patients.Peripheral NeuropathyDisorders of the peripheral nerves commonlycause hypercapnic respiratory failure. Guillain-Barré syndrome (GBS) is the most common causeof acute generalized paralysis after the virtualelimination of poliomyelitis around the world.There are at least four subtypes of GBS; the mostcommon is an acute, inflammatory demyelinatingpolyradiculopathy; others are acute motor andsensory neuropathies (alone or together), acutepandysautonomic forms, and overlap syndromes.GBS is believed to be the result of an aberrant T-cellresponse to a previous infection, leading to a macrophage-mediateddestruction of myelin. Themajority of cases involve either a flu-like illness orgastroenteritis within the previous 6 weeks, andapproximately one fourth of patients with GBS,especially those with axonal forms of the disease,have had a recent infection with Campylobacterjejuni.The lipopolysaccharide of C jejuni cell wallsshares similar antigens with peripheral nerve gangliosides.Patients usually present with ascendingparalysis, paresthesias, and hypoflexia or areflexia.The disorder usually is diagnosed by finding electromyography(EMG) evidence of conductionblock in motor nerves. Cerebrospinal fluid studiestypically show few or no cells, increased proteinafter 1 to 2 weeks, and circulating or cerebrospinalfluid antiganglioside antibodies in some patients.Plasmapheresis and infusion of -globulin appearto be equally effective, and the combination of both130 Hypercapnic Respiratory Failure (Rosen)ACC-PULM-09-0302-008.indd 1307/11/09 12:02:21 AM


treatments confers no additional benefit. There isno established role for therapy with corticosteroidsin this disorder.Critical illness polyneuropathy (CIP) is an acutesensory-motor disorder that mainly affects thelower limbs of critically ill patients. It may be theconsequence of microcirculatory damage causedby decreased perfusion or endothelial injury froma systemic inflammatory cytokine response. Criticalillness myopathy (CIM) is an acute myopathythat causes weakness and paralysis in critically illpatients and that usually occurs with CIP. BecauseCIP and CIM usually overlap, the term critical illnesspolyneuromyopathy is often used. Sepsis, multiorganfailure, and hyperglycemia often are citedas risk factors but cannot be confirmed because ofmethodologic limitations of investigations intothese disorders.These patients often present with “weaningproblems” after sepsis resolves, typically with flaccidparesis that spares the cranial muscles, muscleatrophy, and reduced or absent deep tendonreflexes. EMG shows axonal neuropathic and/ormyopathic patterns, and muscle biopsies (whenperformed) show denervation atrophy. If thepatient survives the critical illness, the syndromemay resolves over weeks to months, but manypatients remain permanently disabled. There is nospecific treatment, but efforts to prevent this disorderinclude minimizing the use of corticosteroids,and some studies have suggested thatrigorous glycemic control is associated with areduced incidence.Unusual causes of neuropathy include diphtheria,porphyria, and tick paralysis. Exposure totoxins also may cause severe neuropathy andparalysis. For example, fugu, a puffer fish, elaboratesa potent neurotoxin, and ingestion ofimproperly prepared fugu sushi accounts forapproximately 50 deaths in Japan each year.Neuromuscular JunctionDisorders/particulars interfering with transmissionof impulses at the neuromuscular junctioninclude the following:Myasthenia Gravis: In this disease, binding ofIgG antibodies to the postsynaptic acetylcholinereceptor induces complement-mediated receptordestruction and skeletal muscle weakness.Most cases are idiopathic, but the disease maybe induced by penicillamine and transientlyexacerbated by corticosteroids. The diagnosis issupported by the following: (1) impaired neuromusculartransmission on EMG, with decrementalresponse to repetitive stimulation; (2) clinicalresponse to cholinergic agents (edrophonium);and (3) the presence of acetylcholine receptor antibodiesin the serum (85 to 90% of patients withgeneralized weakness). Myasthenia gravis is associatedwith both benign and malignant thymoma.Treatments include cholinergic drugs, corticosteroidsfor mild-to-moderate disease (and sometimeswith other immunosuppressive agents),and plasma exchange or IV Ig for patients withsevere weakness. Plasma exchange and IV Ig areequally effective in exacerbations of myastheniagravis, and the role of these treatments in chronicdisease is unproven. Even in the absence of thymoma,thymectomy is generally recommendedfor patients with generalized myasthenia gravisbetween puberty and 60 years of age because it isassociated with a greater likelihood of remissionor improvement.Eaton-Lambert Syndrome: This paraneoplasticdisorder often is associated with small cell carcinomaof the lung. Weakness is mainly proximaland spares the ocular and bulbar muscles. Unlikemyasthenia, EMG in Eaton-Lambert syndromeshows an incremental pattern with repetitivestimulation.Organophosphates: Organophosphates, a com ponentof some insecticides, inhibit choline-esterase.Botulism: The spores of Clostridium botulinumgenerate a potent neurotoxin that causes a potentiallylethal, descending flaccid paralysis. Patientstypically present with oculomotor palsies, progressingto facial muscle paralysis and difficultyswallowing, and then limb and respiratorymuscle weakness and paralysis. Although botulismis best known to be acquired by ingestionof preserved food containing preformed toxin,a similar syndrome may occur when the organismis inoculated into a wound or other devitalizedtissue; it may also be acquired by inhalation,making it a potential bioterrorist weapon. Similarto tetanus, a dramatic increase in reported casesof wound botulism was linked to subcutaneousor IM injection of “black tar” heroin in California.Like Eaton-Lambert syndrome, there is<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 131ACC-PULM-09-0302-008.indd 1317/11/09 12:02:21 AM


augmentation of muscle action potential withrepetitive nerve stimulation, although to a lesserdegree. The diagnosis is established by identifyingthe toxin in the serum, stool, or wound, or theorganism in stool or wound. Treatment includesdebridement, botulinum antitoxin, and supportivecare.Prolonged Muscle Weakness: Prolonged muscleweakness after nondepolarizing neuromuscularblocking agents is common. Discontinuingmechanical ventilation may be delayed, andweakness may persist for months. Two typesof neuromuscular dysfunction occur. Persistentblockade of the neuromuscular junctionmay occur as the result of accumulation of thedrug or active metabolites, especially in patientswith renal failure. This occurs more commonlywith aminosteroid agents (pancuronium andvecuronium), in which renal excretion of activemetabolites is important, than with the benzylisoquinoloneagents (atracurium, doxacurium,and cisatracurium), which are inactivated spontaneouslyin the blood. The second is an acute,generalized, necrotizing myopathy that clinicallyresembles and may be indistinguishable fromCIM in the absence of these agents, except thatthese patients generally have intact sensation,usually have greater creatine kinase levels, andnerve conduction is normal. This complication ismost common in patients who are receiving highdoses of corticosteroids along with neuromuscularblocking agents, including the benzylisoquinolones.Treatment is supportive, but patientsmaybe incapacitated for months after recoveryfrom the acute illness. Minimizing the dose andduration of paralytic agents and corticosteroids isthe best way to prevent this complication.Respiratory MusclestheThe respiratory muscles perform the work ofbreathing. The diaphragm is the most importantinspiratory muscle, responsible for 60 to 90% ofwork of breathing at rest. Contraction of the costalcomponent displaces abdominal viscera downwardand lifts the lower rib cage; the crural componentdisplaces abdominal viscera downward.The external intercostals, scaleni, and sternocleidomastoidscomprise the accessory muscles of inspirationand are used when V . e is increased or whendiaphragm is weak or fatigued. Contraction ofthese muscles elevates the rib cage.The expiratory muscles (internal intercostal andabdominal) are not used during resting breathingbecause exhalation is passive. However, they contractactively when V . e is increased and are essentialto the cough reflex. The internal intercostals depressthe ribs, and the abdominal muscles depress thelower ribs and pull the abdominal wall inward.Muscle WeaknessWeakness is defined as a reduced capacity of arested muscle to generate an expected force. Weakenedinspiratory muscles may be incapable ofperforming the work of breathing, leading tohypercapnia. Expiratory muscle weakness impairsthe cough reflex, promoting retention of secretionsand pneumonia. Causes of muscle weakness arelisted in Table 1.As part of normal aging, respiratory musclestrength generally declines in parallel with overallmuscle performance. Malnutrition causes allmuscles, including those of respiration, to becomeatrophic and weak, causing predisposition to respiratoryfailure. Respiratory muscles may becomeweak as the result of denervation, myopathy, andendocrinopathies. Metabolic disorders, includingsevere abnormalities in serum potassium,magnesium, and phosphate, may interfere withcontractile function, causing inadequate V . e andTable 1. Causes of Respiratory Muscle WeaknessAgingMalnutritionDenervationMyopathyMuscular dystrophyPolymyositisDrug-induced (neuromuscular blocking agents,corticosteroids)Endocrinopathy (hyperthyroidism, Cushing syndrome)MetabolicHypokalemiaHyperkalemiaHypophosphatemiaHypomagnesemiaHypermagnesemiaAcidosisHyperinflation132 Hypercapnic Respiratory Failure (Rosen)ACC-PULM-09-0302-008.indd 1327/11/09 12:02:21 AM


hypercapnia. When acidosis impairs the contractileforce of respiratory muscles, a positive feedbackloop of respiratory muscle weakness and respiratoryacidosis may be established.Hyperinflation places the inspiratory musclesat mechanical disadvantage, largely because of areduction in muscle length. Inspiratory muscleweakness with hyperinflation is more pronouncedin acute disorders like severe asthma than inCOPD. With time, remodeling of sarcomeresrestores them to normal length, and the proportionof slow-twitch muscle fibers increases. These adaptiveresponses improve contractile function andincrease resistance to fatigue. Also, many patientswith COPD lose weight; these patients haveincreased circulating levels of tumor necrosis factor-,which promotes muscle wasting and weakness.The benefits of lung volume reductionsurgery (LVRS) in patients with severe COPD areprobably attributable to enhanced ventilatorypump function; LVRS reduces hyperinflation andair trapping, improving diaphragm mechanics andincreasing V . a.Assessment of Respiratory Muscle StrengthIn clinical practice, the strength of respiratorymuscles usually is measured with maximal pressuresat the mouth against a closed airway (maximuminspiratory pressure, maximum expiratorypressure). The “gold standard” measurement, usedmore in scientific investigations than in clinicalpractice, is transdiaphragmatic pressure (Pdi) duringa maximal inspiratory effort. Pdi is calculatedas the difference between abdominal pressuremeasured with a gastric balloon catheter, and pleuralpressure measured in the esophagus.Respiratory Muscle FatigueFatigue is defined as a reduced capacity togenerate an expected force, which is corrected withrest. Fatigue can be classified as central, transmissional,and contractile. Central fatigue describes areversible decrease in central neural respiratorydrive caused by overuse of muscles. It may bemotivational when strength can be restored byvoluntary effort and nonmotivational when strengthis not restored by increased effort but the muscleresponds normally to electrical stimulation. Centralfatigue is probably a consequence of reflex inhibition,influenced by afferent and cortical signals,perhaps endogenous opioids. Transmission fatigueis a reversible, exertion-induced impairment ofneural transmission; its pathophysiology and significanceare unknown.Contractile fatigue is a reversible impairmentin contractile response to neural impulses notcaused by drugs or alteration in length/tension orforce/velocity relationships. It occurs when energydemands on contracting muscles exceed the energysupply. Energy demand is a function of the workof breathing (V . e, compliance, and resistance) andthe efficiency of the respiratory system. Fatiguetypically occurs when the tension time index is 0.15. The tension-time index is defined as follows:Pdi/maximal Pdi Ti/Tt, where Ti durationof inspiration, and Tt duration of onerespiratory cycle.Put more simply, fatigue occurs whenP breath/ MIP Ti/Tt is excessive, where P breath inspiratory pressure for a given breath, and MIP maximal inspiratory pressure. Therefore, a weakmuscle (where maximal Pdi and MIP are reduced) ismore likely to fatigue and fail than a normal muscle.The efficiency of inspiratory muscle contractionis impaired by lung volume, the presence of neuromusculardisease, and malnutrition. Body positionalso influences contractile efficiency; inspirationrequires less effort in the upright position becausethe abdominal contents are displaced downwardby gravity, unloading the diaphragm. Patients withlung disease and an increased work of breathingusually are more comfortable in an upright comparedwith supine position. This is especially truefor patients with massive obesity, when assumingthe supine position may precipitate respiratoryarrest because the diaphragm cannot overcome theintolerable load imposed by the weight of theabdominal contents.The energy supply to muscles varies directlywith blood flow. Shock often culminates in respiratoryfailure because blood flow and oxygen supplyto contracting respiratory muscles are reduced. Incardiogenic shock, hypercapnic respiratory failureoccurs because pulmonary edema increases thework of breathing, and impaired hemodynamicfunction reduces blood flow to the respiratorymuscles. Hypoxemia may compromise oxygensupply to respiratory muscles, but hypoxemia<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 133ACC-PULM-09-0302-008.indd 1337/11/09 12:02:22 AM


alone does not usually cause muscle fatigue; oxygenconsumption can be maintained by increasedcardiac output and augmented oxygen extractionfrom capillary blood.When respiratory muscles are fatigued, theEMG power spectrum is shifted, with increasedlow-frequency and decreased high-frequency components.Dyspnea and tachypnea are common.Respiratory alternans, the alternating recruitmentand derecruitment of the diaphragm and otherinspiratory muscles, often precedes the onset ofparadoxical abdominal motion with respiration.Treatment of fatigue depends on correcting theunderlying disorders and rest. If the underlyingcondition that provoked muscle fatigue is not readilyreversible, the patient will require mechanicalventilation until adequate rest is achieved. Pharmacologicapproaches to improve respiratorymuscle function have been investigated. Althoughaminophylline, digoxin, and dobutamine have allbeen demonstrated to increase the contractility ofthe diaphragm, their clinical utility in patients withrespiratory muscle fatigue is unproven. Similarly,the role of strength and endurance training of therespiratory muscles has been under investigationfor decades, but there is still no good evidence tosupport its use.Hypercapnia With Oxygen AdministrationExcessive oxygen administration may worsenpreexisting hypercapnia, especially in patientswith COPD. Mechanisms include the following:(1) increased physiologic dead space. This factor isan important one, but the precise explanation forthis phenomenon is not completely understood;(2) attenuation of hypoxic ventilatory drive;patients with chronic hypercapnia become insensitiveto increments in Paco 2; and (3) the Haldaneeffect, in which oxygen releases co 2bound tohemoglobin, increasing Paco 2.Primary Disorders of the Chest WallSevere kyphoscoliosis, obesity, thoracoplasty,and pleural thickening all may cause hypoventilationby one of two general mechanisms, which canbe categorized as “can’t breathe” and “won’tbreathe.” Significant increase in the work of breathingcaused by severe mechanical derangements ofthe chest wall may lead to reduced Vts and increasedV . d/Vt (can’t breathe). Many patients with chestwall restriction have central carbon dioxide nonresponsiveness,either as a primary disorder or tocompensate for increased work of breathing (won’tbreathe). In acute illness, respiratory muscle fatigueimposed by increased work of breathing may precipitatefrank hypercapnic respiratory failure.The interplay of increased work of breathingand diminished respiratory drive is a prominentfeature of the obesity-hypoventilation syndrome(OHS), defined as a combination of obesity (bodymass index 30 kg/m 2 ) and hypercapnia (Paco 2 45 mm Hg) during wakefulness in the absenceof other known causes of alveolar hypoventilation.In many patients, severe obesity is associatedwith chronic daytime hypercapnia and hypoxemia,erythrocytosis, and right ventricular failure.Most (but not all) patients with OHS also haveobstructive sleep apnea, but only a minority ofobese patients with obstructive sleep apnea haveOHS. Patients with OHS have a blunted centraldrive in response to hypercapnia and hypoxemia,possibly related to decreased circulating levels orreceptor hyporesponsiveness to leptin, a hormonethat acts on the hypothalamus to suppress appetiteand probably also on central respiratorycenters to maintain hypercapnic ventilatoryresponses. Obesity-related upper airway narrowing,dependent atelectasis, and excessive loadingof respiratory muscles may all be contributingfactors.Annotated BibliographyCaruana-Montaldo B, Gleeson K, Zwillich CW. Thecontrol of breathing in clinical practice. Chest 2000;117:205–225<strong>Review</strong> of the physiology of each component of the integratedsystems that control breathing, with cases illustrating disordersof these systems.Laghi F, Tobin MJ. Disorders of the respiratory muscles.Am J Respir Crit Care Med 2003; 168:10–48Very thorough review of the role of respiratory muscles inacute respiratory failure; specific neuromuscular diseases;and the influence of chest wall, systemic diseases, and surgeryon respiratory muscle function.Weinberger SE, Schwartzstein RM, Weiss JW.Hypercapnia. N Engl J Med 1989; 321:1223–1231134 Hypercapnic Respiratory Failure (Rosen)ACC-PULM-09-0302-008.indd 1347/11/09 12:02:22 AM


Concise summary of mechanisms, specific disorders, andtherapeutic strategies.Neuromuscular DisordersBunch TJ, Thalzi MK, Pillikka PA, et al. Respiratoryfailure in tetanus: case report and review of a 25-yearexperience. Chest 2002; 122:1488–1492Centers for Disease Control and Prevention. Humanrabies: Alberta, Canada, 2007. MMWR Morb MortalWkly Rep 2008; 57:197–200Describes clinical picture, management, and preventionstrategies.Cherington M. Clinical spectrum of botulism. MuscleNerve 1998; 21:701–710Deem S, Lee CM, Curtis JR. Acquired neuromusculardisorders in the intensive care unit. Am J Respir CritCare Med 2003; 168:735–739Hermans G, Wilmer A, Meersseman W, et al. Impact ofintensive insulin therapy on neuromuscular complicationsand ventilator dependency in the intensive careunit. Am I Respir Crit Care Med 2007; 175:480–489Horowitz BZ. Botulinum toxin. Crit Care Clin 2005;21:825–839Hughes RAC, Cornblath DR. Guillain-Barré syndrome.Lancet 2005; 366:1653–1666Jubelt B, Agre JC. Characteristics and management ofpostpolio syndrome. JAMA 2000; 26:412–414Latronico N, Peli E, Botteri M. Critical illness myopathyand neuropathy. Curr Opin Crit Care 2005; 11:126–132Leatherman JW, Fluegel WL, David WS, et al. Muscleweakness in mechanically ventilated patients withsevere asthma. Am J Respir Crit Care Med 1996;153:1686–1690Muscle weakness occurred in 20 of 69 patients who receivedboth a neuromuscular blocking agent and corticosteroidscompared with none of the 38 patients who received corticosteroidsalone. Weakness correlated with the duration ofparalysis, and the incidence of weakness was not reduced inpatients who received atracurium compared with those whoreceived pancuronium or vecuronium.Pujar T, Spinello IM. A 38-year-old woman with heroinaddiction, ptosis, respiratory failure and proximalmyopathy. Chest 2008; 134:867–870Concise review with table summarizing the differential diagnosisof neuromuscular diseases that cause respiratory failure.Schweickert WD, Hall J. ICU-acquired weakness. Chest2007; 131:1541–1549Schwendimann RN, Burton E, Minagar A. Managementof myasthenia gravis. Am J Ther 2005; 12:262–268These articles are cited because they are relatively currentand offer detailed discussions regarding neuromuscular disordersthat may cause respiratory failure.Werner SB, Passaro D, McGee J, et al. Wound botulismin California, 1951–1998: recent epidemic in heroininjectors. Clin Infect Dis 2000; 31:1018–1024The upsurge in diagnoses have been attributed to subcutaneousdrug injection (skin popping). Abscess formation anddevitalized tissue provide a favorable milieu for growth ofthe organism. This article also provides a concise review ofother paralytic diseases discussed in this chapter.Respiratory MusclesAmerican Thoracic Society/European RespiratorySociety. Skeletal muscle dysfunction in chronic obstructivepulmonary disease: a statement of the AmericanThoracic Society and European Respiratory Society.Am J Respir Crit Care Med 1999; 159(Suppl):S1–S40Comprehensive review of mechanisms.Cohen CA, Zagelbaum G, Gross D, et al. Clinical manifestationsof inspiratory muscle fatigue. Am J Med1982; 73:308–316Progression of clinical findings leading to respiratory failure.Criner G, Cordova FC, Leyenson V. Effect of lung volumereduction surgery on diaphragm strength. Am JRespir Crit Care Med 1998; 157:1578–1585Lando Y, Boiselle PM, Shade D, et al. Effect of lung volumereduction surgery on diaphragm length in severechronic obstructive pulmonary disease. Am J RespirCrit Care Med 1999; 159:796–805Polkey MI, Moxham J. Clinical aspects of respiratorymuscle dysfunction in the critically ill. Chest 2001;119:926–939<strong>Review</strong>s respiratory muscle physiology, evaluation of muscledysfunction, and disorders that lead to prolonged ventilatordependency.Roussos C, Macklem PT. The respiratory muscles. NEngl J Med 1982; 307:786–797Develops the concept of “pump failure,” reviewing the physiologyand clinical features of muscle mechanics, energetics,and fatigue.Shade D, Cordova F, Lando Y, et al. Relationshipbetween resting hypercapnia and physiologic parametersbefore and after lung volume reduction surgeryin severe chronic obstructive pulmonary disease. Am JRespir Crit Care Med 1999; 159:1405–1411These three studies from the same center show that LVRSincreases the length of the diaphragm and improvesits strength, the changes correlate with postoperative<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 135ACC-PULM-09-0302-008.indd 1357/11/09 12:02:23 AM


improvements in exercise capacity and maximum voluntaryventilation, and patients with hypercapnia have reductionsin Paco 2as ventilatory pump function and V . e improve.Hypercapnia After Oxygen AdministrationAubier M, Murciano D, Milic-Emili J, et al. Effects of theadministration of O 2on ventilation and blood gases inpatients with chronic obstructive pulmonary diseaseduring acute respiratory failure. Am Rev Respir Dis1980; 122:747–754Hanson CW, Marshall BE, Frasch HF, et al. Causesof hypercarbia with oxygen therapy in patients withchronic obstructive pulmonary disease. Crit Care Med1996; 24:23–28This study uses a computer model of the pulmonary circulationwith data from the study of Aubier et al to evaluate thefactors contributing to hypercarbia after oxygen administrationin patients with COPD. The discussion offers a lucidexplanation of the pathophysiology of this phenomenon.Robinson TD, Freiberg DB, Regnis JA, et al. The roleof hypoventilation and ventilation-perfusion redistributionin oxygen-induced hypercapnia during acuteexacerbations of chronic obstructive pulmonary disease.Am J Respir Crit Care Med 2000; 161:1524–1529These studies exploring the mechanisms of hypercapnia afteroxygen administration in patients with COPD. The formershows that increased wasted ventilation is the major factorwhereas the latter holds that both increased dead space andreduced total ventilation are responsible.OHSKessler R, Chaouat A, Schinkewitch P, et al. The obesity-hypoventilationsyndrome revisited: a prospectivestudy of 24 consecutive cases. Chest 2001; 120:369–376Most patients with OHS also have obstructive sleep apnea.In addition to daytime hypercapnia, OHS is likely associatedwith pulmonary hypertension and severe hypoxemia.Malhotra A, Hillman D. Obesity and the lung: 3—obesity, respiration and intensive care. Thorax 2008;63:925–931Thorough review of physiology and complications of obesitywith a broad discussion of critical illness in thesepatients.Mokhlesi B, Tulaimat A. Recent advances in obesityhypoventilation syndrome. Chest 2007; 132:1322–1336<strong>Review</strong>s pathophysiology and treatment options.136 Hypercapnic Respiratory Failure (Rosen)ACC-PULM-09-0302-008.indd 1367/11/09 12:02:23 AM


Thoracic ImagingRakesh D. Shah, MD, FCCPObjectives:• <strong>Review</strong> the basic plain radiograph and cross-sectionalanatomy and pathology of the tracheobronchial tree, lobaratelectasis, and the mediastinum• Evaluate the role of CT pulmonary angiography in thediagnosis of both acute and chronic pulmonary thromboembolicdisease• <strong>Review</strong> the current methods for evaluation of solitarypulmonary nodule• Learn to interpret high-resolution CT scan images in patientswith diffuse lung diseaseKey words: chest imaging; CT pulmonary angiography;high-resolution CT scan; positron emission tomographyThis chapter is divided into six broad sections:(1) evaluation of the nonneoplastic diseases ofthe tracheobronchial tree, (2) lobar atelectasis,(3) evaluation of mediastinal structures and pathology,(4) CT pulmonary angiography in the diagnosisof acute and chronic thromboembolic disease,(5) current concepts in the evaluation of solitarypulmonary nodule, and (6) high-resolution CT(HRCT) scan findings in patients with diffuse lungdisease.Tracheobronchial TreeFocal or diffuse lesions of the tracheobronchialtree are produced by a variety of diseases. Theetiologies include infection, malignancy, trauma,collagen vascular disease, and idiopathic entitiessuch as amyloidosis and tracheobronchopathiaosteochondroplastica. These conditions may producesymptoms of cough, dyspnea, wheezing, orstridor. Despite significant symptomatology, airwayabnormalities frequently are not apparent oroften are overlooked on chest radiographs, whichresults in a delay of the diagnosis. CT scan of thechest is the imaging modality of choice if there isa clinical suspicion of tracheobronchial abnormality.The following is a brief description of theimportant nonneoplastic diseases that are associatedwith tracheal narrowing.Tracheal StenosisMost tracheal stenoses are complications oftracheal intubations. Narrowing occurs at the thoracicinlet where the cuff has been inflated fortracheostomy or intubation. As the cuff pressureexceeds the capillary pressure, blood supply to thetracheal mucosa is compromised. This compromiseleads to inflammation, followed by ulceration andnecrosis. Eventually, scarring occurs and leads tostenosis. On CT, 1- to 2-cm long circumferentialnarrowing of the trachea is noted at the thoracicinlet.TracheobronchomalaciaTracheobronchomalacia is a condition definedby excessive expiratory collapse of the trachea andbronchi. It is a result of weakness of the airwaywalls or supporting cartilage. Some cases are congenitalin nature; however, most cases are acquiredand are caused by previous intubation, trauma,infection, or chronic inflammation. On CT, diagnosisis suggested when greater than a 50% decreasein the cross-sectional area of the airway lumen isnoted on dynamic expiratory images.Saber-Sheath TracheaSaber-sheath trachea is characterized bymarked decrease in the transverse diameter of theintrathoracic trachea associated with an increasein its sagittal diameter. Changes in tracheal configurationare a result of abnormal intrathoracictransmural pressures. It is found in men whosmoke and have COPD.Tracheobronchopathia OsteochondroplasticaIt is a rare, benign disease characterized bydevelopment of osseous or cartilaginous noduleswithin the anterolateral walls of the trachea. Theposterior membrane of the trachea is sparedbecause of the absence of cartilage in this area. It<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 137ACC-PULM-09-0302-009.indd 1377/10/09 8:07:12 PM


occurs in men older than 50 years and usually isdetected incidentally. On CT, calcified nodulesprotruding into the tracheal lumen are noted.Resultant tracheal narrowing may be present.Relapsing PolychondritisRelapsing polychondritis is a rare inflammatorydisease that affects the cartilages of the ear,nose, respiratory tract, and joints. It is characterizedby repeated episodes of cartilaginous inflammation,leading to loss of structure and fibrosis. Therespiratory tract is affected in approximately onehalf of all the patients. On CT, thickening of the anterolateraltracheal wall with sparing of the posteriormembrane is noted. Resultant luminalnarrowing and airway collapse are best demonstratedon dynamic expiratory CT scans.AmyloidosisAmyloidosis is a rare condition characterizedby deposition of insoluble protein in the extracellulartissues. It may involve any portion of therespiratory tract. Within the lung parenchyma,abnormal amyloid deposition can appear as singleor multiple nodules or diffuse interstitial opacities.Deposits within the tracheobronchial tree lead toconcentric or nodular thickening of the trachealsubmucosa with resultant narrowing of the lumen.Calcification and/or ossification of the lesions mayoccur.Wegener GranulomatosisIt is a necrotizing granulomatous vasculitis thatinvolves the upper and lower respiratory tract.Involvement of the lung parenchyma shows multiplenodules with or without cavitation. Involvementof the tracheobronchial tree is rare andusually presents late in the disease. On CT, circumferentialthickening, ulceration, and luminal narrowingof the trachea are noted.Mounier-Kuhn syndrome(Tracheobronchomegaly)Tracheobronchomegaly, also referred to asMounier-Kuhn syndrome, is a rare condition characterizedby diffuse dilatation of the trachea and themain bronchi. It is thought to result from atrophyof muscular and elastic tissue found in both tracheaand main bronchi. On CT, thin wall trachea withscalloped or corrugated appearance, increasedtracheal diameter of 3 cm, and diverticulosis arepresent.In summary, diagnosis of the nonneoplasticdiseases of the tracheobronchial tree requiresknowledge of the anatomy, and observation of thefollowing on CT scans:1. Is the trachea dilated or narrowed?2. If there is thickening of the airway wall? If so,is it focal or diffuse? Is the thickening at thethoracic inlet or not?3. Is there stenosis? If so is it focal or diffuse?4. Is calcification present or not?5. Is ulceration present or not?6. Is the posterior membrane involved or spared?7. Is there collapse of the tracheal wall on dynamicexpiratory images?The differential diagnosis can be significantlynarrowed once the above questions have beenanswered.AtelectasisAtelectasis is defined as decrease in volume oflung or a portion of the lung. Resorption, passive,cicatrization, and adhesive are the four types ofatelectasis that can be explained by the mechanismin which the loss of lung volume occurs. Resorptiveatelectasis is the most common type and resultsfrom absorption of gas from the alveoli when thecommunication between the alveoli and the tracheais obstructed by an endobronchial lesion or amucus plug. Passive atelectasis is caused by extrinsicpressure on the lung from a large pleural effusion,pneumothorax, or mass resulting in collapse.Cicatrisation atelectasis occurs in patients with scarringor pulmonary fibrosis. The exact mechanismof adhesive atelectasis is poorly understood but isthought to be caused by widespread collapse ofalveoli. It is usually seen in patients with respiratorydistress syndrome and in those who are recoveringfrom surgery.Lobar atelectasis may be complete or partial.The most common cause of lobar atelectasis isobstruction by a central endobronchial lesion.However, in hospitalized patients and particularly138 Thoracic Imaging (Shah)ACC-PULM-09-0302-009.indd 1387/10/09 8:07:13 PM


in those who are intubated, a mucus plug is themost common cause of partial or lobar atelectasis.Primary and secondary signs on chest radiographhelp identify the atelectasis and site of endobronchialobstruction. The major sign of lobar atelectasisis opacification of the affected lobe due toairlessness and displacement of the interlobarfissure. Secondary signs of atelectasis includedisplacement of the mediastinal structures, elevationof the hemidiaphragm, decrease in the distanceof the intercostals spaces, displacement ofthe hila, and compensatory overinflation of theremaining lung. Discussion of the major andminor radiographic features of the main lobaratelectasis follow.Right Upper Lobe AtelectasisOn the frontal chest radiograph, the rightupper lobe collapses superiorly and medially, creatinga wedge-shaped opacity in the upper righthemithorax. The major fissure is displaced anteriorlyand the minor fissure is displaced upward.Right upper lobe collapse secondary to a centralcarcinoma may produce a characteristic appearanceon the frontal chest radiograph, termed the“reverse S-sign” of Golden. On the lateral projection,the collapsed lobe may appear as a triangularopacity with its apex at the hilum and its base atthe apex of the hemithorax.Left Upper Lobe AtelectasisWhen the left upper lobe collapses, the frontalchest radiograph demonstrates a hazy opacificationin the left perihilar area with partial obscurationof the left heart border. Sometimes, theresultant overinflation of the superior segment ofthe left lower lobe inserts between the apex of theatelectatic upper lobe and the aortic knob, creatinga sharp lucent interface called the Luftsichel sign.On the lateral radiograph, there is forward displacementof the major fissure that is almost parallelto the anterior chest wall.Lower Lobe AtelectasisThe pattern of lower lobe collapse is similar inboth lungs because of the equivalent anatomybilaterally. Both lower lobes collapse posteromediallyand inferiorly. On the frontal radiograph, atriangular opacity is visualized in the paraspinallocation of the lower hemithorax, while on the lateralprojection, increased opacity overlying the lowerthoracic vertebral bodies and loss of visualizationof the posterior left hemidiaphragm are noted.Right Middle Lobe AtelectasisOn the frontal chest radiograph, the rightmiddle lobe collapse shows a vague opacity in thelower right hemithorax with obliteration ofthe right heart border. On the lateral projection, thecollapse is seen as a linear band or triangularopacity overlying the cardiac silhouette as a resultof displacement of the minor fissure inferiorly andmajor fissure superiorly.Rounded AtelectasisIt is an unusual form of passive atelectasis thatis thought to occur from pleural fibrosis as a resultof chronic pleural effusion and/or asbestos-relateddisease. The pleural fibrosis causes folding of theadjacent lung parenchyma that appears as a focal,rounded opacity. On CT, a peripherally situatedrounded opacity with associated pleural thickening,“comet-tail” sign, and volume loss of theaffected lobe are noted.<strong>Pulmonary</strong> Embolism<strong>Pulmonary</strong> embolism (PE) is an importantcause of morbidity and mortality, particularly inthe postsurgical and hospitalized patients. Unfortunately,the diagnosis of acute PE is often delayedor unrecognized because of the nonspecific natureof its symptoms, signs, and laboratory test findings.The chest radiograph is often nonspecific ornormal. Ventilation/perfusion scans have highsensitivity but low specificity, whereas pulmonaryangiography is invasive, underused, and has considerableinterobserver variability.Pathophysiology of PEPE refers to the transport of venous thrombusto the pulmonary artery circulation. Most thrombiarise from clots within the deep venous system of<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 139ACC-PULM-09-0302-009.indd 1397/10/09 8:07:13 PM


the lower extremities. Emboli are often multiple,and they have predilection for the lower lobes ofthe lung. This predilection is likely related toincreased blood flow in the lower lobes.CT TechniqueIn recent years, great advances have been madein the CT technology. Because of its ability todirectly visualize PE noninvasively and with highaccuracy, multirow detector CT scanners havebecome the imaging test of choice for the diagnosisof PE. One of the other potential advantages is theability to diagnosis alternative diagnosis that maymimic PE clinically.On multirow detector CT scanners, approximately80 to 100 mL of iodinated IV contrast isinjected at 4 mL/s after a fixed scan delay or a timingbolus of 20 mL. The patient is scanned in asingle breath hold from the lung apices to the levelof the diaphragm. Approximately 250 to 400images per scan are obtained because the study isperformed at 1.25-mm collimation. Thus, the CTinterpretation is performed on computer workstationby actually scrolling through the images andaltering the window width and level to optimallyevaluate the pulmonary arteries.CT FindingsThe CT findings of acute PE are direct visualizationof low-attenuation thrombus within thecontrast-opacified pulmonary artery. Thrombusmay be identified as partial or complete fillingdefect in the vascular lumen. Secondary signs ofPE include oligemia of the affected segment andpulmonary infarcts.Pitfalls in the diagnosis of PE can be classifiedinto anatomic and technical etiologies. Commonanatomic pitfalls include lymph nodes, pulmonaryveins, impacted bronchi, and volume averaging ofpulmonary arteries. On the computer workstation,if one scrolls in and out from the main pulmonaryartery, following each of the lobar, segmental, andsubsegmental arteries, it is impossible to mistakean artery for a vein, lymph node, or mucoidimpacted bronchi for PE. Technical causes for suboptimalquality examinations are poor enhancementof pulmonary arteries, breathing motionartifact, and excessive noise in large patients. Bolustracking software built into most CT scanners hassignificantly reduced nondiagnostic scans frompoor enhancement. In addition, motion artifact isless of a problem with greater detector scannersbecause of the shorter breath holds needed.Chronic ThromboembolismChronic thromboembolism is scarring of thepulmonary arteries after lysis of thrombi fromeither single or repeated bouts of PE. As a result,webs and occlusions are formed within thebranches of the pulmonary arteries. The globaleffect of these is the overall increase of the pulmonaryvascular resistance and subsequent pulmonaryhypertension.Characteristic CT findings can be classified intothree categories: cardiac, pulmonary arteries, andparenchymal. The cardiac findings include rightsidedchamber enlargement and right ventricularhypertrophy. <strong>Pulmonary</strong> artery findings includeeccentric thrombus, abrupt vessel cut-off, webs,beaded vessels, and enlarged central arteries.Parenchymal findings include alternating hazy andlucent areas in lung called mosaic attenuation andenlarged bronchial arteries.MediastinumThe mediastinum is an anatomic region withinthe thoracic cavity that is bound laterally by thelungs, anteriorly by the sternum, and posteriorlyby the vertebral bodies. A wide variety of focaland diffuse abnormalities occur within the mediastinum.To facilitate the differential diagnosis,the mediastinum has been divided into severalcompartments, primarily based on landmarksnoted on the chest radiographs and CT rather thantrue anatomic fascial planes. They include thefollowing: (1) anterior mediastinum, whichincludes the retrosternal clear space and cardiophrenicangle; (2) the middle mediastinum, whichincludes the retrosternal clear space, subcarinalregion, and retrocardiac clear space; and (3) posteriormediastinum.Retrosternal Clear SpaceThis is the region that is posterior to the sternumand anterior the aorta and great vessels. It140 Thoracic Imaging (Shah)ACC-PULM-09-0302-009.indd 1407/10/09 8:07:13 PM


corresponds to the region of the anterior junctionline noted on the frontal chest radiographs. Normalstructures that are present in this locationinclude the thymus gland, lymph nodes, and fat.The differential diagnoses of lesions in this spaceinclude lesions of thymic origin (thymoma, thymiccyst); lymphoma; teratoma; aortic aneurysm;lipomatosis; and sternal lesions. In adults, thymomasaccount for the majority of the lesions inthis space. On CT, they appear as round or lobulatedsoft-tissue density lesions. Most of themenhance homogeneously after IV contrast administrationand may contain calcification. Mostof the thymomas are encapsulated and consideredbenign, but roughly 30% demonstrate invasionthrough the fibrous capsule and areconsidered malignant. Imaging features that supportmalignancy include the following: (1) invasionof mediastinal structures, including thesuperior vena cava and great vessels; (2) chestwall invasion; and (3) contiguous spread alongpleural surfaces.Cardiophrenic AngleThis region is situated anterior and to the rightof the heart. Typically, this region contains fat andlymph nodes. Most of the lesions occurring in thisspace are benign and include prominent fat,lipoma, pericardial cyst, and Morgagni hernia.Occasionally, they may contain lymph nodes fromlymphoma or metastasis.Retrotracheal Clear SpaceThis region is posterior to the trachea, anteriorto the thoracic spine, and superior to the posteriorportion of the aortic arch. It corresponds to theregion of the posterior junction line as noted on thefrontal chest radiograph. Normal structures thatare present in this space include esophagus andlymph nodes. The differential diagnosis of lesionsin this region includes abnormalities of the esophagus(tumor, achalasia, Zenker’s diverticulum).However, a large percentage of lesions in thisregion are vascular (aberrant right subclavianartery) or thyroid in origin. The majority of thethyroid masses represent goiters, which almostalways extend inferiorly from the thyroid glandinto this space.Subcarinal RegionThis region is inferior to the carina and superiorto the left atrium. Fat, lymph nodes, and theesophagus are normal structures that live in thisspace. Differential diagnoses of lesions in thisregion include lymphadenopathy, bronchogeniccyst, esophageal diverticula, and esophagealtumors. Lymphadenopathy is the most commonlesion in this space. It may be neoplastic, inflammatory,or infectious in nature. CT is very helpfulin detecting and characterizing the lymph nodes.On imaging, the nodes may demonstrate homogeneousenhancement, hyperenhancement, calcificationor low-density (fat or necrosis) center. Whenpresent, these lymph node characteristics significantlyhelp shorten very lengthy differential diagnosisof subcarinal adenopathy.Retrocardiac Clear SpaceThis region is posterior to the heart and anteriorthe thoracic spine. The esophagus, aorta, azygosvein, fat, and lymph nodes are present in thisregion. Differential diagnoses include esophageallesions (duplication cyst, varices, hiatal hernia,tumor); aortic aneurysm; and lymphadenopathy.Posterior MediastinumIt is defined as the area posterior to the anteriormargin of the vertebral bodies. Normal structuresthat are present in this location include the vertebralcolumn, fat, and nerves. The majority of thelesions occurring in this space are neurogenic (congenital,malignant, infection) in nature. MRI is thepreferred imaging modality of choice for evaluationof posterior mediastinal lesions because of itsability to demonstrate intraspinal extension oftumor and detect spinal cord abnormalities.Solitary <strong>Pulmonary</strong> NoduleA solitary pulmonary nodule (SPN) is definedas a well-circumscribed round lesion that is surroundedby lung parenchyma measuring 3 cm.A larger lesion with the same characteristics iscalled a mass. A SPN can be solid, ground glass, orpart solid (containing both ground-glass and solidcomponents). The differential diagnoses are many,<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 141ACC-PULM-09-0302-009.indd 1417/10/09 8:07:13 PM


including benign and malignant neoplasms,granuloma, infection, vascular abnormalities, andintrapulmonary lymph nodes.Once a nodule is detected on plain chest radiographs,comparison with previous films, if available,is recommended. If none are available, thenHRCT scans at 1- to 2-mm thick sections throughthe nodule should be obtained. This scanning ishelpful in detecting calcification, fat, and in definingthe morphology of the nodule.Morphologic characteristics of SPN often arenonspecific. For example, malignant lesions can beround and well-defined, whereas benign lesions canbe irregular and speculated. However, some lesionshave morphologic characteristics typical enough toallow a diagnosis to be made on CT. These includearteriovenous fistula, pulmonary artery pseudoaneurysms,mycetomas, and mucous plugs.Demonstration of specific types of calcificationor fat is the only reliable sign that a SPN is benign.“Benign” types of calcification include those with acentral nidus, “popcorn,” and laminated. Eccentriccalcification can be identified in a small percentageof lung carcinomas that may incorporate adjacentcalcified granulomas, or they may themselves calcify.In these instances, the calcification is oftenstippled or punctuate. Presence of fat within a nodulesuggests the lesion as being a hamartoma.Nodules 8 mm in size and present at the lungbases and in the peripheral aspect of the lung maybe classified as intrapulmonary lymph nodes ifthey fulfill the following criteria: (1) a triangularshape and situated on the fissure and/or peripheralsubpleural location or (2) be well-defined,smooth, round[,] or elliptical in contour, situatedon a interlobular septa, and adjacent to a drainingpulmonary vein. Nodules that fulfill the aforementionedcriteria are benign and do not require furtherworkup.If the nodule is found to be indeterminate,further evaluation should be performed with:(1) nodule enhancement; (2) transthoracic percutaneousneedle (TTPN) biopsy, bronchoscopy, andsurgery; (3) PET scan; or (4) periodic follow-up.Nodule EnhancementAfter noncontrast images, sections through thenodule are obtained after IV contrast administrationfor up to 4 min at 1-min intervals. Attenuation ofthe nodule is measured before and at the peak ofcontrast enhancement. Those nodules that enhanced 15 Hounsfield units (HU) are strongly predictiveof benign lesions. Lesions that enhanced 15 HUare indeterminate because carcinomas, activegranulomas, inflammatory lesions, and hamartomasall can demonstrate enhancement.TTPN BiopsyTTPN biopsy can be performed on most CTvisiblelesions. It has a high sensitivity for thediagnosis of malignant cells (79%) but is less accuratefor benign processes.PET ImagingPET is a physiologic imaging technique inwhich 2-[fluorine-18]-fluoro-2-deoxy-D-glucose(FDG), a D-glucose analog, is labeled with a positronemitter ( 18 F). Increased glucose metabolism inlesions result in increased uptake and accumulationof FDG. Thus, metabolically active lesions willdemonstrate increased FDG uptake. The degree ofFDG accumulation is measured by the use of thestandardized uptake ratio. Typically, lung cancersdemonstrate standardized uptake ratio of 2.5.For nodules 1 cm, FDG-PET has sensitivity ofabout 95%. However, its role in evaluation of SPNshould be approached with caution. The reason isinfectious and inflammatory SPN can demonstrateincreased FDG uptake and thus may give falsepositiveresults. False-negative results may occurwith carcinoid and bronchioloalveolar carcinomas.Those nodules that are too small for the resolutionof PET imaging, are not approachable by biopsy,and do not qualify as intrapulmonary lymph nodesneed periodic follow-up CT scans to ensure theirstability or to assess for interval growth.In summary, SPN is a common radiologic findingthat may require extensive workup to establisha definitive diagnosis. There is no one correct managementapproach. The objective is to use a logicalapproach using clinical history and radiologicfindings.HRCT Features of Diffuse Lung DiseaseHRCT scans provide detailed visualization ofthe lung parenchyma. Indications for HRCT scans142 Thoracic Imaging (Shah)ACC-PULM-09-0302-009.indd 1427/10/09 8:07:13 PM


include the following: (1) suspected diffuse lungdisease, (2) for possible prebiopsy workup, and(3) assessment of disease activity. Knowledge ofthe secondary pulmonary lobule and the flow oflymphatics within the pulmonary interstitium areabsolutely necessary for interpretation of HRCTscans. The secondary lobule is the smallest unit ofthe lung visible on HRCT. It is polyhedral in shapeand measures about 1 to 2.5 cm. The core structuresof the lobule include central pulmonary artery andterminal bronchiole. The septal structures includevenules, lymphatics, and fibrous septa. The pulmonarylymphatic channels flow from the subpleuralinterstitium that is loculated beneath thevisceral pleura into the interlobular septa ofthe secondary pulmonary nodule. It then flows tothe intralobular interstitium and into the peribronchovascularinterstitium before reaching the mediastinallymph nodes.HRCT PatternsSeveral basic patterns of diffuse lung diseaseare characterized on HRCT scans. They include thefollowing: (1) linear, (2) reticular, (3) nodular,(4) ground-glass opacity (GGO), (5) consolidation,and (6) cysts. Recognition of these basic patternsalong with their distribution (central or peripheral,upper, or lower) on HRCT scans can help narrowthe differential diagnosis in large majority ofcases.Linear pattern is defined by the presence ofKerley’s lines and represents thickening of theinterlobular septa. This pattern has a long differentialdiagnosis but is commonly seen in patientswith pulmonary edema and lymphangitic spreadof carcinoma.Reticular pattern consists of interlacing lineshadows that appear as a mesh or net-like. Visualizationof this pattern suggests that the patient haspulmonary fibrosis and “honeycombing.” It is seenin patients with usual interstitial pneumonia,asbestosis, collagen vascular disease, and drugtoxicity.Nodular pattern refers to multiple roundopacities that are 1 cm in diameter. They arefurther classified based on their distribution withinthe lung. They can be present in the perilymphaticspace, centrilobular location, or random distribution.Perilymphatic nodules occur at the pleuralsurfaces, on the interlobular septa, and in theperibronchovascular interstitium. Nodules in perilymphaticdistribution are caused by diseases suchas sarcoidosis and lymphangitic spread of carcinoma.Random nodules are evenly distributedthroughout both lungs and suggest hematogenousspread of disease. Entities include metastasis andinfections such as military tuberculosis. Centrilobularnodules occur in the center of the secondarypulmonary lobule. They are seen in patientswith hypersensitivity pneumonitis and infection.Tree-in-bud opacities are a form of centrilobularnodules and represent dilated and impacted distalterminal bronchioles. Their presence is almostalways indicative of endobronchial spread ofinfection.GGO represents hazy increased attenuation oflung with preservation of the underlying bronchovascularmargins while consolidation representsincreased attenuation of lung with complete obscurationof underlying bronchovascular margins.GGO represents partial filling, whereas consolidationrepresents complete filling of the air spaces bytransudative fluid, infectious material, blood, ormalignancy. Both have a very long differential diagnosis.The presence of GGO with superimposedinterlobular septal thickening has been termed acrazy-paving pattern. It has classically been describedin cases of alveolar proteinosis; however, it can beseen in a number of entities. Most common etiologiesin everyday practice include pulmonary edemaand Pneumocystis jiroveci pneumonia. Other causesinclude hemorrhage, malignancy (such as bronchioloalveolarcarcinoma), and lipoid pneumonia.Cystic pattern refers to thin-walled, welldefinedair lesions that are 1 cm in diameter.Entities manifesting a cystic pattern includelymphangioleiomyomatosis and Langerhans cellhistiocytosis. In summary, knowledge of the pulmonaryanatomy, recognition of HRCT patternsand their distribution are prerequisites in theproper interpretation of diffuse lung disease onHRCT scans.BibliographyGeneral ReferencesFelson B. Chest roentgenology. Philadelphia, PA: W. B.Saunders, 1973<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 143ACC-PULM-09-0302-009.indd 1437/10/09 8:07:14 PM


Fraser RS, Pare PD. Diagnosis of diseases of the chest.4th ed. Philadelphia, PA: W. B. Saunders, 1999Webb WR, Muller NL, Naidich DP. High resolution CTof the lung. 3rd ed. Philadelphia, PA: Lippincott, Williamsand Wilkins, 2001Tracheobronchial TreeGamsu G, Webb WR. Computed tomography of thetrachea and mainstem bronchi. Semin Roentgenol1983; 18:51–60Grillo HC, Donahue DM. Postintubation tracheal stenosis.Chest Surg Clin N Am 1996; 6:725–731Kwong JS, Muller NL, Miller RR. Diseases of thetrachea and main stem bronchi: correlation of CTwith pathologic findings. Radiographics 1992; 12:647–657Mariotta S, Pallone G, Pedicelli G, et al. Spiral CT andendoscopic findings in a case of tracheobronchopathiaosteochondroplastica. J Comput Asst Tomogr 1997;21:418–420Pickford HA, Swenson SJ, Utz JP. Thoracic cross-sectionalimaging of amyloidosis. AJR Am J Roentgenol1997; 168:351–355Screaton NJ, Sivasothy P, Flower CD, et al. Trachealinvolvement in Wegner’s granulomatosis: evaluationusing spiral CT. Clin Radiol 1998; 53:809–815Shin MS, Jackson RM, Ho KJ. Tracheobronchomegaly(Mounier-Kuhn syndrome): CT diagnosis. AJR Am JRoentgenol 1988; 150:777–779Stark P. Radiology of the trachea. New York, NY:Thieme, 1991Webb EM, Elicker BM, Webb WR. Using CT to diagnosenonneoplastic tracheal abnormalities: appearanceof the tracheal wall. AJR Am J Roentgenol 2000;174:1315–1321AtelectasisBatra P, Brown K, Hayashi K, et al. Rounded atelectasis.J Thorac Imaging 1996; 11:187–197Blankerbaker DG. The Luftsichel sign. Radiology 1998;208:319–320Naidich DP, Ettinger N, Leitman BS, et al. CT of lobarcollapse. Semin Roentgenol 1984; 19:222–235Woodring JH, Reed JC. Types and mechanisms of pulmonaryatelectasis. J Thorac Imaging 1996; 11:92–108Woodring JH, Reed JC. Radiographic manifestations oflobar atelectasis. J Thorac Imaging 1996; 11:109–144MediastinumBrown LR, Aughenbaugh GL. Masses of the anteriormediastinum: CT and MR imaging. AJR Am J Roentgenol1991; 157:1171–1180Franquet T, Erasmus JJ, Gimenez A, et al. The reterotrachealspace: normal anatomic and pathologic appearances.Radiographics 2002; 22:S231–S246Jeung M, Gasser B, Gangi A, et al. Imaging of cysticmasses of the mediastinum. Radiographics 2002; 22:S79–S93Nishino M, Ashiku SK, Kocher ON, et al. The thymus:a comprehensive review. Radiographics 2006; 26:335–348Teece PM, Fishman EK, Kuhlman JE. CT evaluation ofthe anterior mediastinum. Radiographics 1994; 14:973–990Whitten CR, Khan S, Munneke GJ, et al. A diagnosticapproach to mediastinal abnormalities. Radiographics2007; 27:657–671<strong>Pulmonary</strong> EmbolismCastaner C, Gallardo X, Ballesteros E, et al. CT diagnosisof chronic pulmonary thromboembolism. Radiographics2009; 29:31–50Garg K, Sieler H, Welsh CH, et al. Clinical validity ofhelical CT being interpreted as negative for pulmonaryembolism: implications for patient treatment. AJR AmJ Roentgenol 1999; 172:1627–1631Kavanagh EC, O’Hare A, Hargaden G, et al. Risk ofpulmonary embolism after negative MDCT pulmonaryangiography findings. AJR Am J Roentgenol 2004;182:499–504Kuzo RS, Goodman LR. CT evaluation of pulmonaryembolism: technique and interpretation. AJR Am JRoentgenol 1997; 169:959–965Patel S, Kazerooni EA. Helical CT for the evaluationof acute pulmonary embolism. AJR Am J Roentgenol2005; 185:135–149Remy-Jardin M, Remy J, Deschildre F, et al. Diagnosisof pulmonary embolism with spiral CT: comparisonwith pulmonary angiography. Radiology 1996;200:699–706Schoepf UJ, Costello P. CT angiography for diagnosisof pulmonary embolism: state of the art. Radiology2004; 230:329–337Schoepf UJ, Holzknecht N, Helmberger TK, et al. Subsegmentalpulmonary emboli: improved detection144 Thoracic Imaging (Shah)ACC-PULM-09-0302-009.indd 1447/10/09 8:07:14 PM


with thin collimation multi-detector row spiral CT.Radiology 2002; 222:483–490Wittram C, Maher MM, Yoo AJ, et al. CT angiographyof pulmonary embolism: diagnostic criteria and causesof misdiagnosis. Radiographics 2004; 24:1219–1238Solitary <strong>Pulmonary</strong> NoduleAsad S, Aquino SL, Piyavisetpat N, et al. False-positiveFDG positron emission tomography uptake in nonmalignantchest abnormalities. AJR Am J Roentgenol2004; 182:983–989Cheran SK, Nielsen ND, Patz EF. False-negative findingsfor primary lung tumors on FDG positron emissiontomography: staging and prognostic implications.AJR Am J Roentgenol 2004; 182:1129–1132Erasmus JJ, Connolly JE, McAdams HP, et al. Solitarypulmonary nodules: Part I. Morphologic evaluation fordifferentiation of benign and malignant lesion. Radiographics2000; 20:43–58Erasmus JJ, Connolly JE, McAdams HP, et al. Solitarypulmonary nodules: Part II. Evaluation of the indeterminatenodule. Radiographics 2000; 20:59−66Erasmus JJ, McAdams HP, Patz EF, et al. Thoracic FDGPET: state of the art. Radiographics 1998; 18:5−20Hiroyuki I, Naoya K, Tetsuro M, et al. Ultrasmallintrapulmonary lymph node: usual high resolutioncomputed tomographic findings with histopathologiccorrelation. J Comput Assist Tomogr 2007; 31:409−413Khouri NF, Stitik FP, Erozan YS, et al. Transthoracicneedle aspiration biopsy of benign and malignantlesions. AJR Am J Roentgenol 1985; 144:281−288MacMahon H, Austin JHM, Gamsu G, et al. Guidelinesfor management of small pulmonary nodules detectedon CT scans: a statement from the Fleischner Society.Radiology 2005; 237:395−400Matsuki M, Satoshi N, Yasumasa K, et al. Thin sectionCT features of intrapulmonary lymph nodes. J ComputAssist Tomogr 2001; 25:753−756Ost D, Fein AM, Feinsilver SH. The solitary pulmonarynodule. N Engl J Med 2003; 348:2535−2542Swensen SJ, Viggiano RW, Midthun DE, et al. Lungnodule enhancement at CT: multi-center study. Radiology2000; 214:73−80Tan BB, Flaherty KR, Kazerooni EA, et al. The solitarypulmonary nodule. Chest 2003; 123:89S−96SWestcott JL. Direct percutaneous needle aspiration oflocalized pulmonary lesions: results in 422 patients.Radiology 1980; 137:31−35Diffuse Lung DiseaseAustin JH, Muller NL, Friedman PJ, et al. Glossaryof terms for CT of the lung: recommendations of thenomenclature committee of the Fleischner Society.Radiology 1988; 167:327–331Bergin CJ, Muller NL. CT of interstitial lung disease:a diagnostic approach. AJR Am J Roentgenol 1987;148:8–15Bergin C, Roggli V, Coblentz C, et al. The secondarypulmonary lobule: normal and abnormal CT appearances.AJR Am J Roentgenol 1988; 15:21–25Hansell DM, Bankier AA, MacMahon H, et al. Fleischersociety: glossary of terms for thoracic imaging. Radiology2008; 246:697–722Mueller-Mang C, Grosse C, Schmid K, et al. What everyradiologist should know about idiopathic interstitialpneumonias. Radiographics 2007; 27:595–615Raoof S, Amchentsev A, Vlahos I, et al. Pictorial essay:multidetector disease; a high resolution CT scan diagnosticalgorithm. Chest 2006; 129:805–815Swenson SJ, Aughenbaugh GL, Myers JL. Diffuse lungdisease: diagnostic accuracy of CT in patients undergoingsurgical biopsy of the lung. Radiology 1997;205:229–234Wittram C, Mark EJ, McLoud TC. CT-Histologic correlationof the ATS/ERS 2002 classification of idiopathicinterstitial pneumonias. Radiographics 2003; 23:1057–1071<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 145ACC-PULM-09-0302-009.indd 1457/10/09 8:07:14 PM


Notes146 Thoracic Imaging (Shah)ACC-PULM-09-0302-009.indd 1467/10/09 8:07:14 PM


Ethics in <strong>Pulmonary</strong> and Critical Care <strong>Medicine</strong>Mark J. Rosen, MD, FCCPObjectives:• Discuss the basic principles of bioethics as they apply tothe practice of medicine (the Georgetown mantra)• Summarize the common ethical issues that apply to allphysicians• Focus on issues that are most pertinent to pulmonary andcritical care physicians, especially regarding ethical issuesat the end of lifeKey words: autonomy; end-of-life care; ethics; futility; preferencesWhen caring for patients, decisions about what isa “right” or “wrong” course of action are notalways based on randomized clinical trials or evenscience. The complex and at times competing interestsof patients, families, the care setting, the payor,society, the law, and physicians often complicatepatient care, and these issues cannot be resolvedby the use of scientific methods. <strong>Pulmonary</strong> andcritical care physicians are on the front lines ofthese dilemmas, but few have formal training inbioethics, communication, conflict resolution, andrelated disciplines. Therefore, we often improvisebased on past experience or a “see one, do one”training pathway. At the same time, physicians asa group (like the rest of humanity), including pulmonaryand critical care physicians, may not wantto confront difficult problems and choices.This reticence was demonstrated by the landmarkStudy to Understand Prognosis and Preferencesfor Outcomes and Risks of Treatments, inwhich the authors attempted to improve end-of-lifedecision making and to reduce the frequency ofprolonged and potentially unwanted and painfulcare at the end of life. Despite interventions thatincluded providing physicians with predictionmodels and decision-making tools, together withtimely reports by trained nurses of patient andsurrogate preferences, there was no improvementin physician communication, knowledge ofpatients’ preferences not to be resuscitated, controlof pain, or duration of ICU stay and mechanicalventilation. This section will review the generalprinciples of bioethics as they apply to patient careand will focus on the issues most pertinent to pulmonaryand critical care physicians.Basic PrinciplesMost medical ethics discussions can be framedin terms of the following four principles, whichoften are referred to as the Georgetown mantra. Someethicists criticize the wide application of these principlesas being simplistic and sometimes irrelevant,but their simplicity and clarity have stood the testof decades of use by frontline clinicians who lackformal training. The four principles are as follows:• Autonomy: The patient has the right to acceptor refuse every treatment;• Beneficence: The clinician should act in the bestinterest of the patient;• Nonmaleficence: “First, do no harm”; and• Justice: The distribution of limited resourcesmust be fair.Many (or most) bedside ethical dilemmas arisewhen two or more of these values are in conflict.However, other conflicts are believed to arise fromethical concerns as a consequence of a lack of communicationamong patients, families, and thehealth-care team. With open communication(which may require the presence of a mediatorwhen communications have broken down), theethical issues often disappear.In addition, two other core values are also commonlyapplied to the analysis of ethical issues, asfollows:• Dignity: Both the patient and the caregiver havea right to dignity; and• Truthfulness and honesty: Clinicians shouldtell the truth.Some criticize the term medical ethics ashaving a physician focus and prefer to use the term<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 147ACC-PULM-09-0302-010.indd 1477/10/09 8:07:31 PM


health-care ethics to make it more comprehensive.That physician focus is obvious in the AmericanMedical Association preamble to their “Principlesof Medical Ethics” (Table 1). Despite the focus onphysicians, these principles are generally accepted.AutonomyThe patient’s right to make an informed anduncoerced decision regarding treatment generallyoverrides the other values, but autonomy has itslimits. First, a decision must be legal; a request foreuthanasia must be denied, even if it comes froman informed patient because it is against the law.Second, the patient must be competent, which isdefined here as having the capacity to make decisionsabout the care (see the section “InformedConsent”). If patients with severe illness do nothave this capacity, then we depend on surrogatedecision makers. Autonomy depends on the properprocess of informed consent, where the risks, benefits,and alternatives are explained honestly.Beneficence and NonmaleficenceA physician is not obligated to provide carethat would not benefit the patient. However,conflicts arise when the patient or surrogateperceives of a benefit that is different from theclinician’s view of their duty to beneficence andnonmaleficence. The obvious example is the commonconflict between a family who wants “everything”despite all evidence that “everything” willnot be effective and the clinician’s view that suchcare will not lead to a benefit, may cause harm,and wastes resources. In most cases, patientautonomy dictates that the benefit must be judgedby the patient’s and surrogate’s preferences, notby those of the team. Frustrating as this often is,it also provides a check against other pressuresthat may influence care (eg, need for an ICU bedor a perceived waste of money).JusticeAlthough justice is one of the four basic tentsof the Georgetown mantra, this should enter bedsidedecision making rarely, if ever, at least in theUnited States. All physicians are obligated to useresources in an appropriate and efficient manner,but the primary role of the physician is as a patientadvocate. Therefore, we are not charged withrationing health-care resources at the bedside.Rather, this is a policy issue. The perennial exerciseof “who gets the last bed in the medical ICU?” followedby a list of candidates often is amusing butusually specious because there are probably otherICU beds in the institution that could receive acritically ill patient temporarily. Unless there is ashortage of beds, physicians must allocate ICUsTable 1. The American Medical Association Principles of Medical Ethics*1. A physician shall be dedicated to providing competent medical care, with compassion and respect for human dignity andrights.2. A physician shall uphold the standards of professionalism, be honest in all professional interactions, and strive to reportphysicians deficient in character or competence, or engaging in fraud or deception, to appropriate entities.3. A physician shall respect the law and also recognize a responsibility to seek changes in those requirements that are contraryto the best interests of the patient.4. A physician shall respect the rights of patients, colleagues, and other health professionals and shall safeguard patient confidencesand privacy within the constraints of the law.5. A physician shall continue to study, apply, and advance scientific knowledge; maintain a commitment to medical education;make relevant information available to patients, colleagues, and the public; obtain consultation; and use the talents ofother health professionals when indicated.6. A physician shall, in the provision of appropriate patient care, except in emergencies, be free to choose whom to serve, withwhom to associate, and the environment in which to provide medical care.7. A physician shall recognize a responsibility to participate in activities contributing to the improvement of the communityand the betterment of public health.8. A physician shall, while caring for a patient, regard responsibility to the patient as paramount.9. A physician shall support access to medical care for all people. Adopted by the AMA’s House of Delegates June 17, 2001.*Available at www.ama-assn.org/ama/pub/physician-resources/medical-ethics/code-medical-ethics/principles-medicalethics.shtml.Accessed June 22, 2009.148 Ethics in <strong>Pulmonary</strong> and Critical Care <strong>Medicine</strong> (Rosen)ACC-PULM-09-0302-010.indd 1487/10/09 8:07:31 PM


eds, dialysis slots, bags of packed RBCs, and anyother unit of resource fairly to all who wouldqualify to receive it. At the same time, we shouldadvocate vigorously for rational health policiesthat would provide proper care to those who needit while educating the public about the limits ofwhat such care can provide.Ethical Decision MakingIdeally, medical decisions should be based onempiric information like the results of randomizedtrials. However, ethical dilemmas usually involvesubjective considerations, including the preferencesof patients, their surrogates, and clinicians,which in turn are often influenced by their experiences.These considerations defy a standard“evidence-based” approach to decision making.There is an emerging and contentious literature on“evidence-based ethics” that attempts to applyprinciples of evidence-based medicine to ethicaldilemmas in clinical medicine.Regardless of the outcome of this controversy,some basic principles apply. One potentially usefulmethod includes first framing the ethical dilemmain the dimensions of autonomy, beneficence, nonmaleficence,and justice. The next step would be toevaluate potential decisions in terms of patientpreferences, the medical benefits and harms, andhow they impact on access to care and societalcosts.Specific IssuesInformed ConsentPhysicians are required to obtain informedconsent from patients before initiating treatment,and informed consent is valid only when a patientis competent to make voluntary choices amongtreatment options. This consent requires that thepatient is capable of understanding the relevantinformation and the consequences of treatmentoptions, can rationally process this information,and can communicate a choice.Competent patients may refuse any treatment,even if that decision is contrary to the physician’srecommendation. Studies of competency inpatients with unstable angina, diabetes mellitus,and HIV infection failed to show an increasedlikelihood of incapacity to make medical decisions.However, impairment in the capacity for makingmedical decisions was found in outpatients withcancer, the elderly, and patients with dementia.Even patients with psychiatric illnesses are notnecessarily incompetent to make treatment decisions;rather, they should be assessed by use of thesame criteria as used for other patients.Often, the great majority of patients in an ICUare deemed to lack the capacity to make decisions;therefore, the determination of capacity is animportant part of ICU management. Obviously,critically ill patients who are deeply sedated orobviously delirious do not have decisional capacity,and their treating clinicians can determineincapacity. Other patients may lack capacitybecause they have cognitive impairments causedby preexisting disease, acute illness, or the effectsof analgesics. Standardized assessment tools forthese conditions may be helpful in assessingdelirium, which is usually underestimated in hospitalizedpatients. Psychiatric consultation is notnecessary to determine whether a patient is incompetent;rather, consultations should be reserved forcases in which the clinician believes that the patientis making an irrational decision, when there isassociated mental illness, or in other situations thatpose uncertainty.Problems often arise in patients who lack thecapacity to make decisions and also lack a surrogate.In most situations, physicians use the principleof “substituted judgment” and proceed witha course that most patients with capacity wouldchoose. Often, hospitals have policies that requireconsultation and approval by an independentpractitioner or an ethics committee. Withdrawinglife support in a patient without capacity and withno surrogate presents a special issue. In one studysurveying seven medical centers, 5.5% of all deathsoccurred among patients in this category; in mostcases, decisions were made by physicians with noinstitutional or judicial review, contrary to theirinstitution’s policies.Informed consent for research also presentsunique issues, especially when it involves subjectswho are critically ill and incapable of providingconsent because of their illness or medications thatimpair cognition. US Federal regulations permitconsent by the subject’s “legally authorized representative,”which is defined as “an individual<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 149ACC-PULM-09-0302-010.indd 1497/10/09 8:07:32 PM


or judicial or other body authorized under applicablelaw to consent on behalf of a prospectivesubject to the subject’s participation in theprocedure(s) involved in the research.” Thisrequirement may be waived if an institutionalreview board determines that the research posesminimal risk, defined in U.S. Federal guidelinesas “the probability and magnitude of harm ordiscomfort anticipated in the research are notgreater in and of themselves than those ordinarilyencountered in daily life or during the performanceof routine physical or psychological examinationsor tests.” However, state laws vary widelyin permitting surrogates to consent for participationin clinical research, and critical care researchis generally not conducted in some states unlessthere is a court-appointed guardian. Others permit“emergency research” in situations such as aftercardiopulmonary resuscitation.Withdrawal of Life SupportIt is widely accepted in modern societies thatpatients and their surrogates have the authority todecline or withdraw any medical intervention, aslong as the decision maker is intellectually andpsychologically competent to make such a decision.Withdrawing ventilatory support is generallydeemed the moral and ethical equivalent of withholdingit, but many families and physicians cannothelp but think and act otherwise. Although itis ethically and legally prohibited to withdraw carewith the specific intent of hastening death, theprinciple of “double effect” supports the use ofmedications intended to prevent or relieve sufferingeven if it incidentally hastens death.The withdrawal of mechanical ventilatory supportmay be undertaken in most locations if thereis an oral advance directive by the patient or withthe agreement of the clinical team and family; therequirement for a written advance directive isunusual. The process should begin only after thereis a consensus of both the medical team and thepatient and family that this treatment is bothunwanted and not likely to lead to a desired patientoutcome. When the decision to limit or withdrawtreatment is reached, the clinician is still responsiblefor treating the patient throughout the dyingprocess and being attentive to the needs of thepatient, family, and the clinical team. The ICU teamneeds to attend the family almost as much as thepatient because they are often under great emotionalstress. This process should be planned andcommunicated to the team and the family, preferablywith an organized protocol including theadministration of analgesics and sedatives titratedto maintain the comfort of the patient.Whether to simply extubate the patient afteradministering and maintaining sedatives (calledterminal extubation) or to gradually reduce support(ie, respiratory rate and fraction of inspired oxygen)with the endotracheal tube in place (called terminalwean) is a topic of ongoing controversy. Promptextubation has the advantages of not prolongingthe dying process, and the goals of care are clearand morally transparent. Gradual withdrawal ofsupport with the endotracheal tube in place reducesthe likelihood that visible distress will develop inpatients from stridor and oral secretions, oftenpromoting their own comfort and that of their familyby reducing anxiety. However, this approachmay prolong the dying process, and some familymembers may misinterpret this process asan attempt to extubate the patient successfully.No randomized trials have compared thesetwo approaches from a patient-centered or familycenteredperspective, and there is no consensusamong clinicians and ethicists on which method ispreferable. Rather, decisions on how to extubatepatients who are expected to die depend on thepreferences of the ICU team in consultation withthe family.Neuromuscular blocking agents have no sedativeor analgesic properties, would mask signs ofdiscomfort after the withdrawal of ventilatory support,and should almost always be discontinuedbefore withdrawing ventilatory support. In addition,some patients survive to be discharged fromthe hospital despite predictions that the withdrawalof support will lead to death; this wouldbe impossible in the presence of neuromuscularblockade. Unusual situations in which continuingtherapy with these agents are warranted duringthe withdrawal of mechanical ventilator supportwould include patients who are certain not to survivemore than a short interval after the withdrawalof support even without this treatment, orif the benefits of waiting for the return of neuromuscularfunction do not outweigh the burdens.150 Ethics in <strong>Pulmonary</strong> and Critical Care <strong>Medicine</strong> (Rosen)ACC-PULM-09-0302-010.indd 1507/10/09 8:07:32 PM


Regardless of the process of extubation, thereis no evidence that increasing the dose of benzodiazepinesand opiates before extubation with thegoal of maintaining patient comfort is associatedwith a shorter time to death after extubation.Indeed, one study showed no relationship betweennarcotic dose and the time to death, and a directrelationship between the dose of benzodiazepinesand the time to death after extubation. Therefore,a judicious use of sedation and analgesics does notappear to hasten death in these patients and shouldbe part of any standard protocol.Brain DeathIt is now almost universally accepted that aperson is considered to be dead when the wholebrain (including the brainstem) is dead; therefore,establishing a precise diagnosis of brain death isoften crucial in decisions about terminating life supportand organ donation. Although the criteria fordiagnosing brain death have evolved, the currentguidelines and the laws in most countries require adetailed clinical assessment that includes the presenceof coma and the absence of brainstem reflexesand apnea over the course of two successive examinations.These criteria must also be fulfilled in theabsence of a complicating condition that may renderclinical examination unreliable, including severehypothermia, hypotension, severe metabolic disturbances,drug intoxication, and the use of neuromuscularblocking agents. A disorder that wouldpreclude performing an apnea test (eg, high cervicalcord injury, in which a patient may be incapable ofbreathing spontaneously) would also make clinicalassessment impossible. In these cases, confirmatorytests of the cessation of brain function are needed.Although EEG is frequently used because it canbe performed at the bedside, the interpretation ofthe findings is subject to interobserver variability,as well as the possibility of false-positive findings(eg, as a result of the use of barbiturates or anesthetics,or undetectable subcortical neuronal activity)and false-negative results (eg, as a result of electricalartifacts in an ICU environment or agonalelectrical activity). Tests of brain flow are morereliable for confirming the diagnosis. Cerebralangiography with findings that show a cessationof blood flow to the brain is considered to be the“gold standard,” and technetium nuclear imagingis also reliable. CT scanning and magnetic resonanceangiography are probably as accurate butare less well validated. Transcranial Doppler examinationsof cerebral blood flow are safe, noninvasive,and can be performed at the bedside, althoughperforming them requires a high level of skill. Thedetermination of a complete absence of flow usingthis examination is unreliable in the diagnosis ofbrain death because false-positive results mayoccur in 10 to 15% of cases as the result of technicalfactors, which are often related to poor imageacquisition.Annotated BibliographyGeneral and Basic TopicsAulisio MP, Arnold RM, Youngner SJ. Health care ethicsconsultation: nature, goals, and competencies—aposition paper from the Society for Health and HumanValues-Society for Bioethics Consultation Task Forceon Standards for Bioethics Consultation. Ann InternMed 2000; 133:59 –69Summary of a task force report delineating the role of ethicscommittees and consultative services, with recommendationson policies, competencies, and processes.Beauchamp T, Childress J. Principles of biomedicalethics. Oxford, UK: Oxford University Press, 2001Late edition of an influential textbook from the authors of theGeorgetown mantra.Snyder L, Leffler C, Ethics and Human Rights Committee,American College of Physicians. Ethics manual,fifth edition. Ann Intern Med 2005; 142:560– 582Concise overview of issues related to medicine, law, and socialvalues that covers issues related to patient care, the practice ofmedicine, and the impact of economics and the government.Strech D. Evidence-based ethics: what it should be andwhat it shouldn’t. BMC Med Ethics 2008; 9:16Discussion of the “evidence-based ethics,” a new conceptthat attempts to synthesize evidence-based medicine principleswith ethical decision making.SUPPORT Principal Investigators. A controlled trialto improve care for seriously ill hospitalized patients:the Study to Understand Prognosis and Preferences forOutcomes and Risks of Treatments (SUPPORT). JAMA1995; 274:1591– 1598Two-year study that shows that physicians in critical careunits are not likely to know patient preferences about endof-lifecare, nor are they likely to change their practice evenwith intensive intervention.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 151ACC-PULM-09-0302-010.indd 1517/10/09 8:07:32 PM


Informed Consent and CompetenceApplebaum PS. Assessment of patients’ competence toconsent to treatment. N Engl J Med 2007; 357:1834–1840Carrese JA. Refusal of care: patients’ well-being andphysicians’ ethical obligations; “but doctor, I want togo home.” JAMA 2006; 296:691–695Detailed case-based discussion, with some useful recommendations.Cohen LM, McCue JD, Green GM. Do clinical and formalassessments of the capacity of patients in the intensivecare unit to make decisions agree? Arch InternMed 1993; 153:2481– 2485Issues and methods to determine competence to make medicaldecisions.Fan E, Shahid S, Kondreddi VP, et al. Informed consentin the critically ill: a two-step approach incorporatingdelirium screening. Crit Care Med 2008; 36:94– 99Luce JM. Informed consent for clinical research involvingpatients with chest disease in the United States.Chest 2009; 135:1061– 1068This is a thorough and thoughtful review of laws and issuessurrounding performing clinical research, especially inpotential subjects who lack decision-making capacity.White DB, Curtis JR, Wolf LE, et al. Life support forpatients without a surrogate decision maker: whodecides? Ann Intern Med 2007; 147:34– 40In most cases of withdrawal of life support in patients withoutcapacity and without a surrogate, physicians made thedecision despite institutional policies that call for outsidereview.Withdrawing Life SupportCampbell ML, Bizek K, Thill M. Patient responses duringrapid terminal weaning from mechanical ventilation:a prospective study. Crit Care Med 1999; 27:73– 77Chan JD, Treece PD, Engelberg RA, et al. Narcotic andbenzodiazepine use after withdrawal of life support:association with time to death? Chest 2004; 126:286–293These two studies deal with the processes of withdrawingmechanical ventilatory support, indicating that the appropriateuse of sedatives and narcotics is associated with minimalpatient discomfort and does not hasten the time to deathafter extubation.Rubenfeld GC. Principles and practice of withdrawinglife-sustaining treatments. Crit Care Clin 2004; 20:435–451Truog RS, Cist AF, Brackett SE, et al. Recommendationsfor end-of-life care in the intensive care unit: the EthicsCommittee of the Society of Critical Care <strong>Medicine</strong>.Crit Care Med 2001; 29:2332– 2348These articles review ethical and practical aspects of withdrawinglife-sustaining treatments.Brain DeathWijdicks EFM. Current concepts: the diagnosis of braindeath. N Engl J Med 2001; 344:1215– 1221Young GB, Shemie SD, Doig CJ, et al. Brief review: therole of ancillary tests in the neurological determinationof death. Can J Anesth 2006; 53:620– 627152 Ethics in <strong>Pulmonary</strong> and Critical Care <strong>Medicine</strong> (Rosen)ACC-PULM-09-0302-010.indd 1527/10/09 8:07:32 PM


Chronic Obstructive <strong>Pulmonary</strong> DiseaseSidney S. Braman, MD, FCCPObjectives:• <strong>Review</strong> the definition of COPD that describes this dis easeas an airway and systemic inflammatory condition• Explore the impact of COPD, including rates of morbidityand mortality• <strong>Review</strong> the risk factors for COPD• Explore the natural history of COPD from its earlierasymptomatic stages to the late stages associated withmorbidity and mortality• Explore the current understanding of the pathophysiologyof COPD: the pathologic consequences of airway inflammationand parenchymal lung destruction.• Explore the systemic consequences of the disease and thecomorbidities associated with COPD• <strong>Review</strong> the current state of pharmacologic and nonpharmacologictherapy for COPD, including preventivemea sures such as smoking cessation.Key words: chronic bronchitis; COPD; emphysema; naturalhistory; risk factors; treatmentDefinition and EpidemiologyInternationally, the accepted definition ofCOPD is as a disease state characterized by chronicairflow limitation attributable to chronic bronchitisand emphysema. Chronic bronchitis has beendefined in clinical terms: the presence of chronicproductive cough for at least three consecutivemonths in two consecutive years. Other causes ofchronic productive cough must been ruled out.Emphysema, on the other hand, has been definedby its pathologic description: an abnormal enlargementof the air spaces distal to the terminal bronchiolesaccompanied by destruction of their wallsand without obvious fibrosis. Contrary to thistraditional view, recent data have shown thatthis destructive process is accompanied by a netincrease in the amount of collagen, suggestingthere is indeed active fibrosis associated with thebreakdown of the elastic framework of the lung.During the past decade, as it became recognizedthat the disease was having a major worldwideimpact, there has been increasing interest in thepathogenesis and management of COPD. It hasbeen calculated that COPD ranks as the twelfthlargestdisease burden in the world, with projectionsthat it will rank as the fifth by the year 2020.In the United States, estimates from national interviewstaken by the National Center for HealthStatistics showed that 16 million people areafflicted with COPD; about 14 million are thoughtto have chronic bronchitis, whereas 2 million haveemphysema. It is likely that these statistics underestimatethe prevalence of COPD by as much as50% because many patients underreport theirsymptoms and remain undiagnosed.In the National Health and Nutrition ExaminationSurvey (NHANES) III the prevalence of COPDin the United States was found to be 23.6 millionadults (13.9% of the adult population). These statisticswere derived using telephone surveys andphysical examinations as well as pulmonary functiontesting of randomly selected subjects. Theaverage number of days of restricted activitiesreported by patients with COPD is very high,ranking this condition among the highest chronicconditions for this measure of morbidity in thenation. In 1996, COPD was listed as the eighthleadingcause of disability-adjusted life-years inmen and the seventh-leading cause of disabilityadjustedlife-years among women.COPD is a leading cause of hospitalizations inthe United States. In 1998, nearly 2% of all hospitalizationswere attributed to COPD, and 7% ofpatients had COPD as a contributing cause of hospitalization.From 1995 to 2000, the trend in COPDhospitalization rates was about the same for menand women. However, the rates were slightlygreater among black than white patients during thissame period. In 2000, the COPD hospitalizationrates were 31.5 and 36.0 per 10,000 persons for whiteand black subjects, respectively. More striking areCOPD statistics regarding the elderly. Nearly 20%of all hospitalizations in patients 65 years hadCOPD as a primary or contributing cause.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 153ACC-PULM-09-0302-011.indd 1537/11/09 1:59:11 PM


Perhaps even more striking and ominous arethe magnitude and trend of the mortality rates forCOPD. During the past few decades, the mortalityrates associated with other common diseases suchas cardiovascular disease and stroke were decreasing.During the same time period, the mortalityrate for COPD was increasing at an alarming rate.For example, the US death rate for COPD rosealmost 33% during the years 1979 to 1991, resultingin COPD becoming the fourth-leading cause ofdeath in this country. Although the death rate formen has reached a plateau in recent years, amongwomen it has continued to increase. In 1998, 54,615men and 51,377 women died from COPD. From1995 to 1998, the death rate for COPD increased9% for US women. In the year 2000, for the firsttime more women died from COPD than men.The cost of COPD has also increased dramatically.In 2002, it was estimated to be $32.1 billion.Direct medical services accounted for $18.0 billion,and the indirect cost of morbidity and prematuremortality was $14.1 billion. Medicare expenses forCOPD beneficiaries were nearly 2.5 times that ofthe expenditures for all other patients. The cost perpatient in other countries around the world variesdepending on how health-care is provided andpaid. There is a striking direct relationship betweenthe severity of COPD and the cost of care.As a result of the impact of COPD on morbidity,mortality, and health-care utilization andfinances, a number of thoracic and respiratorydisease societies offered consensus statementsregarding the assessment and management ofCOPD. The dawn of the twenty-first century bringsanother guideline, the Global Initiative on ChronicObstructive Lung Disease (GOLD) initiative sponsoredby the World Health Organization and theNational Heart, Lung, and Blood Institute (NHLBI).GOLD presents an evidenced-based approach thatcan be followed by all health-care systems throughoutthe world.The term COPD defined by GOLD differsfrom previous consensus statements. It doesnot incorporate the terms chronic bronchitisand emphysema into the definition. Instead, itdefines COPD as a disease state characterized byairflow obstruction that is no longer fully reversibleand is usually progressive. The disease iscaused by the interaction of noxious inhaledagents such as cigarette smoke, industrial, andother environmental pollutants and host factors(genetic, respiratory infections, etc) that result inchronic inflammation in the walls and lumen ofthe airways. The pathology of COPD, as well asthe symptoms, the pulmonary function abnormalities,and complications all can be explainedon the basis of the underlying inflammation.In the 2006 GOLD definition of COPD, thephrase “preventable and treatable” was added,similar to recent American Thoracic Society/EuropeanRespiratory Society (ATS/ERS) recommendations.This was done to encourage a positiveoutlook for patients and the healthcare communitytoward COPD and to stimulate effective managementprograms to treat those with the disease.Although the GOLD document does not specificallyinclude chronic bronchitis and emphysemain the definition of COPD, it is clear that they areconsidered the predominant causes of COPD. Inthe past, asthma was also included under theumbrella term COPD. As asthma became recognizedas an inflammatory disease with a differentcellular and inflammatory mediator profile comparedwith smoking-induced chronic bronchitisand emphysema, it was no longer considered underthe term COPD. This was also justified by the factthat asthma is thought to be distinct: airflowobstruction is predominantly reversible, and theairways of asthmatic subjects are markedly hyperresponsiveto a variety of specific (aeroallergens)and nonspecific (methacholine, histamine, cold air)inhaled substances. Although it is still clinicallyuseful to distinguish asthma from COPD (chronicbronchitis and emphysema), it is important to recognizethree problems with this nosology:• Remission rates for adults with asthma aremuch less than those observed in children, and ifremission does occur, the probability of relapseincreases with increasing age. As a result, mostasthmatics, even those who have never smoked,who are at the age when COPD is most often recognized( 60 years), have evidence of poorlyreversible airflow obstruction on pulmonaryfunction testing. This statement is supportedby data from longitudinal studies that the averageFEV 1in asthmatics decrease over a periodof years. Despite these findings, there are manypatients who show no correlation between theirFEV 1and the duration of their disease.154 COPD (Braman)ACC-PULM-09-0302-011.indd 1547/11/09 1:59:12 PM


• Bronchial hyperresponsiveness, the exaggeratedbronchoconstrictive response to nonspecificagonists such as methacholine, is a constantfeature in asthmatics and is so important toits pathogenesis it has been incorporated intothe definition of asthma. However, increasedresponsiveness to constrictors such as methacholineand histamine (but not indirect bronchoconstrictorssuch as cold air and bradykinin)is found in most middle-aged smokers withCOPD and has even been shown to be a strongpredictor of progression of airway obstructionin those who continue to smoke. In asthma,bronchial hyperresponsiveness is not related tobaseline airway caliber, whereas in COPD theincreased responsiveness to bronchoconstrictorscan be explained entirely by the geometriceffect of fixed airway narrowing.• Unfortunately, many adult patients withasthma are current or former smokers. It islikely that such patients have more than onepathologic process with several pathways ofinflammation. Although studies on the effectsof smoking on the rate of decrease of FEV 1inasthma have given conflicting results, mostsmoking asthmatics will present with somedegree of fixed airflow obstruction. This findinghas raised a complexity of semantic issuesthat have not been solved. One attempt hasbeen to combine two of the major pathologicprocesses and describe such patients using theterm asthmatic bronchitis, but this definitiondoes not have widespread acceptance.Risk Factors for COPDThe major risk factors for the development ofCOPD are listed in Table 1. All of these factors haveone thing in common: they are associated with anaccelerated decrease in FEV 1over time. Cigarettesmoking leads the list and in most countries is theresponsible agent in at least 85 to 90% of cases.Because the basic cellular and molecular mechanismsof disease are still poorly understood, thereasons why only certain individuals with a positiveexposure history become affected are notknown. For example, it is common to see pulmonaryfunction abnormalities reflecting obstructionTable 1. Risk Factors for COPDRiskExternal factorsCigarette smoking: the most important risk factor in 85 to90% of cases; 15 to 20% of one pack per-day smokers and25% of two pack per-day smokers develop COPDPipe and cigarsPassive smokingOccupational exposure: risk in certain professions such ascoal miners, gold miners, grain handlers, cement andcotton workers; interacts with smokingEnvironmental pollution: indoor use of biomass fuels forcooking and heating in underdeveloped countries;unknown particulate matter from urban pollutionHost factorsGenetic factorSocioeconomicBronchial hyperresponsivenessAtopy and asthmaChildhood illnessesDietary influencesCommentHigh risk but lower than cigarette smokersChildren of smoking parents have more respiratory symptomsand lower airway function; unknown significance foradult COPDAAT deficiency, 90% caused by homozygous ZZ phenotype.Gender: men are more at risk than womenMore common with low socioeconomic statusA strong predictor of progressive airway obstruction insmokersMany adult nonsmoking asthmatics develop fixed airwayobstruction; atopy alone not a factorLow birth weight, respiratory infections, and symptomaticchildhood asthmaVitamin C and E (tocopherol) deficiencies<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 155ACC-PULM-09-0302-011.indd 1557/11/09 1:59:12 PM


in the small airways and nonuniform distributionof ventilation in young adults who smoke. Thesechanges are not predictive, because only 15 to 20%of these heavy smokers will develop an accelerateddecrease of their FEV 1over time and clinicallyevident COPD. The differences in susceptibility totobacco smoke injury are likely to be related togenetic factors.It has been more difficult to demonstrate theassociation between COPD and occupational exposuresthan with cigarette smoking. There have beenseveral reasons offered but perhaps the most compellingis that many exposed workers are alsocigarette smokers or have been exposed to secondhandsmoke. An increasing body of literature isproviding longitudinal evidence that coal workers,hard rock miners, tunnel workers, concrete manufacturingworkers, and other non-mining industrialworkers (exposures to dusts, gases and/or fumes)have an excess annual loss of FEV 1of 8 mL/yrcompared with 9 to 11 mL/yr attributable to smoking.Some workers have been shown to have acuteairflow obstruction when exposed to organic andinorganic dusts, isocyanates, and irritant gases.Such events appear to predict later findings ofchronic, fixed obstructive lung disease. Recentestimates by the ATS suggest that for COPD, apopulation-attributable risk of approximately 15to 20% is caused by occupational exposures.Genetic Risk Factors for COPD 1-Antitrypsin Deficiency: One genetic abnormalitythat has been known and well studiedis 1-antitrypsin (AAT) deficiency. Also called 1antiprotease deficiency, this disorder accountsfor 1% of COPD in the United States. AAT isa serum protein made in the liver that is capableof inhibiting the activity of specific proteolyticenzymes such as trypsin, chymotrypsin, and neutrophilelastase. If not inactivated by AAT, neutrophilelastase destroys lung connective tissue,particularly elastin, which leads to the developmentof emphysema.AAT is coded for by a single gene on chromosome14. The serum protease inhibitor phenotype(Pi type) is determined by the independent expressionof two independent alleles. More than 90% ofseverely deficient patients are homozygous for theZ allele. Such patients are designated PiZZ andhave a reduced serum AAT level to approximately20% of the normal level. Other rarely observedphenotypes associated with very low levels of AATare Pi SZ, Pi null-null, and Pi null-Z. The normalM allele phenotype is designated Pi MM. Individualswith this phenotype have AAT levels of150 to 350 mg/dL reported from most commerciallaboratories. Often the true laboratory standardvalues are reported. Normal levels are 20 to48 mol/L.Most patients homozygous with the Z alleleare of northern European descent. Asian and AfricanAmericans are rarely affected. Cigarette smokingis the most important risk factor for in thedevelopment of COPD in patients with AAT deficiency.Symptoms begin early in adult life (usuallyby age 40), and COPD leads to an early death.However, it is rare that PiZZ lifetime nonsmokersever develop COPD. Most never develop any pulmonarysymptoms, have a normal life expectancy,and remain undetected unless they are foundthrough family screening. Heterozygotes that arePi MZ have intermediate levels of [propto]-1 antitrypsindeficiency. Epidemiologic evidence suggeststhat they are not at increased risk fordeveloping COPD when compared with Pi MMphenotypes. It is important to screen patients forAAT deficiency (Table 2) for family counseling andbecause multivariate analyses from the NationalInstitutes of Health Registry suggest that severelydeficient patients who receive augmentationtherapy (IV infusions of human AAT) had decreasedrates of mortality. In addition, thosewho were treated who had an FEV 1percentageTable 2. Diagnosis of AAT DeficiencyFeatures that prompt suspicion for screening1. Onset of COPD in patients 50 years2. COPD without a smoking history or other risk factor3. Family history of AAT deficiency4. Family history of COPD, bronchiectasis, panniculitis5. Young adult asthmatic unresponsive to therapy6. Patient with predominant lower-lobe emphysema7. Necrotizing panniculitis8. C-ANCA vasculitis9. Unexplained liver diseaseLaboratory testing for diagnosis1. Protein electrophoresis reveals absence of 1peak2. Serum AAT levels 2.5 to 7 mol/L (normal 20 to48 mol/L)3. Phenotyping by isoelectric focus; 90% PiZZ156 COPD (Braman)ACC-PULM-09-0302-011.indd 1567/11/09 1:59:12 PM


predicted of 35 to 49% had a slowing in theirdecline in lung function compared with an untreatedgroup. In addition to developing advan cedemphysema, hepatitis and cirrhosis, AAT deficiencypatients are at a 70 to 100% increased riskof developing lung cancer.Other Genetic Factors: It is likely that a numberof other genetically determined abnormal protectivemechanisms against protease, oxidant, andtoxic peptide injury are important in the pathogenesisof COPD. Multiple association studieshave been performed for COPD using case-controland population- and family-based studies. Unfortunately,these studies have often yielded inconsistentresults. Candidate genes that have beenstudied in COPD are those regulating proteasesand antiproteases. For example, polymorphismsof the [propto]-1 antitrypsin gene not associatedwith deficiency of the protein predisposeto the development of COPD. These abnormalitiesare located at the TaqI and HindIII sites of the[propto]-1 antitrypsin gene. There is speculationthat abnormalities at these sites of the gene maynot be normally responsive to inflammatory cytokines,which may lead to inappropriate synthesisand, hence, a relative deficiency of the proteinduring times of stress.Antioxidant genes have also been of interest.Polymorphisms in the gene that controls glutathioneS-transferases have also recently been linkedto the development of COPD. This family of metabolicenzymes may play an important part in cellulardefense by detoxifying various substances intobacco smoke. There is recent evidence thatgenetic polymorphisms of cytochrome p450 andmicrosomal epoxide hydrolase may also be associatedwith the development of COPD, and alsovariants in the gene ADAM33 have been significantlyassociated with COPD. Finally, genes regulatingmucociliary clearance (eg, cystic fibrosistransmembrane regulator and mucins) and genesthat regulate inflammatory mediators (tumornecrosis factor [TNF]-[propto], vitamin D-bindingprotein, and transforming growth factor-1) havebeen of interest.40 years. During this asymptomatic period, thereis deterioration of lung function in marked excessof the normal age-related decline. This was firstshown in population studies on London workingmen and reported in the 1970s. It was discoveredthat the FEV 1decreases continuously and evenlyduring an individual’s life. Nonsmokers withoutrespiratory disease can expect to lose 25 to30 mL/yr of lung function after age 35. The rate ofdecrease appears to accelerate slightly with aging,but sudden large irreversible decreases are veryrare. Although nonsmokers lose FEV 1slowly, theynever develop significant airflow obstruction; thisis also true of most smokers.Susceptible smokers, on the other hand, developvarying degrees of airflow obstruction, as theyoften demonstrate a decrease in lung function of60 mL/yr and more. This ultimately causes exertionaldyspnea when the FEV 1is 40 to 59% of itspredicted value and becomes disabling when theFEV 1decreases by approximately 70%. Such aninexorable decrease may lead to fatality; when theFEV 1decrease to 1 L, the 5-year mortality isapproximately 50%. Smoking cessation in thesusceptible smoker will not result in recovery oflost FEV 1but will slow the rate of decrease tonormal (Fig 1). Smoking cessation in the patientwith early changes of COPD at age 45 can makethe difference between a normal life span andpremature death.A prospective multicenter longitudinal studyof the effects of smoking cessation in patients identifiedwith mild-to-moderate airflow obstructionhas been conducted in this country and supportedby the NHLBI. This Lung Health Study showedNatural History of COPDThe natural history of COPD includes a prolongedpreclinical period of approximately 20 toFigure 1. Natural history of COPD. Accelerated decline inlung function (FEV 1) in smoker and “susceptible” smoker inwhom the progressive disability of COPD develops.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 157ACC-PULM-09-0302-011.indd 1577/11/09 1:59:12 PM


that smokers who stopped smoking with the helpof an aggressive smoking-intervention programsignificantly reduced the decrease in their FEV 1during a 5-year study period when compared withthose who continued to smoke. The inhaled bronchodilatoripratropium did not slow the decline oflung function when given over the same timeperiod.The slowly progressive course of COPD is ofteninterrupted by a sudden change in symptoms andlung function. These acute respiratory illnesses orexacerbations are usually caused by viral or bacterialinfections and are heralded by an increase insymptoms. Patients with milder stages of COPDwill often develop one to two exacerbations peryear; those with more severe disease are likely tohave many more. The exacerbations cause adecrease in lung function for up to 90 days. Thereis evidence that more frequent exacerbations playa causal role in accelerating the development ofchronic airflow obstruction.Pathology and PathophysiologyThe pathologic changes in COPD are found inthe large and small ( 2 mm) airways and in thelung parenchyma (Fig 2). In advanced stages, thereare also changes in the pulmonary circulation, theheart, and the respiratory muscles. Alveolarhypoxia causes medial hypertrophy of vascularsmooth muscle with extension of the muscularislayer into distal vessels that do not ordinarily containsmooth muscle. Intimal hyperplasia alsoFigure 2. Mechanisms underlying airflow limitation inCOPD. Inflammation caused by cigarette smoke and otherirritants causes inflammation in the large and mainly smallairways (chronic bronchitis) and parenchymal destruction(emphysema).occurs in advanced stages. These latter changes areassociated with the development of pulmonaryhypertension and its consequences: right ventricularhypertrophy and dilation. Loss of vascular bedalso occurs with emphysema in association withthe destructive alveolar processes. In some patientswith advanced COPD, there is atrophy of diaphragmaticmuscle, loss of skeletal muscle, andwasting of limb muscles.Changes in the Airways of SmokersEarly structural changes have been describedin otherwise-healthy smokers even as young as20 to 30 years of age. These changes are attributedto the toxic effects of cigarette smoke and possiblythe added effects of environmental pollution.Cigarettes contain a complex mixture of particlesand gases and thousands of chemicals. The smokethat is inhaled contains many free radicals andreactive oxygen species that are injurious to thelungs. The burden of gases and particles that areinhaled into the lungs results in inflammatory andimmune responses that are both innate and adaptive.The innate respiratory defense systemincludes an epithelial cell barrier and mucociliarytransport system. When they are overwhelmed,foreign particles may penetrate the airway, andthe epithelial disruption initiates an immuneresponse. Inflammatory cells migrate into theepithelial layer, including polymorphonuclearcells, eosinophils, macrophages, natural killercells, and mast cells. Uncommitted B cells andCD4 and CD8 lymphocytes also participate in theinnate response. Antigens that are deposited onthe epithelium are transported within the airwayby antigen-presenting cells, the specialized epithelialM cells, and the dendritic cells. The antigensare transported to the bronchial-associatedlymphatic tissue layer and to regional lymphnodes, where B and T lymphocytes initiate thecellular and humoral components of the adaptiveimmune response. This response assists inthe destruction of microbes that may penetratethe airway as the innate immune system is overwhelmedand in the neutralization extracellularmicrobial toxins that are released by these microorganisms.BAL studies in cigarette smokers have shownan increase in the number of neutrophils and158 COPD (Braman)ACC-PULM-09-0302-011.indd 1587/11/09 1:59:13 PM


macrophages. Macrophages likely play an importantrole in perpetuating the inflammatory processof COPD and are increased 5- to 10-fold in the BALwhen the disease is present. They may releaseneutrophil chemotactic factors as well as proteolyticenzymes, such as matrix metalloproteinases,that damage the epithelial barrier. Neutrophils arealso thought to be important in causing the tissuedamage of COPD. They release a number ofmediators, including proteases such as neutrophilelastase and matrix metalloproteinases, oxidantssuch as the oxygen free radical H 2O 2, and toxicpeptides such as defensins. Proteases and freeradicals can damage the epithelium as well as thebasement membrane. The repair process thatensues includes the secretion of antiproteases, suchas secretory leukocyte protease inhibitor and tissueinhibitor of metalloproteinases, by epithelial cellsto regulate the proteolytic process.The process of repair may be impaired inpatients with COPD. In patients with moreadvanced airflow obstruction, T lymphocytes andneutrophils are found in the epithelium, andT lymphocytes and macrophages are found in thesubepithelium of centriacinar airways whereemphysema is most marked. Although CD4 lymphocytesare found in the airways of atopic asthmatics,the T lymphocytes of patients withcigarette-induced COPD are CD8 cells. Thesecells are also thought to cause cell damage. Eosinophilicinflammation of the airways is also a hallmarkof asthma and not usually seen in chronicbronchitis, although there is an increase in eosinophilicbasic proteins (eosinophil cationic proteinsand eosinophil peroxidase) in induced sputum.This may indicate that eosinophils had been presentin the sputum but were destroyed, possibly byneutrophil elastase. During attacks of acute bronchitis,to the contrary, the number of tissue andsputum eosinophils is markedly increased. Bronchialbiopsies from former smokers show similarinflammatory changes as the active smoker, suggestingthat inflammation may persist in the airwayonce established. The inflammatory cellsinvolved in the pathogenesis of COPD are listedin Table 3.A number of extracellular-signaling proteinscalled cytokines are important in the pathogenesisof COPD because they are thought to mediate thetissue damage and repair induced by cigaretteTable 3. The Inflammatory Cells of COPDNeutrophilsPotent chemokines (IL-8) and leukotriene B 4attract neutrophilsto airwaysIncreased numbers in sputum and BAL fluid, few inairwaysNeutrophil numbers correlate with degree of airflowobstructionSecrete proteases: neutrophil elastase, cathepsin, andproteinase-3A causative link to mucus hypersecretionPhagocytic ability of neutrophils impaired by cigarettesmoke (by suppression of caspase-3–like activity)predisposing to respiratory infectionMacrophagesActivated by cigarette smokeIncreased numbers found in sputum, BAL fluid, andbronchial submucosaRelease inflammatory mediators: TNF-, leukotriene B 4,IL-8, reactive oxygen speciesRelease increased amounts of metalloproteinase-1 andmetalloproteinase-9Have defective phagocytic function against apoptoticepithelial cellsNumbers of macrophages increase with increasingCOPD severityT lymphocytesIncreased numbers in lung parenchyma and airwaysGreatest number are CD8 cellsT-cell numbers correlate with amount of alveolardestructionT-cell numbers also correlate with severity of airflowobstructionCD8 cells shown to cause cytolysis and apoptosisof alveolar epithelial cellsEosinophilsUsually not in sputum in COPD, but eosinophil basicprotein is foundHence, they may have been present and degranulated bysputum elastaseIncreased numbers predict corticosteroid responsivenessand also asthmaIncreased in BAL fluid and airway biopsy in acute exacerbationof COPDEpithelial cellsLikely an important source of inflammatory mediators asin asthmaActivated by cigarette smoke to produce IL-1 and IL-8,TNF-, and granulocyte colony-stimulating factorAirway epithelial cells important source of transforminggrowth factor-Transforming growth factor- may the cause of airwayfibrosis in COPDVEGF importantVEGF receptor blockade causes alveolar cell destruction(emphysema)Dendritic cells and B cellsFound in increased numbers in lymphoid follicles inbronchial wallMay reflect an adaptive immune response to infectionMay reflect a role for autoimmunity in COPD<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 159ACC-PULM-09-0302-011.indd 1597/11/09 1:59:13 PM


smoking. Increased quantities of certain proinflammatorycytokines, including interleukin (IL)-8,IL-1, IL-6, and TNF-, and the antiinflammatorycytokine IL-10 have been found in the sputum ofsmokers and even further increased during anacute exacerbation. Mediators found to be elevatedduring a COPD exacerbation are listed in Table 4.Exposure of bronchial epithelial cells to cigarettesmoke causes a release of neutrophilic and monocyticchemotactic factors, with IL-8 and granulocytecolony stimulating factor accounting forincreased neutrophilic activity and monocyte chemotacticprotein-1 causing increased monocyticactivity.Cytokines may also be involved with tissueremodeling in COPD. There are distinct pathologicfeatures of airway remodeling seen with COPD,including inflammatory cell proliferation, epithelialdamage, subepithelial fibrosis, myofibroblast proliferation,increased collagen deposition in the lungparenchyma, smooth muscle hypertrophy, andneovascularization. Increased expression of transforminggrowth factor- and epidermal growthfactor (EGF) may activate fibroblastic proliferationin the airways of patients with COPD. Patients withCOPD have increased levels of total collagen andcollagen I, III, and IV in the lungs, and levels ofcollagen I and III, fibronectin, and laminin havebeen inversely related to FEV 1, suggesting that thedeposition of extracellular matrix proteins contributeto the airflow obstruction observed withCOPD. Vascular endothelial growth factor (VEGF),which is expressed in airway smooth muscle cells,Table 4. Inflammatory Mediators Elevated in AcuteExacerbations of COPDC-reactive peptideCopeptinIL-8IL-6TNF-LeptinEosinophilic cationic proteinMyeloperoxidase 1-AntitrypsinLeukotrienes E 4and B 4FibrinogenMyeloid progenitor inhibitory factor-1<strong>Pulmonary</strong> and activation–regulated chemokineSoluble intercellular adhesion molecule-1Adiponectin (ACRP-30)plays a key role in neovascularization and vascularpermeability and the in the expression of fibroblastgrowth factors. Vascular remodeling changes havealso been inversely related to measures of lungfunction (FEV 1).Other structural changes in the airways ofsmokers include mucus hyperplasia, bronchiolaredema and smooth muscle hypertrophy, and peribronchiolarfibrosis. These changes result in narrowingof the small airways ( 2 mm). Recentstudies suggest that the inflammatory cellularinfiltrate, fibrosis, and muscle in the airway wallshow a progression worsening of pathologicchanges when nonsmokers, smokers with mildCOPD, and smokers with more severe disease arecompared, and there is a progression of the numberof airways 400 m in diameter as the diseaseworsens.Because the total cross-sectional area of thesesmall airways is so much larger than the crosssectionalarea of the central conducting airways(and hence they contribute so little to total airwayresistance), dyspnea and changes in the FEV 1areusually absent until the disease is quite advanced.This zone of the lungs is often referred to as the“silent zone” because these pathologic abnormalitiesusually go undetected by symptom assessmentand routine spirometry. In some smokers, theinflammation and peribronchiolar fibrosis progress,and the pulmonary functional abnormalitiesassociated with the small airway changes and thesymptoms of COPD become progressively worse.However, the presence of chronic bronchitis (coughand sputum production) in cigarette smokers is notuseful in predicting the future development ofairflow obstruction. Progression from GOLD stages0 to 4 has been correlated with the following:(1) the thickness of the airway walls; (2) degree towhich the airway lumen is filled with mucousexudates; and (3) the extent of the inflammatoryresponse measured by the number of airways containingpolymorphonuclear leukocytes, macrophages,and lymphocytes (CD8 cells, CD4 cells,and B cells).In addition to the structural abnormalities seenin the small airways, there is also a loss of alveolarattachments to the airway perimeter, which impairselastic recoil and favors increased tortuosity andearly closure of the small airways during expiration.Several tests, such as frequency dependence160 COPD (Braman)ACC-PULM-09-0302-011.indd 1607/11/09 1:59:13 PM


of compliance, nitrogen slope of the alveolar plateau,the closing volume and V . max at low lungvolume (eg, V . max75) are thought to reflect smallairway narrowing. These pathophysiologic abnormalitiesin airways 2 mm have been referred toas “small airway disease,” implying that it is adistinct clinical entity. It is more important insteadto think of the early inflammatory changes in thesmall airways as the first stage in a protractedprocess that eventually leads to persistent airflowobstruction and progressive decrease in the FEV 1.The pathogenesis of COPD that targets the airways(chronic bronchitis) and lung parenchyma (emphysema)is thought to occur because of a complexinteraction between extracellular signaling proteins,oxidative stress, and proteolytic digestion ofconnective tissue (Fig 3).Pathology of Chronic BronchitisThe presence of a gel-like mucous in the airwaysof healthy people is essential for normalmucociliary clearance. This mucous is swallowedand rarely noticed. Smokers with chronic bronchitisproduce larger amounts of sputum each day,averaging about 20 to 30 mL/d and even as highas 100 mL/d. This production occurs as a result ofan increase in the size and number of the submucosalglands and an increase in the number ofgoblet cells on the surface epithelium. Mucousgland enlargement and hyperplasia of the gobletcells are therefore the hallmark of chronic bronchitis.Goblet cells are normally absent in the smallairways, and their presence there (often referredto as mucous metaplasia) is important to theFigure 3. Pathogenesis of COPD. COPD is caused by oxidativestress and proteolytic breakdown of connective tissueseen in this schema.development of COPD. In the larger airways inchronic bronchitis, there is a reduction in serousacini of the submucosal glands. This depresseslocal defenses to bacterial adherence because theseglands are known to produce deterrents such aslactoferrin, antiproteases, and lysozyme. Otherepithelial alterations are seen in chronic bronchitisare a decrease in the number and length of the ciliaand squamous metaplasia. The mucociliary abnormalitiesof chronic bronchitis cause a continuoussheet or blanket of mucous lining the airwaysinstead of discreet deposits of mucous seen onnormal airways. Pooling of secretions also mayoccur. This provides additional cause for bacterialgrowth, which, in turn, causes a release of toxinsthat are further damaging to the cilia and epithelialcells. Bacterial exoproducts are known to stimulatemucous production, slow ciliary beating, impairimmune effector cell function, and destroy localimmunoglobulins. This cycle is especially seen incurrent as opposed to former smokers.Pathology of EmphysemaEmphysema is a destructive process that occursin the gas-exchanging airspaces: the respiratorybronchioles, alveolar ducts, and alveoli. It resultsin perforations (fenestrae), obliteration of airspacewalls, and coalescence of small distinct air spacesinto much larger ones. There is permanent enlargementof the gas-exchanging units of the lungs(acini). These pathologic changes cause loss ofelastic recoil of the lungs and abnormal gasexchange. Recent studies have shown good correlationbetween physiologic measurements oflung elastic recoil and diffusing capacity andmicroscopic measures of airspace wall per unit ofalveolar volume and alveolar surface area.Two morphologic forms of emphysema havebeen described (Table 5). It should be recognizedthat most smoking-related emphysema has featuresof both and the clinical relevance is limited.The mechanisms by which emphysema and chronicbronchitis cause airflow obstruction are shown inTable 6. The release of large amounts of neutrophilelastase and metalloproteinases from inflammatorycells that overwhelm the antiprotease defenses ofthe lung are the most likely cause of the alveolardestruction of emphysema. In addition, lung repairis inhibited by cigarette smoke.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 161ACC-PULM-09-0302-011.indd 1617/11/09 1:59:13 PM


Table 5. Pathologic Types of EmphysemaCentriacinar (centrilobular): focal destruction of respiratorybronchioles; surrounded by normal lung parenchymaMore severe in upper lobesType most frequently associated with smokingPanacinar (panlobular): uniform destruction of all airspacesbeyond the respiratory bronchiolesMore lower-lobe distributionOccurs in smokers earlier in lifeAssociated with AAT deficiencyPathophysiologyThe physiologic hallmark of COPD is thelimitation of expiratory flow with relative preservationof inspiratory flow. Reductions in the FEV 1and FEV 1/FVC percent are characteristic, whereasthe maximal inspiratory flow rate is normal or nearnormal. The degree of reversibility in COPD isseldom brisk as it is in asthma. Differences in thedegree of reversibility in COPD and asthma maybe explained by the fact that in the latter disease,many inflammatory mediators have been implicated,including potent bronchoconstrictors suchas histamine, kinins, and prostaglandins. In contrast,there are few bronchoconstrictor mediatorsreleased in COPD. One important inflammatorymediator found in the sputum of patients withCOPD is leukotriene B 4. It is a potent chemoattractantfor neutrophils. The term “partial reversibility”has been used in COPD, although this termhas not been adequately defined. Reversibility inCOPD patients can be acute, in response to bronchodilatordrugs, mainly as the result of the releaseTable 6. Mechanisms of Airflow Obstruction in COPDEmphysemaLoss of elastic recoil1. Less driving pressure from alveoli to proximal airways2. As patency in smaller airways depends on elasticrecoil, intraluminal pressure is reduced, and airwayscollapse during forced expirationChronic bronchitisAirway inflammation and hypersecretion of mucous1. Peribronchiolar fibrosis results in narrowing ofperipheral airways, loss of alveolar attachments, andloss of elastic recoil2. Intraluminal mucous may contribute to airflowobstructionof cholinergic tone. It also may be in response tooral or inhaled corticosteroids, presumably as aresult of decreased inflammation. Studies haveshown that most patients with COPD will have asmall but significant degree of reversibility of airflowobstruction (defined as a 12% and at least a200-mL improvement in the FEV 1). Some longitudinalstudies have suggested that increased reversibilityis associated with a greater risk of dyingfrom COPD, whereas others have suggested thatthe presence of reversibility coupled with closeclinical treatment and follow-up is associated withbetter survival rates.Early in the course of COPD, the expiratoryflow-volume curve shows a scooped out lower partof the expiratory limb as a result of abnormal flowat low lung volume. In later stages there isdecreased expiratory flow at all lung volumes.Nonuniform ventilation of the lungs is seen evenin the earlier stages of COPD and can be demonstratedwith the N2 washout test. The multipleinert gas test has been used to quantify the ventilation/perfusion(V . /Q . ) mismatch in the lungs ofpatients with COPD and has also shown theseabnormalities in early stages of COPD. The predominantabnormalities detected by this test arelow V . /Q . regions. This type of mismatch causesarterial hypoxemia. There is little right-to-leftshunting in COPD. Lung and thoracic wall hyperinflationare another consequence of the loss ofelastic recoil. This favors expiratory flow. As lungvolume increases, elastic recoil pressure alsoincreases and causes the airways to enlarge, whichin turn reduces the airway resistance. The hyperinflationmay also have a deleterious effect. As thechest wall volume expands, the thoracic cage anddiaphragm are placed at a mechanical disadvantage.This disadvantage increases the workof breathing and contributes to the sensation ofdyspnea.Systemic Inflammation and COPDThere is growing evidence that the inflammatoryresponse of COPD is not limited to the lungs.Table 7 lists a number of systemic consequencesof the disease that are considered related tosystemic inflammation. Individuals with stableCOPD have significantly increased blood levelsof C-reactive protein (CRP), fibrinogen, TNF-,162 COPD (Braman)ACC-PULM-09-0302-011.indd 1627/11/09 1:59:14 PM


Table 7. Systemic Consequences of COPDCardiovascular diseaseOsteoporosisMetabolic disease (diabetes mellitus)Gastrointestinal diseaseAnorexiaWeight lossSkeletal muscle dysfunctionAnemia of chronic diseaseGeneralized fatigueDepressionAltered sleepCor pulmonaleendothelin-1, IL-6, and circulating leukocytescompared with normal subjects. Levels of VEGF,known to be involved in angiogenesis and inflammation,also have been shown to be increased inpatients with COPD. Elevated levels of inflammatorycytokines in the blood of COPD patientssuggests that there is a persistent level of systemicinflammation, possibly caused by a “spill over”of inflammatory from the inflammation that isoccurring in the airways, the lung parenchyma,and the pulmonary vasculature. This latter theoryhas not been proven. There is some evidence thatindividuals with increased inflammatory markerssuch as fibrinogen and CRP are at risk for anaccelerated decrease in FEV 1, greater rates of hospitalization,and death.An analysis of the relative risk for systemicdisease in patients with COPD followed by generalpractitioners has shown that patients with COPDare at increased risk for glaucoma, angina, fractures,myocardial infarction, osteoporosis, andrespiratory infection when compared with controlsubjects without COPD. Anemia may occur in 10to 15% of patients with severe COPD. Reducedhemoglobin in patients on long-term oxygentherapy has proven to be a strong predictor ofsurvival and is associated with a greater hospitalizationrate and a longer cumulative duration ofhospitalization. In a nationally representativesample of 47 million hospitalizations from 1979 to2001, hospital discharges with a diagnosis ofCOPD were more likely to be hospitalized withpneumonia, hypertension, heart failure, ischemicheart disease, pulmonary vascular disease, thoracicmalignancies, and ventilatory failure whencompared with age-adjusted discharges withoutCOPD. Further, a diagnosis of COPD was associatedwith greater age-adjusted in-hospital rates ofmortality for pneumonia, hypertension, heartfailure, ventilatory failure, and thoracic malignancieswhen compared with hospital discharges withthese comorbidities who did not have a diagnosisCOPD. These results suggest that the burden ofdisease associated with COPD is largely underestimatedbecause having a diagnosis of COPD isassociated with increase risk for hospitalizationand in-hospital mortality from other commondiagnoses.Another marker of inflammation, surfactantprotein D (SP-D), has been recently studied. SP-Dis a glycoprotein produced in the lungs. In additionto regulating surfactant homeostasis, SP-D also hasa role in innate immunity and host defenseresponses to microorganisms and inhaled particles.Lung injury has been shown to increase levels oflung and systemic SP-D and systemic SP-D maybe harmful. It has been associated with an increasedrisk of atherosclerosis and poor clinical outcomeswith a variety of conditions. There is some evidencethat treatment with an inhaled corticosteroidand long-acting -agonist can reduce systemicSP-D levels. Attenuating lung inflammation mayresult in less leakage of SP-D into the systemiccirculation. Further research on this mechanism ofinflammation is needed.The relationship of COPD, systemic inflammation,and cardiovascular disease has gained agreat deal of attention because the presence ofairflow obstruction doubles the risk of cardiovascularmortality irrespective of current smokingstatus. Ischemic heart disease, stroke, and suddencardiac death rates can each be increased by thepresence of COPD. Because systemic low-gradeinflammation is implicated in affecting clot formationand subsequent cardiovascular events,the findings of systemic inflammation in COPDhas a putative link. The direct effects of inflammatorymarkers of inflammation such as the CRPon the pathogenesis of clot formation and discoveryof increased CRP levels in the blood ofpatients with COPD has provided a possiblemechanism. Using the NHANES III populationdatabase, investigators have found that anincreased CRP score coupled with moderate-tosevereairflow obstruction was associated withan increased occurrence of ischemic changes on<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 163ACC-PULM-09-0302-011.indd 1637/11/09 1:59:14 PM


the electrocardiograms of participants. The effectsof inhaled corticosteroids, newer antiinflammatoryagents such as phosphodiesterase-4 inhibitors,and statins on the cardiovascular andoverall rate of mortality of COPD patients areunder active investigation.The prevalence of the metabolic syndrome hasalso been investigated in patients with COPD.Nearly half of all COPD patients in a pulmonaryrehabilitation program were shown to have threeor more features of the metabolic syndrome,including abdominal obesity, elevated triglyceridelevels, elevated high-density lipoprotein cholesterollevels, elevated BP, and high fasting bloodglucose levels. COPD patients with diabetes whohave been hospitalized for an acute exacerbationhave a greater rate of mortality than those withoutdiabetes mellitus.Patients with COPD have an increased risk forosteoporosis. This risk may be attributable to lifestyle,genetics, treatment with corticosteroids,endocrine abnormalities, or the impairment of thebody composition and peripheral skeletal muscles.Some specifically identified risk factors for osteoporosisin patients with COPD are smoking,increased alchohol intake, reduced vitamin D levels,genetic factors, treatment with corticosteroids,reduced skeletal muscle mass and strength, lowBMI and changes in body composition, hypogonadism,reduced levels of insulin-like growth factors,and chronic systemic inflammation.Clinical Aspects of COPDSymptoms of COPDThe symptoms that are associated with COPDare usually not present until the patient has beensmoking a pack of cigarettes a day for at least20 years. The patient usually presents in the fifthdecade with a chronic cough that often is worsenedafter a viral respiratory infection. The second majorsymptom is dyspnea. It usually does not presentuntil the sixth or seventh decade. Dynamic hyperinflationduring exercise is recognized as an importantfactor in exercise limitation in COPD. Theextent of dynamic hyperinflation during exercisein COPD correlates with a reduced resting inspiratorycapacity. This limits the tidal volume responseto exercise. This inability to expand the tidalvolume in response to increasing metabolicdemand is an important contributor to exerciseintolerance in patients with COPD.Dyspnea is a common symptom in the olderpopulation, and often it is difficult to distinguishwhether heart disease or pulmonary disease is thecause or when both are present. High levels ofbrain natruretic protein (BNP) in the blood havebeen very helpful in differentiating cardiac andpulmonary causes of dyspnea in the acute setting.Natruretic peptides are the major hormones of thenatruretic peptide system, which is highly activatedin patients with both left and right heartfailure. BNP is predominantly secreted by the cardiacventricles, and in both acute and chronic leftheart failure serum levels are increased. BNP levelscan be increased in patients with COPD, butthe levels are not nearly as high as those foundwith CHF.In a large clinical trial called the Breathing NotProperly trial, a subgroup analysis of those withpulmonary disease (asthma or COPD) showed thatBNP had a negative predictive value of 97.7% at acut-off of 100 pg/mL. BNP levels have been shownto be associated with diminished exercise toleranceand a poor prognosis in patients with chronic leftheart disease. In addition, BNP levels have beenfound to be increased in patients with pulmonaryarterial hypertension and reflect the clinical andhemodynamic status in this patient population.Elevated BNP levels resulting from pulmonaryhypertension secondary to end-stage lung diseasehave been reported in the absence of left ventricularfailure. They have been shown to be a risk factorfor death, independent of lung functionalimpairment or hypoxemia. Serum BNP can be usedas a prognostic marker as well as a screening toolfor significant pulmonary hypertension in patientswith chronic obstructive lung disease.The diagnosis of acute pulmonary embolism(PE) in patients with COPD is also often difficultto make, especially during an acute exacerbationof COPD when symptoms of the two conditionsmay be indistinguishable. In fact, the diagnosismay not be possible on clinical grounds. Theprevalence of PE in postmortem series of patientswith COPD is extremely high, ranging from 28 to51%. In the Prospective Investigation of <strong>Pulmonary</strong>Embolism Diagnosis study, in a subgroup of164 COPD (Braman)ACC-PULM-09-0302-011.indd 1647/11/09 1:59:14 PM


patients with COPD and suspected PE, 19%received a diagnosis of PE on pulmonary angiography.Risk factors and symptoms were similar tothose with and without pulmonary embolus.Recent studies have reported a 25% prevalence ofPE in a group of patients admitted to the hospitalfor a severe exacerbation of COPD, but the studyresults may not be generalized to most occurrencesof COPD because some patients in this studywere excluded if they had evidence of a potentialinfection.More recent evidence, which excluded thediagnosis of PE on the basis of a normal( 500 g/mL) d-dimer test and negative diagnosticimaging studies, showed a much lowerprevalence of pulmonary embolism. Only 6.2% ofpatients who had a clinical suspicion of pulmonaryembolism and only 1.3% who had no clinicalsuspicion, judged by emergency department physicians,were discovered to have PE. This studyshowed that the prevalence of suspected PE inpatients presenting with a COPD exacerbation isvery low, and routine investigation for PE in thisgroup is not warranted. There are some importantlessons from these recent studies. When there isno suspicion of infection, either viral or bacterial,suspicion for PE in patients with a COPD exacerbationshould be greater, especially in thosepatients who are sick enough require hospitalizationas they my have more severe and debilitatingdisease that predisposes them to PE. Other cluesthat should increase clinical suspicion of pulmonaryembolism in COPD patients are a previoushistory of malignancy and a decrease in the Paco 2compared with baseline values. PE is a seriouscomorbid condition in patients with COPD, anda high clinical suspicion should be maintained inthe appropriate clinical setting.During the last two decades, as a result of anintense antismoking campaign, the rate of smokingin the United States has been reduced fromapproximately one third of the adult populationto its current level of approximately 22%. In manysusceptible smokers who stopped smoking in midlife,the onset of COPD symptoms is delayeduntil late in life. It is not unusual to see a patientwith COPD present for the first time with symptomsat 75 years. The main symptoms andvarious stages of disease severity are presented inTable 8.Table 8. GOLD Guidelines: Classification of Severity of COPD,2006Stage/characteristicsI: Mild COPDFEV 1 80% of predictedWith or without chronic symptoms (cough, sputumproduction)FEV 1/FVC 70%II: Moderate COPDFEV 1/FVC 70%50% FEV 1 80% of predictedDyspnea worsens; cough sputumIII: Severe COPDFEV 1/FVC 70%30% FEV 1 50% of predictedDyspnea worsens; cough sputumIV: Very severe COPDFEV 1/FVC 70%FEV 1 30% of predicted or respiratory/heart failureWeight loss and anorexiaHemoptysisCough syncopeCoughing spellsDepression and or anxietyCor pulmonalePhysical Examination and Laboratory TestsThe signs of COPD are well known: slow andprolonged expiration, wheezing, chest wall hyperinflation,limited diaphragmatic motion on auscultation,distant breath sounds, and heart sounds andoften coarse early inspiratory crackles. The accessorymuscles of respiration may be in use, and thepatient may be using pursed-lip breathing. Signsof cor pulmonale include pedal edema, a tendercongested liver, and neck vein distention. Cyanosismay be present as well as asterixis, which is associatedwith severe hypercapnia. The roentgenographicsigns of COPD include a low flat diaphragm,increased retrosternal airspace, and a teardropshapedheart. Pruning of the pulmonary arterialvessels and bullae are seen in emphysema. Thehigh-resolution CT scan has a much better sensitivityand specificity but is not recommended forroutine use. It may be helpful when surgery forlarge bullae is contemplated.<strong>Pulmonary</strong> function studies are important forthe diagnosis and staging of COPD. Because thebest correlate with rates of morbidity and mortalityis the FEV 1, proposals for staging have usedthis test. The 1995 ATS standards suggested a<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 165ACC-PULM-09-0302-011.indd 1657/11/09 1:59:15 PM


severity scoring on the basis of the severity ofairflow obstruction: stage I is FEV 1 50% of predicted;stage II is FEV 135 to 49% of predicted, andstage II is FEV 1of 35% of predicted. Patientswith stage I disease usually have little impairmentin their quality of life. Patients with stageII COPD have a significant impairment in thequality of their lives, which results in a large percapitahealth-care cost. Patients with stage IIIdisease are profoundly impaired, are often hospitalizedwith severe exacerbations of disease,and therefore are frequent users of health-careresources.A new severity scoring system was introducedby the GOLD guidelines (Table 8). It began stagingwith an “at-risk” group, stage 0, to emphasize thepoint that patients with risk factors (cigarettesmoking, occupations) who have symptoms coughand sputum need to be counseled about COPDeven though their screening spirometry is normal.The GOLD guidelines were updated in 2006(www.goldcopd.com). The spirometric classificationof severity of COPD now includes only fourstages: stage I, mild; stage II, moderate; stage III,severe; stage IV, very severe. The category stage 0,at risk, is no longer included as a stage of COPD.The GOLD committee believed that there wasincomplete evidence that the individuals who meetthe criteria for this stage (chronic cough and sputumproduction, normal spirometry) necessarilyprogress on to stage I. Other GOLD stages aremodifications of the previous (1995) ATS criterianow adopted in the new ATS/ERS Statement.Although the measurement of lung functionremains the best way to diagnose COPD, the routineuse of spirometry for case finding in adultswith exposure to risk factors (such as cigarettesmoking) or for those with persistent respiratorysymptoms remains controversial. This controversyis, in part, because no randomized clinical trial hasdemonstrated that early detection of COPDchanges the course of disease or increases the rateof smoking cessation. A conclusion not to recommendspirometry screening was reached througha systematic review of the evidence by a taskforce of the Agency for Health-Care Research andQuality. The Agency for Health-Care Researchand Quality conclusion was that the evidence doesnot justify recommending spirometry as a routinetool in the practice of primary care.Differentiation From AsthmaTable 9 lists the most common distinguishingfeatures between asthma and COPD. Epidemiologicstudies have shown that as many as 30% ofpatients with fixed airflow obstruction have a historyof asthma. Differentiating asthma from COPDmay be difficult and at times not possible, especiallyin older patients. The most useful discriminator,the response to an inhaled short-acting-agonist bronchodilator, may not differentiateasthma from COPD. When the postbronchodilatorFEV 1returns to normal, the diagnosis is asthma(no fixed airflow obstruction). On the one hand,many asthmatics, especially those with longstandingdisease, may not show any response, andsome patients with COPD will show a brisk bronchodilatorresponse. On the other hand, asthmaticswith fixed airflow obstruction show other featuresthat are distinct from COPD caused by cigarettesmoking.When compared with patients with COPD ofcomparable age, patients with asthma have significantlymore eosinophils in the peripheralblood, sputum, and BAL fluid on bronchial biopsy,have greater numbers of neutrophils in the sputumand BAL fluid, have greater ratios of CD4 /CD8 T cells infiltrating the airway, and have greaterthickness of the airway basement membrane thanpatients with COPD. Other distinguishing featuresinclude greater levels of exhaled nitric oxide, loweremphysema scores on high-resolution CT scans,and greater diffusing capacities. Because many ofthese tests are impractical, not available for widespreaduse, or not discriminatory enough forclinical use, the most recent ATS/ERS StatementTable 9. Distinguishing Features of COPD and AsthmaCOPDOnset in midlifeSymptoms slowly progressiveLong history of tobaccosmokingDyspnea during exerciseLargely irreversible airflowlimitationAsthmaOnset early in life (oftenchildhood)Symptoms vary from day todayAllergic rhinitis and/oreczema historySudden dyspnea after acuteexposureLargely reversible airflowlimitation166 COPD (Braman)ACC-PULM-09-0302-011.indd 1667/11/09 1:59:15 PM


on COPD stated: “Some patients with asthmacannot be distinguished from COPD with thecurrent diagnostic tests. The management of thesepatients should be similar to that of asthma.”COPD: Outpatient ManagementOnce the diagnosis of COPD is established,pulmonary function testing can be helpful by stagingthe disease. Pharmacologic intervention isoffered according to disease severity and thepatient’s tolerance for specific drugs (Fig 4). Thegoals of therapy are to induce bronchodilation,decrease airway inflammation, and improve mucociliarytransport. Because COPD is basically anirreversible condition, disease management ofdaily symptoms and other impacts on quality oflife are most essential. The patient who still smokesshould be encouraged to quit. Preventive therapywith a pneumococcal vaccine and a yearly influenzavaccine is recommended. Vaccination withthe flu shot has been shown to result in 52% fewerhospitalizations for pneumonia and influenza inpatients with COPD. Vaccinated patients also havefewer outpatient visits for respiratory symptoms.Smoking CessationCigarette smoking compromises airway functionby damaging airway epithelial cells, increasingmucous viscosity, and slowing mucociliary clearance.There is a greater bacterial adherence tooropharyngeal epithelial cells in smokers comparedwith nonsmokers. Smokers are prone toFigure 4. Treatment plan from the GOLD guidelines.attacks of acute bronchitis. As a result, they showincreased rates of acute exacerbations, visits to theemergency department, work absences, and frequencyof medication use. Smokers also showgreater decrease in lung function, worsening ofrespiratory symptoms, and lower quality of lifescores than a comparable group of nonsmokingpatients. The presence of respiratory illness suchas COPD is not a motivator for smoking cessation.Physician-delivered smoking cessation interventionscan significantly increase smoking abstinencerates. Smoking cessation interventions for COPDshould include the following: (1) physician intervention(set a quit date), (2) group smoking cessationclinics, and (3) pharmacologic therapy withnicotine replacement is needed in highly dependentsmokers. These patients can be identified asthose who smoke a pack a day or more, requiretheir first cigarette within 30 min of arising in themorning, and find it difficult refraining from smokingin places where it is forbidden. Therapy withthe antidepressant bupropion hydrochloride hasalso been shown to be effective. The Lung HealthStudy showed that despite intensive antismokingefforts, middle-age smokers with COPD are notlikely to quit. Only 22% were sustained quittersafter a 5-year follow-up.Pharmacologic TherapyShort-Acting 2-Agonists: Short-acting 2-agonistscan be used for the following:• Achieve variable degrees of bronchodilatation,ie, rapid onset of action with duration of 4 to6 h;• Improve symptoms and, in most studies, exercisecapacity; and• Metered-dose inhaler use of 2-selective agentsis safe tid or qid; greater doses may causehypokalemia, cardiac arrhythmia, and reducedarterial oxygen tension.Long-Acting 2-Agonists: Long-acting 2-agonistshave the following advantages:• Maximal effect comparable with short-acting 2-agonists with longer duration of action (12 h)for agents such as salmeterol and formoterol;• Improved symptoms and quality-of-life measures;no tachyphylaxis found, and<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 167ACC-PULM-09-0302-011.indd 1677/11/09 1:59:15 PM


• Good safety profile with recommended doses:occasional tremor is noted that improves afterseveral days of use.Anticholinergic TherapyShort-acting agents (ie, ipratropium) affordbronchodilatation and relieve of symptoms for 4to 6 h. The long-acting anticholinergic tiotropiumoffers improved bronchodilatation for 24 h andtherefore can be given once a day. No tachyphylaxisis noted with these agents, but their effects onlong-term prognosis of COPD are being studied.Tiotropium has been shown to reduce exacerbationrates and hospitalizations caused by exacerbations.Finally, they have mild side-effects, such as occasionaldry mouth. 2-Agonists Plus Anticholinergic Drugs: Combiningboth agents provides a small additionalbenefit to either drug alone, and no additionalside effects from combination therapy are noted.Theophylline: Theophylline causes small changesin FEV 1with chronic use and an improvementin symptoms and exercise capacity. Frequentadverse effects, including nausea, diarrhea, headache,and irritability, are noted with therapeuticblood levels. Seizures and cardiac arrhythmias areconsiderably (ie, 10 to 15 times) more common inelderly with toxic blood levels.Systemic Corticosteroids: Twenty to thirty percentof patients with chronic COPD symptomsimprove if administered oral steroid therapy.Responders have more eosinophils in induced sputumand bronchial biopsies and likely have concomitantasthma. Long-term treatment with oralcorticosteroids is not recommended for patientswith COPD. The treatment of hospitalized patientswith high doses results in fewer treatment failuresand shorter stays; 2 weeks of therapy is sufficientafter hospital discharge. Complications includecataracts, osteoporosis, secondary infection, anddiabetes and skin damage.Inhaled Corticosteroids: No significant benefitin long-term decline in FEV 1has been noted forinhaled corticosteroids. They may provide smallbut significant improvements in lung function and6-min walk test results. Studies have also demonstratedfewer exacerbations of COPD and impro vedquality of life indexes. Benefits are enhanced whencoupled with long-acting -agonist treatment.Guidelines recommend with FEV 1 50% predictedand frequent exacerbations.Mucolytic Drugs: The use of mucolytic drugshas a variable effect in patients with COPD. Randomizedcontrolled trials suggest that they areineffective at shortening the course or improvingoutcomes of patients with acute exacerbations.Antibiotics: The results of multiple trials favorthe use of antibiotics for acute exacerbations ofCOPD that include worsening dyspnea, increasedsputum volume, and sputum purulence. However,there is little evidence to support their prophylacticuse.Oxygen Therapy: Several controlled studieshave been completed: oxygen vs no oxygen (MedicalResearch Council [MRC]) and continuous vsnocturnal oxygen (Nocturnal Oxygen TherapyTrial). In patients with hypoxemia and congestiveheart failure, death rates are significantly lowerand quality of life indexes are improved whenoxygen is in chronic use. Patients who used oxygenfor at least 15 h/d had a significant decrease intheir pulmonary artery pressures and an increasein cardiac output.Oxygen should be prescribed when the followingoccur: (1) the arterial Pao 2is 55 mm Hg orarterial oxygen saturation 88%, and (2) Pao 2is56 to 59 mm Hg with ECG evidence of P pulmonale,pedal edema, and/or secondary erythrocytosis.Long-term oxygen therapy does not improvesurvival in patients with moderate hypoxemia(Pao 256 to 65 mm Hg) or in patients with onlynocturnal hypoxemia. Supplemental oxygen duringexercise training may improve dyspnea andallow patients with COPD to tolerate greater levelsof exercise. However, studies evaluating quality oflife in patients who show hypoxemia duringexercise have not supported oxygen use for thisindication.<strong>Pulmonary</strong> Rehabilitation: <strong>Pulmonary</strong> rehabilitation(PR) is a multidisciplinary programthat attempts to return the patient with COPDto the highest functional capacity possible.PR addresses deconditioning, social isolation,anx iety and depression, muscle wasting, andweight loss. Evidence supports the use of lowerextremityexercise training because it improvesexercise tolerance, and upper-extremity strengthand endurance training is recommended as well.168 COPD (Braman)ACC-PULM-09-0302-011.indd 1687/11/09 1:59:16 PM


PR improves dyspnea, improves quality of lifescores, and reduces the number of hospitalizationsand days in the hospital; however, the effects onsurvival are not definite. PR has been shown tohelp patients requiring long-term mechanical ventilationby improving overall strength, weaningoutcomes, and functional status.Nutrition and COPDMalnutrition occurs in one quarter to one thirdof patients with moderate-to-severe COPD and isan independent risk factor for mortality. Both fatmass and fat-free mass are depleted, the lattercaused by depressed protein synthesis. Because theproinflammatory cytokines IL-6 and TNF- havebeen shown to be elevated, tt is believed that weightloss, particularly skeletal muscle mass loss, is associatedwith a systemic inflammatory response inmalnourished patients with COPD. Skeletal muscleloss in COPD is probably multifactorial in origin.There is a link to systemic inflammation even withoutweight loss.Studies have shown that leptin, an adipocytederivedhormone involved in the control of bodyweight, is decreased in patients with COPD. Serumleptin levels have been correlated with TNF-levels, thus creating a link between nutritionalstatus and systemic inflammation in COPD. Restingenergy expenditure is elevated and contributesto the negative energy balance. Nutritional supplementsalone do not reverse the loss; results withanabolic steroids, growth hormone, and the progestationalagent megestrol acetate show someeffect on appetite and body weight; however,improved exercise tolerance and respiratorymuscle function do not always follow. A combinationof nutritional support and exercise as ananabolic stimulus appears to be the best approachto obtaining marked functional improvementNoninvasive Positive Pressure VentilationA theoretical advantage has been shown to restingchronically fatigued muscles at night. Patientswith COPD have a high prevalence of sleep apneaand hypoxemia and hypercarbia at night. Theyhave less REM sleep and shorter sleep times thannormal subjects. It has been theorized that noninvasivepositive pressure ventilation (NPPV) wouldhelp. However, trials have shown conflicting conclusions.Patients without CO 2retention show littlegain; patients with severe CO 2retention, especiallythose that have severe nocturnal O 2desaturation,appear to respond most favorably.The most recent Cochrane evidenced-basedreview on this topic concluded that nocturnalNIPPV for at least 3 months in hypercapnic patientswith stable COPD had no consistent clinically orstatistically significant effect on lung function, gasexchange, respiratory muscle strength, sleep efficiency,or exercise tolerance. However, the smallsample sizes of these studies precluded a definiteconclusion regarding the effects of NPPV inpatients COPD.Surgery for COPDLung volume reduction surgery (LVRS) hasemerged as a treatment for far advanced COPDcaused by emphysema. LVRS involves resection offunctionless areas of emphysematous lung toimprove lung elastic recoil and lung and chest wallhyperinflation. The overall rate of mortality withthe procedure is 0 to 6% within 30 days of surgeryand 0 to 8% at 6 months, although in patients withvery severe degrees of obstruction the rates aremuch greater. The range of FEV 1improvementbefore and after surgery is approximately 250 to350 mL. Borg scale dyspnea scores on the 6-minwalk test show improvement as do the distanceswalked. Sixteen percent to 42% of patients no longerrequire oxygen after surgery.The recently completed National EmphysemaTreatment Trial (NETT) defined a subgroup ofpatients a very high risk for mortality. Interimanalysis showed that if the FEV 1was 20% predictedand the CT scan of the lungs showed homogeneousemphysema or the carbon monoxidediffusing capacity was 20%, there was a high riskof death after surgery and little chance of functionalimprovement. Among those not in the highriskgroup, there was a 2.2% mortality with surgeryat 30 days compared with a 0.2% mortality in themedically treated group. After 1 month, 28% of thesurgery group were hospitalized, living in a nursinghome or rehabilitation facility, or unavailablefor interview. Changes in exercise capacity, pulmonaryfunction, quality of life measures, and dyspneaat 6 to 24 months all favored surgery. Further<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 169ACC-PULM-09-0302-011.indd 1697/11/09 1:59:16 PM


subgroup analysis showed that those with upperlobeemphysema on CT scans and with lowexercise capacity at the beginning of the trial hadthe most favorable outcomes.Given the potential risks of median sternotomyand thoracoscopy, the usual surgical approachesto LVRS, less-invasive procedures are being studied,including bronchoscopic LVRS, which is performedwith the placement of endobronchialone-way valves. Preliminary studies have shownimprovement in exercise capacity and dynamichyperinflation with this procedure.Resection of large bullae is occasionally necessaryfor dyspnea, chest pain, or recurrent pneumothorax.Other complications that can occur withlarge bullae are bleeding and infection. The bestresults occur when the bullae occupy more thanone third of the hemithorax. Factors that do notfavor surgical resection are the presence of multiplesmall bullae and diffuse emphysema in the remainderof the lungs. No prospective clinical trials havebeen performed. Case series have shown improvementin dyspnea and even reversal of respiratoryfailure after bullectomy. However, in those withunderlying emphysema there is a progressivedecline in the FEV 1over time and return of cripplingdyspnea in many. Bullectomy can be performedby thoracoscopic surgery in selectedpatients.Lung transplantation can be life-saving forpatients with very advanced disease complicatedby respiratory failure and secondary pulmonaryhypertension. Patients are considered for single ordouble lung transplantation when the FEV 1is 25%, and/or the Paco 2is 55 mm. Hg. Theprocedure is costly, limited by lack of organs, andrequires prolonged immunosuppression. Recentstudies have suggested that unlike patients withcystic fibrosis and idiopathic pulmonary fibrosis,patients with COPD do not fair well after transplantation,with limited exercise capacity becauseof skeletal muscle dysfunction. In January 2008,The International Society for Heart and LungTransplantation has reported survival figures forCOPD of 85% at 1 year and 68% at 3 years.Acute Exacerbation of COPDUnfortunately, many definitions of the acuteexacerbation or “flare” of COPD exist, and manyauthors use substantially different criteria. Awidely quoted description of a COPD exacerbationincludes “the Winnipeg criteria,” which have alsoincluded a staging proposal. Patients with COPDexacerbation have the following three changes intheir clinical condition: (1) worsening of dyspnea,(2) an increase in sputum volume, and (3) sputumpurulence. Patients with type I exacerbations haveall of the aforementioned symptoms. Patients withtype II exacerbations have two of the three symptoms.Type III exacerbations are characterized byat least one of these symptoms and also one of thefollowing clinical criteria: upper respiratory infectionin the past 5 days, fever without other apparentcause, increased wheezing, increased cough,and increased respiratory rate or heart rate by 20%above baseline. A more simple and practical definitionhas been offered by the ATS as “an event inthe natural course of the disease characterized bya change in the patient’s baseline dyspnea, coughand/or sputum beyond day-to-day variabilitysufficient to warrant a change in management.”The classification of severity offered with thisdefinition ranks the clinical relevance of the episodeand its outcome: level I, treatment at home;level II, requiring hospitalization; and level III,respiratory failure.Before the onset of an exacerbation of COPD,there is usually a prodromal period of up to 7 dayswhen symptoms of increased dyspnea, cough, sorethroat, and the common cold occur. These symptomsare not accompanied with a decrease in lungfunction. On the day of the exacerbation, there isa small but significant decrease in lung function,including peak flow rates, FEV 1, and FVC. Peakflow rate recovery to baseline values is completein 75% of patients by 1 month. Approximately 7%of patients with a COPD exacerbation do not returntheir peak flow rates to baseline by 3 months. Thisfinding suggests that a substantial proportion ofexacerbations of COPD are associated with a permanentloss of lung function. If this were to occurrepeatedly, it could lead to an accelerated decreasein FEV 1over years. Whether repeated respiratoryinfections or exacerbations of COPD do lead to amore rapid decrease of lung function over time hasnot been proven.The most frequent cause of an acute exacerbationof COPD is thought to be respiratory infection.Sputum cultures in patients with mild to170 COPD (Braman)ACC-PULM-09-0302-011.indd 1707/11/09 1:59:16 PM


moderately severe COPD often show nonpathogenicbacteria (Gram-negative and Gram-positivesuch as Streptococcus viridans, Neisseria sp, orCorynebacterium sp). Pathogenic bacteria includeHaemophilus influenzae (22%), Pseudomonas aeruginosa(15%), Streptococcus pneumoniae (10%), andMoraxella catarrhalis (9%). Sputum samples mayhave limited validity considering the possiblecontamination from oropharyngeal secretions.Techniques that potentially preclude contaminationinclude transtracheal aspirates, BAL, andprotected brush sampling during bronchoscopy.Studies in which the authors used these techniqueshave shown that patients requiring hospitalizationcompared with those treated as outpatients for anexacerbation of COPD and those outpatients withthe most severe degree of airflow obstruction aremore likely to harbor Gram-negative enteric organisms.The presence of bacterial colonization in thestable state is associated with an increased exacerbationfrequency, and patients colonized withH influenzae have an increased total symptom countand more sputum purulence at the time of exacerbation.When molecular typing is used, it has alsobeen shown that the acquisition of new strains ofbacteria such as H influenzae, M catarrhalis, andS pneumoniae are associated with an exacerbationof COPD in as many as one third of cases.The role of viral infection has been recentlyrevisited with the use of sophisticated methodssuch as polymerase chain reaction viral detectionas well as more standard methods such as serologictesting and culture. Viruses may be detected innearly 40% of COPD exacerbations when thesetechniques are used, including rhinovirus, respiratorysyncytial virus, coronavirus, influenza, andparainfluenza. Evidence for Mycoplasma infectionhas been found in up to one third of the acuteexacerbations of chronic bronchitis in some studies;however, in most patients Mycoplasma and Chlamydiapneumonia are seen in 10% of cases. ACOPD exacerbation associated with a virus is likelyto produce more severe symptoms, and the exacerbationis likely to last longer (13 days). There isalso more laboratory evidence of an acute-phaseresponse (eg, elevated levels of IL-6 and fibrinogen)with an associated viral infection.Clearly, environmental factors are also important.Hospital admissions for COPD appear to behigh when air pollution levels in the ambientTable 10. Factors Associated With Increased Risk forExacerbationsIncreased ageSeverity of airway obstruction(FEV 1impairment)Chronic bronchial mucus hypersecretionFrequent past exacerbationsDaily cough and wheezePersistent symptoms of chronic bronchitisHypercapniaHypoxiaComorbid heart diseaseRight ventricular failureLow serum albuminSingle nucleotide polymorphism in CCL1 geneenvironment are high. Small increases in SO 2andairborne particles have been shown to causeincreases in emergency room visits for COPD duringwinter and summer seasons by 6% and 9%,respectively. Whatever the cause of an acute exacerbationof COPD, it is clear that bronchial inflammationis enhanced during this event. Increasedlevels of myeloperoxidase have been found in thesputum, indicating neutrophilic activity, and highlevels of IL-8 and leukotriene B 4, well-known neutrophilchemoattractants, are also seen. Duringrecovery, levels of sputum chemoattractants andinflammatory markers rapidly decrease. Risk factorsassociated with COPD exacerbations are listedin Table 10.Diagnostic and TherapeuticApproachesOften, patients present to a local emergencydepartment or hospital with an acute exacerbationof COPD. It is useful to do a chest roentgenogram.The chest roentgenogram may be positive inapproximately 15% of cases, and in 25% of suchpatients there is a change of management promp tedby the chest roentgenogram result. Spirometry, onthe other hand, has not been helpful in decisionmakingfor patients with an acute exacerbation ofCOPD. There is poor correlation between arterialblood gases and the FEV 1at the time of presentation.The FEV 1has not been helpful in predictingneed for hospital admission. Although severeabnormalities of arterial blood gases and history ofprevious relapses are helpful, there are no predictivemodels of adequate sensitivity and specificity<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 171ACC-PULM-09-0302-011.indd 1717/11/09 1:59:16 PM


on which to rely. The best indicators for the needfor mechanical ventilation include the blood gasvalues on admission and the degree of change inpH after initial oxygen therapy. Significant predictorsof hospital mortality include older age, lowerbody mass index, poor functional status beforeadmission, lower ratio of arterial partial pressureof oxygen (Pao 2) over the fraction of inspired oxygenratio, history of congestive heart failure, lowserum albumin, and presence of cor pulmonale.Most therapeutic interventions for the treatmentof the acute exacerbation of COPD have beenstudied through randomized controlled trials. Ingeneral, mucolytic agents are ineffective in shorteningthe course and improving the outcome ofpatients. Chest physiotherapy and mechanicalpercussion of the chest are also ineffective andmay be potentially harmful. 2-agonists and anticholinergicagents appear to be equivalent in theirusefulness and are superior to all parenterallyadministered bronchodilators, including methylxanthines.The addition of a methylxanthine toinhaled bronchodilators has also been carefullyexamined. One study showed a statistical trendtoward lower hospitalization rates for patientsgiven IV aminophylline in the emergency department,whereas two other studies showed no significantadvantage. The effect of adding twoinhaled bronchodilators has also been carefullystudied. There appears to be only marginalimprovement to the use of both agents. The sideeffectsof anticholinergic therapy are generallyfewer and mild. The adverse effects of albuterol inthe acute setting include tremors, headache, andpalpitations. Changes in heart rate, BP, and ECGtracings are also possible.The use of systemic steroids for patients withan acute exacerbation of COPD who require hospitalizationis helpful. A change to po therapycontinues to have a positive effect, but only for2 weeks after hospitalization. The most commonadverse effect was hyperglycemia. Antibiotictherapy has been a controversial issue. Patientswith more severe exacerbations are more likely tobenefit. For instance, antibiotics appear to be effectivewhen two of the three symptoms of increaseddyspnea, sputum volume, and purulence arepresent. However, there is still some controversyconcerning the role of antibiotics for COPDexacerbations, especially in studies performed inprimary care, where generally patients with milderdisease are treated. Treatment with antibiotics inaddition to oral corticosteroids has been associatedwith a longer time to the next exacerbation and adecreased risk of developing a new exacerbation.The most consistently measured end point with apositive result is an improvement in peak expiratoryflow rate that improves a mean of 11 L/min,more in patients with the antibiotic treatment.Arterial blood gas measurements should bedone if a patient does not show prompt improvementwith initial measures. The initiation ofmechanical ventilation should be considered when,despite optimal medical therapy and oxygenadministration, there is acidosis (pH 7.35) andhypercapnia (Paco 245–60 mm Hg) and a persistentlyelevated respiratory rate 24 breaths/min.Mechanical ventilation can be administered viainvasive ventilation (use of an endotracheal tube)or noninvasive (nasal or face masks). NPPV isbeneficial for patients who have a high likelihoodof respiratory failure and who may require invasivemechanical ventilation. In the first hours,NPPV requires the same level of supervision asconventional mechanical ventilation. The use ofNPPV provides a statistically significant differencein the need for intubation reported in severalclinical trials. Arterial blood gases improve becauseof an increase in alveolar ventilation without significantmodifications in the alveolar V . /Q . mismatchingand gas exchange in the lungs.The addition of NPPV to standard care inpatients with an acute exacerbation of COPD havedecreased the rate of endotracheal intubation 28%,length of hospital stay 4.57 days, and in-hospitalmortality rate 10%. Subgroup analyses have shownthat these beneficial effects occurred only inpatients with severe exacerbations (pH 7.30), notin those with milder exacerbations. Optimal NPPVdelivery methods, optimal duration of treatment,and appropriate delivery pressures are additionalquestions that need to be addressed in the literature.In general, a bilevel ventilatory support withcontinuous positive airway pressure of 4 to8 cmH 2O and pressure support ventilation of 10 to15 cmH 2O provides the most effective mode ofNPPV.Contraindications for NPPV include the following:respiratory arrest; cardiovascular instability(hypotension, arrhythmias, myocardial infarction);172 COPD (Braman)ACC-PULM-09-0302-011.indd 1727/11/09 1:59:17 PM


impaired mental status causing an inability tocooperate, copious and/or viscous secretions withhigh aspiration risk, recent facial or gastroesophagealsurgery, craniofacial trauma and/or fixednasopharyngeal abnormality, burns, and extremeobesity. Patients meeting these exclusion criteriashould be considered for immediate intubation.Patients who are hospitalized with an acuteexacerbation of COPD have a mortality rate ofapproximately 3%. Almost half of all patientsadmitted with hypercarbic respiratory failure willbe readmitted to the hospital within the next6 months. There is a significant decrement infunctional status and quality of life and the likelihoodin such patients. Mortality for those patientswho require treatment in an ICU may be closeto 50%.Outcomes of COPDCOPD is the only major cause of mortality withan increasing incidence worldwide, and it isexpected to become the third-leading cause ofdeath by 2020. Traditionally, clinical staging ofseverity and risk of mortality has relied on themeasurement of FEV 1. Staging criteria for COPDwith use of the percentage of predicted FEV 1canbe useful in predicting impairment in healthrelatedquality of life. Even patients with milddisease show substantially compromised healthrelatedquality of life. Comorbid conditions arecommon, and they also negatively influencehealth-related quality of life.Other measures of COPD such as the BMI anddyspnea scores have proved useful in predictingoutcomes such as survival. BMI is easily obtainedby dividing weight (in kilograms) over height(in m 2 ). Values 21 kg . m 2 are associated withincreased rates of mortality. Functional dyspneacan be assessed by the MRC dyspnea scale as follows:0, not troubled with breathlessness exceptwith strenuous exercise; 1, troubled by shortnessof breath when hurrying or walking up a slighthill; 2, walks slower than people of the same agebecause of breathlessness or has to stop for breathwhen walking at own pace on the level; 3, stopsfor breath after walking approximately 100 m orafter a few minutes on the level; and 4, too breathlessto leave the house or breathless when dressingor undressing. A simple grading index, the BODETable 11. Outcome Measures for COPD (Morbidity/Mortality)FEV 1Dyspnea (MRC scale)BMIExercise capacityBODE indexArterial blood gasesCOPD exacerbationsQuality of life health statusindex, has been proposed as a useful tool to gradeCOPD severity and predict outcomes. The systemuses the body mass index (B), the degree of airflowobstruction measured as the FEV 1as percent predicted(O), dyspnea measured by the MRC dyspneascale (D), and exercise capacity measured bythe six-minute walk test (E). Patients with greaterBODE scores were at greater risk for death. Thescore proved to be a better discriminator than lungfunction alone. A summary of indexes of COPDoutcomes is listed in Table 11.SummaryIn the United States, 16 million adults haveCOPD. As the United States population ages, theprevalence for this disease will increase evenmore. COPD currently accounts for approximately110,000 deaths per year. There are 500,000 hospitalizationsin this country for COPD and 16million office visits to physicians. The directhealth costs are $18 billion/yr. It is likely thatthe care of patients with COPD will consume agrowing proportion of the pulmonologist’stime in the years to come. Differentiation fromanother chronic airway disease, asthma, is oftendifficult.Annotated BibliographyAgusti AG, Gari PG, Sauleda J, et al. Weight loss inchronic obstructive pulmonary disease: mechanismsand implications. Pulm Pharmacol Ther 2002; 15:425–432Weight loss is a common occurrence in COPD, and whensevere, it is labeled the pulmonary cachexia syndrome. It is aprognostic indicator of poor outcome. The mechanism is notentirely understood, but loss of skeletal muscle mass is themmain cause of weight loss. Loss of fat is also contributory.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 173ACC-PULM-09-0302-011.indd 1737/11/09 1:59:17 PM


Poor nutritional intake, high metabolic rate, sedentary lifestyle,tissue hypoxia, and a systemic inflammatory responsefrom high levels of TNF have all been proposed as causative.It is likely that oxidative stress also enhances muscle proteolysis.Alsaeedi A, Sin DD, McAlister FA. The effects ofinhaled corticosteroids in chronic obstructive pulmonarydisease: a systematic review of randomized placebo-controlledtrials. Am J Med 2002; 113:59–65This article presents a systematic review of inhaled corticosteroidsrelating to the exacerbation rate with COPD. It reviewsnine randomized trials and concludes that inhaled corticosteroidsdo reduce the exacerbation rate in COPD. It remindsus that side-effect profiles are also increased. Oropharyngealcandidiasis and skin bruising were more likely to occur. Noeffects on overall mortality were observed in these trials.American Thoracic Society/European RespiratorySociety statement. Standards for the diagnosis andmanagement of individuals with alpha-1 antitrypsindeficiency. Am J Respir Crit Care Med 2003; 168:818–900This is an updated comprehensive review of the manyaspects of AAT deficiency and COPD. Augmentation therapywith purified human AAT concentrate is supported bytwo large observational studies suggesting that the progressionof emphysema may be slowed in patients with moderateemphysema (FEV 131 to 65% predicted). The rate of mortalitymay be decreased in patients with a lower FEV 1. Adversereactions to the concentrate are rare occurrences.Anthonisen NR, Connett JE, Kiley JP, et al. Effects ofsmoking intervention and the use of an inhaled anticholinergicbronchodilator on the rate of decline of FEV 1:the lung health study. JAMA 1994; 272:1497–1505This is a landmark study by a lung health study researchgroup. This lung health study, in a prospective way, substantiatedwhat had been learned from the Fletcher and Petostudy of the 1970s. An aggressive smoking-intervention programsignificantly reduced the age-related decrease in FEV 1in middle-age smokers with COPD. Unfortunately, the useof the inhaled anticholinergic bronchodilator ipratropiumdid not show similar improvement. Of importance in thisstudy is that with intense smoking intervention, only about20% of the group compromised sustained smokers duringthe 5-year study.Balmes J, Becklake M, Blanc P, et al. Occupationalcontribution to the burden of airway disease: an officialstatement of the American Thoracic Society. Am JRespir Crit Care Med 2003; 167:787–797This statement reviewed the evidence implicating occupationalfactors in the pathogenesis of obstructive airwaydiseases (asthma and COPD) and quantified the contributionof work-related risk to the burden of these diseases in thegeneral population. The review demonstrated that approximately15% of both asthma and COPD is likely to be workrelated, and a conservative estimate of the annual costs ofthis occupational asthma and COPD is nearly $7 billion inthe United States alone.Balmes JR. Occupational contribution to the burden ofchronic obstructive pulmonary disease. J Occup EnvironMed 2005; 47:154–160This is an excellent review of the contribution of occupationalexposures to the occurrence of COPD. Mechanisms,diagnosis, and management are each discussed.Barnes PJ. Emerging pharmacotherapies for COPD.Chest 2008; 134:1278–1286With better understanding of the inflammatory and destructiveprocesses in the pathophysiology of COPD, the authorargues that several new targets have been identified to treatthe disease. Although mediator antagonists tested in patientswith COPD have so far been disappointing, CXC receptor-2antagonists that block pulmonary neutrophil and monocyterecruitment may be more promising. Broad-spectrum antiinflammatorydrugs, including inhibitors of the enzymes phosphodiesterase4, p38 mitogen-activated protein kinase, nuclearfactor-kappaB, and phosphoinositide-3-kinase-gamma, maybe more effective. The author argues that the most promisingapproach is the reversal of corticosteroid resistance throughincreasing histone deacetylase-2 activity. This might beachieved by theophylline-like drugs, more effective antioxidants,and nonantibiotic macrolide agents.Barnes PJ. Mediators of chronic obstructive pulmonarydisease. Pharmacol Rev 2004; 56:515–548This is a very long and comprehensive review of the inflammatorycascade of COPD. For those who are interested in amore in-depth review of this topic, this comprehensive discussionwill be most helpful.Berry JK, Baum C. Reversal of chronic obstructive pulmonarydisease-associated weight loss: are there pharmacologicaltreatment options? Drugs 2004; 64:1041–1052This is a review of the current approaches to weight loss inCOPD. To date, interventional studies that have looked atnewer pharmacotherapy such as growth hormone and anabolicsteroids have been disappointments. Currently, earlyidentification of patients at risk, aggressive nutritional supplementation,and a vigorous exercise program has demonstratedthe greatest benefit.Bhowmik A, Seemungal TAR, Sapsford RJ, et al. Relationof sputum inflammatory markers to symptomsand lung function changes in COPD exacerbations.Thorax 2000; 55:114–120174 COPD (Braman)ACC-PULM-09-0302-011.indd 1747/11/09 1:59:17 PM


Patients with multiple exacerbations of COPD have highermean stable sputum levels of IL-6 and IL-8 than those withfew exacerbations. During acute exacerbations, cytokine levelscorrelate with sputum cells counts of inflammatory cells.The cell counts and cytokine levels do not predict the sizeand duration of lung function changes in the exacerbation.Biskobing DM. COPD and osteoporosis. Chest 2002;121:609–620This review discusses the problem of osteoporosis in advancedCOPD. Bone loss factors are diverse and include smoking,vitamin D deficiency, low mass index, hypogonadism, sedentarylifestyle, and the use of corticosteroids. Effectivestrategies to prevent bone loss including calcium and vitaminD are discussed. The use of calcitonin and bisphosphonateadministration is also reviewed. Because up to 60% ofpatients with advanced COPD have osteoporosis, the causesand approaches to osteoporosis in COPD are extremelyimportant.Calverley PM, Anderson JA, Celli B, et al. Salmeteroland fluticasone propionate and survival in chronicobstructive pulmonary disease. N Engl J Med 2007;22:775–789The results of the Towards a Revolution in COPD Health(TORCH) trial were published in early 2007. This was arandomized, double-blind trial comparing salmeterol plusfluticasone propionate at a dose of 500 µg twice daily (combinationregimen), administered with a single inhaler, comparedwith placebo, salmeterol alone, or fluticasone propionatealone for a period of 3 years. The reduction in deathfrom all causes (the primary outcome) among patientswith COPD in the combination-therapy group did notreach the predetermined level of statistical significance.There were significant benefits in all other outcomes amongthese patients. When compared with placebo, the combinationregimen reduced the annual rate of exacerbations from1.13 to 0.85 and improved health status and spirometricvalues (p 0.001 for all comparisons with placebo). Therewas no difference in the incidence of ocular or bone sideeffects.Celli BR, Cote CG, Marin JM, et al. The body-massindex, airflow obstruction, dyspnea, and exercisecapacity index in chronic obstructive pulmonary disease.N Engl J Med 2004; 350:1005–1012Traditionally the FEV 1has been used as the main indicatorof clinical outcomes including rate of mortality in COPD.This article explores other predictors for the risk of deathin COPD. These predictors, BMI (B), the degree of airflowobstruction (O), dyspnea (D), and exercise capacity (E), offeran index of severity, the BODE index, that offers a superiorgrading system for COPD mortality.Celli BR, MacNee W, Agusti A, et al. Standards forthe diagnosis and treatment of patients with COPD: asummary of the ATS/ERS position paper. Eur Respir J2004; 23:932–946This is the latest American Thoracic Society statement andposition paper on COPD. It is a comprehensive review of allaspects of the disease and a follow-up to the 1995 statement.Classification of disease and recommendations are similar tothe GOLD international guidelines.Celli BR, Thomas NE, Anderson JA, et al. Effect ofpharmacotherapy on rate of decline of lung functionin chronic obstructive pulmonary disease: results fromthe TORCH study. Am J Respir Crit Care Med 2008;178:322–323The primary outcome of the TORCH trial, death from anycause, did not reach statistical significance (p 0.052; seeaforementioned Calverley PM, et al. N Engl J Med 2007;356:775). Before treatment unblinding of the TORCH data,the authors of this Celli study tested the hypothesis thatpharmacotherapy would modify the rate of decline of postbronchodilatorFEV 1compared with placebo and exploredfactors that could affect this decline. The primary outcomewas the rate of postbronchodilator FEV 1decline. The rateof decline of FEV 1was slowest in patients on combinationtherapy (39 mL/yr), fastest in those randomized to the placeboarm (55 mL/yr) and intermediate (42 mL/yr) in bothsingle treatment arms. A slower rate of decrease in absolutemilliliters per year was observed in former smokers, females,patients 65 years, and those with FEV 1 30% predictedand patients with a BMI 25.The decline in FEV 1has been accepted as a key marker forprogression of COPD. The decline in postbronchodilatorFEV 1in this study was 55 mL/yr, similar to that seen inthe Lung Health Study 1 (52 mL), Lung Health Study 2(47 mL), Bronchitis Randomized on NAC Cost-Utility(54 mL), and Inhaled Steroids in Obstructive Lung Diseasestudies (59 mL). Before this study, smoking cessationwas the only intervention conclusively shown to alter therate of decline in FEV 1with COPD. In this study, there wasan association between exacerbation frequency and FEV 1decline, supporting previous observations. However, inpatients who had no exacerbations during the study, the rateof decline was significantly faster in the placebo group comparedwith active treatments (56 mL/yr vs 2731 mL/yr),which suggests that the effect of treatment on exacerbationswas not the sole mechanism responsible for the reduced rateof decline with active treatment. The authors suggest thatmaintenance of airway patency and reduction in hyperinflation,improvements in mucociliary clearance, or decreases inairway inflammation may contribute singly or together to<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 175ACC-PULM-09-0302-011.indd 1757/11/09 1:59:18 PM


produce the observed functional changes. Further studies areneeded to determine the mechanisms for the lung functionimprovements shown by TORCH.Chung KF, Adcock IM. Multifaceted mechanisms inCOPD: inflammation, immunity, and tissue repair anddestruction. Eur Respir J 2008; 31:1334–1356The mechanisms of COPD are numerous and complex. Theyare discussed in the excellent review. The authors suggestthat COPD severity and clinical phenotyping needs to becorrelated with cellular and pathological processes. Smallairways obstruction and emphysema are associated withcellular inflammation and structural remodeling. Other featuresinclude apoptosis as well as proliferation of cells andboth tissue repair and lack of tissue repair. Metalloproteaserelease, together with that of apoptotic factors, may underliethe emphysema and, conversely, fibrosis of the small airwaysmay be accounted for by the effects of growth factoractivation. In advanced disease, they suggest that influentialfactors include the development of autoimmunity, withactivation of dendritic cells and T-helper cells. An inabilityof macrophages to ingest apoptosed cells and bacteria mayexacerbate inflammatory responses. A discussion of how systemicinflammation may reflect the effect of cigarette smokeon nonpulmonary cells.Clini EM, Ambrosino, N. Nonpharmacologic treatmentand relief of symptoms in COPD. Eur Respir J 2008;31:1114–1124An excellent review of the nonpharmacologic therapies availableto the COPD patient, including pulmonary rehabilitation,long-term oxygen therapy, surgery, and noninvasiveventilation strategies.Connors AF Jr, Dawson NV, Thomas C, et al. Outcomesfollowing acute exacerbation of severe chronic obstructivelung disease. Am J Respir Crit Care Med 1996;154:959–967This is another landmark study from the Study to UnderstandPrognoses and Preferences for Outcomes and Risksof Treatments (SUPPORT) group. This five-center studydescribed the outcomes of patients hospitalized with anacute exacerbation of hypercarbic ( 55 mm Hg) COPD.The report shows that only 11% of the patients died duringtheir hospital stay; however, the 60-day, 180-day, 1-year, and2-year rates of mortality were extremely high (20%, 33%,43%, and 49%, respectively). The mean cost of hospital staywas $7,100. After discharge, there was a considerably highreadmission rate. During the next 6 months, almost half thepatients were readmitted at least once. Some of the patientswere readmitted several times. At 6 months, only 26% wereboth alive and well and able to report a good, very good, orexcellent quality of life.Eller J, Ede A, Schaberg T, et al. Infective exacerbationsof chronic bronchitis: relation between bacteriologicetiology and lung function. Chest 1998; 113:1542–1548This important paper looked at colonization of the airwaysof patients with severe COPD. It determined that there wasa correlation between deterioration of lung function (FEV 1)and the bacteria isolated from patients with infected exacerbationsof COPD. During an acute exacerbation, Enterobacteriaceaeand Pseudomonas sp were the predominant bacteriain patients with an FEV 1 35% of the predicted value.Fabbri LM, Romagnoli M, Corbetta L, et al. Differencesin airway inflammation in patients with fixed airflowobstruction due to asthma or chronic obstructive pulmonarydisease. Am J Respir Crit Care Med 2003;167:418–424This article explores the differences in airway inflammationin a cohort of older patients (age 65) with fixed airflowobstruction caused by asthma vs COPD. Subjects withasthma had significantly more eosinophils in peripheralblood, sputum, BAL, and airway mucosa. There were alsogreater CD4 /CD8 ratios of T cells in the airway mucosaand a thicker basement membrane in the asthmatic subjects.Asthmatics had greater diffusing capacities, residual volumes,and exhaled nitric oxide levels. High-resolution CTscan also differentiates the two conditions because patientswith COPD had a greater emphysema score. The authorsbelieve that older subjects with fixed airway obstructionshould be differentiated into those with a history of asthmaand those with distinct features of COPD. Although thesedistinctions are important from a research point of view, thedifferentiation may, at times, be impractical because of thehigh expense of the testing, lack of availability of these tests,and finally the fact that the tests are not always discriminatorybetween the two diseases.Ferguson GT, Funck-Brentano C, Fischer T, et al. Cardiovascularsafety of salmeterol in COPD. Chest 2003;123:1817–1824This study evaluated the cardiovascular safety of salmeterolfrom a pooled analysis of safety data from 17 studies. Theincidence of adverse cardiovascular events was no differentfrom placebo. The incidence increased with age and concurrentcardiovascular condition. There were no episodes of sustainedventricular tachycardia.Fletcher C, Peto R. The natural history of chronic airflowobstruction. BMJ 1977; 1:1645–1648This is a classic article. It is a study of London working menshowing that the FEV 1gradually decrease over the course ofa lifetime in all nonsmokers. The same occurs in most smokers.Nonsmokers and most smokers never develop clinicallysignificant airflow obstruction. In susceptible smoking people,176 COPD (Braman)ACC-PULM-09-0302-011.indd 1767/11/09 1:59:18 PM


irreversible obstructive changes occur. These changes may leadto a rapid rate of loss of FEV 1and high rates of morbidity andmortality. If the susceptible group stopped smoking, their rateof decrease returned to the normal decrease seen with aging.Gan WQ, Man SFP, Senthilselvan A, et al. Associationbetween chronic obstructive pulmonary diseaseand systemic inflammation: a systematic review and ameta-analysis. Thorax 2004; 59:574–580This review examines the association between COPD and systemicinflammation. Markers of systemic inflammation includingC-reactive protein, fibrinogen levels, circulating leukocytes,and TNF- levels are elevated in patients with COPD.Garcia-Aymerich J, Monso E, Marrades RM, et al.Risk factors for hospitalization for a chronic obstructivepulmonary disease exacerbation. Am J Respir CritCare Med 2001; 164:1002–1007This group measured the risk factors for exacerbations ofCOPD by using a case-control approach. Multivariate logisticregression showed that low FEV 1, underprescription ofoxygen therapy, and current smoking were all factors thatwould predict frequent exacerbation rates.Hanania NA, Darken P, Horstman D, et al. The efficacyand safety of fluticasone propionate (250 g)/salmeterol(50 g) combined in the diskus inhaler for thetreatment of COPD. Chest 2003; 124:834–843These investigators show that treatment with fluticasoneand salmeterol delivered by the diskus device and given twicedaily improved lung function when compared with either ofthese two components alone. Similar studies have also shownimprovements in quality of life indices and dyspnea score.Hogg JC, Chu F, Utokaparch S, et al. The nature ofsmall-airway obstruction in chronic obstructive pulmonarydisease. N Engl J Med 2004; 350:2645–2653Tissue from surgically resected lung specimens frompatients with COPD was evaluated for signs of inflammation.Patients were categorized into GOLD stages ofCOPD. The progression of COPD from GOLD stage 0 toGOLD stage 4 was strongly associated with an increase inthe volume of the tissue in the wall and the accumulationof inflammatory mucus exudates in the lumen of the smallairways. As COPD severity progressed, the percentage ofthe airways that contained polymorphonuclear neutrophils,macrophages, CD4 cells, CD8 cells, B cells, and lymphoidaggregate increased.Hogg JC, Senior RM. Chronic obstructive pulmonarydisease 2: pathology and biochemistry of emphysema.Thorax 2002; 57:830–834This article offers an excellent review of the pathogenesis ofemphysema and particularly factors responsible for alveolarseptal destruction in emphysema.Hogg JC. Pathophysiology of airflow limitation inchronic obstructive pulmonary disease. Lancet 2004;364:709–721This review article introduces the topic of innate and adaptiveimmune responses to inhaled irritants such as cigarettesmoke. There is an excellent discussion of the pathology ofCOPD with specific reference to the changes of emphysemaand chronic bronchitis. This is an excellent tutorial on thepathophysiology of COPD.Ionescu AA, Schoon E. Osteoporosis in chronic obstructivepulmonary disease. Eur Respir J Suppl. 2003;46:64s–75sOsteoporosis is one of the systemic effects associated withCOPD. Potential risk factors of osteoporosis may be attributableto lifestyle, genetics, treatment with corticosteroids,endocrine abnormalities, or the impairment of the bodycomposition and peripheral skeletal muscles. Evidence forthe possible contribution of such factors is reviewed in thisarticle.Johnson M, Rennard S. Alternative mechanisms forlong-acting (beta)2-adrenergic agonists in COPD.Chest 2001; 120:258–270This is a review of the nonbronchodilating effects of longactingagonists. These agents have important effects on theadherence of bacteria to the epithelium, capillary leak, mucusclearance, and inflammatory mediator release. It offers a reasonfor use of long-acting beta agonists in addition to theirbronchodilating properties.Jones AP, Rowe BH. Bronchopulmonary hygiene physicaltherapy for chronic obstructive pulmonary diseaseand bronchiectasis. Cochrane Database Syst Rev 2002;(2):CD000045This review showed that there is not enough evidence to supportor refute the use of bronchial hygiene physical therapyin people with COPD. The evidence-based report on acuteexacerbation of COPD presented in Chest in 2001 clearlydemonstrated that chest physiotherapy has no role in theacute exacerbation of COPD.Kanner RE, Anthonisen NR, Connett JE. Lower respiratoryillnesses promote FEV 1decline in current smokersbut not ex-smokers with mild chronic obstructivepulmonary disease. Am J Respir Crit Care Med 2001;164:358–364This important work is from the Lung Health Study showingthat patients with lower-respiratory tract infections whoare continuous smokers had greater decreases of lung functionthan the smokers who had no lower-respiratory tractinfections.Keenan SP, Sinuff T, Cook DJ, et al. Which patients withacute exacerbation of chronic obstructive pulmonary<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 177ACC-PULM-09-0302-011.indd 1777/11/09 1:59:18 PM


disease benefit from noninvasive positive-pressureventilation? A systematic review of the literature. AnnIntern Med 2003; 138:861–870This study examined selected randomized control trialsexamining the effects of noninvasive NPPV in patients withan acute exacerbation of COPD. The addition of this modalityto standard care was shown in this systematic review todecrease the rate of endotracheal intubation by 28%, lengthof hospital stay 4.5 days, and reduce mortality rate by 10%.The subgroup analysis shows that the beneficial effectsoccurred only in patients with severe exacerbations (pH 7.30) and not in those with milder exacerbation).Kierszniewska-Stepien D, Pietras T, Gorski P, et al.Serum vascular endothelial growth factor and itsreceptor level in patients with chronic obstructive pulmonarydisease. Eur Cytokine Netw 2006; 17:75–79The authors investigated a cytokine VEGF known to beinvolved in angiogenesis in the serum of 20 patients withmild COPD and 10 patients with very severe COPD. Theyfound that the concentration of VEGF and its soluble receptor(sVEGF R1) in the serum of patients with mild COPD wassignificantly greater (665.31 ± 102.20 pg/mL) in comparisonto the control group (318.94 ± 51.40 pg/mL; p 0.05), andthere was a strong negative correlation with FEV1 (r –0.859;p 0.001). These results suggest that VEGF and sVEGF R1receptor are involved in the development of abnormal pulmonaryvascular remodeling in patients with COPD.Kolodziej MA, Jensen L, Rowe B, et al. Systematicreview of noninvasive positive pressure ventilation insevere stable COPD. Eur Respir J 2007; 30:293–306The authors present a systematic review of the effectivenessof bilevel NPPV in the management of chronic respiratoryfailure attributable to severe stable COPD. Randomizedcontrolled trials did not find improved gas exchange; lunghyperinflation and diaphragmatic work of breathing werereduced in a nonrandomized subset of patients. There is littleevidence to support the use of NPPV in the management ofstable hypercarbic patients with COPD.Lacasse Y, Goldstein R, Lasserson TJ, et al. <strong>Pulmonary</strong>rehabilitation for chronic obstructive pulmonary disease.Cochrane Database Syst Rev 2006; (4):CD003793This evidence-based review concludes that pulmonary rehabilitationrelieves dyspnea and fatigue, improves emotionalfunction, and enhances patients’ sense of control over theircondition. These improvements are moderately large andclinically significant.Leuchte HH, Baumgartner RA, El Nounou M, et al.Brain natriuretic peptide is a prognostic parameter inchronic lung disease. Am J Respir Crit Care Med 2006;173:744–750Natruretic peptides are the major hormones of the natrureticpeptide system that is highly activated in patients with bothleft and right heart failure. This study evaluated circulatingBNP levels as a parameter for the presence and severityof pulmonary hypertension in patients with chronic lungdisease. Patients with significant pulmonary hypertensionhad elevated BNP levels and a lower 6-min walk distance.The authors showed that an elevated BNP is a risk factorfor death that is independent of lung functional impairmentor hypoxemia in patients with COPD. Serum BNP can beused as a prognostic marker as well as a screening tool forsignificant pulmonary hypertension in patients with chronicobstructive lung disease.Liesching T, Kwok H, Hill NS. Acute applications ofnoninvasive positive pressure ventilation. Chest 2003;124:699–713This is an excellent review of the use of NPPV to treat theacute exacerbation of COPD.Mallia P, Johnston SL. How viral infections cause exacerbationof airway diseases. Chest 2006; 130:1203–1210During the last decade, studies have established that respiratoryviruses are a major cause of exacerbations of asthma andCOPD and considerably create an impact on rates of morbidity,mortality, and health-care costs. The most prevalentviruses detected during exacerbations are the rhinoviruses.This article discusses the mechanisms of virus-induced exacerbationsof airway diseases. Exacerbations are associatedwith increased airway inflammation in patients with bothasthma and COPD. The authors conclude that identifyingthe key inflammatory mediators involved in exacerbationsholds the promise of developing diagnostic and prognosticmarkers of exacerbation. In addition, such studies can identifynew therapeutic targets for the development of noveldrugs for the prevention and treatment of exacerbations.Mannino DM, Gagnon RC, Petty TL, et al. Obstructivelung disease and low lung function in adults inthe United States: data from the National Health andNutrition Examination Survey, 1988–1994. Arch InternMed 2000; 160:1683–1689This NHANES III data suggests that 12.5% of current smokersand 9.4% of former smokers had evidence of obstructivelung disease. Sixty-three percent of patients with low lungfunction had no previous or current reported diagnosis ofany obstructive lung disease.McEvoy CE, Ensrud KE, Bender E, et al. Associationbetween corticosteroid use and vertebral fractures inolder men with chronic obstructive pulmonary disease.Am J Respir Crit Care Med 1998; 157:704–709This is an important reminder that vertebral fractures arecommon in older men with COPD, and the likelihood of178 COPD (Braman)ACC-PULM-09-0302-011.indd 1787/11/09 1:59:18 PM


these fractures is greatest in men who continuously use systemicsteroids.McKay SE, Howie CA, Thomson AH, et al. Value oftheophylline treatment in patients handicapped bychronic obstructive lung disease. Thorax 1993; 48:227–232This older article reminds us of the advantages of using theophyllinefor COPD, including improvements in peak flow,trapped gas volumes, vital capacity, distance walk, breathlessnessin everyday activities, and fatigue.Meyers BF, Patterson GA. Chronic obstructive pulmonarydisease 10: bullectomy, lung volume reductionsurgery, and transplantation for patients with chronicobstructive pulmonary disease. Thorax 2003; 58:634–638This is an excellent review of current surgical procedures forCOPD, giving the indications for bullectomy and transplantation.The American view of lung volume reduction surgeryis reported in other articles related to the NETT trial.National Emphysema Treatment Trial ResearchGroup. Patients at high risk of death after lung volumereductionsurgery. N Engl J Med 2001; 345:1075–1083This is an early report from the NETT Research Groupwarning that lung volume-reduction surgery in patientswho have a very low FEV 1, homogenous emphysema, or avery low carbon monoxide diffusing capacity are at veryhigh risk of death after surgery.National Emphysema Treatment Trial Research Group.A randomized trial comparing lung-volume-reductionsurgery with medical therapy for severe emphysema.N Engl J Med 2003; 348:2059–2073These are the long-awaited results of the NETT. It identifiedthose patients that were most likely to benefit from lungvolume-reduction surgery. Patients with upper-lobe emphysemaand very poor exercise tolerance are those who had thebest outcomes. After surgery, rate of mortality was lower,exercise capacity higher, and overall quality of life improvedcompared withthe medically treated group.NHLBI/WHO workshop report. Global strategyfor the diagnosis, management, and prevention ofchronic obstructive pulmonary disease. Available at:http://www.goldcopd.com. Accessed May 7, 2009This is the World Health Organization/NHLBI internalstandard guideline for the prevention, recognition, andtreatment of COPD.Niewoehner DE, Erbland ML, Deupree RH, et al. Effectof systemic glucocorticoids on exacerbations of chronicobstructive pulmonary disease. N Engl J Med 1999;340:1941–1947This article received a great deal of press. It showed thattreatment with systemic corticosteroids results in moder ateimprovement in clinical outcomes among patients hospitalizedfor exacerbations of COPD. The maximum benefit ofcorticosteroids, however, was seen only in the first 2 weeksafter hospitalization. Hyperglycemia was a common problem.Nonoyama ML, Brooks D, Guyatt GH, et al. Effect ofoxygen on health quality of life in patients with chronicobstructive pulmonary disease. Am J Respir Crit CareMed 2007; 176:343–349Long-term use of supplemental oxygen has been shown tobe beneficial in patients with COPD and resting hypoxemia,and it is now standard of care. Short-term ambulatory oxygenfor COPD patients who meet the hypoxemia criteria onlyduring exercise has offered some improvement in exerciseperformance, but it is not clear whether the long-term useof oxygen in such patients would be clinically beneficial andworth the expense. Previous randomized controlled trials toanswer this question have produced mixed results. The currentstudy was performed as a series of individual randomizedcontrolled trials (n-of-1) to measure the effect of oxygenin patients with COPD who do not meet the criteria of restinghypoxemia. The use of oxygen improved patients’ 5-minwalk test results (427 vs 412 steps, p 0.04), but quality oflife questionnaires did not show any difference between theoxygen and placebo. However, 2 of the 27 patients showedconsistent improvement of dyspnea measured on the ChronicRespiratory Questionnaire. This study does not support thelong-term application of oxygen to patients with exerciseinducedhypoxemia and normal resting oxygen levels. Thecosts of such therapy are great to the health-care system,and the therapy results in an imposition and discomfort tothe patient. These are compelling reasons why this practiceshould be discouraged for most patients.O’Donnell DE, Hernandez P, Aaron S, et al. CanadianThoracic Society COPD guidelines: summary ofhighlights for family doctors. Can Respir J 2003; 10:183–185This is a review of the Canadian guidelines for COPD.O’Donnell DE, Revill SM, Webb KA. Dynamic hyperinflationand exercise intolerance in chronic obstructivepulmonary disease. Am J Respir Crit Care Med 2001;164:770–777This article looked at the role of dynamic hyperinflationand exercise limitation in COPD. During exercise, 80%of the patients showed significant dynamic hyperinflationabove resting values. The extent of dynamic hyperinflationcorrelated best with the resting inspiratory capacity. Becausedynamic hyperinflation curtails the tidal volume response toexercise, it was thought that this was an important factorin exercise limitation in response to increasing metabolicdemands.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 179ACC-PULM-09-0302-011.indd 1797/11/09 1:59:19 PM


Patel IS, Seemungal TAR, Wilks M, et al. Relationshipbetween bacterial colonisation and the frequency,character, and severity of COPD exacerbations. Thorax2002; 57:759–764This study shows a relationship between lower bacterial colonizationand exacerbation frequency in COPD.Plant PK, Owen JL, Elliott MW. Early use of non-invasiveventilation for acute exacerbations of chronicobstructive pulmonary disease on general respiratorywards: a multicentre randomised controlled trial. Lancet2000; 355:1931–1935This large trial of BiPAP for the treatment of the acute exacerbationof COPD was performed on a general respiratoryward. Patients were mild-to-moderately acidotic (pH 7.25to 7.25) and were received standard therapy or the additionof NPPV. The use of NPPV significantly reduced the needfor intubation (27% vs 15%) and inpatient rate of mortality(20% vs 10%). NPPV led to more rapid improvement of pHduring the first hour and a greater decrease in the respiratoryrate and the duration of breathlessness.Puhan MA, Vollenweider D, Latshang T, et al. Exacerbationsof chronic obstructive pulmonary disease:when are antibiotics indicated? A systematic review.Respir Res 2007; 8:30This review discusses the unresolved debate about adequateprescription of antibiotics for patients suffering from exacerbationsof COPD. The study analyzed randomized controlledtrials investigating the clinical benefit of antibioticsfor COPD exacerbation. Antibiotics did not reduce treatmentfailures in outpatients with mild to moderate exacerbations(pooled OR 1.09, 95% CI 0.75 to 1.59, I2 18%).Inpatients with severe exacerbations had a substantialbenefit on treatment failure rates (pooled OR of 0.25,95% CI 0.16 to 0.39, I2 0%; number-needed to treat of4, 95% CI 3 to 5) and on mortality (pooled OR of 0.20,95% CI 0.06 to 0.62, I2 0%; number needed to treat of14, 95% CI 12 to 30). The authors concluded that antibioticseffectively reduce treatment failure and mortality ratesin COPD patients with severe exacerbations. For patientswith mild-to-moderate exacerbations, antibiotics may not begenerally indicated and the authors felt that further researchis needed to guide antibiotic prescription in these patients.Qui Y, Zhu J, Bandi V, et al. Biopsy neutrophilia, neutrophilchemokine and receptor gene expression insevere exacerbations of chronic obstructive pulmonarydisease. Am J Respir Crit Care Med 2003; 168:968–975In this study, bronchial biopsies were performed on patientsintubated for an acute exacerbation of COPD and those withstable COPD. The acute exacerbation is associated with anincrease in IL-8 and marked airway neutrophilia.Rana JS, Mittleman MA, Sheikh J, et al. Chronic obstructivepulmonary disease, asthma, and risk of type 2 diabetesin women. Diabetes Care 2004; 27:2478–2484Inflammation plays a key role in COPD and asthma.Increasing evidence points toward a role of inflammation inthe pathogenesis of type 2 diabetes The authors investigatedthe incidence of type 2 diabetes in patients with COPD byusing data from the The Nurses’ Health Study. The risk oftype 2 diabetes was significantly greater for patients withCOPD than those without (multivariate relative risk 1.8,95% confidence interval [95% CI] 1.1 to 2.8). By contrast,the risk of type 2 diabetes among asthmatic patients was notincreased (1.0, 0.8 to 1.2).Reed CE. The natural history of asthma in adults: theproblem of irreversibility. J Allergy Clin Immunol 1999;103:539–547This is an important article because it documents how manyelderly asthmatic patients develop irreversible obstructivechanges on their spirogram despite apparently adequatetherapy. It suggests that chronic asthma does lead to chronicnonreversible airflow obstruction (COPD).Rennard SI. Clinical approach to patients with chronicobstructive pulmonary disease and cardiovascular disease.Proc Am Thorac Soc 2005; 2:94–100This review discusses the interactions between cardiac andpulmonary disease.Ries AL, Bauldoff GL, Carlin BW, et al. <strong>Pulmonary</strong>rehabilitation: joint <strong>ACCP</strong>/AACVPR evidence-basedclinical practice guidelines [review]. Chest 2007;(5 Suppl)131:4S–42SThis document provides a systematic, evidence-based reviewof the pulmonary rehabilitation literature that updates the1997 guidelines published by the American College of ChestPhysicians and the American Association of Cardiovascularand <strong>Pulmonary</strong> Rehabilitation. The new evidence strengthensthe previous recommendations supporting the benefits oflower- and upper-extremity exercise training and improvementsin dyspnea and health-related quality-of-life outcomesof pulmonary rehabilitation. Additional evidence supportsimprovements in health-care utilization and psychosocialoutcomes. There are few additional data about survival.Some new evidence indicates that longer term rehabilitation,maintenance strategies following rehabilitation, and theincorporation of education and strength training in pulmonaryrehabilitation are beneficial. Current evidence does notsupport the routine use of inspiratory muscle training, anabolicdrugs, or nutritional supplementation in pulmonaryrehabilitation. Evidence does support the use of supplementaloxygen therapy for patients with severe hypoxemia at restor with exercise. Noninvasive ventilation may be helpful for180 COPD (Braman)ACC-PULM-09-0302-011.indd 1807/11/09 1:59:19 PM


selected patients with advanced COPD. Finally, pulmonaryrehabilitation appears to benefit patients with chronic lungdiseases other than COPD.Roede BM, Bresser P, Bindels PJ, et al Antibiotic treatmentis associated with reduced risk of a subsequentexacerbation in obstructive lung disease: an historicalpopulation based cohort study. Thorax 2008; 63:968–973The risk of a subsequent exacerbation after treatment of anexacerbation with oral corticosteroids without (OS) or with(OSA) antibiotics was evaluated in a historical populationbased cohort study comprising patients using maintenancemedication for obstructive lung disease. Treatment withantibiotics in addition to oral corticosteroids was associatedwith a longer time to the next exacerbation, and a decreasedrisk of developing a new exacerbation.Rutschmann OT, Cornuz J, Poletti PA, et al. Shouldpulmonary embolism be suspected in exacerbation ofchronic obstructive pulmonary disease. Thorax 2007;62:121–125The diagnosis of acute PE in patients with COPD is oftendifficult to make especially during an acute exacerbation ofCOPD when symptoms of the two conditions may be indistinguishable.The prevalence of PE in postmortem seriesof patients with COPD is extremely high, ranging from28 to 51%, and one recent study (Ann Intern Med 2006;144:390396) showed a 25% prevalence of PE in a groupof patients admitted to the hospital for a severe exacerbationof COPD, may not be generalized as patients in this studywere excluded if they had a potential infection. This studyby Rutschmann et al included consecutive patients admittedto the emergency department with a COPD exacerbation.<strong>Pulmonary</strong> embolism was excluded on the basis of anormal ( 500 µg/mL) d-dimer test and negative diagnosticimaging studies. 6.2% of the 48 patients who had a clinicalsuspicion of pulmonary embolism were discovered to havePE and in only 1.3% of the remaining patients who had noclinical suspicion. This study showed that the prevalence ofsuspected pulmonary embolism in patients presenting witha COPD exacerbation is very low and routine investigationfor PE in this group is not warranted.Saetta M, Turato G, Maestrelli P, et al. Cellular andstructural bases of chronic obstructive pulmonary disease.Am J Respir Crit Care Med 2001; 163:1304–1309This review article discusses the two pathological hallmarksof COPD: (1) inflammation of the peripheral airways and(2) destruction of the lung parenchyma (emphysema). In“normal” smokers, studies have shown that T lymphocytesand macrophages infiltrate the airway wall of the central airways,and that neutrophils in the airway wall are increasedin the airway lumen. In the peripheral airways, the cellularinfiltrate in the airway consists of mononuclear cells andmacrophages. The peripheral airways are the major site ofincreased airway resistance. In “normal” smokers there isno sign of parenchymal destruction. In smokers with establishedCOPD, there is a further increase in macrophages andT lymphocytes in the airway walls. Peripheral airways showmore inflammatory changes of all cells, including neutrophilsin severe disease. There is evidence of centriacinar andpanacinar emphysema and some fibrosis in the lung parenchyma.In severe COPD, airflow limitation progressivelyworsens, and neutrophils in the bronchial wall increase.Their increase is correlated with the degree of airflow limitation.Data indicate that when emphysema is severe, lossof elastic recoil assumes overwhelming importance as amechanism of airflow limitation, thus masking the effects ofperipheral airway abnormalities.Seemungal T, Harper-Owen R, Bhowmik A, et al.Respiratory viruses, symptoms, and inflammatorymarkers in acute exacerbations of stable acute exacerbationsof chronic obstructive pulmonary disease. AmJ Respir Crit Care Med 2001; 164:1618–1623This study prospectively investigated the association of theexacerbation of COPD in viral infection. Polymerase chainreaction methodology was used to detect viral infection. Theauthors found that 64% of exacerbations were associatedwith a cold occurring up to 18 days before the exacerbation.The authors found that respiratory virus infections are associatedwith more severe and frequent exacerbations and maycause chronic infection in COPD.Seemungal TA, Wilkinson TM, Hurst JR, et al. Longtermerythromycin therapy is associated with decreasedchronic obstructive pulmonary disease exacerbations.Am J Respir Crit Care Med 2008; 178:1139–1147Long-term erythromycin therapy is associated withdecreased exacerbations of COPD. There has been muchinterest in the use of prophylactic antibiotics for the preventionof exacerbations of COPD. The results of a numberof older studies were disappointing, but this study ismore optimistic. Macrolides have airway antiinflammatoryactions as well as antimicrobial properties, and erythromycinwas chosen to determine whether regular therapy withmacrolides reduces exacerbation frequency. Erythromycinwas administered at 250 mg bid to patients with COPD overthe course of 12 months, with the primary outcome variablebeing the number of moderate and/or severe exacerbations(treated with systemic steroids, treated with antibiotics, orhospitalized). The rate ratio for exacerbations for the macrolide-treatedpatients compared with placebo-treated patientswas 0.648 (95% CI 0.489 to 0.859; p 0.003), and these<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 181ACC-PULM-09-0302-011.indd 1817/11/09 1:59:19 PM


patients had shorter-duration exacerbations compared withplacebo-treated patients. There were no differences betweenthe macrolide and placebo arms in terms of stable FEV 1,sputum IL-6, IL-8, myeloperoxidase, bacterial flora, serumC-reactive protein, or serum IL-6 or in changes in theseparameters from baseline. The authors concluded that macrolidetherapy was associated with a significant reduction inexacerbations compared with placebo.Seemungal TAR, Donaldson GC, Bhowmik A, et al.Time course and recovery of exacerbations in patientswith chronic obstructive pulmonary disease. Am JRespir Crit Care Med 2000; 161:1608–1618This group of authors studied a cohort 101 patients withmoderately severe COPD (mean FEV 142%) during a 2.5-year period. Patients recorded daily morning peak flow ratesand changes in respiratory symptoms. Exacerbation rateswere followed. Before the onset of exacerbation there wasdeterioration in symptoms of dyspnea, sore throat, cough,and symptoms of the common cold but not a worsening oflung function. On the day of the attack, peak flow rates diddecrease by a mean of 8.6 L/min. The FEV 1decreased by24 mL. Recovery of peak flow rates to baseline values wascomplete in only 72% of exacerbations at day 35. In 7% ofexacerbations peak flow rate had not returned to baseline evenat 3 months. The authors conclude that symptom changesduring an exacerbation of COPD do not closely reflect thoseof lung function study, but their increase may predict anexacerbation. Recovery is incomplete in a significant proportionof COPD exacerbations.Seersholm N, Kok-Jensen A. Clinical features andprognosis of lifetime non-smokers with severe 1-antitrypsindeficiency. Thorax 1998; 53:265–268This study documented that it is rare for homozygote ZZAAT--deficient patients to develop COPD if they do notsmoke.Sethi S, Evans N, Grant BJB, et al. New strains of bacteriaand exacerbations of chronic obstructive pulmonarydisease. N Engl J Med 2002; 347:465–471These investigators have shown an association between anexacerbation of COPD and the isolation of a new strain ofbacterial pathogen. This supports a causative role of bacteriain exacerbations of COPD.Sethi S. Infectious etiology of acute exacerbations ofchronic bronchitis. Chest 2000; 117:3808–3858This review discusses the role of infection in exacerbationsof COPD. Respiratory viruses are associated with approximately30% of exacerbations of COPD. A typical bacterium,mostly Chlamydia and Mycoplasma sp, have been implicatedin 10% of cases. There is emerging evidence thatsupports the role of bacteria in the exacerbation of COPD.Concomitant infections by more than one infectious pathogenoccur in 10 to 20% of patients.Sevenoaks MJ, Stockley RA. Chronic obstructive pulmonarydisease, inflammation and co-morbidity: acommon inflammatory phenotype? Respir Res 2006;7:70–78A link has been identified between COPD and other systemicdiseases such as cardiovascular disease, diabetes, and osteoporosis.These authors discuss how proinflammatory cytokines,in particular TNF-, may be the driving force behindthe disease process. The roles of inflammation and these proinflammatorycytokines may extend beyond the lungs andplay a part in the systemic effects of the disease and associatedco-morbidities. This article describes the mechanismsinvolved and proposes a common inflammatory TNF-a phenotypethat may, in part, account for the associations.Sin DD, Anthonisen NR, Soriano JB, et al. Mortalityin COPD: role of comorbidities. Eur Respir J 2006;28:1245–1257This article reviews the role of comorbidities in the rate ofCOPD mortality, the putative underlying pathogenic linkbetween chronic obstructive pulmonary disease and comorbidconditions (ie, inflammation), and the tools used to predictthe rate of COPD mortality.Sin DD, Man SF, Marciniuk DD, et al. The effects offluticasone with or without salmeterol on systemicbiomarkers of inflammation in chronic obstructivepulmonary disease. Am J Respir Crit Care Med 2008;177:1207:1207–1214The cause of systemic inflammation (as evidenced by biomarkers)is not known, but it has been suggested that it isthe result of a “spillover” from inflammation in the lungs.Whether treating lung inflammation can reduce systemicinflammation has not been established. The aim of this randomized,double-blind, multicenter study was to determinewhether the inhaled corticosteroid (fluticasone) with or withouta long-acting beta agonist (salmeterol) could reduce bloodinflammatory markers such as C-reactive protein, the primaryoutcome, and also IL-6 and SP-D. In this 4-week trial (meanFEV 1 48% predicted), neither fluticasone nor the combinationtherapy caused a significant reduction in C-reactiveprotein or IL-6, but it did significantly reduce SP-D levelscompared with placebo (p 0.002). Health status scores andlung function tests during the trial improved significantlyand were related to circulating SP-D levels. The authors concludedthat these treatments improved lung-specific but notgeneralized biomarkers of systemic inflammation.Sin DD, Man SFP. Why are patients with chronicobstructive pulmonary disease at increased risk of cardiovasculardiseases? The potential role of systemic182 COPD (Braman)ACC-PULM-09-0302-011.indd 1827/11/09 1:59:19 PM


inflammation in chronic obstructive pulmonary disease.Circulation 2003; 107:1514–1519The NHANES III database was used by these investigatorsto determine whether C-reactive protein and other systemicinflammatory markers are present in patients with COPDand are associated with cardiac injury. Participants withsevere airflow obstruction had a significant increase inmarkers of inflammation, including circulating leukocyte,platelet, and fibrinogen levels, as well as levels of C-reactiveprotein. Moderate airflow obstruction had similar but smallerincreases. Both moderate and severe airflow obstruction wereassociated with an increased occurrence of ischemic changeson the electrocardiograms of participants.Sin DD, McAlister FA, Man SF, et al. Contemporarymanagement of chronic obstructive pulmonary disease:scientific review. JAMA 2003; 290:2301–2312This is a careful review of the current evidence for the managementof outpatients with COPD.Singh SMD, Amin A, Yoon K, et al Long-term use ofinhaled corticosteroids and the risk of pneumonia inchronic obstructive pulmonary disease. Arch InternMed 2009; 169:219–229Recent studies have suggested a possible associationbetween pneumonia and the use of inhaled corticosteroids.The authors performed a systematic search to ascertain therisk of pneumonia with long-term inhaled corticosteroiduse among patients with COPD. They included randomizedcontrolled trials of any inhaled corticosteroid withat least 24 weeks of follow-up and reporting of pneumoniaas an adverse event. Eighteen randomized controlled trialsshowed that inhaled corticosteroids were associated with asignificantly increased risk of any pneumonia (relative risk[RR] 1.60; 95% CI 1.33 to 1.92, p .001m I 2 16%) andserious pneumonia (1.71; 1.46 to 1.99, p .001, I 2 0%)but without a significantly increased risk of pneumoniarelatedmortality (1.27, 0.80 to 2.03, p .31, I 2 0%) oroverall mortality (0.96, 0.861.08, p 0.51],I 2 0%). Theauthors concluded that patients with COPD and inhaledcorticosteroid use for at least 24 weeks have a significantlyincreased risk of serious pneumonia without a significantlyincreased risk of death.Sobieraj DM, White CM, Coleman CI. Benefits andrisks of adjunctive inhaled corticosteroids in chronicobstructive pulmonary disease: a meta-analysis. ClinTher 2008; 30:1416–1425The authors of this article conducted a metaanalysis toexamine the benefits and risks associated with adjunctiveinhaled corticosteroids treatment in patients with severeor very severe COPD. The authors found that the additionof an inhaled corticosteroid to a long-acting -agonistwas associated with a reduced risk for exacerbations butan increased risk for pneumonia and oral candidiasis comparedwith long-acting bronchodilator monotherapy in thismetaanalysis of 9 randomized controlled trials. Althoughmeasured patient-perceived health and well-being increasedto a statistically significant level, this did not translate intoa clinically meaningful level for all patients with combinationtreatment. A lower risk of study withdrawal wasobserved in patients administered adjuvant inhaled corticosteroids.Soriano JB, Visick GT, Muellerova H, et al. Patterns ofcomorbidities in newly diagnosed COPD and asthmain primary care. Chest 2005; 128:2099–2107The authors quantified baseline rates of comorbiditiesin COPD patients and compared them with the risks inthe general population. The authors showed that COPDpatients are at increased risk for glaucoma, angina, fractures,myocardial infarction, osteoporosis, and respiratoryinfection when compared with control patients withoutCOPD.Studer SM, Levy RD, McNeil K, et al. Lung transplantoutcomes: a review of survival, graft function, physiology,health-related quality of life and cost-effectiveness.Eur Respir J 2004; 24:674–685This article reports on the success of lung transplantationthat improved over time. There are better long-term survivalsand functional outcomes. Despite this success, there arenumerous problems and complications that may develop overthe life of a lung transplant recipient. Significant improvementfor the overall outcomes of lung transplantation willonly occur when better methods exist to prevent or effectivelytreat chronic rejection.Tashkin DP, Celli B, Senn S, et al; UPLIFT study investigators.A 4-year trial of tiotropium in chronic obstructivepulmonary disease. N Engl J Med 2008; 359:1543–1554This study was a randomized, double-blind trial comparing4 years of therapy with either tiotropium or placebo inpatients with COPD who were permitted to use all respiratorymedications except inhaled anticholinergic drugs.Therapy with tiotropium was associated with improvementsin lung function, quality of life, and exacerbations duringa 4-year period but did not significantly reduce the rate ofdecrease in FEV 1.The 1-Antitrypsin Deficiency Registry Study Group.Survival and FEV 1decline in individuals with severedeficiency of 1-antitrypsin. Am J Respir Crit Care Med1998; 158:49–59This is a report from the NHLBI Registry showing thatsubjects who received augmentation therapy had a lower<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 183ACC-PULM-09-0302-011.indd 1837/11/09 1:59:20 PM


ate of mortality. Those with an FEV 1between 35%and 49% predicted also showed a slowing of the FEV 1decline.van Gestel YR, Hoeks SE, Sin DD, et al. Effect of statintherapy on mortality in patients with peripheral arterialdisease and comparison of those with versuswithout associated chronic obstructive pulmonarydisease. Am J Cardiol 2008; 102:192–196The role of statin therapy in addition to standard treatmentfor COPD in a topic of intense interest, and the answer tothis question will likely be forthcoming in the near future.Statins are widely used therapies to control elevated bloodlipids and may show other benefits in patients with greaterlevels of inflammation, as measured by blood inflammatorymarkers such as CRP. Systemic inflammation isincreased in COPD and has been linked to a number ofcomorbidities such as peripheral muscle wasting, whichin turn plays a role in exercise intolerance. The purposeof this study was to determine whether the administrationof a statin, pravastatin, 40 mg per day, was effective inimproving exercise capacity in a group of patients withstable COPD and whether baseline levels or changes inCRP levels (measures as high sensitivity-CRP) predictedclinical outcomes. Previous studies have shown an inverserelationship between CRP levels and exercise capacity inpatients with COPD. This is the first randomized controlledtrial to show that treatment with statin therapy canimprove exercise tolerance in patients with COPD, regardlessof the cholesterol level attained. The improvement inexercise time and degree of breathlessness was observedin those patients who showed greater levels of inflammationat baseline, evidenced by high hs-CRP levels and alsoelevated IL-6 levels, and in those with a greater decreasein levels of these inflammatory markers after statin treatment.In this study, as in previous studies, the individualresponses of CRP to statin therapy were highly variable.The mechanism proposed for improved exercise tolerancein these COPD patients is that the statins reduced muscleinflammation and hence exercise capacity.Wedzicha JA, Calverley PMA, Seemungal TA, et al. Theprevention of chronic obstructive pulmonary diseaseexacerbations by salmeterol/fluticasone propionate ortiotropium bromide. Am J Respir Crit Care Med 2008;177:19–26Investigating New Standards for Prophylaxis in ReducingExacerbations is the first large-scale trial to comparethe clinical outcomes of two frequently used treatments forCOPD. The authors of this study made this comparisonduring a 2-year treatment period in a multicenter studypatients with COPD and an FEV 1 50% predicted andreversibility 10%. They found no difference between thetwo study groups in the primary efficacy end point, exacerbationrates that required treatment with oral corticosteroidsand/or antibiotics, or required hospitalization. Morepatients failed to complete the study with tiotropium (TIO);salmeterol/fluticasone propionate (SFC)treated patientshad a small statistically significant benefit with quality oflife scores, an unexpectedly lower death rate (3% vs 6%),and a greater rate of pneumonia. However, exacerbationsrequiring antibiotics occurred more frequently in patientstreated with SFC (SFC, 0.97 per year; TIO, 0.82 per year;p 0.028), but those requiring systemic corticosteroidswere less frequent than in the TIO-treated patients (SFC,0.69 per year; TIO, 0.85 per year; p 0.039). Mortality wassignificantly lower in the SFC treatment (p 0.032), withan estimated 52% reduction in the risk of on-therapy allcausemortality. However, because more patients who wereadministered TIO withdrew from the study, this differencemay have led to a healthy survivor effect. A clinical diagnosisof pneumonia was reported in 8% of patients administeredSFC and 4% of patients administered TIO, and thehazard ratio for time to reported pneumonia was 1.94 (95%CI 1.19 to 3.17, p 0.008) for SFC compared with TIO overthe 2 years.Wijkstra PJ, Lacasse Y, Guyatt GH, et al. A metaanalysisof nocturnal noninvasive positive pressureventilation in patients with stable COPD. Chest 2003;124:337–343A metaanalysis was conducted to determine whethernocturnal NPPV was helpful to patients with COPDand stable respiratory failure. The authors concludedthat this type of ventilatory support did not improvelung function, gas exchange, or sleep efficiency in suchpatients. Some patients do improve their 6-min walk testresults.Wijkstra PJ, Lacasse Y, Guyatt GH, et al. Nocturnalnon-invasive positive pressure ventilation for stablechronic obstructive pulmonary disease. CochraneDatabase Syst Rev 2002; (3):CD002878This evidence-based review by the Cochrane group concludesthat randomized controlled trials in stable patientswith COPD have not convincingly proven the benefitsof NPPV. There was no consistent clinically or statisticallysignificant effect on lung function, gas exchange,respiratory muscle strength, sleep efficiency, or exercisetolerance with this modality. Larger clinical trials wouldbe necessary in the future to approve its usefulnessconclusively.184 COPD (Braman)ACC-PULM-09-0302-011.indd 1847/11/09 1:59:20 PM


Zielinski J, Tobiasz M, Hawrytkiewicz I, et al.Effects of long-term oxygen therapy on pulmonaryhemodynamics in COPD patients: a 6-year prospectivestudy. Chest 1998; 113:65–70This important article looks at the effects of long-term oxygentherapy in patients with COPD and pulmonary hypertension.It followed patients over the course of 6 years.<strong>Pulmonary</strong> hypertension showed a small reduction afterthe first 2 years, followed by a return to initial values andsubsequent stabilization over 6 years. The article concludedthat long-term stabilization of pulmonary hypertensionoccurs despite progression of the airflow limitation androom air hypoxemia in these patients.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 185ACC-PULM-09-0302-011.indd 1857/11/09 1:59:20 PM


Notes186 COPD (Braman)ACC-PULM-09-0302-011.indd 1867/11/09 1:59:20 PM


<strong>Pulmonary</strong> Function TestingDarcy D. Marciniuk, MD, FCCPObjectives:• <strong>Review</strong> the indications, conduct, and interpretation ofpulmonary function testing• Highlight the limitation(s) of various testing methods,and emphasize the value of both laboratory and testingstandardization and quality assurance• <strong>Review</strong> characteristic pulmonary function test resultsassociated with common respiratory disordersKey words: interpretation; pulmonary function laboratory;pulmonary function testing; quality assurance; spirometryThe performance and interpretation of pulmonaryfunction testing (PFT) is a defining competencyof our specialty, and the PFT laboratory isa fundamental and indispensable resource for thepulmonologist. A thorough understanding of theindications, conduct, interpretation, and limitationsof testing is essential. In addition, becauseour colleagues and patients are dependent on andtrust the results derived from the laboratories wesupervise, pulmonologists must also be experts inissues related to quality assurance, standardization,and testing procedures in the PFT laboratory.Although these two roles are related, the independentimportance of each must be emphasized. Inthis regard, increasing attention is being focusedon mastery of the technical aspects of testing andthe laboratory, which will be reviewed in this chapter.A complete review of this topic is beyond theintent of this course and syllabus. Readers wantingmore complete information are asked to review thereferenced source documents listed at the end ofthis chapter.Clinical Indications for PFTThe indications for pulmonary function testingare varied, and they depend on the clinicalsetting and question(s) to be addressed. Gener allyaccepted clinical indications are listed in Table 1.Important Considerations for TestingThe PatientTesting is demanding, and results will be lessmeaningful in patients who are not physicallycapable of providing consistent, optimal effort. Itis recommended that testing not be performedwithin 1 month of an acute coronary syndrome ormyocardial infarction. 1 Similarly, patients withacute chest, abdominal, or facial pain, as well as inthose with confusion or significant dementia, arelikely to generate suboptimal test results.Patient weight (in street clothes and withoutshoes) and height (without shoes) should be accuratelymeasured and recorded. In patients withkyphoscoliosis or lower-limb amputations, armspan from fingertip to fingertip should be used asan estimate of height (regression equations areavailable). 2 Alternatively, knee height also can beused for handicapped individuals where arm spanis difficult to assess. 3Dentures are often best left in place duringtesting unless they are loose or poorly fitting. Itwould be prudent for patients to refrain fromsmoking or undergoing vigorous exercise for 1 h,eating a large meal for 2 h, or consuming alcoholfor 4 h before testing. 1Table 1. Clinical Indications for PFT• Evaluate symptoms, signs, or abnormal investigations• Assess and monitor the effect of disease/intervention/exposure on pulmonary function and respiratorysymptoms• Early detection of individuals at risk of pulmonarydisease or impairment• Assess preoperative risk• Objectively assess impairment• Understand severity and prognosis<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 187ACC-PULM-09-0302-012.indd 1877/10/09 10:16:37 PM


The Laboratory and EquipmentIt is essential that unintended factors do notcontaminate the results and subsequently lead tomisinterpretation regarding the patient, disease,or therapy. These factors are the most overlookedconsiderations in clinical PFT laboratories today,and a thorough understanding of these issues isespecially necessary for PFT laboratory directors.Although specific important issues will be discussedin sections to follow and a number ofoverall principles deserve highlight, the topic isreviewed in detail elsewhere. 4Measurements of temperature and barometricpressure must be actually recorded, and it is alsoimportant to ensure accuracy of the instrumentused to make those measurements. The time of dayshould be noted, and serial testing should ideallybe made at similar times of the day to minimizevariation.Although the order of testing is not necessarilyrigid, the order in a specific laboratory should bekept constant. Consideration should also be givento the effects of bronchodilator administration onlung volume determination, or factors influencingthe diffusing capacity (discussed in Table 7). Asuggested order for performing lung function testsis noted in Table 2.Laboratory safety is important. Issues relatingto building and general facilities; accident, fire, andevacuation procedures; compressed gas storageand use; electrical safety; and procedures andpractices for tending to patient urgencies andemergencies must be addressed. Appropriate proceduresmust be established, understood, andpracticed by all staff.Infection control measures are also necessaryfor the protection of patients and staff. Althoughthe risk of infection is small, the potential is real,and the consequences are serious. These risks canbe minimized by the following steps:Table 2. Suggested Order for Conducting Lung Function Tests*• Spirometry and flow-volume curves• Lung volumes• Bronchodilator administration• Diffusing capacity• Repeat spirometry and flow-volume curves*Adapted from Miller et al. 11. Proper hand-washing should be performed andbarrier devices such as gloves should be used.2. Reuseable mouthpieces, valves, mouthpieces,and manifolds must be appropriately disinfectedor sterilized.3. Disposable equipment, sensors, and devicesshould be discarded, never reused, even if disinfectedor sterilized.4. Sterilizing and disinfecting techniques shouldbe strictly adhered to and should be establishedin consultation with the manufacturer’s recommendationsand local or hospital infectioncontroldivisions.5. In patients with known or suspected transmissibleinfectious diseases, additional precautionsshould be undertaken. These would include:(a) the use of equipment solely reserved for usein this clinical setting;(b) testing patients at the end of the day toallow for complete spirometry disassemblyand disinfection; and(c) testing patients in their own rooms or inrooms with enhanced capabilities (ie, negativepressure ventilation, etc).The use of inline disposable filters is controversial.Although some significant differences betweenmeasurements with and without filters have beendemonstrated, the effects are not considered to beclinically significant, and they do not lead to misinterpretation.5,6 However, because the benefits offilters have not been clearly identified, their use isnot mandatory, particularly if all other precautionsare strictly followed. Overall, most laboratories useinline filters, perhaps to reassure patients and staffthat their safety and protection are a high priority.It deserves emphasis that the use of inline filtersshould not be viewed as a shortcut for appropriateinfection control, and their use does not eliminatethe need for regular cleaning and decontaminationof lung function equipment.Laboratory PersonnelAll staff must be appropriately trained tounderstand the fundamentals of testing, to befamiliar with signs and symptoms of commonrespiratory disease, and to properly execute allaspects of testing. Training requirements varyacross regions, and many available opportunities188 <strong>Pulmonary</strong> Function Testing (Marciniuk)ACC-PULM-09-0302-012.indd 1887/10/09 10:16:37 PM


Table 3. Evaluation and Feedback for <strong>Pulmonary</strong> FunctionLaboratory Technicians*• Information regarding acceptable maneuvers and nonreproducibletests• Specific corrective actions the technicians can undertake toimprove the quality and number of acceptable maneuvers• Positive feedback for excellence and good performance• Feedback regarding system setup and reporting results• Asking the technician to comment on current laboratoryprocedures and testing, specifically inquiring aboutopportunities for improvement.*Adapted from Miller et al. 1for training courses are available. A period of mentorshipin the laboratory under the experience ofseasoned technicians is a crucial one for new staff.Opportunities for ongoing professional developmentshould be availed to staff, ensuring that trainingand skills remain current, and upgraded whenappropriate. This development is particularlyimportant with alterations to equipment or changesin pulmonary function testing standards. Althoughmanufacturers frequently provide training for newequipment or significant upgrades, laboratoriesshould not solely rely on these methods as a guaranteethat technician skills are up-to-date.The value of continuous in-house feedback andevaluation of laboratory technicians has beenvalidated. 7 This feedback helps to promote the collectionof high-quality data, and it also works toensure that staff remains well motivated and enthusiastic.Appropriate content for regular and formalizedfeedback to technicians is noted in Table 3.Medical directors of PFT laboratories requireall of the aforementioned attributes and skills, butthey are also responsible for all aspects of qualityassurance, selection of reference values, and practicaloperational issues relating to the day-to-dayconduct of the laboratory.Reference ValuesThe interpretation of PFTs is based on comparisonsof data measured in an individual patientwith reference values derived from a representativepopulation of healthy subjects. Unlike manyphysiologic variables, normal values of pulmonaryfunction vary with age, height, sex, and race. Moreover,because the range of normal is considerable(for example 80 to 120% of the predicted value),significant changes in pulmonary function canoccur while values still remain within the normalrange. These factors serve to complicate the choiceof the most appropriate reference value regressionequations to use in the clinical laboratory.In general, a number of observations regardinglung function measurements in the population areknown, and include the following:1. Male patients have larger lung function thanfemale patients.2. Lung function values plateau when patientsare 20 to 35 years of age. 83. FEV 1decreases approximately 30 mL/yr.4. Vital capacity decreases, whereas the residualvolume increases with age, leaving total lungcapacity the same. The diffusing capacitydecreases with age.5. Taller individuals have larger lung volumesand greater maximal flow rates.6. African-American patients have spirometricvalues that are lower than white patients of thesame age, height, and sex. If predicted valuesfor a healthy population of the same ethnicbackground are not available, predicted valuesfor white patients should be corrected by 0.88for African-American, Asian, and East-Indianpatients. The FEV 1/FVC ratio should not becorrected for race.Throughout the years, various referencevalues have been both developed and recommended,which this has led to uneasiness and,frequently, inappropriate reference value selection.However, a new coordinated recommendation hasrecently been put forth for the Third NationalHealth and Nutrition Examination Survey referenceequations to be used for spirometry in theUnited States. 9 Specific recommendations for lungvolume and diffusing capacity reference values areless pointed (Table 4). 10 From a practical point ofview, it is apparent that the discussions and debateregarding reference equations in the pulmonaryfunction laboratory will not end soon. However,there is agreement that the reference equationchosen should reflect a similar age range, sex, andethnic background of patients in the laboratory,and that all spirometric data should use the samesource for reference values.Although reference values are crucial forthe correct first-time interpretation of pulmonary<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 189ACC-PULM-09-0302-012.indd 1897/10/09 10:16:38 PM


Table 4. Reference Values in the PFT Laboratory*General commentsPredicted values should be obtained from normal subjects with the same anthropometric and ethnic characteristics of thepatient being testedHeight and weight should be measured for each patient at each testingAll parameters should be taken from the same reference source when possibleExtrapolation beyond the size and age of the reference population should be avoided when possibleValues below the 5th percentile of the frequency distribution of values from the reference population should be consideredas below the lower limit of normal.SpirometryIn the United States, ethnically appropriate NHANES III reference equations published in 1999 for those patients 8 to80 years of age are recommended. 10Lung volumesNo specific set of reference equations is recommended, however a list of potentially suitable reference equations is available. 9Diffusing capacityNo specific set of reference equations is recommended, however a list of potentially suitable reference equations is available. 9*Adapted from Pelligrino et al. 9 NHANES III Third National Health and Nutrition Examination Survey.function test results, if serial testing is undertaken,comparison with the patient’s previous values ismore appropriate and meaningful.Specific values for “normalcy” are controversial,but in general, there is a move toward reportingthe normal range in terms of a lower limit ofnormal (LLN; spirometry, lung volumes, diffusingcapacity) and an upper limit of normal (lung volumesand diffusing capacity). The use of values such asthe 80% of predicted as the lower range of normalis useful in some instances, but it does result inmany false-positive and false-negative results. It isrecommended that the LLN be determined fromthe fifth percentile (ie, mean, 1.96 SD; two-tailedt test, p 0.05). In addition, categorizing results asborderline abnormal, if they encroach on the ULNor LLN, is an acceptable clinical concept. Althoughsoftware is available to readily supply these values,clinical comfort and acceptance with these recommendedchanges in reporting and interpretation isstill evolving. Overall guiding principles for theuse of reference values in the PFT laboratory arelisted in Table 4.SpirometrySpirometry is the most commonly performedpulmonary function test. It is used to measure theexhaled volume of air vs time. Volume or flow maybe directly measured, and in many instances, bothinhaled and exhaled maneuvers are undertaken.The basic variables of interest are the FVC andFEV 1, which both are expressed at body temperature,ambient pressure saturated with water vapor(BTPS).EquipmentBasic equipment requirements are a spirometercapable of measuring volume for at least 15 s andat least 8 L, with an accuracy of 3% or 0.05 L,whichever is greater, with flows between 0 and14.0 L/s. 11 Total resistance to airflow, at up to14 L/s, should be 1.5 cm H 2O/L/s, including alltubing, filters, and valves. Real-time display of flowvs volume is useful for assessing the magnitude ofeffort during the initial portion of the maneuver,whereas a volume vs time display provides informationduring the remainder of the maneuver fordetermining a satisfactory end of test. The spirometermust be appropriately calibrated at least daily,which establishes the relationship between sensordeterminedvalues for flow or volume and theactual flow or volume. Calibration syringes musthave an accuracy of 0.5% of the full scale(ie, 15 mL for a 3-L syringe). It is also importantto ensure that the calibration syringe functionsappropriately, and that monthly leak tests areconducted. The syringes themselves should beserviced, recalibrated, and verified on a regularbasis. Finally, regular daily checks for system leaksmust be undertaken and are easily performed.190 <strong>Pulmonary</strong> Function Testing (Marciniuk)ACC-PULM-09-0302-012.indd 1907/10/09 10:16:38 PM


More detailed instructions regarding calibrationare discussed elsewhere. 4,11Testing ProcedureA detailed review of testing procedures isbeyond the scope of this <strong>Board</strong> <strong>Review</strong>, but indepthinformation on the subject is available elsewhere.4,11 A number of important fundamentalswill be highlighted here. There is significant unfamiliaritywith assessing the “acceptability” and“repeatability” of spirometry. Inherent in optimalquality control is the acceptability of each individualspirogram, as well as the repeatability of anumber of acceptable spirograms, recognizing thatan adequate test requires at least three acceptablemaneuvers.Acceptability criteria are met if efforts are freefrom artifacts (cough, glottis closure, early termination,submaximal effort, leak, or obstructedmouthpiece); they have good starts (extrapolatedvolume is 5% of the FVC or 0.15 L, whichever isgreater); and they show satisfactory exhalationduration of 6 s (or a plateau in the curve isnoted). Within maneuver acceptability criteria arelisted in Table 5.Repeatability criteria are only applied after theacceptability of individual spirograms has beenmet, as outlined previously. The two largest values(from three acceptable individual spirograms) forboth the FEV 1and FVC must be within 0.15 L ofeach other to achieve repeatability criteria. If FEV 1and FVC repeatability criteria are not achieved,testing may be terminated after a total of eight“acceptable” individual maneuvers have beenperformed (Table 5) or if the patient cannot orTable 6. Between-Maneuver Reproducibility Criteria for Spirometry*• After three acceptable spirograms have been achieved,testing can be concluded if the two largest values forboth the FEV 1and the FVC are within 0.15 L of eachother• After three acceptable spirograms have been achieved,testing is still continued if the two largest values for boththe FEV 1and the FVC are not within 0.15 L of each other,until both criteria are met with analysis of additionalacceptable spirograms; or until a total of eight tests havebeen performed; or until the patient cannot/should notcontinue.• All satisfactory maneuvers should be saved*Adapted from Miller et al. 11should not continue. Between-maneuver criteriaare listed in Table 6.The largest FVC and the largest FEV 1shouldbe reported after examining all acceptable curves,even if they do not come from the same curve.The “best test” curve should be from the trial withthe largest sum of the FEV 1and FVC, and otherflow parameters should be derived from this besttest curve. These procedures are summarized inFigure 1.Table 5. Within-Maneuver Acceptability Criteria for Spirometry*• Individual spirograms are acceptable if they are free fromartifacts (cough during first second of exhalation, glottisclosures, early termination, submaximal effort, leak,obstructed mouthpiece)• Efforts have good starts (extrapolated volume 5% ofFVC or 0.15 L, whichever is greater)• Efforts show satisfactory exhalation (duration of 6 s,or a plateau in the volume-time curve, or subjectcannot/should not continue)*Adapted from Miller et al. 11Figure 1. Required steps for performing spirometry.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 191ACC-PULM-09-0302-012.indd 1917/10/09 10:16:38 PM


Maximal Flow-Volume CurveThe maximal flow-volume curve is helpful inquality assurance, in detecting mild airflowobstruction, and in detecting central airwayobstruction. For best results, both inspiratory andexpiratory loops are obtained. The subject, whiletidal breathing at rest, is typically asked toinspire to total lung capacity (TLC) and thenexpire with maximal force until empty, down toresidual volume (RV), followed by a maximuminspiration back up to TLC. Evaluation of theefforts is the same as noted previously, althoughthe inspiratory portion of the curve is oftenassessed only qualitatively. A volume vs timeplot and a flow vs volume plot for the samemaneuver are shown in Figure 2. Other measurementsalso are possible (volume vs timeplot shown in Fig. 3, where measured volumesare shown in clear bars and summed capacitiesare shown in shaded bars).The vital capacity (VC) is the volume of gasthat can be expelled from full inspiration to completeexpiration, whereas the FVC is the samemeasurement when the patient exhales withmaximal speed and effort. The VC is occasionallygreater than the FVC, the difference being relatedto the degree of obstruction. The inspiratorycapacity (IC) is the volume from a positionof passive resting end-tidal expiration (endexpiratorylung volume [EELV]) to full inspiration.The IC has been demonstrated to correlatewith lung hyperinflation at rest, and it is responsiveto changes in EELV with activity and pharmacologicinterventions. These measurementsare not forced, but at the same time, exhalationor inspiration should not be unduly slow. Toprovide absolute placement of these volumes andcapacities (and to estimate RV), actual measurementof functional residual capacity (FRC) mustbe undertaken (see “Lung Volumes” in thischapter for further discussion). As always, a tightseal between the lips and mouthpiece and a comfortable,constant seated position with appropriateheight adjustment of the mouthpiece isnecessary to ensure validity and repeatability.Unlike other measurements, for IC should bereported as the average of at least 3 maneuvers.The mean coefficient of variation for the IC inCOPD is estimated to be 5 3%.Figure 2. Volume vs time and flow vs volume plots.Reversibility TestingReversibility testing after administration of abronchodilator is frequently undertaken in thePFT laboratory. Baseline testing is followed byadministration of a short-acting bronchodilator,after which repeat testing is repeated. Either ashort-acting β 2-agonist (ie, salbutamol, 400 μg, withrepeat testing at least 10 min afterward) or anticholinergic(ie, ipratropium, 160 μg, with repeat testingat least 30 min afterward) can be used. 11 Currentrecommendations are for increased doses of thebronchodilator (four puffs vs two puffs) as notedpreviously, unless there is concern for the patient’s192 <strong>Pulmonary</strong> Function Testing (Marciniuk)ACC-PULM-09-0302-012.indd 1927/10/09 10:16:39 PM


Figure 3. Lung volumes and capacities.heart rate or tremor. An increase in either the FEV 1or the FVC of 12% and 200 mL is now considereda “positive” bronchodilator response. 9 Itshould be recognized that the utility of a bronchodilatorresponse is dependent on the clinical setting,and over-reliance on the result often leads toinappropriate clinical decision making. For example,a lack of a bronchodilator response does notnecessarily predict lack of a therapeutic response.In this regard, it is important to know when thepatient last received any medication that mightaffect the response. It is recommended that patientswithhold taking short-acting bronchodilators forat least 4 h before testing and any long-acting orsustained-release bronchodilators for at least 12 hbefore testing.Peak Expiratory FlowsPeak expiratory flow (PEF) is the highest flowachieved from a maximal forced expiratorymaneuver from TLC. When obtained from a spirometer,PEF is normally expressed at BTPS inL/s, whereas most portable monitoring instrumentsdisplay results in L/min. The measurement(and derived/related variables) is highly dependenton lung volume and effort. After achievingfull inflation, the patient should deliver the blowwithout hesitation. A delay of as little as 2 s canalter tracheal viscoelastic properties enough tocause a reduction in the PEF by as much as10%. 12Work on the measurement and clinical utilityof the PEF are ongoing. These measurements arenow embedded in asthma management strategies,but difficulties have been experienced with toomuch trust in the measurement. There are manyother related and derived variables from the PEF(Fig. 4), but they have not been proven to be equalor superior to spirometry in the individual clinicalsetting. 11Maximum Voluntary VentilationThe maximum voluntary ventilation is themaximum volume of air a subject can breatheover a period of time (typically 12 to 15 s), and itis expressed in L/min at BTPS. Its clinical contributionhas largely been superseded by the FEV 1and FVC and is now rarely measured. It may beuseful in some clinical settings in which ventilatorycapacity may be impaired by mechanismsdifferent than those directly affecting the FEV 1,for example, neuromuscular disorders and centralairway obstruction. Although it has beenused as an estimate of maximal ventilation duringexercise, there are significant limitations withthis approach. (This will be discussed in furtherdetail in the chapter “Cardiopulmonary ExerciseTesting.”)<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 193ACC-PULM-09-0302-012.indd 1937/10/09 10:16:40 PM


Figure 4. Peak expiratory flow and derived variables in flowvs volume [A] and volume vs time [B] plots.Lung VolumesMeasurement of absolute lung volumes is morechallenging and less available than assessing basiclung volumes by spirometry. However, althoughtheir utility in a variety of clinical settings continuesto be studied, their measurement remains thegold standard for establishing a restrictive abnormality.There are a variety of available techniquesin the pulmonary function laboratory that may beused to measure absolute lung volumes, includingbody plethysmography, nitrogen washout, andinhaled inert gas dilution. The lung volumes ofinterest are noted in Figure 3. In addition to the VCand IC (discussed and defined previously), othermeasurements are as follows:• FRC is the volume of gas present in the lungat end-expiration during tidal breathing and isthe absolute volume measured in the PFT laboratory.• Expiratory reserve volume (ERV) is the volumeof gas maximally exhaled from EELV duringtidal breathing.• Inspiratory reserve volume is the volume of gasthat can be maximally inhaled from the endinspiratorylung volume during tidal breathing.• Tidal volume is the volume of gas inhaled orexhaled at rest with each breath.• Total lung capacity (TLC) is the volume of gasin the lungs after maximal inspiration (the sumof all volumes).The determination of FRC is the key componentin the measurement of lung volumes. Regardlessof the technique used to measure FRC, TLCand RV will be estimated from the FRC and themeasurements of IC and ERV. Although opinionvaries on the best technique, the preferred methodis to measure ERV immediately after the FRC measurement,followed by a slow inspiratory vitalcapacity (IVC) maneuver, all performed with thepatient remaining on the mouthpiece until thecompletion of the maneuver (Fig. 5).Assuming technically satisfactory measurements,the reported value for FRC is the mean ofFRC measurements associated with the ERV andIVC maneuvers used to calculate the RV and TLC.The reported value for the RV is the reported FRCminus the ERV linked to the acceptable FRC. Thereported value for the TLC is the reported RV plusthe largest of the IVCs. 13 An alternative recommendedmethod involves determination of ICimmediately after the measurement of FRC toestimate TLC. The VC measurement can then eitherlinked to this maneuver, or it can be performedseparately after an ERV maneuver as describedpreviously.Body Plethysmography TechniqueThe term thoracic gas volume refers to theplethysmographic measurement of intrathoracicgas volume at the time of airflow occlusion, ie, the194 <strong>Pulmonary</strong> Function Testing (Marciniuk)ACC-PULM-09-0302-012.indd 1947/10/09 10:16:40 PM


Figure 5. Recommended method for measuring lung volumes.compressible gas within the thorax. In normalindividuals, FRC measured by plethysmographand gas dilution techniques yield similar results.In patients with obstructive lung disease and gastrapping,FRC determined by plethysmography istypically greater than measured by gas dilutiontechniques.Plethysmographic testing is performed by havingthe patient sit in a sealed box and pant againsta closed shutter. During the inspiratory phase ofthe pant, thoracic volume increases slightly,decompressing the volume of air in the lungs whileslightly compressing the volume of air in the box.Conversely, during the expiratory phase of thepant, thoracic volume decreases slightly, compressingthe volume of air in the lungs while slightlydecompressing the volume of air in the box. UsingBoyle’s law, where at a given temperature, theproduct of gas volume and pressure is constant:(1) V 1 P 1 V 2 P 2where the initial pressure and volume at FRCare P 1and V 1. Pressure and volume at the end ofthe inspiratory phase of the pant are P 2(P 1 P)and V 2(V 1 V). Thus:(2) V 1 P 1 (V 1+ ΔV) (P 1+ P)P 1 P are measured at the mouth, assumingthat mouth and alveolar pressures are equal, andΔV is measured using the change in box pressure,thus allowing for V 1to be solved.Body plethysmography measured lung volumesare considered the gold standard, but strictattention to testing methods, equipment upkeepand quality control, and panting technique andfrequency is essential.Gas Dilution TechniquesGas dilution techniques measure the gas volumein the lungs that communicates via the airwaysby use of a mass balance approach. A massbalance equation uses both the initial volume andtracer gas concentration and the final tracer concentrationto calculate the volume in the patient’slungs at the moment the tracer gas breathingbegan. The assumption that the tracer gas is largelyinsoluble, inert, and is well mixed in the lung isfundamental. For example, incomplete gas mixingwould cause lung volumes to be underestimated,whereas use of a soluble gas would lead cause lungvolumes to be overestimated. Either helium (He)rebreathing or nitrogen (N 2) washout techniques<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 195ACC-PULM-09-0302-012.indd 1957/10/09 10:16:42 PM


may be used, although both tend to underestimatelung volume when airway obstruction is present.For nitrogen washout, the washout is satisfactorywhen the N 2concentration is 1.5% for at least3 successive breaths. 13 For He rebreathing, a typical10% He concentration with 25 to 30% oxygen (O 2)gas mixture inspirate is breathed until equilibrationis reached, when the change in He concentrationis 0.02% for at least 30 s. 13 If only room air isadded to the 10% He gas mixture, it is importantto ensure adequate O 2replacement during the test.Both measurements must be corrected to BTPS,and the volume of the equipment dead spaceshould be subtracted.Diffusing CapacityThe carbon monoxide diffusing capacity of thelung (Dlco) assesses gas exchange by measuringthe rate of carbon monoxide (CO) transfer from thelungs to the blood, with values reported inmL/min/mm Hg (standard temperature [0C],pressure [760 mm Hg], dry). A number of bothstructural and functional properties influence thecapacity of the lung to exchange gas across thealveolar-capillary border. These include the volumeand distribution of ventilation, mixing and diffusion,the composition of the gas, characteristics ofthe alveolar membrane and lung parenchyma, thevolume of alveolar capillary plasma, the concentrationand binding properties of hemoglobin (Hgb),and the gas tensions in blood entering the alveolarcapillaries.As a test of gas transfer, CO uptake by the lungcan be measured by a number of techniques(ie, steady-state, intrabreath, rebreathing, andsingle-breath techniques), but the most commonlyused methodology is the single-breath technique.Testing begins with the patient exhaling down toRV, at which time a valve opens and the patientrapidly inhales the test gas to TLC. The patientholds his or her breath for 10 s and then breathesout rapidly while the exhaled gas is collected.The gas inspirate usually consists of a mixtureof N 2, 0.3% CO, an inert tracer gas such as He, and19 to 21% O 2. An alveolar sample of the exhaledgas is considered after anatomic and mechanicaldead space gas is discarded from the collected gas(∼0.75 to 1.0 L, but as little as 0.5 L if the patient’svital capacity is 2.0 L). A submaximal inspiredvolume (V I) can affect CO uptake. Therefore, it isimportant that the V Ibe as close to the known VCas possible. Because a reduction in V Iof 15% maysignificantly reduce the measured Dlco, a V Itargetof 85% of the largest VC is suggested. 14 To minimizevariability in the measurement, inspiration shouldbe rapid ( 4 s), breath-hold should be 10 ± 2 s, andexpiration should ideally also be 4 s in duration.At least 4 min should transpire between repeattesting to allow for adequate elimination of the testgas from the lungs. This time period should beextended (∼10 min) in patients with significantobstruction.Conditions affecting the Dlco are listed inTable 7. Other factors known to affect the Dlcoare diurnal variation (increased in the morning,and reported to decrease 1 to 2%/h throughoutthe day), 15 menstrual cycle (up to 13% change,increased premenses), alcohol consumption(increased after ingestion), and bronchodilatoradministration (increased up to 6% after use).Adjusting the DLCO MeasurementThe Dlco measurement is dependent on anumber of physiologic conditions, specificallyTable 7. Factors Affecting DLCO*Conditions that reduce DlcoSubmaximal inspiration or respiratory muscle weaknessAnemiaValsalva maneuverCOHgbIncreased fraction of inspired O 2Lung resectionChronic COPDInterstitial lung diseases<strong>Pulmonary</strong> vascular disease, ie, emboli, pulmonaryhypertension, vasculitisConditions that increase DlcoPolycythemiaLeft-to-right shuntAsthma<strong>Pulmonary</strong> hemorrhageMuller maneuverExerciseSupine positionObesityDecreased fraction of inspired O 2*Adapted from MacIntyre et al. 14196 <strong>Pulmonary</strong> Function Testing (Marciniuk)ACC-PULM-09-0302-012.indd 1967/10/09 10:16:42 PM


changes with Hgb, lung volume, carboxyhemoglobin(COHgb), fraction of inspired oxygen, exerciseand body position, in addition to age, sex, height,and possibly race. 14 Specific adjustments should bemade for Hgb, COHgb, and fraction of inspiredoxygen before interpretation.Adjusting for HemoglobinBecause COHgb binding is fundamental to themeasurement of Dlco, Hgb concentration cansubstantially affect its measurement. ExpressingHgb in g/dL, the recommended equation foradjusting the predicted Dlco in adolescents andadult men is as follows 14 :(3) Dlco (adjusted) Dlco (predicted) (1.7Hgb/[10.22 Hgb ])The recommended equation for adjusting thepredicted Dlco in women is as follows 14 :(4) Dlco (adjusted) Dlco (predicted) (1.7Hgb/[9.38 Hgb ])Adjusting for PAO 2Pao 2also affects the measurement of Dlcoeither because of supplemental O 2(higher Fio 2) orbecause of testing at altitude (lower Fio 2). TheDlco will change by 0.35% per mm Hg increasechange in Pao 2, or by 0.31% per mm Hg decreasechange. The recommended equation for adjustingthe predicted Dlco in a subject on supplementalO 2, and assuming a room air Pao 2at sea level of100 mm Hg is:(5) Dlco (adjusted) Dlco (predicted)/(1.0 0.0035[Pao 2 100])The recommended equation for adjusting thepredicted Dlco for altitude, and assuming a P IO 2at sea level of 150 mm Hg is:(6) Dlco (adjusted) Dlco (predicted)/(1.0 0.0031[P IO 2 150])Adjusting for COHgb Concentration and COBack PressureCOHgb may acutely reduce the Dlco 16 byoccupying Hgb binding sites and by reducing drivingpressure for CO transport from alveolar gas tocapillary blood. Typical exposure to environmentalCO and endogenous production results in measuredCOHgb of 1 to 2%, which is already incorporatedinto reference values based on healthynonsmoking subjects. Cigarette smoke and otherenvironmental sources of CO can lead to furtherappreciable increases in measured CO back pressureand COHgb that should be considered.Increases in COHgb also occur with repeated measurementsbecause CO is inspired in the Dlco test.For example, five tests will increase COHgb by∼3.5%, which will decrease the measured Dlco by3.0 to 3.5%. It is therefore recommended that nomore than five repeated measurements be undertakenat any sitting.Accounting for both the back pressure andCOHgb effects, a 1% increase in COHgb reducesthe measured Dlco by 0.8 to 1% from both knowneffects. This is represented by the following equation14 :(7) Dlco (adjusted) Dlco (predicted) (102% COHgb %)Adjusting for Lung VolumeAdjusting the Dlco for lung volume is bothcomplex and confusing. It is complex because thereason(s) responsible for a reduced lung volumevary, as do the effects of the specific underlyingreason on the resultant measured Dlco. Adjustingthe Dlco for lung volume is also confusing becausein some instances the relationship is proportional,whereas in other settings it is unpredictable andvaried. In view of these realities, drawing definitiveclinical conclusions based on a volume-correctedDlco should be made with caution. As recommendedin international guidelines, 14 further studyand understanding is required before more specificvolume-adjustment recommendations for theDlco can be made and uniformly adopted.The average of at least two acceptable testsshould be reported for interpretation of the Dlco.Reporting should include the unadjusted Dlco,predicted and percent predicted Dlco, predictedand percent predicted Dlco/ alveolar volume, anda list of any adjustments made to the value becauseof corrections (ie, Hgb, COHgb, Pao 2, or lungvolume). As always, technician comments regardingthe quality of the measurements should be noted.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 197ACC-PULM-09-0302-012.indd 1977/10/09 10:16:43 PM


Respiratory Muscle PressuresRespiratory muscle strength can be assessed bydetermining the maximal inspiratory pressure(MIP) and the maximal expiratory pressure (MEP).The MIP reflects the strength of the diaphragm andother inspiratory muscles, whereas the MEP reflectsthe strength of the abdominal and other expiratorymuscles. Clinical indications for assessing respiratorymuscle strength are varied, but they relate toany clinical setting in which respiratory muscleweakness or neuromuscular disorders affectingrespiratory status are suspected. Testing can behelpful in the diagnosis of respiratory muscleweakness, in assessing the severity of respiratorymuscle weakness, in predicting clinical outcomes,and in the longitudinal follow-up of patients diagnosedwith respiratory muscle weakness.The equipment necessary for measuring theMIP and MEP is simple, and similar techniques areused in both measurements. A pressure gauge connectedto a mouthpiece can be readily assembledor, alternatively, commercially available systemsmay be used. In either case, there must be a smallhole of approximately 1 mm in the system to allowfor an air leak, which minimizes the likelihood ofthe patient generating pressure with the use of theircheek muscles.The MIP measurement is performed with thepatient maintaining a tight seal between the lipsand the mouthpiece, exhaling down to RV, andthen breathing in as hard and fast as they can. Thepatient should keep up the maximal inspiratorypressure for ∼ 2 s, with the largest negative pressuresustained for 1 s being recorded. A longer continuousinspiratory effort is not required nor recommended.After resting for ∼ 1 min, the procedureis repeated for a total of at least three times, withthe intent of obtaining repeat measurements witha variability of 10 cm H 2O. It must be emphasizedthat testing is very dependent on a patient’s effortand cooperation, which may be enhanced by thetechnician initially demonstrating the maneuverto the patient and by the use of a practice test. Otherpotential sources of error in the measurementsinclude pressure gauge accuracy, the patient-deviceinterface, and leaks.The MEP measurement is similarly performedexcept that the patient is instructed to inhale toTLC and then blow out as hard and fast as he orshe can. In similar fashion to the MIP, the largestpositive pressure sustained for 1 s is recorded. Afterresting for ∼ 1 min, the procedure is repeated for atotal of at least three times, with the intent ofobtaining repeat measurements with a variabilityof 10 cm H 2O.In both instances, the maximum value obtainedfrom the three suitable maneuvers should bereported. Definitive reference ranges for the MIPand MEP have not been agreed on, although manyare available. 17,18 In general, a number of keyremarks are appropriate:1. Values decrease with age, and are approximatelyone third lower in women when comparedwith men.2. In adults 18 to 65 years of age, the MIP shouldexceed 90 cm H 2O in men and 70 cm H 2O inwomen. In adults 65 years, the MIP shouldgenerally exceed 65 cm H 2O in men and45 cm H 2O for women.3. In adults, the MEP should exceed 140 cm H 2Oin men and 90 cm H 2O in women.4. A MIP less than on third of normal predictshypercarbic respiratory failure. 195. A MEP 60 cm H 2O predicts a weak cough anddifficulty clearing secretions. 20A normal MIP reliably excludes inspiratorymuscle weakness (ie, good negative predictivevalue), but a low MIP does not reliably confirminspiratory muscle weakness. The poor positivepredictive value reflects the high frequency offalsely low MIP measurements caused by unrecognizedpoor effort or technique. 21 In normal healthyindividuals, 95% of the test-to-test variation has amagnitude of 25 cm H 2O. 22 Moreover, it has beendemonstrated that improvements in the MIPof −13 cm H 2O and MEP of 24 cm H 2O do not necessarilycorrelate with symptomatic improvement. 23The VC or the FVC often are used as alternativesor complimentary measurements to the MIPand MEP in clinical practice. Comparisons of theMIP and the VC demonstrate similar sensitivityand specificity for detecting hypercarbic respiratoryfailure. In the upright position, a reduced FVCseems to be less specific for detecting respiratorymuscle abnormalities than the MIP. However, adecrease in the FVC from upright to the supineposition appears to detect inspiratory muscleweakness more reliably than the MIP. 24198 <strong>Pulmonary</strong> Function Testing (Marciniuk)ACC-PULM-09-0302-012.indd 1987/10/09 10:16:43 PM


Airway ChallengesAssessing airway responsiveness is most commonlyundertaken to either confirm or exclude asuspected diagnosis of asthma, recognizing thatairway hyperresponsiveness (AHR) to exogenousstimuli is a characteristic feature of the disease.Testing can be performed by the use of a numberof various techniques, using either selective (typicallyspecific allergens, but uncommonly usedoutside of the research laboratory) or nonselectiveagents. Nonselective stimuli can be either direct(for example, methacholine and histamine) orindirect (for example, exercise, eucapnic voluntaryhyperventilation, cold air and mannitol).Methacholine is the most commonly performeddirect challenge. It is typically undertaken with twodifferent techniques: the tidal breathing method orthe dosimeter method. Recent studies have questionedthe sensitivity of the dosimeter methodusing TLC inhalations in patients with mild airwayhyperresponsiveness. Thus, the use of the tidalbreathing method or a modified dosimeter methodis now recommended. 25 A more detailed descriptionof the testing methods is available elsewhere.26,27Methacholine responsiveness is usually reportedas the provocation concentration (or dose)causing a 20% decrease in the FEV 1(ie, PC 20or PD 20).The PC 20is log-normally distributed in the population,can be measured into the normal range, andis a relatively imprecise measurement with a repeatabilityof 1 doubling concentration. The resultsare typically interpreted as follows 26 : normalPC 20 16 mg/mL; borderline PC 20 4 to 16 mg/mL;mild AHR PC 20 1 to 4 mg/mL; moderate AHRPC 20 0.25 to 1 mg/mL; and marked AHR PC 20 0.25 mg/mL. An example of results consistentwith moderate AHR is demonstrated in Figure 6.From a diagnostic point of view, the test has astrong negative predictive value, and thereforefunctions best when normal to rule out currentasthma. Several caveats should be remembered.First, methacholine challenges should not be performedwith TLC inhalations to maintain diagnosticsensitivity. Second, it is critical that symptomsbe current (ie, within the past few days). It followsthat a normal (“negative”) methacholine challengetest in a currently asymptomatic patient cannotrule out asthma.Most patients with exercise-associated bronchoconstriction(which when clinically occurringin isolation is frequently a manifestation of mildasthma) will have a positive direct inhalationalchallenge. However, it is now recognized thatsome patients, in particular high-performanceathletes, may have positive exercise challenges buta negative direct (ie, methacholine) challenge. TheFigure 6. Methacholine challenge test study results.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 199ACC-PULM-09-0302-012.indd 1997/10/09 10:16:43 PM


implication is that a negative methacholinechallenge will not completely rule out exerciseassociatedbronchoconstriction in this population.Exercise challenge testing will be discussed infurther detail in the chapter “CardiopulmonaryExercise Testing.”Another important and more frequent cause ofa false-negative methacholine challenge is recentexposure to bronchodilating agents, including bothβ 2-agonists and anticholinergics. The suggestedtime periods these agents should be withheld beforetesting are listed in Table 8. The specificity of direct(ie, methacholine and histamine) challenge testingis affected by underlying airflow obstruction, whichreflects resting airway geometry and not asthma. 28Airflow obstruction is a clinical feature that reducesthe specificity of the methacholine challenge, but itdoes not directly influence the sensitivity.In summary, the clinical utility of the methacholinechallenge test is primarily to exclude currentasthma in a symptomatic patient when thetidal breathing methacholine PC 20 16 mg/mL. Apositive methacholine challenge test is not synonymouswith a diagnosis of asthma, particularlyin the presence of significant airflow obstruction(ie, FEV 1 70% predicted). It is also suggested that5 to 15% of normal individuals and 20 to 40% ofpatients with rhinitis may also have a false-positivemethacholine challenge test. 27General Considerations in InterpretingPFTs<strong>Pulmonary</strong> function testing requires appropriateequipment, calibration, testing procedures,Table 8. Common Medications That Decrease BronchialResponsiveness*Short-acting β 2agonistsShort-acting anticholinergicsLong-acting β 2agonistsLong-acting anticholinergicsShort-acting theophyllinesSustained-release theophyllines*Adapted from Crapo et al. 26Suggested TimeInterval to WithholdAdministration8 h24 h48 hUp to 7 d24 h48 hassessment of acceptability and repeatability, interpretation,and integration with the patient’s clinicalpresentation. Interpretation of PFT results shouldbe meaningful, ie, a rendition of only which resultsare normal or abnormal can readily be providedby a machine. Requisition forms for PFTs shouldbe designed to encourage the requesting physicianto provide as much clinical information as is reasonableto enable the interpretation to be valuable.The key principles for diagnosing common abnormalventilatory disorders are listed in Table 9.Typical examples of commonly encountered flowvolumecurves are shown in Figure 7.Early evidence of airflow obstruction is qualitativelydemonstrated by a concave shape of theexpiratory flow-volume curve. Quantitative evidenceis often inferred by a reduction in expiratoryflows at mid-lung volumes (ie, FEF 25–75 and others).However, a reduction in the mid-volume expiratoryflows is not specific and not necessarily diagnosticfor small airways disease in an individualpatient. As airways disease becomes more pronounced,a disproportionate reduction in the FEV 1compared with the FVC becomes evident and, asnoted, this is more certain for a diagnosis of airwayobstruction.Pneumothoraces and bullae are clinical situationsoften characterized by a normal FEV 1/FVCand TLC measured in a body plethysmograph,but reduced FEV 1and FVC values. In this instance,Table 9. Diagnosing Common Respiratory Disorders*ObstructionFEV 1/VC 5th percentile of predictedReduced flows at low lung volumes are not specific forsmall airways disease in individual patients.A combined reduction in both FEV 1and VC is mostcommonly related to suboptimal effort. This patternmay occasionally indicate airflow obstruction, butconfirmation requires absolute lung volume measurement.Measurement of absolute lung volume measurement mayassist in diagnosing and assessing COPD and asthma.RestrictionTLC 5th percentile of predictedSpirometry can not accurately diagnose restriction.A reduced TLC from a single-breath technique should notbe used to diagnosis restriction.Mixed obstruction/restrictionFEV 1/VC and TLC 5th percentile of predicted.*Adapted from Pelligrino et al. 9200 <strong>Pulmonary</strong> Function Testing (Marciniuk)ACC-PULM-09-0302-012.indd 2007/10/09 10:16:44 PM


Figure 7. Flow-volume curves in common respiratory disorders.TLC assessed by gas dilution techniques will below. This difference between TLC measured in abody plethysmograph vs by gas dilution is alsoseen in patients with significant airflow obstruction,often accompanied by a significantlyincreased RV.Although the FVC is often used in place of theVC, it is suggested to use the largest measured VC,<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 201ACC-PULM-09-0302-012.indd 2017/10/09 10:16:45 PM


whether it is the IVC, the slow expiratory vitalcapacity, or the FVC. The forced expiratory volumein 6 s may be substituted for the FVC, if appropriatereference equations are used.Attention should always be paid to possibletechnical issues or problems, as well as to the commentsfrom the performing technicians. In additionto common obstructive, restrictive, and mixeddisorders, suspicion of possible neuromusculardisorders and central airway obstruction shouldalways be present. Although they infrequentlyare found, these conditions should be activelyexcluded.Typical flow-volume plots observed with differingtypes of physiologic central airway obstructionare shown in Figure 8. Poor or submaximalefforts should always be considered in this setting.It is important that repeated flow-volume maneuversare maximal and reproducible, and the technicianshould confirm this in the report. In thissetting, a printout of each individual maneuver canbe very helpful.Serial testing is routinely undertaken in clinicalpractice to monitor results over time, and to evaluatechange following intervention(s). An understandingof test variability is required before onecan conclude that any observed change relates tothe underlying process or to an intervention. Thereproducibility of variables can be presented inmany different ways, and there is ongoing debateabout which method is best. For example, theAmerican Thoracic Society/European RespiratorySociety recommends that in normal individuals,the year-to-year change in the FEV 1should exceed15% before confidence can be given that a meaningfulclinical change has occurred. 9 The coefficient ofvariation (CV) is also commonly used for this purpose,and is calculated as:CV (%) SD divided by the mean 100It is assumed that a difference of 2 CV% isrequired to conclude a clinically significant change.Under ideal conditions, the mean CV for commonlymeasured lung function tests is as follows: 29,30FEV 1 3.2 to 7.5%, FVC 4.5 to 7.9%, TLC (plethysmography) 2.2 to 4.3%, TLC (He) 1.5to 3.7%, FRC (plethysmography) 3.5 to 6.8%,FRC (He) 4.9 to 10.4%, RV (plethysmography) 8.9 to 12.4%, RV (He) 2.4 to 14.0%, and Dlco 4.1to 4.5%.The use of severity classification systems iscurrently fraught with difficulty and frustrationfor the practicing clinician. First, none have beenspecifically validated—as stated by others, “thenumber of categories and exact cut-off points arearbitrary.” 9 Second, conflicting classificationsFigure 8. Flow-volume curves in physiologic central airway obstructions. A variable intrathoracic obstruction; B variableextrathoracic obstruction; C physiologically fixed central airway obstruction.202 <strong>Pulmonary</strong> Function Testing (Marciniuk)ACC-PULM-09-0302-012.indd 2027/10/09 10:16:46 PM


Table 10. Classifying Severity of Airflow Obstruction*ATS/ERS Standardization ofATS/ERS COPD Guidelines, 31Lung Function Testing 9 Canadian Thoracic Society, 32and GOLD 33Degree of severity (FEV 1% predicted) (FEV 1% predicted)Mild 70 80Moderate 60–69 50-9Moderately severe 50–59 —Severe 35–49 30-49Very severe 35 30*ATS/ERS American Thoracic Society/European Respiratory Society; GOLD Global Initiative for Chronic ObstructiveLung Disease.schemas exist (see Table 10). 31−33 It is thereforerecommended that clinicians choosing to use aspecific classification system should recognizethese limitations, and make note of the specificschema utilized in their interpretation.Similarly, a severity classification scheme hasbeen proposed for the Dlco (mild 60% and


testing. Standardization of the single-breath determinationof carbon monoxide uptake in the lung.Eur Respir J 2005; 26:720−73515. Cinkotai FF, Thomson ML. Diurnal variation inpulmonary diffusing capacity for carbon monoxide.J Appl Physiol 1966; 21:539−54216. Graham BL, Mink JT, Cotton DJ. Effects of increasingcarboxyhemoglobin on the single breath carbonmonoxide diffusing capacity. Am J Respir CritCare Med 2002; 165:1504−151017. Harik-Khan RI, Wise RA, Fozard JL. Determinantsof maximal inspiratory pressure: the BaltimoreLongitudinal Study of Aging. Am J Respir CritCare Med 1998; 158:1459−146418. Enright PL, Kronmal RA, Manolino TA, et al.Respiratory muscle strength in the elderly. Correlatesand reference values. Am J Respir Crit CareMed 1994; 149:430−43819. Lyall RA, Donaldson N, Polkey MI, et al. Respiratorymuscle strength and ventilatory failurein amyotrophic lateral sclerosis. Brain 2001;124:2000−201320. Man WD, Kyroussis D, Fleming TA, et al. Coughgastric pressure and maximum expiratory mouthpressure in humans. Am J Respir Crit Care Med2003; 168:714−71721. Aldrich TK, Spiro P. Maximal inspiratory pressure:does reproducibility indicate full effort? Thorax1995; 50:40−4322. Maillard JO, Burdet L, van Melle G, et al. Reproducibilityof twitch mouth pressure, sniff nasalinspiratory pressure, and maximal inspiratorypressure. Eur Respir J 1998; 11:901−90523. Goswami R, Guleria R, Gupta AK, et al. Prevalenceof diaphragmatic muscle weakness and dyspnoeain Grave’s disease and their reversibilitywith carbimazole therapy. Eur J Endocrinol 2002;147:299−30324. Lechtzin N, Wiener CM, Shade DM, et al. Spirometryin the supine position improves the detectionof diaphragmatic weakness in patients with amyotrophiclateral sclerosis. Chest 2002; 121:436−44225. Cockcroft DW, Davis BE, Todd DC, et al. Methacholinechallenge: comparison of two methods.Chest 2005; 127:839−84426. Crapo RO, Casaburi R, Coates AL, et al. Guidelinesfor methacholine and exercise challengetesting—1999. Am J Respir Crit Care Med 2000;161:309−32927. Cockcroft DW. Bronchial challenge testing. In:Adkinson NF, Bochner BS, Busse WW, et al, eds.Middleton’s allergy. 7th ed. Philadelphia, PA: ElsevierInc., 2009; 1295−130828. Verma VK, Cockcroft DW, Dosman JA. Airwayhyperresponsiveness to inhaled histamine inchronic obstructive airways disease: chronic bronchitisvs emphysema. Chest 1988; 94:456−46129. Punjabi NM, Shade D, Patel AM, et al. Measurementvariability in single-breath diffusing capacityof the lung. Chest 2003; 123:1082−108930. Hankinson JL, Stocks J, Peslin R. Reproducibilityof lung volume measurements. Eur Respir J 1998;11:787−79031. Celli B, MacNee W. ATS/ERS Task Force. Standardsfor the diagnosis and treatment of patientswith COPD: a summary of the ATS/ERS positionpaper. Eur Respir J 2004; 23:932−94632. O’Donnell DE, Aaron S, Bourbeau J, et al. Canadianthoracic society recommendations for managementof chronic obstructive pulmonary disease—2007update. Can Respir J 2007; 14(suppl B):5B−32B.33. Rabe KF, Hurd S, Anzueto A, et al. Global strategyfor the diagnosis, management, and preventionof chronic obstructive pulmonary disease: GOLDexecutive summary. Am J Respir Crit Care Med2007; 176:532−555204 <strong>Pulmonary</strong> Function Testing (Marciniuk)ACC-PULM-09-0302-012.indd 2047/10/09 10:16:47 PM


Bronchoscopy and Interventional PulmonologyDavid Feller-Kopman, MD, FCCPObjectives:• <strong>Review</strong> the indications for flexible and rigid bronchoscopy• <strong>Review</strong> the diagnostic yield from various bronchoscopicprocedures• <strong>Review</strong> the procedures available to the interventionalpulmonologist• Stress the need for specific training in advanced proceduresKey words: airway obstruction; bronchoscopy; hemoptysis;interventional pulmonology; laser; lung cancer; pleuroscopy;stent; thoracoscopyGustav Killian has been described as the father ofbronchoscopy. 1 Using a metallic tube, electric light,and topical cocaine anesthesia, Killian removed apork bone from a farmer’s airway in 1897. 2 Beforethe development of rigid bronchoscopy, more thanone half of the patients who aspirated foreign bodiesdied as the result of pneumonia. Rigid bronchoscopywith foreign body removal quickly evolvedinto the treatment of choice in these patients, witha clinical success rate 98%. 3 Over the subsequentyears, Killian published extensively and lecturedthroughout Europe and the United States. He wenton to develop bronchoscopes, laryngoscopes, andendoscopes, and he described techniques such asthe use of fluoroscopy and x-ray to define endobronchialanatomy and endoluminal radiotherapy(brachytherapy).Chevalier Jackson, an otolaryngologistwho practiced in Philadelphia, is credited withadvancing the field of bronchoscopy and airwayinterventions in the United States. During the next150 years, bronchoscopic techniques and instrumentscontinued to be refined. In 1966, ShigetoIkeda presented the first prototype flexible fiberopticbronchoscope at the Ninth InternationalCongress on Diseases of the Chest in Copenhagen.In 1968, Machita and Olympus both introducedcommercially available fiberoptic bronchoscopes. 4In 1980, Dumon presented his use of the Nd:YAGlaser via the fiberoptic bronchoscope, and sincethat time, the flexible bronchoscope has beenwidely used as both a diagnostic and therapeutictool for both diseases of the parenchyma andcentral airways. With the miniaturization of electronicdevices, the first video bronchoscope wasintroduced in 1987. This development allowedendoscopic pictures to be printed out and sharedand, even more importantly, physicians no longerneeded to look through an eyepiece. Instead, theendoscopic image could be projected onto monitors,allowing everyone in the room to see whatwas happening in the airway.Little has changed in the appearance of bronchoscopessince 1968. The external diameter of theflexible bronchoscope ranges from 2.8 to 6.3 mmin diameter. The diameter of the working channelranges from 1.2 to 3.2 mm. A working channel 2.8 mm is recommended for more therapeuticflexible bronchoscopy because it allows for bettersuction and the passage of larger instruments. It isimportant to note the relative anatomy at the tip ofthe bronchoscope. By convention, as viewed fromthe operator’s perspective, the camera is at 9:00,suction at 6:00, and the working channel at 3:00.These landmarks play a role when navigating theairways because the bronchoscope may need to berotated to visualize or sample the intended target.There are several outstanding texts and reviewarticles in which the authors summarize the currentstate of the art for bronchoscopy and interventionalpulmonology. 5–9 This chapter will review some ofthe key factors associated with these fields.Thoracic AnatomyIt is crucial that the bronchoscopist become anexpert in airway and thoracic anatomy, includingknowledge of the nasopharnyx and oropharynxin addition to the bronchial tree and mediastinalstructures. Several excellent chapters/texts 10,11 areavailable to assist one with this process. The segmentalanatomy of the lungs from both an externaland internal perspective are required, and oneshould become familiar with the name (ie, superior<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 205ACC-PULM-09-0302-013.indd 2057/10/09 8:25:47 PM


segment of the right lower lobe), as well as the numbersystem (ie, RB6) as initially proposed by Jacksonand Huber 12 and Boyden. 13 It is also important toappreciate mediastinal anatomy, including intrathoracicvessels and lymph nodes, as well as theirrelationship to endobronchial landmarks. These willserve as reference points for transbronchial needleaspiration (TBNA) and endobronchial ultrasound(EBUS) and help avoid injury to the patient shouldthe bronchoscopist use therapy such as the Nd:YAGlaser or brachytherapy. The use of endoscopicsimulators has been associated with a more rapidacquisition of bronchoscopic expertise. 14Diagnostic BronchoscopyThere are many indications for diagnosticbronchoscopy. The most common indication isfor the diagnosis of suspected lung cancer. Otherindications include evaluation of diffuse lung disease,infiltrates in the immunocompromised host,hemoptysis, and cough. Additionally, bronchoscopywith bronchoalveolar lavage (BAL) and/orbrushing may be useful for the diagnosis of ventilator-associatedpneumonia. 15,16Cancer Diagnosis and StagingLung cancer is the leading cause of cancerdeaths in the United States, and it will account forapproximately 161,840 deaths in 2008, more thancolon, breast, and prostate cancers combined. 17 Itis estimated that there were 215,020 new cases oflung cancer in 2008, and the incidence and numberof lung cancer deaths continues to increaseamong women in the United States. 18 Bronchoscopyprovides a minimally invasive approachfor the diagnosis tumors in the central airways.Although the yield is somewhat lower for solitaryparenchymal lesions, advances in navigation suchas CT fluoroscopy and electromagnetic and virtualbronchoscopic navigation have significantlyimproved the yield for peripheral tumors. Severalcomprehensive reviews discussing the diagnosisand staging of lung cancer, as well as the role ofbronchoscopy, are available. 19–23The bronchoscopic evaluation of patientswith suspected malignancy is guided by clinicalsymptoms as well as radiographic findings.Depending on the history and chest CT/PET-CTfindings, bronchoscopy may be the diagnosticmodality of choice because it can be used to makethe diagnosis and stage the patient at the samesitting. If appropriate staging is to be performed,biopsy of the lesion that will place the patient inthe highest clinical stage should occur before otherbiopsies. For example, if a patient presents with aleft lower lobe mass and an enlarged subcarinal(station 7) and right paratracheal (station 4R)lymph node, the appropriate procedure would bea bronchoscopy with TBNA of the 4R node becausethis procedure would stage the patient with N3disease, making the patient unresectable. If a diagnosiswas not obtained from 4R, station 7 should besampled next because this would be considered anN2 node if involved with cancer. Only if samplingof this node is nondiagnostic would one need toproceed with addressing the left lower lobe mass.Initial biopsy of the mass could contaminate theworking channel and make the TBNA a false-positivefinding, precluding the patient from curativesurgery.There are several available tools by which toobtain specimens during bronchoscopy, includingforceps biopsy, brushing, bronchial wash/lavage,and TBNA. Electrocautery snare forceps is anexcellent tool for the removal of a pedunculatedairway lesion because it can open the airway andalso provide excellent tissue for the pathologist. 24The choice of the aforementioned modalitiesis primarily determined by the location of thepathology. However, data support the use of thecombination of techniques to improve diagnosticyield, as opposed to using them in isolation. 19,20,25,26Additionally, endobronchial needle aspirationshould be used with all visible lesions because itsuse in combination with conventional techniqueshas been associated with an improvement in theyield from 65 to 96%. 27If the lesion is not visible endoscopically, BALcan be performed. 28–30 In brief, the bronchoscope iswedged in the target segmental or subsegmentalbronchus leading to the lesion. Aliquots of normalsaline solution are instilled and then aspirated. Ideally,the return should be 40 to 60% of the instilledvolume; however, this result is dependent onthe segment lavaged, with better return comingfrom less-dependent locations. 31 BAL is thoughtto sample approximately 1 million alveoli, andthe cellular and noncellular contents of the lavage206 Bronchoscopy and Interventional Pulmonology (Feller-Kopman)ACC-PULM-09-0302-013.indd 2067/10/09 8:25:48 PM


fluid have been shown to closely correlate with theinflammatory nature of the entire lower respiratorytract. 30,32–34 Transbronchial biopsy (TBBx), brushing,and TBNA can also be performed for peripherallesions.The diagnostic yield of BAL for peripheralcancer ranges from 4 to 68%. 35–37 Without advancedguidance (discussed in the next paragraph), theyield for TBBx ranges from 49 to 77%; for brushing,it is 26 to 57%. However, again, the combinationof all three increases the yield to 68%. 35 TBNAof peripheral nodules has been shown to havea greater diagnostic yield than other samplingtechniques, and it should be used for the biopsyof peripheral lesions. 38,39 Kawaraya et al 37 recommendwashing the tools used to obtain samplesand sending the wash for cytologic analysis, witha combined diagnostic yield of 93%.As for peripheral nodules and masses, despiteTBNA being introduced 25 years ago, it remainsan underused technique, with only 12% of pulmonologistsreporting its routine use for the evaluationof patients with suspected lung cancer. 40,41 The technique,however, is incredibly useful. A study of 365patients with documented malignancy found thatthe use of standard TBNA (without EBUS) wouldpreclude further invasive surgery in 29% of patientsand was able to establish exclusively a diagnosisin 18% of cases. 42,43 The yield with TBNA hasbeen associated with tumor cell type (small cell non-small cell lymphoma), lymph node size, andlymph node location. 44The use of CT fluoroscopy to guide TBNAand TBBx has several advantages over standardfluoroscopy. As opposed to standard fluoroscopy,which is typically only used in two dimensions, CTprovides the ability to visualize the target in threedimensions. With standard fluoroscopy, either thepatient or the C-arm needs to be rotated to confirmthe biopsy tool is not anterior or posterior to thetarget. CT fluoroscopy provides real-time threedimensionalconfirmation of the appropriate andinappropriate biopsy sites. The use of CT fluoroscopyhas been associated with an accuracy of 88%,including patients who have previously undergonea nondiagnostic bronchoscopy with standardTBNA. 45 The one frustration that can occur withCT-guided bronchoscopy is that one may be ableto see the lesion but be unable to steer the bronchoscope/catheterinto the correct airway.EBUS is another important modality used tohelp improve the accuracy of sampling peripheralnodules, as well as in guiding TBNA. The earliestEBUS system used a radial probe with a 20-MHztransducer that could be inserted via a 2.8-mmworking channel. Despite not allowing for realtimeguidance, early studies 46–48 with radial-probeEBUS found significant improvements in the yieldof TBNA, sampling of peripheral nodules, anddistinguishing tumor invasion vs compression inthe central airways. The sonographic characteristicsof peripheral nodules also have been shown tocorrelate well with histology findings. 49,50More recently, a bronchoscope with a dedicatedconvex 7.5-MHz linear array transducer anda distinct working channel has been developedand has the major benefit of providing real-timeguidance for TBNA of mediastinal and hilarlymph nodes, with excellent results. 51–53 Despitethe high sensitivity and specificity of EBUS-TBNA, it is crucial to understand that a nondiagnosticresult is not equivalent to a negativeresult. The false-negative rate for EBUS-TBNA canbe as high as 20%. Therefore, it is mandatoryfor the physician to follow all nondiagnosticprocedures with a confirmatory surgical biopsyor appropriate clinical follow-up before concludingthat the test produced a true-negativeresult. 54Autofluoresence bronchoscopy can be used forthe early detection of cancer in the central airways,primarily carcinoma in situ (CIS), and squamouscell carcinoma. When exposed to light in the violetbluespectrum (400 to 450 nm), the normal airwayfluoresces green. As submucosal disease progressesfrom normal, to metaplasia, to dysplasia, to CIS,there is a progressive loss of the green autofluoresence,causing a red-brown appearance of theairway wall. Several studies 54a-b have shown thatthe use of autofluoresence increases the detectionof early-stage lung cancer by up to sixfold whencompared with white light bronchoscopy. Themajor limitation of autofluoresence lies in its lowspecificity, and one study 55 suggests that only onecase of CIS is identified per 100 bronchoscopies.Other new technologies such as narrow-bandimaging, optical coherence tomography, and confocalimaging may also help in the diagnosis ofearly-stage lung cancer, although definitive dataare still pending.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 207ACC-PULM-09-0302-013.indd 2077/10/09 8:25:48 PM


Diffuse Parenchymal Lung DiseaseDiffuse parenchymal lung disease describes agroup of infectious and inflammatory and fibroticdisorders that may involve the interstitial, alveolar,bronchial, and vascular structures of the tracheobronchialtree. The most commonly used samplingtechniques for patients with diffuse disease includeBAL bronchial brushing, TBBx and, occasionally,endobronchial biopsy and TBNA.Although associated with a low morbidityand mortality, TBBx should be used only whenthe potential results will have an impact on treatmentdecisions. Diseases in which TBBx can provediagnostic or when it has been shown to significantlyincrease the diagnostic yield as comparedwith less-invasive means include lymphangiticcarcinomatosis, sarcoidosis, rejection after lungtransplantation, hypersensitivity pneumonitisand, sometimes, mycobacterial and invasive fungalinfection. 29,56,57 The overall diagnostic yield forTBBx clearly depends on the disease entity as wellas severity. For example, the yield for sarcoidosiscan approach 90% in patients who have stage IIand III disease. 58Aside from providing a specific diagnosis incases of cancer or infection, the results of BAL canserve to limit the differential diagnosis considerably.56 For example, BAL findings with lymphocytepredominance suggest granulomatous disease suchas sarcoidosis, berylliosis, or a lymphoproliferativedisorder. Neutrophil predominance suggestsbacterial infection, acute interstitial pneumonia,and may also be found in patients with asbestosisor usual interstitial pneumonitis. Eosinophils areseen in patients with eosinophilic pneumonias,hypereosinophilic syndromes, or Churg-Strausssyndrome. The BAL fluid of patients with pulmonaryalveolar proteinosis has a unique appearancethat is described as milky or opaque. The alveolarmacrophages are filled with pulmonary alveolarproteinosis-positive material, and lamellar bodiescan be seen with electron microscopy.Infectious DiseasesCommunity-Acquired PneumoniaThe role of bronchoscopy in communityacquiredpneumonia remains controversial. 59When used, BAL and protected brush are the maindiagnostic procedures, and the specimens shouldideally be sent for quantitative culture, with athreshold of 10 4 cfu in the BAL and 10 3 cfu in theprotected brush. A study by van der Eerden andcolleagues 60 found bronchoscopy to be of additivediagnostic value in 49% of patients who wereunable to raise sputum for Gram stain and cultureand in 52% of patients for whom treatment failed.As expected, pretreatment with antibiotics maysignificantly reduce the yield, 61 although Feinsilveret al 62 found an 86% yield in patients with nonresolvingpneumonia who had already receivedantibiotics. Another important indication for bronchoscopyin patients with community-acquiredpneumonia is the ruling out of an obstructingendobronchial lesion. Obviously, it is crucial toavoid contamination with upper airway secretionswhen performing bronchoscopy in patientswith pneumonia. Key procedural aspects includeminimizing suctioning, as well as minimizing theinstillation of saline solution or lidocaine, becausesecretions in the working channel will be flushedback into the airways, and high concentrationsof lidocaine can be bacteriostatic. Baselski andWunderink 59 reviewed the indications, proceduraldetails, mechanisms for specimen transport/handling,and interpretation of results.Health Care and Ventilator-AssociatedPneumoniaBy using the techniques of evidence-basedmedicine to guide their recommendations,the American Thoracic Society and InfectiousDiseases Society of America have recentlyreviewed these topics in great detail. 63 Some majorpoints included the collection of a lower respiratorytract culture before the initiation of antibiotics,the use of semiquantitative or quantitativeculture data, and the use of negative culture datato discontinue antibiotics in patients who have nothad changes in their antibiotic regimen within thelast 72 h. 63 Additionally, the use of a bronchoscopicstrategy was supported as a way to reduce 14-daymortality. 15,63 More recently, an invasive strategywas found to have no impact on mortality. 16 Amajor criticism of this study was the fact that theauthors excluded patients with pseudomonas andStaphylococcus pneumoniae.208 Bronchoscopy and Interventional Pulmonology (Feller-Kopman)ACC-PULM-09-0302-013.indd 2087/10/09 8:25:48 PM


The Immunocompromised HostThe early diagnosis and initiation of the appropriateantibiotic administration is the cornerstone ofsuccessful treatment of the immunocompromisedpatient with pneumonia. Additionally, it is importantto note that multiple diagnoses often can bepresent simultaneously in these patients 64 and thatnoninfectious conditions may have a similar presentationof cough, dyspnea, fever, and an infiltrateon imaging. Bronchoscopy is an excellent methodof evaluating these patients because the use ofless-invasive techniques can miss the diagnosis inapproximately 30% of patients. 34 In addition, earlierdiagnoses may improve rates of mortality. 65Helmers and Pisani 30 suggest three broad categoriesof immunocompromised patients: thosewho are pharmacologically immunosuppressed,those with hematologic malignancy/malfunction,and those with immunodeficiency. The reason forthese categories is because each type of immunodeficiencypredisposes the patient to distinct infections.For example, infection with S pneumoniae andHaemophilus influenzae is common in patients withhypogammaglobulinemia, whereas Aspergillussp is more likely to cause disease in patients withneutropenia. Cryptococcus sp, Histoplasma sp, andviruses such as cytomegalovirus are seen in patientswith defects in cell-mediated immunity. 66BAL, which has an overall diagnostic yield ofapproximately 60 to 70%, is the most commonlyused bronchoscopic technique used to obtain adiagnosis in immunocompromised patients. 66 Incomparative studies, the sensitivity of TBBx hasbeen shown to be roughly the same, 67,68 althoughthe combined use of both techniques may increasethe yield. 69 The results of BAL have been shownto change disease management in up to 84% ofcases. 70 Unfortunately, although bronchoscopychanged disease management in patients whohave undergone bone marrow transplantation, therate of mortality has remained extremely high. 64,71Bronchoscopy with BAL remains a safe procedure,even in the immunocompromised host. Brushingand TBBx have been associated with a greaterincidence of bleeding complications in patientswho are thrombocytopenic. 72 The “gold standard”remains open-lung biopsy, which can yield a specificdiagnosis in 62% of patients and result in asignificant increase in survival. 73HemoptysisThere are many causes of hemoptysis, includinginfectious and inflammatory, vascular, andneoplastic processes. 74 Although one would thinkthat the use of bronchoscopy can often aid in thediagnosis of a radiographic occult neoplasm in apatient who presents with hemoptysis, a bronchoscopicdiagnosis of malignancy is made in 5%.The incidence of cancer is increased in patientswho present with hemoptysis and normal chestimaging if they are 40 years of age, are male,and have a 40 pack-year smoking history. 75–77As such, bronchoscopy is recommended in thispatient population. Although the appropriatetiming for bronchoscopy is controversial, thereis a greater likelihood of identifying the bleedingsource when performed within the first 48 h ofsymptoms. 78 The combined use of bronchoscopyand CT is also recommended. 74 If patients areclinically stable, I would recommend performingthe CT first because it can serve as a road map toguide bronchoscopy. 79Therapeutic BronchoscopyAlthough many advanced techniques can beperformed with flexible bronchoscopy, the rigidbronchoscope is the preferred tool in many situations.80,81 The rigid bronchoscope provides an airwayin which to oxygenate/ventilate the patient aswell as pass multiple devices, including large-boresuction catheters, laser/argon plasma coagulation/electrocautery fibers, the microdebrider blade, andlarger biopsy instruments. The rigid bronchoscopeitself can also be used to achieve relative lungisolation, dilate stenosis, and “core-out” tumortissue from the central airways. Unfortunately,only approximately 4% of pulmonologists currentlyperform rigid bronchoscopy, 82 and its use istypically not taught in pulmonary and critical carefellowship programs. 83Massive HemoptysisThe definition of massive hemoptysis rangesfrom 100 to 1,000 mL expectorated during a 24-hperiod. 84,85 Because the majority of patients withmassive hemoptysis die from asphyxia, not exsanguination,and because the anatomic dead space<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 209ACC-PULM-09-0302-013.indd 2097/10/09 8:25:49 PM


is approximately 150 mL, I consider any amount 100 mL in 24-h period as massive. By using adefinition of 500 mL/24 h, Hirshberg et al 74 foundthat 14% of 208 patients presenting with hemoptysishad massive hemoptysis.In addition to identifying the source andcause of bleeding, bronchoscopy clearly plays animportant therapeutic role in patients with massivehemoptysis. If available, rigid bronchoscopyis generally considered the airway procedure ofchoice in these patients. If rigid bronchoscopy isnot available, the options include intubation witha single-lumen endotracheal tube, a double-lumentube, or placement of a bronchial blocker. Becausedouble-lumen endotracheal tubes, or specializedtubes that come with an endobronchial blocker (ie,Univent; Vitaid; Williamsville, NY), can be moredifficult to place, especially in the setting of massivehemoptysis, my practice is to place a largersingle-lumen tube.The main role of flexible bronchoscopy in thepatient with massive hemoptysis lies in helping toobtaining lung isolation by guiding the endotrachealtube or bronchial blocker and providing sometherapeutic aspiration to protect the “good” lung.All bronchoscopists should become familiar withbronchial blockers (ie, Arndt Bronchial Blocker;Cook Critical Care; Bloomington, IN), which canbe passed in parallel to the bronchoscope. Thesecatheters are guided to the culprit segmental,lobar, or mainstem bronchus, and the balloon isinflated to the recommended volume/pressure.After a maximum of 24 h, the balloon should bedeflated under bronchoscopic visualization. 86–88Other bronchoscopic techniques used to controlhemoptysis include the topical application of icedsaline, epinephrine (1:20,000), and thrombin/thrombin-fibrinogen/or cyanoacrylate solutions,although the data supporting these techniques areextremely limited. 89–91Advanced Bronchoscopic TechniquesArgon Plasma CoagulationArgon plasma coagulation (APC) is a noncontactmethod that uses ionized argon gas (plasma) toachieve tissue coagulation and hemostasis. Becausethe plasma is directed to the closest groundedsource, APC has the ability to treat lesions lateralto the probe. The depth of penetration for APC isapproximately 2 to 3 mm and, hence, the risk ofairway perforation is less when compared withlasers. Gas flow through the APC fiber should bekept to the lowest setting possible (ie, 0.5 L/min) toavoid the potential of gas emboli formation. 92Laser TherapyThe Nd:YAG laser is the most widely usedlaser in the lower respiratory system and has beenused for both benign and malignant disease. 93The primary advantage of laser photoresectionis that it provides rapid destruction/vaporizationof tissue. Care must be taken because thedepth of penetration can approach 10 mm, andairway perforation with resultant pneumothorax,pneumomediastinum, and vascular injury hasbeen reported. As such, it should be used onlyby experienced interventional bronchoscopists.Despite this caveat, the safety record of laserbronchoscopy is excellent, with an overall complicationrate of 1%. 93 Lesions most amenableto laser therapy are central, intrinsic, and short( 4 cm), with a visible distal endobronchiallumen. 94 When lesions meet these criteria, patencycan be reestablished in 90% of cases. Of note, thetissue destruction obtained with laser therapy isoften more than that is visually appreciated at thetime of the procedure.CryotherapyCryotherapy is a safe and effective tool for avariety of airway problems. 95,96 By releasing nitrousoxide or carbon dioxide stored under pressure,the tip of the cryoprobe rapidly cools to - 89C.Cryotherapy is especially useful for the removalof organic foreign bodies with high water content,such as grapes and vegetable matter, in additionto facilitating the removal of tenacious mucus orblood clots when performing flexible bronchoscopy.Compared with other techniques for tumordestruction, cryotherapy has a delayed effect and,thus, requires a repeat bronchoscopy to remove thenecrotic tissue. The distinct advantage of cryotherapylies in the fact that the normal airway tissue isrelatively cryoresistant. Additionally, unlike “hot”thermal techniques, cryotherapy is not associatedwith a risk of airway fires.210 Bronchoscopy and Interventional Pulmonology (Feller-Kopman)ACC-PULM-09-0302-013.indd 2107/10/09 8:25:49 PM


ElectrocauteryElectrocautery uses alternating current at highfrequency to generate heat, which cuts, vaporizes,or coagulates tissue depending on the power.Electrocautery is a contact mode of tissue destruction,and a variety of cautery probes are available,including blunt tip probes, knifes, and snares. Thesnare is an ideal tool for pedunculated lesionsbecause the stalk can be cut and coagulated whilepreserving the majority of the tissue for pathologicinterpretation. 97 The knife is also an excellent toolfor producing radial incisions before dilation ofweb-like tracheal stenosis. As with laser therapy, therisks of electrocautery include airway perforation,airway fires, and damage to the bronchoscope.Photodynamic TherapyPhotodynamic therapy (PDT) involves the IVinjection of a photosensitizing agent, and thenactivating the drug with a nonthermal laser toproduce a phototoxic reaction and cell death.Because the laser is not a heat source, airway firesare not an issue. Because tumor cells retain the druglonger than other tissues, waiting approximately48 h after drug injection will lead to preferentialtumor cell death as compared with normal tissueinjury. Similar to cryotherapy, maximal effects aredelayed, and a repeat, “clean-out” bronchoscopyshould be performed 24 to 48 h after drug activation.The primary side effect from PDT is systemicphototoxicity, which can last up to 6 weeks afterinjection. Newer drugs are being developed withthe hope of increasing tumor selectivity andreducing the duration of skin phototoxicity. PDThas been shown to be curative for early-stage lungcancer of the airways and is an especially attractiveoption for patients with endobronchial CISare not surgical candidates as the result of othercomorbidities. 98,99BrachytherapyBrachytherapy refers to endobronchial radiation,primarily used for the treatment of endobronchialmalignant airway obstruction. 100,101 The mostcommonly used source of radiation is 192 Ir, whichis delivered through a catheter inserted bronchoscopically.Brachytherapy may be delivered byeither low-dose rate, intermediate-dose rate, orhigh-dose rate (HDR) methods, with most authorscurrently recommending the afterloading HDRtechnique. This technique allows the bronchoscopistto place the catheter in the desired location andthe radiation oncologist to deliver the radiation ina protected environment. The main advantage ofHDR is patient convenience because each sessionlasts 30 min; however, multiple bronchoscopiesare required to achieve the recommended total1,500 cGy dose. The low-dose rate technique maybe appropriate for patients who live far from thehospital or who are otherwise hospitalized becauseit only requires one bronchoscopy; however, thecatheter has to stay in place for 20 to 60 h, whichis often poorly tolerated by the patient. The mainadvantage of brachytherapy as compared withexternal-beam radiation is the fact that less normaltissue is exposed to the toxic effects of radiation.The most common side effects include intoleranceof the catheter, radiation bronchitis, airway perforationand, occasionally, massive hemorrhage.The greatest incidence of hemorrhage occursduring the treatment of tumors in the right andleft upper lobes because they are located near thegreat vessels.Airway StentsMontgomery is credited as initiating the widespreaduse of airway stents after his developmentof a silicone T tube in 1965 for use in patients withtracheal stenosis. 102 Dumon, however, introduced thefirst completely endoluminal airway stent in 1990. 103Airway stents are the only technology that can alleviateextrinsic airway compression. They are commonlyused in conjunction with the other modalitiesfor patients with intrinsic or mixed disease.During the last 15 years, there has been anexplosion in both stent design and in the number ofendoscopists who place airway stents. As with anyprocedure, it is crucial to understand the indicationsand contraindications of the procedure as wellas be able to anticipate, prevent, and manage theassociated complications. Unfortunately, the idealstent has not yet been developed. This stent wouldbe easy to insert and remove, yet not migrate; ofsufficient strength to support the airway, yet beflexible enough to mimic normal airway physiologyand promote secretion clearance; biologically<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 211ACC-PULM-09-0302-013.indd 2117/10/09 8:25:49 PM


inert to minimize the formation of granulation tissue;and available in a variety of sizes.There are currently two main types of stents:metal, which generally is composed of NitinolSuper-Elastic Archwire (3M Unitek; Monrova,CA), and silicone. Although metal stents are easyto place, they can be extremely difficult to extract.They are available in covered (typically withsilicone or polyurethane) and uncovered varieties.For malignant airway obstruction, the only appropriatemetal stents are covered models, whichminimize tumor in-growth. Some authors believethat there is no indication for an uncovered metalstent. The overuse of metallic stents in patientswith nonmalignant airway obstruction has led tothe US Food and Drug Administration creatinga black-box warning recommending their use asa modality of last resort in these patients. 104 Themain advantage of metal stents is their larger internal/externaldiameter ratio as compared with siliconestents. Although silicone stents require rigidbronchoscopy for placement, they are more easilyremoved and are significantly less expensive.In addition to malignant airway obstruction,airway stents can be helpful in patients withtracheobronchomalacia and tracheoesophagealfistula. In patient with tracheoesophageal fistula,double stenting of the esophagus and airway isrecommend to maximally prevent aspiration. 105Future Directions in BronchoscopyBronchoscopic Lung Volume ReductionIn the patient with upper-lobe-predominantdisease and low exercise capacity, lung volumereduction surgery has been shown to improve bothquality and length of life. 106 The major drawback tosurgical lung volume reduction is its invasivenessand cost. As such, several techniques to achievelung volume reduction bronchoscopically arecurrently under investigation, although no suchdevice has been approved by the US Food andDrug Administration.Bronchial ThermoplastyBronchoscopic thermoplasty is a technique inwhich controlled thermal energy (radiofrequencyenergy at 65C) is applied directly to visible airwaysthrough the bronchoscope. The goal is to achieveablation of the airway smooth muscle withoutcausing scarring and stenosis. In a preliminarystudy, 107 the use of this technique in the treatmentof patients with mild-to-moderate asthma led toreduction in airway hyperresponsiveness, whichwas associated with improvements in daily symptomsand peak expiratory flow for a period of2 years. In a study 108 of 112 subjects in whom asthmacontrol was impaired, bronchial thermoplastyreduced the rate of mild exacerbations, and at12 months, there were significantly greater improvementsin the bronchial-thermoplasty group thanin the control group. Widespread clinical applicationof this technique requires validation by otherongoing multicenter studies.Enhanced NavigationThe use of electromagnetic navigation, especiallywhen used in conjunction with radial-probeEBUS, is a novel technology that has been shownto allow accurate sampling of peripheral solitarypulmonary nodules 2 cm in diameter. 109,110 Inbrief, a virtual bronchoscopy is generated from thepatient’s CT scan. Anatomic landmarks that areeasily identified, such as the carina, are marked,as is the target. At the time of bronchoscopy, thepatient lies in an electromagnetic field and a steerable,locatable guide is placed through an extendedworking channel of the bronchoscope. The locationof the guide in the electromagnetic field is accurateto 5 mm in the x, y, and z axes, as well as yaw,pitch, and roll. The points previously identified inthe virtual bronchoscopy are then marked with thelocatable guide that in essence fuses the CT scanwith the patient’s anatomy.Navigation is then performed by examiningthe CT in the axial, saggital, and coronal planes,and the guide is steered toward the target. Oncefound, the guide is removed, and standard bronchoscopictools such as TBNA needles, forceps,and brushes are placed through the extendedworking channel. This technology can be used forboth diagnostic bronchoscopy as well as therapeuticmodalities. For example, if a patient witha 2.5-cm, stage Ia non-small cell cancer is not anoperative candidate, this technology may allow forbronchoscopic treatment by either radiofrequencyablation or the implantation of fiducials to allow212 Bronchoscopy and Interventional Pulmonology (Feller-Kopman)ACC-PULM-09-0302-013.indd 2127/10/09 8:25:49 PM


stereotactic radiosurgery. 111 Other techniques ofvirtual bronchoscopic navigation are also beingdeveloped that do not require the presence of anelectromagnetic field and rely on high-definitionvirtual bronchoscopic road maps with ultrathinbronchoscopes. 112–114Other Interventional <strong>Pulmonary</strong>ProceduresPleural InterventionsUltrasound has revolutionized the evaluationand management of patients with pleuraldisease. 115 Although randomized trials have notbeen performed, the use of ultrasound appears tobe associated with an increased success as well asa reduction in the pneumothorax rate for patientsundergoing thoracentesis. Ultrasound allows forthe precise placement of small-bore chest tubesin loculated effusions and has been shown to beextremely accurate in ruling out pneumothorax. 116Medical thoracoscopy describes the insertionof a semirigid or rigid thoracoscope with the intentof draining pleural fluid, obtaining biopsies ofthe parietal pleura, and/or instilling an agent toachieve pleurodesis. 66,117,118 Whereas video-assistedthoracic surgery is typically performed in the operatingroom with general anesthesia and lung isolationwith two or three ports, medical thoracoscopyis commonly performed in an endoscopy suite withmoderate sedation on a spontaneously breathingpatient with one access port. Common indicationsfor medical thoracoscopy include a recurrent exudativeeffusion with no clear diagnosis or a knownmalignant effusion requiring pleurodesis.Indwelling, tunneled pleural catheters are anexcellent treatment option for patients with dyspneacaused by a recurrent malignant pleural effusion.Although the treatment of choice for patientswith lung entrapment, they may also be used forpatients with expandable lung. Pleural palliationcan be achieved in close to 90% of patients, andside effects are relatively few. 119Additional Airway ProceduresMany interventional pulmonologists routinelyperform percutaneous dilational tracheostomy. Ascompared with open tracheostomy, percutaneoustracheostomy is at least as safe, and it may beassociated with reduced infectious and bleedingcomplications. It is also less expensive and doesnot require transport of a critically ill patient tothe operating room. 120ComplicationsBronchoscopy is generally considered a relativelysafe procedure. 5 However, it is common thatpatients in whom bronchoscopy is performed canhave multiple comorbidities, and therefore, the bronchoscopistneeds to be aware of the patient’s medicalhistory, including medications and allergies. Additionally,the bronchoscopist needs to have a thoroughunderstanding of the effects and side effects of themedications he or she will use to achieve adequatetopical anesthesia and moderate sedation and to beskilled in airway management should the patientexperience hypoxemia/hypercapnia.Transient fever is the most common adverseevent after BAL is performed, and it can be seen inup to one-third of patients. Major adverse eventsincluding hypoxemia, pneumothorax, bleeding,and death may also occur, and both patient andprocedure-related factors are important determinants.The British Thoracic Society has publishedtheir recommendations concerning the performanceof diagnostic flexible bronchoscopy. Theserecommendations include steps that should betaken to minimize complications. 5 Although stillrelatively safe, complications associated with themore advanced procedures including rigid bronchoscopy,pleuroscopy/thoracoscopy, indwellingpleural catheters, and percutaneous tracheostomyare slightly greater as a result of the increased acuityof the patient population as well as the invasivenessof the interventions. 8,9TrainingFlexible bronchoscopy is one of the most commonlyperformed procedures by the pulmonologist.Although 96% of fellowship programs offerthe recommended number of diagnostic flexiblebronchoscopies, the percentage of programsachieving the recommended number was only 69%for TBNA and decreased dramatically for moreadvanced procedures. 83 As such, recommendationsare forthcoming that a year of advanced training<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 213ACC-PULM-09-0302-013.indd 2137/10/09 8:25:49 PM


should be required for the majority of proceduresconsidered essential to the interventional pulmonologists.Clearly, it is not necessary to obtain aformal additional year of training to become anexpert in any one technique, such as EBUS. To fullyunderstand the diagnostic approach and availabletherapeutic modalities as well as obtain the necessaryskills to anticipate, prevent, and managecomplications in patients with complex centralairway or pleural disease, it is best to obtain formaltraining in the subspecialty of interventionalpulmonology.References1. Zollner F. Gustav Killian, father of bronchoscopy.Arch Otolaryngol 1965; 82:656–6592. Killian G. Ueber direkte Bronchoskopie. MuenchMed Wochen 1898; 27:844–8473. Becker HD, Marsh BR. History of the rigid bronchoscope.In: Bolliger CT, Mathur PN, eds. Interventionalbronchoscopy. Basel: Karger, 2000; 2–154. Miyazawa T. History of the flexible bronchoscope.In: Bolliger CT, Mathur PN, eds. Interventionalbronchoscopy. Basel: Karger, 2000; 16–215. British Thoracic Society guidelines on diagnosticflexible bronchoscopy. Thorax 2001; 56:i1–i216. Ernst A, Feller-Kopman D, Becker HD, et al. Centralairway obstruction. Am J Respir Crit Care Med2004; 169:1278–12977. Wahidi MM, Herth FJF, Ernst A. State of the art:interventional pulmonology. Chest 2007; 131:261–2748. Bolliger CT, Mathur PN, Beamis JF, et al. ERS/ATS statement on interventional pulmonology:European Respiratory Society/American ThoracicSociety. Eur Respir J 2002; 19:356–3739. Ernst A, Silvestri GA, Johnstone D. Interventionalpulmonary procedures: guidelines from theAmerican College of Chest Physicians. Chest 2003;123:1693–171710. Cortese DA, Prakash UBS. Anatomy for the bronchoscopist.In: Prakash UBS, ed. Bronchoscopy.New York: Raven Press, 1994; 13–4211. Becker HD. Atlas of bronchoscopy: technique,diagnosis, differential diagnosis, therapy. Philadelphia;BC Decker, 199112. Jackson CL, Huber JF. Correlated anatomy of thebronchial tree and lungs with a system of nomenclature.Dis Chest 1943; 9:319–32613. Boyden EA. Segmental anatomy of the lung. NewYork, McGraw Hill, 195514. Ost D, DeRosiers A, Britt EJ, et al. Assessment ofa bronchoscopy simulator. Am J Respir Crit CareMed 2001; 164:2248–225515. Fagon JY, Chastre J, Wolff M, et al. Invasive andnoninvasive strategies for management of suspectedventilator-associated pneumonia: a randomizedtrial. Ann Intern Med 2000; 132:621–63016. Heyland D, Cook D, Dodek P, et al. A randomizedtrial of diagnostic techniques for ventilator-associatedpneumonia. N Engl J Med 2006; 355:2619–263017. American Cancer Society. Cancer facts and figures2008. Available at: http://www.cancer.org/downloads/STT/2008CAFFfinalsecured.pdf.AccessedApril 22, 200918. National Cancer Institute. SEER cancer statisticsreview 1975–2005. Available at: http://seer.cancer.gov/csr/1975_2005/results_merged/sect_15_lung_bronchus.pdf. Accessed April 22, 200919. Mazzone P, Jain P, Arroliga AC, et al. Bronchoscopyand needle biopsy techniques for diagnosisand staging of lung cancer. Clin Chest Med 2002;23:137–58, ix.20. Arroliga AC, Matthay RA. The role of bronchoscopyin lung cancer. Clin Chest Med 1993; 14:87–9821. Hyer JD, Silvestri G. Diagnosis and staging of lungcancer. Clin Chest Med 2000; 21:95-ix22. Rafanan AL, Mehta A. Role of bronchoscopy inlung cancer. Semin Respir Crit Care Med 2000;21:405–42023. Wang KP. Staging of bronchogenic carcinoma bybronchoscopy. Chest 1994; 106:588–59324. Hooper RG, Jackson FN. Endobronchial electrocautery.Chest 1985; 87:712–71425. Saita S, Tanzillo A, Riscica C, et al. Bronchial brushingand biopsy: a comparative evaluation in diagnosingvisible bronchial lesions. Eur J CardiothoracSurg 1990; 4:270–27226. Mak VH, Johnston ID, Hetzel MR, et al. Value ofwashings and brushings at fibreoptic bronchoscopyin the diagnosis of lung cancer. Thorax 1990;45:373–376214 Bronchoscopy and Interventional Pulmonology (Feller-Kopman)ACC-PULM-09-0302-013.indd 2147/10/09 8:25:50 PM


27. Dasgupta A, Jain P, Minai OA, et al. Utility of transbronchialneedle aspiration in the diagnosis ofendobronchial lesions. Chest 1999; 115:1237–124128. Anzueto A, Levine SM, Jenkinson SG. The techniqueof bronchoalveolar lavage: a guide to sampling theterminal airways and alveolar space. J Crit Illn 1992;7:1817–182429. Kvale PA. Bronchoscopic biopsies and bronchoalveolarlavage. Chest Surg Clin N Am 1996; 6:205–22230. Helmers RA, Pisani RJ. Bronchoalveolar lavage.In: Prakash UB, ed. Bronchoscopy. New York, NY:Raven Press, 1994; 155–18231. Baughman RP. Technical aspects of bronchoalvoelarlavage: recommendations for a standardprocedure. Semin Respir Crit Care Med 2007;28:475–48532. Hunninghake GW, Gadek JE, Kawanami O, et al.Inflammatory and immune processes in the humanlung in health and disease: evaluation by bronchoalveolarlavage. Am J Pathol 1979; 97:149–20633. Goldstein RA, Rohatgi PK, Bergofsky EH, et al.Clinical role of bronchoalveolar lavage in adultswith pulmonary disease. Am Rev Respir Dis 1990;142:481–48634. Reynolds HY. Bronchoalveolar lavage. Am RevRespir Dis 1987; 135:250–26335. Bilaceroglu S, Kumcuoglu Z, Alper H, et al. CTbronchus sign-guided bronchoscopic multiple diagnosticprocedures in carcinomatous solitary pulmonarynodules and masses. Respiration 1998; 65:49–5536. Shiner RJ, Rosenman J, Katz I, et al. Bronchoscopicevaluation of peripheral lung tumours. Thorax1988; 43:887–88937. Kawaraya M, Gemba K, Ueoka H, et al. Evaluationof various cytological examinations by bronchoscopyin the diagnosis of peripheral lung cancer. BrJ Cancer 2003; 89:1885–188838. Katis K, Inglesos E, Zachariadis E, et al. The role oftransbronchial needle aspiration in the diagnosisof peripheral lung masses or nodules. Eur Respir J1995; 8:963–96639. Shure D, Fedullo PF. Transbronchial needle aspirationof peripheral masses. Am Rev Respir Dis 1983;128:1090–109240. Prakash UB, Offord KP, Stubbs SE. Bronchoscopyin North America: the <strong>ACCP</strong> survey. Chest 1991;100:1668–167541. Haponik EF, Shure D. Underutilization of transbronchialneedle aspiration: experiences of currentpulmonary fellows. Chest 1997; 112:251–25342. Harrow EM, Abi-Saleh W, Blum J, et al. The utilityof transbronchial needle aspiration in the stagingof bronchogenic carcinoma. Am J Respir Crit CareMed 2000; 161:601–60743. Salathe M, Soler M, Bolliger CT, et al. Transbronchialneedle aspiration in routine fiberoptic bronchoscopy.Respiration 1992; 59:5–844. Sharafkhaneh A, Baaklini W, Gorin AB, et al. Yieldof transbronchial needle aspiration in diagnosis ofmediastinal lesions. Chest 2003; 124:2131–213545. Garpestad E, Goldberg SN, Herth F, et al. CT fluoroscopyguidance for transbronchial needle aspiration:an experience in 35 patients. Chest 2001;119:329–33246. Herth FJ, Becker HD, Ernst A. Ultrasound-guidedtransbronchial needle aspiration: an experience in242 patients. Chest 2003; 123:604–60747. Herth F, Becker HD, Ernst A. Conventional vsendobronchial ultrasound-guided transbronchialneedle aspiration: a randomized trial. Chest 2004;125:322–32548. Herth F, Ernst A, Schulz M, et al. Endobronchialultrasound reliably differentiates between airwayinfiltration and compression by tumor. Chest 2003;123:458–46249. Kurimoto N, Murayama M, Yoshioka S, et al.Analysis of the internal structure of peripheralpulmonary lesions using endobronchial ultrasonography.Chest 2002; 122:1887–189450. Herth FJ, Ernst A, Becker HD. Endobronchial ultrasound-guidedtransbronchial lung biopsy in solitarypulmonary nodules and peripheral lesions.Eur Respir J 2002; 20:972–97451. Yasufuku K, Chiyo M, Sekine Y, et al. Real-timeendobronchial ultrasound-guided transbronchialneedle aspiration of mediastinal and hilar lymphnodes. Chest 2004; 126:122–12852. Yasufuku K, Nakajima T, Motoori K, et al. Comparisonof endobronchial ultrasound, positronemission tomography, and CT for lymph nodestaging of lung cancer. Chest 2006; 130:710–71853. Herth FJ, Eberhardt R, Vilmann P, et al. Real-timeendobronchial ultrasound guided transbronchialneedle aspiration for sampling mediastinal lymphnodes. Thorax 2006; 61:795–798<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 215ACC-PULM-09-0302-013.indd 2157/10/09 8:25:50 PM


54. Detterbeck FC, Jantz MA, Wallace M, et al. Invasivemediastinal staging of lung cancer: <strong>ACCP</strong>evidence-based clinical practice guidelines (2ndedition). Chest 2007; 132(3 suppl):202S–220S54a. Lam S, Kennedy T, Unger M, Miller YE, GelmontD, Rusch V et al. Localization of bronchialintraepithelial neoplastic lesions by fluorescencebronchoscopy. Chest 1998;113(3):696–70255b.Weigel TL, Yousem S, Dacic S, Kosco PJ, SiegfriedJ, Luketich JD. Fluorescence bronchoscopicsurveillance after curative surgical resectionfor non-small-cell lung cancer. Ann Surg Oncol2000;7(3):176–18055. Haussinger K, Becker H, Stanzel F, et al. Autofluorescencebronchoscopy with white light bronchoscopycompared with white light bronchoscopyalone for the detection of precancerous lesions: aEuropean randomised controlled multicentre trial.Thorax 2005; 60:496–50356. Poletti V, Chilosi M, Olivieri D. Diagnostic invasiveprocedures in diffuse infiltrative lung diseases.Respiration 2004; 71:107–11957. Ensminger SA, Prakash UB. Is bronchoscopic lungbiopsy helpful in the management of patients withdiffuse lung disease? Eur Respir J 2006; 28:1081–108458. Gilman MJ, Wang KP. Transbronchial lung biopsyin sarcoidosis: an approach to determine the optimalnumber of biopsies. Am Rev Respir Dis 1980;122:721–72459. Baselski VS, Wunderink RG. Bronchoscopic diagnosisof pneumonia. Clin Microbiol Rev 1994;7:533–55860. van der Eerden MM, Vlaspolder F, de Graaff CS,et al. Value of intensive diagnostic microbiologicalinvestigation in low- and high-risk patientswith community-acquired pneumonia. Eur J ClinMicrobiol Infect Dis 2005; 24:241–24961. Hohenthal U, Sipila J, Vainionpaa R, et al. Diagnosticvalue of bronchoalveolar lavage in communityacquiredpneumonia in a routine setting: a studyon patients treated in a Finnish university hospital.Scand J Infect Dis 2004; 36:198–20362. Feinsilver SH, Fein AM, Niederman MS, et al. Utilityof fiberoptic bronchoscopy in nonresolvingpneumonia. Chest 1990; 98:1322–132663. Guidelines for the management of adults with hospital-acquired,ventilator-associated, and healthcare-associatedpneumonia. Am J Respir Crit CareMed 2005; 171:388–41664. Campbell JH, Blessing N, Burnett AK, et al. Investigationand management of pulmonary infiltratesfollowing bone marrow transplantation: an eightyear review. Thorax 1993; 48:1248–125165. Rano A, Agusti C, Jimenez P, et al. <strong>Pulmonary</strong> infiltratesin non-HIV immunocompromised patients: adiagnostic approach using non-invasive and bronchoscopicprocedures. Thorax 2001; 56:379–38766. Hersh CP, Feller-Kopman D, Wahidi M, et al. Ultrasoundguidance for medical thoracoscopy: a novelapproach. Respiration 2003; 70:299–30167. Abramson MJ, Stone CA, Holmes PW, et al. Therole of bronchoalveolar lavage in the diagnosis ofsuspected opportunistic pneumonia. Aust N Z JMed 1987; 17:407–41268. Stover DE, Zaman MB, Hajdu SI, et al. Bronchoalveolarlavage in the diagnosis of diffuse pulmonaryinfiltrates in the immunosuppressed host.Ann Intern Med 1984; 101:1–769. Hsu AA, Allen DM, Yeo CT, et al. Bronchoscopyin immunocompromised host with pulmonaryinfiltrates. Ann Acad Med Singapore 1996; 25:797–80370. Hohenadel IA, Kiworr M, Genitsariotis R, et al.Role of bronchoalveolar lavage in immunocompromisedpatients with pneumonia treated with abroad spectrum antibiotic and antifungal regimen.Thorax 2001; 56:115–12071. White P, Bonacum JT, Miller CB. Utility of fiberopticbronchoscopy in bone marrow transplant patients.Bone Marrow Transplant 1997; 20:681–68772. Dunagan DP, Baker AM, Hurd DD, et al. Bronchoscopicevaluation of pulmonary infiltrates followingbone marrow transplantation. Chest 1997;111:135–14173. White DA, Wong PW, Downey R. The utility ofopen lung biopsy in patients with hematologicmalignancies. Am J Respir Crit Care Med 2000;161:723–72974. Hirshberg B, Biran I, Glazer M, et al. Hemoptysis:etiology, evaluation, and outcome in a tertiaryreferral hospital. Chest 1997; 112:440–44475. Jackson CV, Savage PJ, Quinn DL. Role of fiberopticbronchoscopy in patients with hemoptysisand a normal chest roentgenogram. Chest 1985;87:142–14476. Poe RH, Israel RH, Marin MG, et al. Utility of fiberopticbronchoscopy in patients with hemoptysisand a nonlocalizing chest roentgenogram. Chest1988; 93:70–75216 Bronchoscopy and Interventional Pulmonology (Feller-Kopman)ACC-PULM-09-0302-013.indd 2167/10/09 8:25:50 PM


77. Freimanis AS. Ultrasound and thoracentesis [letter].JAMA 1977; 238:163178. Gong H Jr., Salvatierra C. Clinical efficacy of earlyand delayed fiberoptic bronchoscopy in patientswith hemoptysis. Am Rev Respir Dis 1981;124:221–22579. McGuinness G, Beacher JR, Harkin TJ, et al.Hemoptysis: prospective high-resolution CT/bronchoscopic correlation. Chest 1994; 105:1155–116280. Ayers ML, Beamis JF Jr. Rigid bronchoscopy inthe twenty-first century. Clin Chest Med 2001;22:355–36481. Helmers RA, Sanderson DR. Rigid bronchoscopy:the forgotten art. Clin Chest Med 1995; 16:393–39982. Colt HG, Prakash UBS, Offord KP. Bronchoscopyin North America: survey by the American Associationfor Bronchology, 1999. J Bronchol 2000;7:8–2583. Pastis NJ, Nietert PJ, Silvestri GA. Variation intraining for interventional pulmonary proceduresamong US pulmonary/critical care fellowships:a survey of fellowship directors. Chest 2005;127:1614–162184. Dweik RA, Stoller JK. Role of bronchoscopy inmassive hemoptysis. Clin Chest Med 1999; 20:89–10585. Cahill BC, Ingbar DH. Massive hemoptysis:assessment and management. Clin Chest Med1994; 15:147–16786. Jean-Baptiste E. Clinical assessment and managementof massive hemoptysis. Crit Care Med 2000;28:1642–164787. Jougon J, Ballester M, Delcambre F, et al. Massivehemoptysis: what place for medical and surgicaltreatment. Eur J Cardiothorac Surg 2002; 22:345–35188. Karmy-Jones R, Cuschieri J, Vallieres E. Role ofbronchoscopy in massive hemoptysis. Chest SurgClin N Am 2001; 11:873–90689. Bhattacharyya P, Dutta A, Samanta AN, et al. Newprocedure: bronchoscopic endobronchial sealing:a new mode of managing hemoptysis. Chest 2002;121:2066–206990. Conlan AA, Hurwitz SS. Management of massivehaemoptysis with the rigid bronchoscope andcold saline lavage. Thorax 1980; 35:901–90491. Dupree HJ, Lewejohann JC, Gleiss J, et al. Fiberopticbronchoscopy of intubated patients withlife-threatening hemoptysis. World J Surg 2001;25:104–10792. Feller-Kopman D, Lukanich JM, Shapira G, et al.Gas flow during bronchoscopic ablation therapycauses gas emboli to the heart: a comparative animalstudy. Chest 2008; 133:892–89693. Cavaliere S, Foccoli PO, Toninelli C, et al. Nd:YAG laser in lung cancer: an 11 year experiencewith 2,253 applications in 1,585 patients. J Bronchol1994; 1:105–11194. Mehta AC, Lee F, DeBoer G. Flexible bronchoscopyand the use of lasers. In: Wang KP, MehtaAC, eds. Flexible bronchoscopy. Oxford, MA:Blackwell Science, 1995; 247–27495. Maiwand MO, Mathur PN. Endobronchialcryotherapy. Semin Respir Crit Care Med 1997;18:545–55496. Maiwand MO, Homasson JP. Cryotherapy fortracheobronchial disorders. Clin Chest Med 1995;16:427–44397. Hooper RG, Jackson FN. Endobronchial electrocautery.Chest 1988; 94:595–59898. Moghissi K, Dixon K, Stringer M, et al. The placeof bronchoscopic photodynamic therapy inadvanced unresectable lung cancer: experience of100 cases. Eur J Cardiothorac Surg 1999; 15:1–699. Cortese DA, Edell ES, Kinsey JH. Photodynamictherapy for early stage squamous cell carcinomaof the lung. Mayo Clin Proc 1997; 72:595–602100. Macha HN, Wahlers B, Reichle C, et al. Endobronchialradiation therapy for obstructing malignancies:ten years’ experience with iridium-192high-dose radiation brachytherapy afterloadingtechnique in 365 patients. Lung 1995; 173:271–280101. Kelly JF, Delclos ME, Morice RC, et al. High-doserateendobronchial brachytherapy effectivelypalliates symptoms due to airway tumors: the 10-year M. D. Anderson Cancer Center experience.Int J Radiat Oncol Biol Phys 2000; 48:697–702102. Montgomery WW. T-tube tracheal stent. ArchOtolaryngol 1965; 82:320–321103. Dumon JF. A dedicated tracheobronchial stent.Chest 1990; 97:328–332104. FDA public health notification: complicationsfrom metallic tracheal stents in patients withbenign airway disorders. Available at: www.fda.gov/cdrh/safety/072905-tracheal.html. AccessedApril 22, 2009105. Freitag L, Tekolf E, Steveling H, et al. Managementof malignant esophagotracheal fistulas withairway stenting and double stenting. Chest 1996;110:1155–1160<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 217ACC-PULM-09-0302-013.indd 2177/10/09 8:25:50 PM


106. Fishman A, Martinez F, Naunheim K, et al. A randomizedtrial comparing lung-volume-reductionsurgery with medical therapy for severe emphysema.N Engl J Med 2003; 348:2059–2073107. Cox G, Miller JD, McWilliams A, et al. Bronchialthermoplasty for asthma. Am J Respir Crit CareMed 2006; 173:965–969108. Cox G, Thomson NC, Rubin AS, et al. Asthmacontrol during the tear after bronchial thermoplasty.N Engl J Med 2007; 356:1327–1337109. Eberhardt R, Anantham D, Herth F, et al. Electromagneticnavigation diagnostic bronchoscopy inperipheral lung lesions. Chest 2007; 132:930–935110. Eberhardt R, Anantham D, Ernst A, et al. Multimodalitybronchoscopic diagnosis of peripherallung lesions: a randomized controlled trial. Am JRespir Crit Care Med 2007; 176:36–41111. Anantham D, Feller-Kopman D, ShanmughamLN, et al. Electromagnetic navigation bronchoscopy-guidedfiducial placement for robotic stereotacticradiosurgery of lung tumors: a feasibilitystudy. Chest 2007; 132:930–935112. Hofmeister CC, Czerlanis C, Forsythe S, et al.Retrospective utility of bronchoscopy after hematopoieticstem cell transplant. Bone MarrowTransplant 2006; 38:693–698113. Asano F, Matsuno Y, Tsuzuku A, et al. Diagnosisof peripheral pulmonary lesions using a bronchoscopeinsertion guidance system combined withendobronchial ultrasonography with a guidesheath. Lung Cancer 2008; 60:366–373114. Shinagawa N, Yamazaki K, Onodera Y, et al. Virtualbronchoscopic navigation system shortensthe examination time: feasibility study of virtualbronchoscopic navigation system. Lung Cancer2007; 56:201–206115. Feller-Kopman D. Ultrasound-guided thoracentesis.Chest 2006; 129:1709–1714116. Lichtenstein D, Meziere G, Biderman P, et al. Thecomet-tail artifact: an ultrasound sign ruling outpneumothorax. Intensive Care Med 1999; 25:383–388117. Colt HG. Therapeutic thoracoscopy. Clin ChestMed 1998; 19:383–394118. Lee P, Colt HG. State of the art: pleuroscopy. JThorac Oncol 2007; 2:663–670119. Tremblay A, Michaud G. Single-center experiencewith 250 tunnelled pleural catheter insertions formalignant pleural effusion. Chest 2006; 129:362–368120. De LP, Bedert L, Delcroix M, et al. Tracheotomy:clinical review and guidelines. Eur J CardiothoracSurg 2007; 32:412–421218 Bronchoscopy and Interventional Pulmonology (Feller-Kopman)ACC-PULM-09-0302-013.indd 2187/10/09 8:25:51 PM


<strong>Pulmonary</strong> Complications of HIV InfectionMark J. Rosen, MD, FCCPObjectives:• <strong>Review</strong> the types of pulmonary disorders that occur inpatients with HIV infection• Understand how demographic factors, the degreeof immunosuppression, and the application of anti-Pneumocystis prophylaxis and antiretroviral therapyinfluence the risk of the development of different types ofpulmonary disorders• Summarize the clinical, radiographic, and laboratoryfeatures of infections, neoplasms, and other pulmonarydisorders in patients with HIV infection• Discuss the diagnostic evaluation and treatment of HIVassociatedpulmonary disordersKey words: bacterial pneumonia; cytomegalovirus; fungi;HIV; Kaposi sarcoma; mycobacteria; Pneumocystis jirovecipneumonia; pulmonary diseasesFor the last two decades, the AIDS pandemichas remained the most serious infectious globalhealth problem. In 2007, every day, 7,400 personsbecome infected with HIV, and 5,400 diefrom AIDS; an estimated 33.2 million people areliving with HIV infection worldwide. Aroundthe world, the pandemic seems to have leveledoff as the result of intensive public health effortsat prevention; in developed nations, the rates ofopportunistic infections and death in personswith AIDS have decreased dramatically since 1995because of advances in the treatment of HIV. Thereis now a general perception that AIDS may beconsidered to be not uniformly fatal but a chronicand controllable disorder. By the end of 2007, 1million cases of AIDS had been reported in theUnited States, of whom approximately 583,000have died. Approximately 56,000 Americans areinfected with HIV each year; in 2006, persons 13to 29 years of age accounted for approximatelyone third of new HIV infections. African Americanscomprised approximately half of the newcases in the past few years. At the end of 2007,an estimated 732,000 people in the United Stateswere living with HIV infection, of whom 468,000had AIDS. However, the US death rate in personswith AIDS decreased 17% from 2004 to 2007.In many urban communities, AIDS-related diseasesare still the leading causes of death amongyoung adults.The lung is a frequent site of opportunisticinfection in immunocompromised patients, andnoninfectious pulmonary disorders associatedwith HIV infection and antiretroviral treatmentsare increasingly common. Table 1 lists the infectious,neoplastic, and inflammatory diseases thatTable 1. HIV-Associated Respiratory DisordersBacterial pneumoniaS pneumoniaeH influenzaeP aeruginosaStaphylococcus aureusMoraxella catarrhalisR equiM tuberculosisMACOther nontuberculous mycobacteriaFungal infectionsP jiroveciC neoformansH capsulatumAspergillus fumigatusCoccidioides immitisBlastomyces dermatitidisProtozoal infectionsStrongyloides stercoralisToxoplasma gondiiViral infectionsCMVAdenovirusHerpes simplexMalignanciesKSNon-Hodgkin lymphoma (NHL)Primary effusion lymphoma (PEL)Carcinoma of the lungOther disordersSinusitisBronchitisBronchiectasisEmphysemaLymphocytic interstitial pneumoniaNonspecific interstitial pneumoniaCryptogenic organizing pneumoniaPrimary pulmonary hypertensionImmune reconstitution inflammatory syndrome<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 219ACC-PULM-09-0302-014.indd 2197/10/09 8:26:54 PM


occur in patients with HIV infection, and theirtypical radiographic patterns are summarized inTable 2.HIV-Associated <strong>Pulmonary</strong> DisordersThe pulmonary disorders associated with HIVinfection range from asymptomatic and mildabnormalities in pulmonary function to fulminatingopportunistic infections. The risk for the developmentof each of these disorders is stronglyinfluenced by the severity of immunosuppression,the patient’s demographic characteristics, the placeof current or previous residence, and whether thepatient is using prophylaxis against commonHIV-associated infections. Genetic factors probablyalso are important but have been less preciselydefined.Influence of Immune FunctionThe severity of the abnormality in host defenseis the primary determinant of the risk of thedevelopment of specific pulmonary disorders.Early in the course of HIV infection, when theimmune system is not severely compromised,respiratory disorders occur that are generallysimilar to those found in the general population.Opportunistic infections occur only with severeimmunodeficiency.The CD4 lymphocyte count is still the mostreliable surrogate marker for immune function, therisk of opportunistic infection, and the risk of progressionof HIV disease. The measurement of HIVactivity with serum HIV RNA (called the viral load)is used routinely to assess the response to treatmentwith antiretroviral agents and to stratify patientsby risk of the progression of disease. However, itis a less reliable predictor than CD4 count indetermining the risk of the development of specificHIV-related diseases. In a survey of the medicalrecords of 18,000 HIV-infected subjects whoreceived care in 100 sites in 10 US cities, commondisorders such as sinusitis, bronchitis, and pharyngitisoccurred at all strata of CD4 cell counts(Table 3). With lower counts, different pulmonaryinfections occurred with increasing frequency.Some opportunistic infections tend to occur onlyTable 2. HIV Infection: Chest Radiographic Patterns and Common EtiologiesRadiographic PatternsEtiologiesFocal opacitiesMediastinal lymphadenopathyBacteria M tuberculosisM tuberculosis MACP jiroveci KSFungi LymphomaFungiImmune reconstitutionDiffuse opacitiesPleural effusionP jiroveci BacteriaM tuberculosis M tuberculosisKS KSBacteria Lymphoma (including primary effusion lymphoma)FungiCMV CardiomyopathyImmune reconstitution HypoproteinemiaImmune reconstitutionDiffuse nodules/cavitiesKS (large nodules) M tuberculosis (high CD4 + count)M tuberculosis (miliary) P jiroveci (low CD4 + count)Fungi (small nodules) P aeruginosa (low CD4 + count)P jiroveci (small nodules) FungiR equiPneumothoraxP jiroveci220 <strong>Pulmonary</strong> Complications of HIV Infection (Rosen)ACC-PULM-09-0302-014.indd 2207/10/09 8:26:55 PM


Table 3. Select Disease Conditions Listed by Usual CD4 atTime of Diagnosis*CD4 500 cells/µLSinusitis/mastoiditis/otitisBronchitisPharyngitisLung cancerCD4 400 cells/µLBacterial pneumonia<strong>Pulmonary</strong> M tuberculosisCardiomyopathyCD4 200 cells/µLPCPKSBacterial sepsisDisseminated M tuberculosisCD4 100 cells/µLDisseminated MACCMV diseaseDisseminated fungal infections*Adapted from Hanson et al.with severe immunodeficiency (ie, CD4 count, 100 cells/µL), including disseminated nontuberculousmycobacterioses, disseminated fungal infections,CNS toxoplasmosis, and cytomegalovirus(CMV) disease. Therefore, common respiratoryproblems such as sinusitis and bronchitis mayoccur at any CD4 count, and bacterial pneumoniaand tuberculosis (TB) often occur before AIDSdefiningopportunistic infections and neoplasms.Decreasing immune function increases the risk forall HIV-associated respiratory diseases.Demographic FactorsThe increasing proportion of injection drugusers in the HIV-infected population was accompaniedby the recognition of bacterial pneumonia andTB as important HIV-associated infections. Bacterialpneumonia occurs more frequently than Pneumocystispneumonia (PCP) in all HIV-infected groups,and the risk of bacterial pneumonia is significantlygreater in injection drug users than in other personswith HIV infection. Race and ethnicity may alsoinfluence the risk of the development of bacterialpneumonia and TB, but these associations are confoundedby differences in access to health care, thegreater prevalence of TB in minority communities,and the disproportionately high numbers of injectiondrug users who are black or Hispanic.ResidenceThe place of residence strongly influences therisk of the development of specific infections. Thehigh incidence of PCP in the United States andEurope contrasts sharply with that in Africa, whereit is much less common. It is likely that both geneticand environmental factors account for the lowerincidence of PCP in Africa. In the United States,the incidence of HIV-associated TB is greatest inthe Northeast. The geographic distribution ofendemic fungi is a strong determinant of the riskof those infections developing; disseminated histoplasmosisand coccidioidomycosis are commonin patients with AIDS who live in endemic areas.These infections also may occur as a reactivationdisease after HIV-infected persons move to otherareas and immunocompromise develops.Use of Prophylaxis and AntiretroviralTherapyThe risk of specific opportunistic infectionsdecreases with the use of prophylaxis. Even beforethe availability of combination active antiretroviraltherapy (ART), the incidence of PCP and TB, alongwith their attendant mortality rates, was decreasing,partly attributable to the use of prophylaxis bysusceptible persons. After the widespread use ofART, beginning around 1996, the incidence ofopportunistic infections and death among HIVinfectedpersons decreased dramatically becausesuccessful ART inhibits viral replication andrestores immune function. However, immunereconstitution is associated with the occurrence ofother clinical syndromes.Bacterial PneumoniaHIV infection impairs humoral immunitythrough quantitative and functional defects inCD4 lymphocytes, which in turn increases the riskof bacterial infection, including sinusitis and pneumonia,and the risk of pneumonia developingincreases as the CD4 lymphocyte count decreases.Injection drug users are at a greater risk than othergroups, and neutropenia is an independent riskfactor. Anti-Pneumocystis prophylaxis with trimethoprim(TMP)-sulfamethoxazole (SMX) appearsto reduce the risk of bacterial pneumonia. Bacterial<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 221ACC-PULM-09-0302-014.indd 2217/10/09 8:26:55 PM


pneumonia also may accelerate the course of HIVdisease because it is an independent predictor ofprogression to AIDS and mortality.Streptococcus pneumoniae and Haemophilus influenzaeare the most frequent bacterial pathogens(Table 1). Pneumonia caused by Mycoplasma,Legionella, and Chlamydia is described but seemsto be relatively uncommon, especially in patientswith severe immunosuppression. Rhodococcus equi,an aerobic Gram-positive acid-fast bacillus, maycause focal consolidation, endobronchial disease,and cavitation, usually in patients with advancedHIV disease. Pneumonia caused by Pseudomonasaeruginosa may develop in patients with very lowCD4 lymphocyte counts (typically 50 cells/µL),even in the absence of risk factors such as neutropenia,corticosteroid use, and hospital-acquiredinfection. Nocardia asteroides may cause nodules,consolidation, cavitation, pleural effusions, empyema,and intrathoracic lymphadenopathy in HIVinfectedpersons.The diagnosis and treatment of bacterial pneumoniain persons with HIV infection is in almostall respects the same as that for patients who arenot infected. Patients usually present with fever,chills, productive cough, and localized areas ofconsolidation on chest radiographs. Although thisclinical picture strongly suggests bacterial pneumonia,it may also occur with TB and fungal infection.Conversely, patients with bacterial pneumoniamay have diffuse pulmonary opacities that resemblePCP.Polyvalent pneumococcal vaccine is recommendedfor all HIV-infected people, although thosewith low CD4 counts are less likely to mount anadequate antibody response. A vaccine against Hinfluenzae type B is available, but its use in patientswith HIV infection is limited because mostinfections are caused by other strains. Althoughadministration of the influenza vaccine is alsoreco mmended, there are no data indicating thatpatients with HIV infection are at an increased riskof contracting influenza or that the illness is moresevere than in the general population.Pneumocystis jiroveci PneumoniaPneumonia caused by P jiroveci (formerly classifiedas P carinii) was the first opportunistic infectiondescribed in patients with AIDS and hasalways been a major cause of illness and death. Theterm PCP has been used for decades, and ratherthan changing our terminology to PJP to reflect thenew nomenclature, a consensus exists that PCP beused to refer to the term PCP. Once thought to bea parasite, genomic analysis revealed that P jiroveciis in fact a fungus that infects only humans,whereas P carinii is pathogenic only in immunodeficientrats. The organism cannot be cultured reliablyoutside the lung, and its source is still notidentified; therefore, the precise route of transmissionis elusive.Although ART and effective prophylaxis forPCP have existed for years, this infection stilloccurs frequently for the following reasons: manypatients do not know that they have HIV infectionuntil an opportunistic infection develops, otherpatients know that they have HIV but are notreceiving medical care, and some patients arereceiving care and are not prescribed PCP prophylaxisor ART. Adherence to complex regimens withintolerable side effects often is problematic, andthe development of resistant strains of HIV is common.Some patients receive prophylaxis for PCPbut are still so profoundly immunocompromisedthat it is ineffective.DiagnosisPatients with PCP usually present with feverand gradually increasing cough and dyspnea fora few weeks, but the disease sometimes presentsas an acute illness with rapid deterioration overthe course of a few days. The chest radiographusually shows diffuse granular opacities, whichstrongly suggest the diagnosis. Some patients withPCP have nodular densities, lobar consolidation,or normal radiographic findings. Cystic abnormalitiesand spontaneous pneumothoraces inpatients with known or suspected HIV infectionare usually caused by PCP.Adjunctive testing may support the diagnosisbut by themselves do not establish a diagnosis. PCPis unlikely in a patient who had a CD4 cell countof 200 cells/µL in the preceding 2 months in theabsence of other HIV-associated symptoms.Approximately 90% of patients with PCP have anelevated serum lactate dehydrogenase level, butthis increase may occur in the presence of otherpulmonary diseases, especially mycobacterial and222 <strong>Pulmonary</strong> Complications of HIV Infection (Rosen)ACC-PULM-09-0302-014.indd 2227/10/09 8:26:55 PM


fungal infections. Oxygen desaturation with exerciseis a relatively sensitive and specific test inpatients who are suspected to have PCP, but it isnot diagnostic. 67 Ga and 111 In lung scans are highlysensitive indicators of PCP, but isotope uptake alsooccurs in other pulmonary infections and thereforethey seldom are useful.Microbiological DiagnosisP jiroveci cannot be cultured in vitro; therefore,the diagnosis of PCP can be confirmed only bydemonstrating organisms in a lung-derivedspecimen, either in sputum induced by the inhalationof hypertonic saline solution or by bronchoscopy.Although establishing a diagnosis is notdifficult, many clinicians treat patients with suspectedPCP empirically, reserving bronchoscopyfor patients who do not respond to treatment. Theexamination of BAL fluid establishes the diagnosisin 90% of cases, and some centers have reportedthat bronchoscopic lung biopsy increases the diagnosticyield not only for PCP but also for otherinfections (especially Mycobacterium tuberculosis)and noninfectious disorders. The optimal approachto the diagnosis of HIV-associated pulmonarydisorders (including PCP) can be determined onlyby a prospective randomized clinical trial withoutcome measures that include economic andsurvival analyses, but no trial has yet been performed,nor is such a trial likely as the incidenceof PCP is now low compared with earlier in theAIDS epidemic.TreatmentAntimicrobial Agents: TMP-SMX is the preferredtreatment for PCP in patients who have nothad an adverse reaction to this drug, regardlessof the severity of disease, and even in patientsin whom PCP develops despite TMP-SMX prophylaxis.It is consistently the most effective incomparative studies and is also inexpensive andavailable in both oral and IV preparations. Patientswith severe PCP who do not respond to or whodo not tolerate TMP-SMX usually are given pentamidine,but this drug is associated with adversereactions that are more serious than the adversereactions associated with TMP-SMX. Trimetrexateleucovorinis not as effective as TMP-SMX but isbetter tolerated than pentamidine. Mild-to-moderatePCP can be treated in the outpatient settingwith dapsone/TMP, clindamycin/primaquine, oratovaquone. The optimal duration of treatment isunknown, but most clinicians treat the patient for21 days, followed by prophylaxis.Adjunctive Corticosteroids: Animal modelsof PCP have shown that the clinical severity ofinfection correlates more closely with markersof inflammation than with the burden of organisms,suggesting that the immune response andits attendant inflammation account for the clinicalmanifestations of pneumonia. Respiratorycompromise is associated with the presence ofactivated CD8 cells and neutrophils in the lung,and corticosteroids are thought to attenuate theseeffects. The administration of corticosteroids atthe start of antipneumocystis treatment reducesthe likelihood of respiratory failure, the deteriorationof oxygenation, and death in patientswith moderate-to-severe pneumonia. Patientswho are likely to benefit from therapy have aPao 2of 70 mm Hg or an arterial- alveolar oxygenpressure difference of 35 mm Hg. Nobenefits have been shown with less severe gasexchange abnormalities at the start of therapy orin patients to whom corticosteroids were administered 72 h after antipneumocystis treatmentwas started. Adverse reactions to adjunctivecorticosteroid therapy occur infrequently; lifethreateningsuperinfections have been described,but they are uncommon. Patients in whom pulmonarysymptoms with diffuse radiographicopacities develop shortly after apparently successfultreatment of PCP should be evaluated forthe presence of another opportunistic infection,especially CMV.Respiratory FailureBecause the incidence of PCP decreased in thelast decade, fewer patients with PCP have beenadmitted to the ICU. Nevertheless, when thetreatment of PCP is delayed or ineffective,hypoxemic respiratory failure may develop; theclinical and radiographic features of severe PCPresemble those of ARDS, and the supportivetreatment is similar, including intubation, mechanicalventilation with low tidal volume (6 mL/kgideal body weight), and the application of positive<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 223ACC-PULM-09-0302-014.indd 2237/10/09 8:26:56 PM


end- expiratory pressure. Survival after respiratoryfailure in PCP seems to have improved since theearly days of the AIDS epidemic and is perhapsrelated to improved strategies for mechanicalventilation. However, the development of respiratoryfailure in patients after several days of appropriatetherapy for PCP or of a pneumothorax whilereceiving mechanical ventilation portend a poorerprognosis.PreventionProphylaxis can be discontinued safely inpatients who achieve a sustained increase in CD4 lymphocyte count to 200 cells/µL after startingART. In those patients who have a suboptimalresponse to ART, lifelong anti-Pneumocystistherapy is recommended for all HIV-infectedpatients with CD4 cell counts of 200 cells/µLor for patients with HIV-related symptoms, includingunexplained persistent fever (temperature, 100˚F [37.8˚C]) for 2 weeks, oropharyngeal candidiasisthat is unrelated to antibiotic or corticosteroidtherapy, and unexplained weight loss.Prophylaxis with TMP-SMX is most effective butis associated with more adverse events requiringthe discontinuation of therapy with aerosolizedpentamidine and dapsone. Therapy with TMP-SMX is also inexpensive and prevents other infections,including cerebral toxoplasmosis andinfection with some strains of pathogenic bacteria.Current expert opinion recommends the administrationof one double-strength tablet (160 mg ofTMP/800 mg of SMX) daily or a single-strengthtablet daily if the former is not tolerated. A lack ofresponse to TMP-SMX prophylaxis is associatedwith nonadherence to treatment and with severeimmunosuppression, as measured by very lowCD4 lymphocyte counts.TBCoinfection with HIV and M tuberculosis isdiscussed in detail in another section. Modestreductions in cell-mediated immunity increase therisk of the reactivation of latent TB, and the riskincreases as the CD4 cell counts decrease. In HIVinfectedpersons, TB often occurs before the occurrenceof opportunistic infections, probably becauseM tuberculosis is more virulent. In patients withmild immunodeficiency, the clinical presentationis similar to TB in HIV-negative patients. Atypicalpulmonary presentations, including the presenceof diffuse infiltrates, miliary patterns, intrathoraciclymphadenopathy, or normal chest radiographfindings occur more frequently in patients withadvanced immunosuppression (Table 2). Thesepatients also have a high incidence of extrapulmonaryinfection, including infections in thepleura, lymph nodes, GI tract, bone marrow, andblood.The diagnosis of TB may be difficult in HIVinfectedpersons. Cutaneous anergy is moreprevalent as CD4 + cell counts decrease, makingtuberculin skin tests less useful. Radiographic cluesto the diagnosis include cavitation, hilar and mediastinallymphadenopathy, and pleural effusions.When there is cavitation, acid-fast smears andcultures of sputum usually are positive. In patientswho do not expectorate spontaneously, sputummay be induced with hypertonic saline solution.Bronchoscopy with BAL fluid, transbronchialbiopsy, and postbronchoscopy sputum often arediagnostic. Biopsy specimens enhance the immediatediagnostic yield of bronchoscopy in the diagnosisof pulmonary TB compared with BAL fluidalone.Despite an appropriate evaluation, acid-fastsmears of sputum and bronchoscopic specimensmaybe negative, and cultures may not be positivefor several weeks. The early treatment of TBimproves the outcome and reduces the transmissionof the disease to others; therefore, initialempiric therapy is warranted for patients withradiographic abnormalities that are consistent withTB, unless another disorder is identified.Atypical MycobacteriaMycobacterium avium complex (MAC) causesdevastating complications and death in patientswith severe immunosuppression. Patients mayhave persistent fever, wasting, and diarrhea. However,MAC is rarely a pulmonary pathogen inpatients with AIDS. Its isolation in patients withsymptomatic pulmonary disease usually occurs inassociation with that of another pathogen, such asPneumocystis. Other nontuberculous mycobacteriaalso cause pulmonary infections in patients withHIV infection. However, if acid-fast bacilli are224 <strong>Pulmonary</strong> Complications of HIV Infection (Rosen)ACC-PULM-09-0302-014.indd 2247/10/09 8:26:56 PM


identified from sputum or bronchoscopic specimens,patients should be treated presumptivelyfor M tuberculosis until the mycobacterial speciesis identified, mainly because TB is more common,has a better response to treatment, and the earlytreatment of active TB is an essential public healthmeasure.CMVIn patients with CD4 + lymphocyte counts of 50 cells/µL, CMV commonly causes retinitis,esophagitis, gastritis, colitis, hepatitis, encephalitis,pneumonia, and death. Curiously, CMV pneumonitisis uncommon in patients with AIDS. Althoughthe virus often is isolated in cultures of BAL fluid,it is not usually pathogenic. <strong>Pulmonary</strong> infectioncan be inferred when typical intranuclear or intracytoplasmicinclusions are found in BAL fluid orbiopsy material. The likelihood that CMV is a pulmonarypathogen is also greater when CMV infectionis found at other sites. It also seems to berelatively common in patients whose conditionsdeteriorate after receiving corticosteroid treatmentfor PCP.CMV pneumonitis usually occurs in patientswho have had previous AIDS-defining illnesses.They present with a clinical syndrome similar toPCP, with dyspnea, nonproductive cough, fever,and diffuse pulmonary opacities noted on chestradiographs. Unilobar radiographic involvement,cavitation, nodules, and pleural effusions alsohave been described. CMV pneumonitis is treatedin the same way as infection in other sites, with IVganciclovir or foscarnet, and the responses totherapy are similar to those of patients with CMVretinitis or GI disease. Because CMV infectionoccurs only in patients with very severe immunosuppression,the long-term prognosis is verypoor.Fungal PneumoniasFungal infections have been discussed in detailelsewhere. Life-threatening fungal disease mayoccur in HIV-infected patients, either by new infectionor the reactivation of latent disease. The typesof fungal infections depend on the severity ofimmunodeficiency and whether the patient haslived in endemic areas.CryptococcosisCryptococcus neoformans is distributed throughoutthe world and is the most common funguscausing life-threatening illness in patients withAIDS. The meninges are the most common siteof infection, and cryptococcal meningitis is oftenthe first manifestation of AIDS. With cryptococcalpneumonia, the chest radiograph usually showsdiffuse infiltrates, similar to PCP, but localizedinfiltrates, nodules, cavitation, pleural effusions,miliary patterns, and lymphadenopathy also areseen. Most patients with cryptococcal pneumoniahave meningitis and disseminated disease, andCD4 + lymphocyte counts typically are 100cells/µL. The diagnosis is established by theidentification of the organism from sputum, BALfluid, pleural fluid, or lung biopsy specimens. Ahigh titer of cryptococcal antigen in serum isstrongly suggestive of and an antigen titer of 1:8 in BAL fluid is diagnostic of cryptococcalpneumonia.HistoplasmosisHistoplasma capsulatum is endemic in the Ohioand Mississippi river valleys, Central and SouthAmerica, and the Caribbean islands. Disseminateddisease may develop in patients with HIV infectionwho come from endemic areas when immunodeficiencypermits the reactivation of latentinfection. The clinical presentation is usually subacute,and the chest radiograph typically shows adiffuse or miliary pattern, although localized infiltratesmay occur. The diagnosis is established bythe identification or culture of the organism fromblood, lung-derived specimens, bone marrow, orliver.Amphotericin B is the treatment of choice formost cases of HIV-associated histoplasmosis. Theliposomal form is less toxic and more effectivethan the standard deoxycholate formulation.Itraconazole maybe used for patients who cannottolerate amphotericin B or have milder diseaseand should be used as a long-term suppressivetherapy for the life of the patient after the primaryinfection is controlled. Voriconazole is notrecommended in treatment or prevention ofhistoplasmosis.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 225ACC-PULM-09-0302-014.indd 2257/10/09 8:26:56 PM


AspergillosisLife-threatening pulmonary aspergillosis maydevelop in patients with advanced immunosuppression.The following two common patterns ofdisease have been identified: an invasive parenchymalinfection, which is usually fatal, and apredominantly bronchial disease presenting withdyspnea and airway obstruction. The classic risksfor Aspergillus infection, namely, prolonged neutropeniaand treatment with high-dose corticosteroids,often are absent. Aspergillosis probablyoccurs in patients with advanced AIDS because ofdefects in neutrophil or alveolar macrophage function.The CD4 + lymphocyte count is typically 30cells/µL, and the previous use of corticosteroidsand neutrophil counts of 500 cells/µL increasethe risk. Disseminated disease is common, especiallyin the brain.Clues to the diagnosis of invasive pulmonaryaspergillosis include upper-lobe disease withcavitation and hemoptysis. This diagnosis hastraditionally required histologic proof, becauseAspergillus is ubiquitous, and its presence in nasopharyngealsecretions, sputum, and BAL fluid mayrepresent contamination or colonization. However,studies in patients with severe immunosuppression,including AIDS, have indicated that the isolationof Aspergillus in BAL fluid correlates stronglywith histologic proof of tissue invasion. Althoughvoriconazole has not been studied for the treatmentof invasive aspergillosis in AIDS patients, the Centersfor Disease Control and Prevention recommendit as the first choice for therapy, withamphotericin B (liposomal or deoxycholate) as analternative. Caspofungin has been approved forpatients who cannot tolerate standard therapy,but it has not been studied in patients with HIVinfection.Other Fungal InfectionsIn endemic areas, disseminated coccidioidomycosisand blastomycosis may occur in patientswith AIDS, usually as a complication of advancedimmunosuppression. The reactivation of a previousinfection may develop in some patients aftermoving from an endemic area. These infectionsusually involve the lung; patients present withcough, fever, dyspnea, and the appearance ofnodular, focal, cavitary, or diffuse disease. Thediagnosis is established by demonstrating the presenceof the organism by microscopy or culture inrespiratory specimens.Neoplastic Diseases of the LungsKaposi SarcomaKaposi sarcoma (KS) is the most commonmalignancy in persons with HIV infection, and theskin is the major site of involvement. KS is causedby human herpesvirus (HHV)-8. This virus infectsmany healthy adults and can be isolated commonlyin saliva, prostate tissue, and semen. It is probablytransmitted by sexual contact and causes diseasewhen activated by HIV-associated immunosuppression.This hypothesis helps to explain why KSis much more common among HIV-infected gaymen than in other transmission groups.Visceral involvement with KS is common inpatients with advanced disease and may involvethe airways, lungs tissue, mediastinal lymphnodes, and pleura. Patients with thoracic KS usuallyhave obvious mucocutaneous lesions, but thelung may be the only site of disease in up to 15%of cases. The involvement of the airways, parenchyma,pleura, and intrathoracic lymph nodescauses a diverse range of symptoms and radiographicfindings. The majority of patients withpulmonary KS that was diagnosed antemortemhad cough, dyspnea, and fever. In the airways, KSlesions usually are asymptomatic but sometimescause obstruction or hemoptysis. The finding oftypical lesions on inspection of the airways is usuallyconsidered to be diagnostic. A histologic diagnosismay be difficult because the yield of theforceps biopsy is low. Some clinicians believe thata forceps biopsy of KS lesions places the patientat significant risk of bleeding, but this notion iscontroversial.Parenchymal involvement with KS is suggestedby bronchial wall thickening, nodules,Kerley B lines, and coexisting pleural effusions,especially in patients with cutaneous disease. Theuse of bronchoscopy can determine whetherdiffuse radiographic opacities are caused by KS oran opportunistic infection. The yield of bronchoscopiclung biopsies in the diagnosis of KS is low,and even open-lung biopsy is nondiagnostic in226 <strong>Pulmonary</strong> Complications of HIV Infection (Rosen)ACC-PULM-09-0302-014.indd 2267/10/09 8:26:56 PM


approximately 10% of cases because of the focaldistribution of the lesions. Therefore, the diagnosisof pulmonary parenchymal KS is usually inferredin patients with cutaneous disease, chest radiographfindings, or CT scan findings that suggestthis disorder; visual confirmation of airway lesions;and no evidence of opportunistic infection in BALfluid or bronchoscopic lung biopsy specimens.Patients with parenchymal opacities who havetypical lesions in the airways and no identifiedpulmonary infection are assumed to have parenchymalKS.When KS involves the pleura, effusions are usuallyexudative and sanguineous, but the cytologicexamination is nondiagnostic. Closed pleuralbiopsy specimens are rarely positive for KS due tothe focality of the pleural lesions and the predominantinvolvement of the visceral pleura, rather thanthe parietal pleura. Because establishing a diagnosisusually necessitates a thoracoscopic or openpleural biopsy, the presence of pleural involvementwith KS is usually inferred in a patient with cutaneousdisease and a serosanguineous effusion withouta reasonable alternative explanation.Patients with disseminated disease may benefitfrom cytotoxic chemotherapy. The most promisingapproach may be the use of ART, which maylead to the regression of KS lesions. In many cases,ART leads to a reduction in HHV-8 load, which isassociated with regression inhibition of angiogenesisand of KS tumor growth.LymphomaNon-Hodgkin B-cell lymphoma is associatedwith HIV infection and, unlike most other HIVrelateddisorders, it continues to occur despite theuse of highly active ART (ie, HAART). Althoughpulmonary involvement is usually clinicallyinnocuous, the lung is a common site of extranodaldisease. HIV-associated primary pulmonary lymphomais usually a high-grade B-cell tumor occurringin the setting of advanced HIV infection. Ifsymptoms occur, they usually do so late in thecourse of HIV disease and simulate commonopportunistic infections. Even in patients withlymphoma, lung involvement is usually a latefeature of HIV disease. It may present with lobarconsolidation, nodules, reticular opacities, andmasses. The biopsy is established by bronchoscopicor open biopsy; BAL fluid has a very low diagnosticyield. In contrast to non-AIDS patients, mediastinaland hilar lymphadenopathy are generallynot prominent in patients with non-Hodgkin B-celllymphoma and HIV infection. Pleural involvementis characterized by effusions and pleural thickeningand the airway involvement by atelectasis. Thediagnosis is established by biopsy or cytologicanalyses of pleural fluid.Primary effusion lymphoma or “body cavitylymphoma” is a disorder characterized by pleural,pericardial, or peritoneal growth of malignantlymphoma cells, usually in the absence of solidtumors. In almost all cases, tumor DNA showsHHV-8, which is also the etiologic agent of KS andmulticentric Castleman disease, a dysplastic lymphoiddisorder that may progress to plasmablasticlymphoma. The virus has a strong tropism for Bcells, and in some patients HHV-8-infected clonesproliferate and transform into a large cell “liquidlymphoma” within serous cavities. In HIV-seropositivepersons, the prognosis of primary effusionlymphoma is no better, and is probably worse, thanin most other forms of lymphoma.Carcinoma of the LungThe incidence of lung cancer is now recognizedto be increased in persons with HIV infection,especially in the years after the use of ART as thestandard of care. Although persons in most populationswith HIV infection are current or formersmokers, their risk of the development of lungcancer appears to be increased even when incidenceratios are adjusted for smoking. The prevalenceof microsatellite alternations reflectinggenomic instability also occurs with greatlyincreased frequency in HIV-associated lung cancers,possibly playing a role in their pathogenesis.Patients with lung cancer in the setting of HIVinfection tend to be relatively young at presentation(mean age, 45 years) and have mild or moderateimmunosuppression. Similar to age-matchedcontrol subjects, 75 to 90% of patients present withstage III or IV disease. Adenocarcinoma is the mostcommon histologic type, and the prognosis appearsto be worse in patients with HIV infection. Untilfurther studies are performed, the diagnostic andtherapeutic approaches to patients with HIVassociatedlung cancer are the same as for other<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 227ACC-PULM-09-0302-014.indd 2277/10/09 8:26:57 PM


patients, although the efficacy and toxicity of chemotherapyand radiotherapy may be different.Other <strong>Pulmonary</strong> DisordersObstructive Lung DiseaseThere is a propensity for the development ofchronic bronchitis and bronchiectasis in patientswith advanced HIV infection, even if they do notsmoke. The CD4 count is usually low ( 100cells/µL). Therapy with standard antimicrobialagents is usually effective, but symptoms are likelyto recur, especially when P aeruginosa is isolatedfrom the sputum. The role and efficacy of therapywith bronchodilators and antiinflammatory agentsin patients with HIV-associated airway diseasehave not been studied.HIV infection also appears to accelerate theonset of smoking-related emphysema, possiblythrough cytotoxic lymphocyte activity, lung capillaryendothelial injury, and increased oxidativestress. The incidence of emphysema appears to beincreased in patients with HIV infection independentof smoking status.Idiopathic Inflammatory Disorders<strong>Pulmonary</strong> disorders without a defined infectiousor neoplastic etiology occur in HIV-infectedpersons. Lymphocytic interstitial pneumonitis(LIP) and nonspecific interstitial pneumonitis (NIP)are believed to comprise a spectrum of inflammatorychanges in response to HIV infection in thelung itself. LIP is most common in HIV-infectedchildren and persons of African descent. It mayoccur as part of a systemic CD8 + lymphoproliferativesyndrome, with lymphadenopathy, bloodlymphocytosis, and the involvement of otherorgans. NIP is very common in persons with lowCD4 + counts but is rarely diagnosed because itusually causes no symptoms. When symptomatic,both LIP and NIP are treated with corticosteroids.Bronchiolitis obliterans organizing pneumonia hasa similar clinical presentation in patients with andwithout HIV infection. Lung biopsy, either transbronchialor open, is necessary for the diagnosis.This disorder often improves dramatically withcorticosteroid therapy.<strong>Pulmonary</strong> Hypertension<strong>Pulmonary</strong> hypertension occurs more commonlyin HIV-infected patients than in the generalpopulation. It may eventually lead to cor pulmonaleand death. Some evidence links HHV-8 (theetiologic agent of KS) with pulmonary hypertensionin patients without HIV infection, perhapsthrough the dysregulation of endothelial cellgrowth or growth-factor signaling. An alternativeexplanation is that the presence of HIV antigens inthe pulmonary endothelium may stimulate apoptosis,growth, and proliferation. HIV-associatedpulmonary hypertension appears to occur at allstages of HIV infection, and the approach to diagnosisand treatment is the same as that for idiopathicpulmonary arterial hypertension inHIV-uninfected persons. The effects of startingART during a course of treatment for pulmonaryhypertension are not known.Immune Restoration SyndromesThe effectiveness of ART in restoring immunefunction has given rise to a new disorder, theimmune reconstitution inflammatory syndrome(IRIS), which is defined as a paradoxical deteriorationin clinical status that is attributable to therecovery of the immune system during ART.When treatment with ART is successful, the numberof blood CD4 + lymphocytes increases, as doestheir activity. Increased immune function thenleads to inflammation that may have been otherwiseclinically silent, leading to overt clinicalillness. The diagnostic criteria for IRIS include thediagnosis of AIDS, treatment with anti-HIVmedications that lead to a reduction in HIV viremia(followed by an increase in the CD4 + lymphocytecount), symptoms consistent with aninfectious or inflammatory condition thatappeared while the subject was receiving ART,and symptoms that cannot be explained by anewly acquired infection, by the expected clinicalcourse of the disease, or by the side effects oftherapy.Paradoxical worsening of TB was described inpatients in whom transient worsening ofTB-related symptoms developed after ART. Thepresence of fever that worsened or the emergence228 <strong>Pulmonary</strong> Complications of HIV Infection (Rosen)ACC-PULM-09-0302-014.indd 2287/10/09 8:26:57 PM


of cervical intrathoracic lymphadenopathy, pulmonaryinfiltrates, pleural effusions, or othertuberculous lesions develop in these patientsshortly after starting ART, which is associated withthe restoration of cutaneous reactivity to skintest antigens. Respiratory failure caused by IRISwas reported in several cases after the introductionof ART after the successful treatment of PCP.Other patients had apparent exacerbations ofdisease after the introduction of ART after orduring infection with MAC, cryptococcosis, CMV,herpes zoster, hepatitis B and C viruses, andthe agent that causes progressive multifocalleukoencephalopathy.IRIS is a diagnosis of exclusion, which is suggestedby a compatible clinical syndrome in apatient who is recovering from an infection andhas begun receiving ART in the prior severalmonths. The plasma HIV load and CD4 + lymphocytecounts usually are improved compared withprior measurements, but the circulating CD4 +count may be unchanged because these cells arecompartmentalized to sites of active inflammation.Opportunistic infection is usually a diagnosticconsideration but, in the absence of a severeillness and in the proper clinical setting, invasiveprocedures can usually be avoided and thepatient observed. If bronchoscopy is performed,the mean CD4/CD8 ratio is significantly greater(0.54) than in HIV-infected patients who areundergoing bronchoscopy for other respiratorycomplaints (0.07). Most patients should justreceive palliative therapy for symptoms, such asantipyretic agents for fever. Systemic corticosteroidshave been used successfully in the treatmentof severe inflammatory disease causing significantend-organ damage.It is not known when to start therapy withantiretroviral agents in patients with active opportunisticinfection. However, there is a consensusthat in patients who have an HIV-associated disorderfor whom there is no alternative treatment(such as cryptosporidiosis and progressive multifocalleukoencephalopathy), the benefits of ARTwould outweigh the risks. However, many (butnot all) experts would defer starting ART inpatients with active treatable disorders like PCPand tuberculosis until there is a response to theinfection.Annotated BibliographyGeneralCenters for Disease Control and Prevention. HIV/AIDS surveillance report: 2007. (Vol 19). Atlanta: U.S.Department of Health and Human Services, Centersfor Disease Control and Prevention, 2009; 1−63. Availableat: http://www.cdc.gov/hiv/topics/ surveillance/resources/reports/. Accessed April 4, 2009An updated report describing the epidemiology of AIDSin the United States. It documents the decreasing rates ofopportunistic infections and death in addition to the trendsin the incidence of AIDS in different risk groups. It is troublingthat the estimated number of newly diagnosed cases ofHIV/AIDS increased 15% from 2004 to 2007, with 15% ofall of these cases among persons 40 to 44 years of age.Centers for Disease Control and Prevention. Guidelinesfor prevention and treatment of opportunisticinfections in HIV-infected adults and adolescents:recommendations from CDC, the National Institutesof Health, and the HIV <strong>Medicine</strong> Association of theInfectious Diseases Society of America. Available at:http://www.cdc.gov/mmwr/preview/mmwrhtml/rr58e324a1.htm?s_cid=rr58e324a1_e. Accessed April 5,2009Latest version of comprehensive guidelines for the preventionof opportunistic infections in persons with HIV infection.They emphasize the prevention of PCP, TB, toxoplasmosis,and MAC.Hanson DL, Chu SY, Farizo KM, et al. Distribution ofCD4 T lymphocytes at diagnosis of acquired immunodeficiencysyndrome-defining and other humanimmunodeficiency virus-related illnesses. Arch InternMed 1995; 155:1537−1542Stratifies risk of developing HIV disorders by CD4 + lymphocytecount.Palella FJ, Delaney KM, Moorman AC, et al. Decliningmorbidity and mortality among patients withadvanced human immunodeficiency virus infection.N Engl J Med 1998; 38:853−860Mortality rate in a cohort of 1,255 patients with advancedHIV infection decreased from 29.4 per 100 person-years in1995 to 8.8 per 100 person-years in 1997, with a correspondingdramatic decrease in opportunistic infections, althoughthe rate of use of prophylaxis against PCP and MAC diseaseremained similar. This study demonstrated that combinationART in an outpatient cohort dramatically reduced morbidityand mortality related to HIV infection.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 229ACC-PULM-09-0302-014.indd 2297/10/09 8:26:57 PM


UNAIDS. 2007 AIDS epidemic update. Available at:http://data.unaids.org/pub/EPISlides/2007/2007_epiupdate_en.pdf. Accessed April 30,2008 Provides adetailed summary of the HIV/AIDS pandemic around theworld.Wallace JM, Rao AV, Glassroth J, et al. Respiratory illnessin persons with human immunodeficiency virusinfection. Am Rev Respir Dis 1993; 148:1523−1529Multicenter prospective study of HIV-infected subjects whodid not have AIDS (1987 definition) at the time of studyenrollment showed that the most frequent respiratory diagnoseswere also common in the general population (ie, upperrespiratory tract infections, acute bronchitis, and sinusitis).Bacterial pneumonia and PCP occurred significantly morefrequently in subjects whose CD4 + count at study enrollmentwas 250 cells/µL, and bacterial pneumonia was morecommon in injection drug users compared with others.Bacterial PneumoniaBaron AD, Hollander H. Pseudomonas aeruginosa broncho-pulmonaryinfection in late human immunodeficiencyvirus infection. Am Rev Respir Dis 1993;148:992−996<strong>Review</strong>s the clinical features and risks for the developmentof P aeruginosa lower respiratory tract infection in HIVinfectedpersons.Hirschtick RE, Glassroth J, Jordan MC, et al. Bacterialpneumonia in patients infected with human immunodeficiencyvirus. N Engl J Med 1995; 333:845−851Describes the risk factors and bacteriology in a prospectivecohort study. Mortality during the follow-up period wasalmost fourfold greater in subjects who had an episode ofbacterial pneumonia than in others. Prophylaxis with TMP-SMX was associated with a significant reduction in the incidenceof bacterial pneumonia.PCPHuang L, Morris A, Limper AH, et al. Proc Am ThoracSoc 2006; 3:655−664<strong>Review</strong> of the biology, host defense mechanisms, epidemiologyand clinical features, diagnosis, and treat mentof Pneumocystis, recommending directions for futureinvestigations.Morris A, Lundgren JD, Masur H, et al. Current epidemiologyof Pneumocystis pneumonia. Emerg InfectDis 2004; 110:1713−1720Trends in the epidemiology and clinical features of PCP inthe era of ART.National Institutes of Health, University of California.Consensus statement on the use of corticosteroid asadjunctive therapy for Pneumocystis pneumonia inthe acquired immunodeficiency syndrome: NationalInstitutes of Health-University of California ExpertPanel for Corticosteroid as Adjunctive Therapyfor Pneumocystis Pneumonia. N Engl J Med 1990;323:1500−1504<strong>Review</strong> of studies of adjunctive corticosteroid therapy forPCP, with recommendations for clinical use.Rosen MJ. Intensive care of patients with HIV infection:time to take another look. J Intensive Care Med2005; 20:312−315<strong>Review</strong> of trends in critical care of patients with HIV infection,including the declining incidence of PCP and theincrease in mortality from non-HIV-associated disorders.Viral and Fungal PneumoniaLortholary O, Meyohas MC, Dupont B, et al. Invasiveaspergillosis in patients with acquired immunodeficiencysyndrome: report of 33 cases. Am J Med 1993;95:177−187A large series of patients is the basis of this description of theclinical features of invasive aspergillosis in AIDS patientsdemonstrating that a positive culture finding in BAL fluidis diagnostic of the disease.Rodriguez-Barradas MC, Stool E, Musher D. Diagnosingand treating cytomegalovirus pneumonia inpatients with AIDS. Clin Infect Dis 1996; 23:76−81The diagnosis of CMV pneumonia is difficult to establishwith certainty because it is commonly found in associationwith other pathogens and may be isolated from lung-derivedspecimens without histologic evidence of disease. This studydescribes the clinical and laboratory features of patients withconfirmed CMV pneumonia.Noninfectious <strong>Pulmonary</strong> DiseasesAscoli V, Lo-Coco F. Body cavity lymphoma. Curr OpinPulm Med 2002; 8:317−322<strong>Review</strong>s the pathogenesis and clinical features of primaryeffusional lymphoma.Barnett CF, Hsue PY, Machado RF. <strong>Pulmonary</strong> hypertension:an increasingly recognized complication ofhereditary hemolytic anemias and HIV infection.JAMA 2008; 299:324−331Concise case-based review of epidemiology, mechanisms,diagnosis, and treatment.230 <strong>Pulmonary</strong> Complications of HIV Infection (Rosen)ACC-PULM-09-0302-014.indd 2307/10/09 8:26:57 PM


Engels EA, Brock MV, Chen J, et al. Elevated incidenceof lung cancer among HIV-infected individuals. J ClinOncol 2006; 24:1383−1388Herida M, Mary-Krause M, Kaphan R, et al. Incidenceof non-AIDS defining cancers before and during thehighly active antiretroviral therapy era in a cohort ofhuman immunodeficiency virus-infected patients. JClin Oncol 2003; 21:3447-3453The incidence of lung cancer is increasing in the era of ART.It is unknown whether this is simply related to increasedlongevity leading to increased cancer rates, is an effectof HIV or ART on oncogenesis, or is a failure of immunesurveillance.Huang L, Schnapp LM, Gruden J, et al. Presentation ofAIDS-related Kaposi’s sarcoma diagnosed by bronchoscopy.Am J Respir Crit Care Med 1996; 153:1385−1390The clinical, laboratory, radiographic, and bronchoscopicfeatures of pulmonary KS are described in a large series ofpatients.Petrache I, Diab K, Knox KS, et al. HIV associated pulmonaryemphysema: a review of the literature andinquiry into its mechanism. Thorax 2008;63:463−469Recent review of clinical features and mechanisms.Travis WD, Fox CH, Devaney KO, et al. Lymphocyticpneumonitis in 50 adult HIV-infected patients: lymphocyticinterstitial pneumonitis versus nonspecificinterstitial pneumonitis. Hum Pathol 1992; 23:529−541The clinical and pathologic features of these inflammatorydisorders were studied. They probably represent differentmanifestations of HIV infection of the lung.Wistuba II, Behrens C, Milchgrub S, et al. Comparisonof molecular changes in lung cancers in HIVpositiveand HIV-indeterminate subjects. JAMA 1998;279:1554−1559Lung cancers in patients with HIV infection have a greaterincreased incidence of microsatellite alterations than in otherpatients with lung cancer. These differences reflect genomicinstability and provide evidence for a link between HIVinfection and lung cancer.Immune Restoration SyndromesNaccache JM, Antoine M, Wislez M, et al. Sarcoid-likepulmonary disorder in human immunodeficiencyvirus-infected patients receiving antiretroviral therapy.Am J Respir Crit Care Med 1999; 159:2009−2013Narita M, Ashkin D, Hollender ES, et al. Paradoxicalworsening of tuberculosis following antiretroviraltherapy in patients with AIDS. Am J Respir Crit CareMed 1998; 158:157−161Shelburne SA, Montes M, Hamill R. Immune reconstitutioninflammatory syndrome: more answers, morequestions. J Antimicrob Chemother 2006;57:167−170.<strong>Review</strong>s the definition and mechanisms of IRIS.Wislez M, Bergot E, Antoine M, et al. Acute respiratoryfailure following HAART introduction in patientstreated for Pneumocystis pneumonia. Am J Respir CritCare Med 2001; 164:847−851<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 231ACC-PULM-09-0302-014.indd 2317/10/09 8:26:58 PM


Notes232 <strong>Pulmonary</strong> Complications of HIV Infection (Rosen)ACC-PULM-09-0302-014.indd 2327/10/09 8:26:58 PM


<strong>Pulmonary</strong> Complications of CardiothoracicSurgery and TraumaBruce P. Krieger, MD, FCCPObjectives:• Apply physiologic parameters preoperatively to predictpostoperative pulmonary complications following thoracicresection or cardiac surgery• Explain the pathophysiology of hypoxemia in patients followingcardiac surgery and chest trauma• Describe the initial assessment of the patient who has sufferedblunt chest trauma• Recognize and treat the most common causes of postcardiacsurgery respiratory failureKey words: diaphragm dysfunction; exercise testing; sternotomy;thoracic surgery; thoracic trauma; thoracoscopybeing whether the cancer can be fully resected andcured (resectability), and the second being whetherthe patient will survive the operation and be ableto function in a “reasonable manner” afterwards(operability). A large number of thoracic surgerypatients have underlying lung diseases that areassociated with an increase in the morbidity andmortality from surgery, especially in the elderly.Therefore, preoperative evaluation is critical beforedeciding on the advisability of thoracic surgery.Table 1. Thoracic Surgical ProceduresIndication Approach ProceduresMany recently trained chest surgeons have moreexperience in cardiac surgery than in thoracicsurgery. For the pulmonologist, a surgeon withexpertise in thoracic surgery is essential for patientsundergoing complicated pulmonary procedures.Thoracocardiac surgery encompasses proceduresinvolving the lungs, chest wall, mediastinum,thoracic aorta, and heart (Table 1). This article willconcentrate on those problems most frequentlyencountered by a practicing pulmonologist, includingpreoperative evaluation prior to lung resectionsurgery, pulmonary complications following thoracicand cardiac surgery, diagnostic thoracoscopy,and pulmocardiac complications following chesttrauma. Lung transplantation and lung volumereduction surgery (LVRS) are discussed elsewherein this course.<strong>Pulmonary</strong> ComplicationsFollowing Lung Resection SurgeryLung resection surgery is performed in theUnited States at least 30,000 times annually, usuallyin an attempt to cure bronchogenic carcinoma. Thepulmonologist is often consulted preoperatively toevaluate the advisability of lung resection. Thereare two general aspects to this evaluation, the firstResectional Thoracotomy Pneumonectomy,lobectomy,segmentectomy,wedge resection, riband chest wallresectionThoracoscopy Wedge resection,lobectomyDiagnostic Thoracotomy Biopsy of lung, pleura,pericardium,mediastinal structures,adenopathyThoracoscopy Biopsy of lung, pleura,pericardium,mediastinal structures,nodes; diagnosediaphragmatic injuryTherapeutic Thoracotomy Control hemorrhage/pneumothorax/pleualeffusion/empyemas/granulomatousinfections; lungtransplanats; LVRS;others*Thoracoscopy Controlpneumothoraxpleuraleffusion/empyema,granulomatous lunginfection; LVRS;others**Others = pericardial window, ligation of thoracic duct,esophageal surgery, vagotomy.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 233ACC-PULM-09-0302-015.indd 2337/10/09 8:27:22 PM


The decreases in lung function that occur afterthoracic surgery are more complicated than thechanges that occur following upper- abdominal orcardiac surgery because a functional lung is removed.Following laparotomy, there is an approximately50% decrease in FVC for the first few postoperativedays. There are similar predictable decreases in lungfunction following cardiac surgery, as describedlater in this article. The effect of thoracic surgeryon postoperative pulmonary function is not onlypermanent but also more difficult to predict since itdepends on the amount of viable lung parenchymathat remains following surgery. If the patient haslimited pulmonary reserve preoperatively, furtherworsening following lung resection can result inpostoperative atelectasis, pneumonia, prolongedmechanical ventilator support (MVS), or evendeath. Over the past 5 decades, various pulmonaryfunction tests have been utilized to predict postresectionpulmonary function and thus screen forpatients who are at increased risk for postoperativecomplications.BronchospirometrySixty years ago, thoracic surgery was most commonlyperformed to control tuberculosis. Duringthat time, bronchospirometry was introduced asa means of predicting postoperative pulmonaryfunction. Bronchospirometry measures minuteventilation (V • e) and oxygen uptake (V • o 2) from eachindividual lung. This is accomplished by placing adouble-lumen endotracheal tube (under conscioussedation) that essentially isolates the right from theleft lung, therefore allowing measurements to bemade unilaterally. Predicted postpneumonectomyfunction is calculated by multiplying the total preoperativepulmonary function by the percentage oflung that will remain postoperatively.Unilateral <strong>Pulmonary</strong> Artery OcclusionUnilateral pulmonary artery occlusion wasused 50 years ago as a means of assessing whether apatient could withstand a pneumonectomy. A catheterwas placed into the pulmonary artery of the lungthat was going to be resected, and a large (50 mL)balloon was then inflated, thus directing all bloodflow to the contralateral lung. If the mean pulmonaryartery pressure increased to 35 mm Hg or the Pao 2decreased to 45 mm Hg, the candidate was deemedinoperable because of the risk of postoperative death.However, this technique is difficult to perform (26%technical failure), and pulmonary artery cathetersare no longer equipped with large balloons. Giventhe improvement in other less invasive testingtechniques for predicting postoperative pulmonaryfunction, the temporary unilateral pulmonary arteryocclusion is rarely used in modern times.Standard <strong>Pulmonary</strong> Function TestingThe history of using pulmonary function testing(PFT) as a predictor of postoperative morbidity andmortality dates back to 1955, when the combinationof a preoperative maximal breathing capacity(maximal voluntary ventilation [MVV]) 50% ofpredicted combined with a FVC 50% of normalpredicted a 40% mortality rate in patients undergoingresection surgery or thoracoplasty for tuberculosis.Over the next 50 years, other parameters wereproposed to be more accurate prognostic indicatorsof postoperative pulmonary failure. Theseincluded FEV 1 2 L for pneumonectomy or 1 L forlobectomy, and FEV 1/FVC ratio 50%. In the 1980s,diffusing capacity of the lung for carbon monoxide(Dlco) was proposed to be a reliable predictor ofmortality when 50% of predicted. In the early1980s, the accepted criteria for pneumonectomyincluded an FEV 1 2 L along with a MVV 55%,and for a lobectomy a MVV 40% of predicted withan FEV 1 1 L. However, most of these parameterswere derived from studies that included mainlymen in their sixth to eighth decades. As such, anFEV 1of 2 L would be distinctly abnormal for a manbut not necessarily for a small, female octogenarian.In addition, these parameters lacked sufficientspecificity for the individual patient, and thereforeattention was directed to more accurate predictorsof postoperative pulmonary function based on theamount of predicted lung parenchyma that wouldremain following resection. Radionuclide quantitativelung scanning coupled with preoperativepulmonary function testing was introduced as away of refining postoperative lung function by couplinganatomic and physiologic relationships. Thepostpneumonectomy or postlobectomy FEV 1andFVC are predicted by multiplying the percentageof total radioactivity that would remain after lungresection surgery by the preoperative FEV 1and234 <strong>Pulmonary</strong> Complications of Cardiothoracic Surgery and Trauma (Krieger)ACC-PULM-09-0302-015.indd 2347/10/09 8:27:22 PM


FVC. Initially, an absolute lower limit of acceptablepostoperative lung function was established to be 0.8 L based on an accepted dictum that hypercarbiafrequently occurred in patients with COPDwhen the FEV 1was 0.8 L. However, Olsen andassociates suggested that further accuracy couldbe obtained by this technique if the predicted postoperativeFEV 1was corrected for the patient’s age,height, and sex (ie, based on percentage of predictedrather than an absolute number).Markos and coworkers prospectively studied55 consecutive patients utilizing quantitative lungscanning, PFT based on percentage of predicted,and incremental load cycle ergometry. Threedeaths occurred, all in patients with postoperativepredicted FEV 1and Dlco 35%. Based on thesedata and other studies, “acceptable” criteria in the1990s included either a postoperative predictedFEV 1 40% or a combination of an FEV 1postoperativepredicted 30% along with a postoperativepredicted Dlco 40%. This combined approach isapplicable to pneumonectomy or lobectomy. CombiningLVRS with small resections has extended thelimits of thoracic surgery, since postoperative pulmonaryfunction may actually improve after LVRSwhile successfully removing small bronchogeniccarcinomas.Exercise Testing as a Predictor of PostoperativeMorbidity and MortalityThe history of establishing a global assessmentof pulmonary reserve utilizing exercise testingdates back to simple walking and stair-climbingprotocols. Refinement that utilized incrementalexercise testing has shown that a preoperativemaximum V • o 2(V • o 2max) 10 mL/kg/min or 40%of predicted was associated with a high percentageof fatal postoperative pulmonary complications,whereas few postpneumonectomy complicationsoccurred when the V • o 2max was 20 mL/kg/minor 75% of predicted. Lobectomy is considered to berelatively safe when the V • o 2max is 15mL/kg/min. Some algorithms for assessing the operabilityof a lung resection patient suggest utilizing formalexercise testing before undergoing quantitativelung scanning when either the preoperative FEV 1or Dlco are 80% of predicted. Recent studies havecorrelated symptom-limited stair climbing with V • o 2in patients with COPD. The inability to climb twoflights of stairs preoperatively had an 82% positivepredictive value for the development of a postoperativecardiopulmonary complication. However,the simple shuttle (6 min) walk test was not usefulin distinguishing lung cancer surgical outcome.Cardiac Risk FactorsCardiac complications are the second-mostcommondifficulty encountered after thoracotomy.Therefore, exercise testing is theoretically desirablesince it stresses the combined cardiopulmonary systems.Recently, a “risk index” that combines cardiacand pulmonary signs and tests has been advocatedas a guide to reliably predict postoperative risks.From these studies, it is apparent that no singleparameter can be used to exclude surgery in theindividual patient. Rather, a combination of parameterscan be helpful in predicting a statistically highrisk of postoperative morbidity and mortality sothat the patient, with the counseling of their physicians,can make an informed decision as to whetherto proceed with potentially curative lung resectionsurgery. For many patients, the relatively highlikelihood of a poor quality of life postoperativelymay be a more important concern than a limitedchance of surviving for 5 years following resectionof a pulmonary carcinoma. For other patients, an“acceptable“ postoperative mortality is worth therisk if the alternative is close to 100% mortality. Aguide to help the prediction of postoperative pulmonaryfunction and thus quality of life is shownin Figure 1.Figure 1. Stepwise approach for evaluating whether a patienthas the pulmonary reserve to undergo thoracic resectionalsurgery.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 235ACC-PULM-09-0302-015.indd 2357/10/09 8:27:23 PM


<strong>Pulmonary</strong> Complications FollowingCardiac SurgeryPulmonologists are frequently asked to participatein the postoperative management of patientsfollowing median sternotomy for cardiac surgery.Respiratory failure is one of the most feared complicationsfollowing this surgery. Ryan et al reportedthat patients who required ICU care for 14 days followingcardiac surgery had a 43.5% in-hospital mortality,especially if the requirement for mechanicalventilation was due to a nonpulmonary cause. Asnoted in Table 2, there are some causes of poststernotomypulmonary complications that are uniqueto this surgical approach. These include diaphragmaticdysfunction, chest wall (sternal instability),and various forms of noncardiogenic pulmonaryedema. In addition, there is a predictable decline inpulmonary function following cardiac surgery thatin some respects is more severe than the changesthat occur following upper-abdominal laparotomy(Table 3). Despite this, preoperative PFT has notbeen as accurate in predicting postoperative complicationsas it has in thoracic procedures, probablybecause lung volumes are decreased by pulmonaryedema. Rady and Ryan analyzed the hospital outcomeof 11,330 patients and reported the followingpredictors of extubation failure: age 65 years;the presence of COPD, pulmonary hypertension,severe left ventricular dysfunction, or cardiac shock;requirement for hospitalization preoperatively; surgicalparameters (multiple transfusions, prolongedcardiopulmonary bypass time, redo operations, andprocedures involving thoracic aorta); and variouslaboratory values (hematocrit 34%, elevated BUN,lower albumin concentrations, and systemic oxygendelivery 320 mL/min/m 2 ). Although abnormalPFT results did not predict extubation failure in thisstudy, they quantitated the patient’s physiologicpulmonary reserve and thus helped to determinethe patient’s ability to mount an appropriateresponse to postoperative processes that increasecarbon dioxide production and V • e requirements.In an analysis of 1,112 patients, reduced FEV 1wasa statistically significant risk factor for prolongedintubation following coronary artery bypass grafting(CABG). Another retrospective multivariatesurvival analysis of 4,863 consecutive patients whounderwent CABG also identified preoperativerespiratory or cardiac instability as a risk factorTable 2. Causes of <strong>Pulmonary</strong> Complications Following Cardiac Surgery*CategoryParenchymal lung injuryAirways diseaseChest bellows diseaseDepressed respiratory drivePleural disease<strong>Pulmonary</strong> vascular diseaseNonpulmonary diseaseCauses<strong>Pulmonary</strong> edema (cardiogenic and noncardiogenic)AtelectasisPneumoniaHemorrhageCOPD, asthmaBronchorrhea (bronchitis, infectious, recurrent aspiration)Diaphragm weakness, paresis, paralysis or flutterSternal instabilityMorbid obesityProlonged effect of muscle relaxantsPain and chest tubesDrugsNeurological complicationsPleural effusions and hemithoraxPneumothorax<strong>Pulmonary</strong> emboli<strong>Pulmonary</strong> hypertension (valvular surgery)SIRS (increased metabolic demands)Cardiac dysfunctionNeurogenic hyperventilationPsychosisPostlaparotomy*SIRS = systemic inflammatory response syndrome, which includes sepsis, septic syndrome, bacteremia, and septic shock.236 <strong>Pulmonary</strong> Complications of Cardiothoracic Surgery and Trauma (Krieger)ACC-PULM-09-0302-015.indd 2367/10/09 8:27:23 PM


Table 3. Postcardiac Surgery <strong>Pulmonary</strong> Function*POD 2 POD 4 POD 50VC (% preop)IMA 36 56 74SVG 45 63 84FRC (% preop)IMA 59 69SVG 66 72 88*POD = postoperative day; VC (% preop) = vital capacityexpressed as a percent of the preoperative value; FRC(% preop) = functional residual capacity expressed as apercent of the preoperative; value; IMA = internal mammaryartery graft; SVG = saphenous vein graft.for MVS. Preoperative renal failure or mechanicalventilation for 24 h after CABG identified patientsat risk for readmission to the ICU.Yende et al prospectively studied 400 patientsand identified tumor necrosis factor (TNF) genepolymorphisms and angiotensin-convertingenzyme gene polymorphisms preoperatively andshowed that the presence of specific haplotypes wasassociated with prolonged mechanical ventilationafter CABG. Another recent study involving 325consecutive patients noted that an elevated lactatelevel ( 3 mmol/L) was associated with a significantlyelevated risk for morbidity and mortalityafter cardiac surgery.As noted in Table 3, postoperative declines inlung volume are more marked when the internalmammary artery is utilized rather than saphenousveins for grafts. Adverse effects may be even moresevere when patients have pulmonary diseasessuch as COPD or pulmonary fibrosis. Oxygenationdecreases by approximately 12 mm Hg postoperativelydue to a combination of shunting and lowventilation/perfusion units. These effects havebeen shown to be secondary to bibasilar atelectasisthat can be imaged on CT scanning postoperatively.Hypoxemia is exaggerated when pulmonary edema(cardiogenic or noncardiogenic) or pneumonia arepresent. In addition, thoracoabdominal discoordinationhas been shown to contribute to the physiologicchanges that occur poststernotomy.<strong>Pulmonary</strong> Edema Following Cardiac SurgeryThe most common cause of postoperative pulmonaryedema is poor left ventricular function thatsometimes is exaggerated by the negative inotropiceffects of anesthesia. In addition, diastolic dysfunctionmay occur, which can be difficult to recognize.It is underappreciated that the pulmonary arteryocclusion pressure may not necessarily reflectthe propensity for cardiogenic edema to developunless other factors, such as hypoalbuminemia, leftventricular compliance, and elevated intrathoracicpressure from positive pressure ventilation, are considered.In addition, alveolar-capillary membranepermeability is often altered by cardiopulmonarybypass, and therefore capillaries may “leak” withoutan exceptionally high wedge pressure reading.Poststernotomy noncardiogenic pulmonary edemahas been recognized since the 1960s, when it wascalled “pump lung.” Most of the theories that havebeen proposed to explain this form of ARDS revolvearound the effect of extracorporeal membraneoxygenation (ECMO) on neutrophils and alveolarepithelial cells. The occurrence of ARDS followingcardiac surgery carries a 15% mortality risk and hasbeen linearly correlated with the length of cardiopulmonarybypass. Studies in animals and humanshave shown activation of polymorphonuclearleukocytes (PMNs) that results in an increase inmarkers of toxic oxygen radical production andadherence of PMNs to pulmonary endothelium.The sequestration of PMNs activates the complementand coagulation systems as well as other proteolyticenzymes, resulting in injury to pulmonaryendothelial cells. Proinflammatory cytokines (TNF,interleukins-6 and -8) and intracellular and vascularcell adhesion molecules measured in the plasma oron platelets and PMNs have been demonstrated tobe increased in pulmonary venous blood comparedto right-atrial and preoperative blood. In addition,it has been hypothesized that surfactant productionmay be depressed during bypass because ofinadequate blood supply to the alveolar epitheliumthat results in atelectasis and ARDS. Despite theseeffects, ARDS has been reported in 2% of patientsundergoing cardiopulmonary bypass. Previous cardiacsurgery, shock, and the need for blood productswere associated with the development of ARDS.Atelectasis and Diaphragmatic DysfunctionPartial atelectasis of the left lung has beenreported in up to 88% of all patients who haveundergone CABG. In addition, partial atelectasis<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 237ACC-PULM-09-0302-015.indd 2377/10/09 8:27:23 PM


of the right lung has also been reported in a highpercentage (61%) of patients following CABG. Inthe 1980s, the term frostbite injury to the left phrenicnerve was introduced to describe the phrenicnerve palsy that was associated with the use ofcold (specifically, ice slush) cardioplegic solutionsintraoperatively. However, phrenic nerve conductionstudies subsequently showed that only 20% ofpatients with left-lower-lobe atelectasis followingsternotomy had evidence for phrenic neuropathy.Even so, most surgeons use a protective insulationblanket prior to the insertion of cold cardioplegicsolutions to minimize any physical effects on thephrenic nerve. If the phrenic nerve is severelydemyelinated, axonal repair requires up to 9 monthsbefore diaphragm function is restored. Mild demyelinationrepair may occur within 6 to 28 days. Otherpossible explanations for the frequent occurrenceof left-lower-lobe atelectasis include intraoperativecompression of lung tissue, pulmonary endothelialdamage with resultant loss of surfactant, ischemiaduring ECMO, and possibly decreased ventilationof the left lower lobe due to a combination ofcardiomegaly and the supine operating position.Atelectasis of the left lung has been associated withthe number of coronary grafts, the length of surgery,the use of the left anterior mammary artery for agraft, and low body temperatures. However, thepresence of atelectasis does not appear to increasethe length of hospital stay.Diaphragmatic flutter has been described ina small number of patients following sternotomy.This is defined as an involuntary contraction of thediaphragm at rates 40 times per minute that is notassociated with contraction of other respiratorymuscles. Diaphragmatic flutter is not suppressiblevoluntarily or by mechanical hyperventilation.It appears to be associated with sternal complications,and it can be an unrecognized cause forprolonged MVS. It is important to distinguish therapid-shallow breathing of diaphragmatic flutterfrom a similar pattern that occurs when patientshave increased V • e requirements (such as duringsepsis or neurogenic hyperventilation). Since flutteris an involuntary diaphragmatic contraction, it willbe associated with discoordination of the thoraciccage and therefore thoracoabdominal asynchronythat can be detected on physical examination orwith respiratory inductive plethysmography ormagnetometry.The adverse effect of median sternotomy on thechest bellows is often underappreciated. A studyby Locke et al using magnetometers correlated theuncoordinated rib cage expansion that occurredfollowing sternotomy with the restrictive ventilatorychanges seen on PFT. Although air flow andabdominal wall motion remained coordinated,there was a delay in rib cage expansion. If increasedventilatory demands are required, this delay wouldact as a significant preload to inspiration with resultingincrease in work of breathing. Postoperativebone scans have detected occult rib fractures orcostochondral dislocations in two-thirds of patientsundergoing sternotomy. The worst postoperativelung function is noted in patients undergoing internalmammary artery grafting that is believed to bedue to loss of blood flow to the sternum. However,detailed studies show that a significant decrease inblood supply occurs in only approximately 5% ofpatients.The combination of diaphragmatic dysfunctionand chest bellows disease results in globalrespiratory muscle strength deterioration followingmedian sternotomy. This amounts to an approximately17% drop in the maximal inspiratory pressuremeasured at the mouth and a decrease of themaximal expiratory pressure of 47%, which reverseswhen measured 6 weeks postoperatively.Other <strong>Pulmonary</strong> Complications FollowingCardiac SurgeryLeft-sided pleural effusions occur in 50 to80% of patients undergoing left internal mammaryartery grafting and in 35% receiving only saphenousvein grafting. These effusions are bilateralin approximately 10% of all CABG patients. Theeffusions are characterized by having a markedlymphocytosis ( 82%) and a potential to have atrapped lung due to fibrosis. The effusions occurless frequently when valvular surgery alone isperformed. Usually, pleural effusions are smallto moderate in volume but contribute to a moremarked reduction in FVC in patients followingCABG then when effusions are not present (38%decrease compared to 26%). Because pleurotomy isrequired when internal mammary artery grafting isused, the occurrence of pleural effusions 3 weeksafter CABG is more common in these patients.Delayed massive effusions have also been reported.238 <strong>Pulmonary</strong> Complications of Cardiothoracic Surgery and Trauma (Krieger)ACC-PULM-09-0302-015.indd 2387/10/09 8:27:24 PM


Etiologies for these effusions include perioperativehemorrhage or contusion, pulmonary thromboemboli,and postcardiac injury syndrome. Postcardiacinjury syndrome (Dressler syndrome) usuallyoccurs 1 to 12 weeks after surgery and is associatedwith a combination of pleuritis, pericarditis, andpneumonitis, along with fever, leukocytosis, andelevation of the sedimentation rate. The diagnosisis one of exclusion but is important to establish sincethis syndrome requires treatment with nonsteroidalantiinflammatory drugs or systemic corticosteroids.The incidence of this syndrome appears tobe declining.Deep venous thrombosis and pulmonaryemboli occur less commonly following cardiacsurgery than after other major surgical procedures.It is believed that the use of anticoagulants duringECMO may protect the patient from the developmentof a nidus for clot formation intraoperatively.The incidence of pulmonary emboli reported afterCABG is approximately 1%, although when pulmonaryembolism occurs there is a significant ( 30%)postoperative mortality. This may reflect the findingthat when pulmonary thromboembolic diseaseoccurs in patients after CABG, it usually is in thesetting of other postoperative complications.Infections significantly affect pulmonary functionafter sternotomy and thus contribute to morbidityand mortality. Postoperative pneumoniahas been associated with preexisting pulmonarydysfunction and older patients. However, woundinfections occur more commonly than postoperativepneumonias. When sternal infections occur,significant thoracic instability results in deleteriouseffects reflected in decreased lung volumes andrespiratory muscle endurance.Respiratory Management FollowingCardiac SurgeryThe most important aspect of treating patientswith respiratory failure following cardiac surgeryis to determine the etiology or etiologies fortheir inability to support their required V • e postoperatively.Astute assessment of the patient fordiaphragmatic dysfunction, thoracic instability,pulmonary edema (which may be radiographicallymasked by the use of high-level positivepressure ventilation) and nonpulmonary sourcesof increased ventilatory requirements is important.Often, the most important approach to “weaning”the patient postoperatively is to place the patientback on full MVS while improving their mechanics,decreasing their V • e requirements, and controllingtheir pain. The need for reintubation after CABG isassociated with prolonged hospitalization, infectiouscomplications, and increased mortality. Riskfactors for prolonged MVS include a reduced FEV 1,renal failure, a positive fluid balance 24 h postoperatively,and the presence of specific polymorphismsin the promoter region of the TNF gene.ThoracoscopyTechnological advances, such as thoracoscopy,have allowed access to body cavities without theneed for traditional surgery. By using fiberoptictechnology and miniaturized cameras, diagnosticand therapeutic procedures are now commonlyperformed without requiring large incisions andprolonged recovery time. This type of surgery hasbeen termed minimally invasive, and when performedby thoracic surgeons usually utilizes videoassistance (video-assisted thoracic surgery [VATS]).Proponents of VATS consider this procedure to havepotentially revolutionary effects on the fields ofpulmonology and cardiothoracic surgery.The history of thoracoscopy dates back to 1910,when Jacobaeus diagnosed two cases of tuberculouspleuritis. Thoracoscopy was used during the firsthalf of the 20th century exclusively for the lysis ofpleural adhesions by means of cautery. “Medical”thoracoscopy, performed under conscious sedationusing nondisposable rigid instruments, is stillcommonly used in Europe as a means of diagnosingpleural disease. However, in North America, VATSis more commonly utilized, not only for diagnosisbut also for resection and therapeutic procedures.The advantage of VATS over thoracotomy isthe ability to access the thoracic cavity withoutrib spreading or rib resection, resulting in lesspostthoracotomy pain, neuralgia, and delayedrecovery. The procedure still requires generalanesthesia, unilateral lung ventilation, and lack ofsignificant pleural adhesions that would preventsafe insertion of instruments through small (2 cm)intercostal incisions. With perfection of the technique,complications are rare. This has become thediagnostic procedure of choice for patients withdiffuse interstitial lung disease who require lung<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 239ACC-PULM-09-0302-015.indd 2397/10/09 8:27:24 PM


iopsy and the therapeutic treatment of choice forpleural diseases such as persistent pneumothoraxor empyema. In addition, as discussed elsewhere inthe syllabus, it has become one of the most commonapproaches to LVRS.Chest TraumaTrauma remains the major cause of accidentaldeath in the United States. Thoracic injuries accountfor approximately 37,000 deaths per year and significantlycontribute to another 37,000 trauma-relateddeaths. Fatal hemorrhage or an inability to securean adequate airway are the major causes of deathin one third of these injuries. Other life- threateningconditions associated with chest trauma are listedin Table 4. The severity of the injury relates to themagnitude and duration of the applied force, thevelocity and contact area of that force, body compressionthat occurs in blast injuries and falls, andthe hyperdynamic stress response that followstrauma. This stress response is manifested by anincrease in cardiac output, V • o 2, and carbon dioxideproduction, along with a decrease in systemicvascular resistance and oxygen extraction relativeto oxygen delivery. It is believed that the posttraumatic“stress” results from cytokine release frommacrophages that stimulates increased toxic oxygenradical production, PMN activation, and the releaseof endotoxin from ischemic tissues. The pulmonologistis often involved after the initial resuscitationto deal with problems such as hypoxemia (Fig 2)associated with fractures, flail chest, and pulmonarycontusions. In addition, the pulmonologist/intensivistmay also be asked to evaluate the patient formyocardial injuries or tracheobronchial tears.Tracheobronchial TearsTable 4. Life-Threatening Chest TraumaAcutelyAirway obstructionTension pneumothoraxOpen pneumothoraxFlail chestMassive hemothoraxPotentiallyTracheobronchial tearLung contusionMultiple rib fracturesCardiac tamponadeAortic ruptureFigure 2. Factors that contribute to hypoxemia after chesttrauma.Although uncommon, tracheobronchial tearsmay be life threatening and are frequently overlooked.The diagnosis is suspected when persistentbarotrauma and air leaks persist followingblunt chest trauma. Hemoptysis is usually, but notalways, present along with concomitant fracture ofthe first and second ribs. The tear is usually within2.5 cm of the carina and often involves the membranoustrachea.Bronchoscopic findings include lacerations,transections, and bared cartilage that may be hiddenbelow “heaped-up” mucosa. Early surgical repairis usually required except for small tears (less thanone third the circumference of the bronchus ortrachea). Stabilization often requires placement ofa double-lumen tube or use of high-frequency jetventilation prior to repair.BarotraumaPneumothorax and hemothorax are potentiallylife-threatening complications of chest trauma.Pneumomediastinum is reported in 5 to 10% ofblunt chest trauma victims. This form of barotraumacan result from tracheobronchial tears, pneumothoraxwith or without associated rib fractures, or theMacklin effect. The Macklin effect involves alveolarrupture that results in dissection of air along thebronchovascular sheath (pulmonary interstitialemphysema) and then into a mediastinum. TheMacklin effect has been identified by chest CTscanning as the cause of the pneumomediastinumin approximately 40% of blunt trauma cases andwas associated with a significantly prolonged stayin the ICU.Flail ChestRib or sternal fractures are caused by suddendecompression forces. Children’s ribs are more240 <strong>Pulmonary</strong> Complications of Cardiothoracic Surgery and Trauma (Krieger)ACC-PULM-09-0302-015.indd 2407/10/09 8:27:24 PM


Table 5. Treatment of Rib Fractures*CPAP Mask CPAP via ETT(n = 36) (n = 33)Age (yrs) 46 58CPAP (cm H 2O) 5.4 6.6Barotrauma 3 3Pneumonia 5 16 †Septicemia 0 3Death 0 2LOS (days) 8.8 14.6* CPAP continuous positive airway pressure;ETT endotracheal tube; LOS length of stay. Adaptedfrom Bollegir and van Eeden (Chest 1990; 97:943–948)†p 0.05 compared with CPAP mask group.flexible and therefore are less frequently fracturedby the same degree of force compared to older victims’ribs. When first or second ribs are fractured,suspicion is raised for injury to great vessels or to thetracheobronchial tree. Prior to 1960, various methodswere used to try to stabilize these fractures. After1960, “internal stabilization” using positive pressuremechanical ventilation with an endotrachealtube was the treatment of choice. Presently, internalstabilization via continuous positive airway pressure(CPAP) mask is the desired approach if the patient’sventilatory status does not require MVS (Table 5).For a flail chest to occur, two or more contiguous ribsneed to be fractured in two or more places that resultsin an unstable segment of the chest wall that paradoxesinward during inspiration because of negativeintrathoracic pressure. However, diagnosis of a flailchest can be delayed if the patient is only examinedwhile receiving ventilation with positive pressure.The hypoxemia that frequently accompanied a flailchest was thought to be due to “internal rebreathing”(Pendelluft). However, in a canine model of flail chestthat specifically avoided any underlying pulmonarycontusion, no significant changes in rib cage distortionor oxygenation occurred in the experimentalanimals vs controls despite multiple rib fractures.Therefore, it was concluded that the hypoxemiathat accompanies a flail chest is due to underlyingpulmonary contusion and other associated injuriesand not to internal rebreathing (Fig 2).<strong>Pulmonary</strong> Contusion From Blunt Chest TraumaExternal forces applied to the pulmonaryparenchyma during blunt chest trauma, such asacceleration/deceleration during motor vehicleaccidents, or blast injury from explosions, canresult in disruption of the lung parenchyma. Thisdisruption is amplified if the glottis is closed at thetime that airway pressure is suddenly increased.It causes disruption of alveolar-capillary membranes,intraparenchymal hemorrhage, an increasein ventilation/perfusion mismatch, shunt, anddead space as well as increased permeability andextravascular lung water. <strong>Pulmonary</strong> edema followingterrorist incidents is often complicated byother tertiary effects, such as crush, shrapnel, orinhalational injuries. Nosocomial pneumonia andARDS can occur early or late. The treatment forpulmonary contusion is similar to treating otherforms of ARDS except that concomitant injuries,such as barotrauma and neurological injuries, mayrequire modification of ventilatory strategies (suchas having to avoid permissive hypercarbia) thatwould otherwise be used. If the patient survives hisor her injuries, lung recovery is usually completeexcept if there has been significant permanent chestwall damage.Nonpulmonary Effects of Blunt Chest TraumaDisruption of the esophagus, diaphragm, orgreat vessels can also occur. However, myocardialinjuries are more frequent. At the present time, themost accurate test for the diagnosis of myocardialinjury is either surface or transesophageal echocardiography.Other modalities, such as cardiacenzymes, are too nonspecific to be diagnostic. Anechocardiogram is able to image wall motion as wellas diagnose potentially life-threatening pericardialeffusion and valvular damage.The pulmonologist’s role in the ICU has dramaticallyexpanded over the past decade. Therefore,familiarity with traditional surgical fields, such astrauma, cardiac surgery, and thoracic surgery isessential for the care of our patients.Annotated BibliographyPostthoracotomy <strong>Pulmonary</strong> FunctionDatta D, Lahiri B. Preoperative evaluation for patientsundergoing lung resection surgery. Chest 2003; 123:2096–2103<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 241ACC-PULM-09-0302-015.indd 2417/10/09 8:27:25 PM


Outlines the basis for the recommended preoperative physiologicassessment of a patient prior to undergoing thoracotomyas outlined in Figure 1.Girish M, Trayner E, Dammann O, et al. Symptomlimitedstair climbing as a predictor of postoperativecardiopulmonary complications after high-risk surgery.Chest 2001; 120:1147–1151The inability to climb two flights of stairs was associated withan 82% positive predictive value for the development of apostoperative cardiopulmonary complication.Markos J, Mullan BP, Hillman DR, et al. Preoperativeassessment as a predictor of mortality and morbidity afterlung resection. Am Rev Respir Dis 1989; 139:902–910Used combination of postoperative predicted FEV 1 35% andDLCO 35% to predict an unacceptably high risk of mortalityfollowing lung resection.Marshall MC, Olsen GN. The physiologic evaluation ofthe lung resection candidate. Clin Chest Med 1993;14:305–320<strong>Review</strong> of the history and scientific data of how to determinepost-thoracic surgery pulmonary complications by one of thepioneers in this field (Dr. Olsen).Reilly JJ. Evidence-based preoperative evaluation ofcandidates for thoracotomy. Chest 1999; 116:474S–476SSummary of criteria that can be used to predict postoperativemorbidity and mortality, including combined cardiacpulmonaryrisk index.Weisman IM. Cardiopulmonary exercise testing in preoperativeassessment for lung resection surgery. SeminThorac Cardiovasc Surg 2001; 13:116–125Concise review of how to use exercise testing to help determinethe advisability of allowing borderline candidates to undergolung resection. V • O 2max 40% of predicted or 10 mL/kg/minprecludes resection surgery.Win T, Groves AM, Wells FC, et al. Relationship ofshuttle walk test and lung cancer surgical outcome. EurJ Cardiothorac Surg 2004; 26:1216–1219The shuttle (6 min) walk distance was not predictive of a poorsurgical outcome.Zeihar BG, Gross TJ, Kern JA, et al. Predicting postoperativepulmonary function in patients undergoing lungresection. Chest 1995; 108:68–72Reports an underestimation of postoperative FEV 1when thenumber of segments (times 5.2% per segment) to be removedare used to predict postoperative lung function.Chest TraumaBolliger CT, van Eeden SF. Treatment of multiple ribfractures: randomized trial comparing ventilatory withnonventilatory management. Chest 1990; 97:943–948Patients who did not have concomitant severe pulmonarycontusion or did not require associated laparotomies recoveredbetter when CPAP was delivered noninvasively.Cappello M, Yuehua C, de Troyer A. Rib cage distortionin a canine model of flail chest. Am J Respir Crit CareMed 1995; 151:1481–1485In a canine model of flail chest without underlying pulmonarycontusion, there were no significant harmful effects on breathingpattern, ventilation, or oxygenation.Demling RH, Pomfret EA. Blunt chest trauma. NewHoriz 1993; 1:402–421<strong>Review</strong> article with multiple references.DePalma RG, Burris DG, Champion HR, et al. Blastinjuries. N Engl J Med 2005; 352:1335–1342The onslaught of terrorism acts has made this review articleall too relevant.Hirshberg B, Oppenheim-Eden A, Pizoo R, et al. Recoveryfrom blast lung injury: one-year follow-up. Chest1999; 116:1683–1688Despite requiring mechanical ventilation because of severelung injury, victims of blast injuries frequently recovered tonormal lung function.Kshettry VR, Bolman RM III. Chest trauma: assessment,diagnosis, and management. Clin Chest Med 1994;15:137–153<strong>Review</strong> article.Newman PG, Feliciano DV. Blunt cardiac injury. NewHoriz 1999; 7:26–34<strong>Review</strong> article.Pretre R, Chilcott M. Blunt trauma to the heart and greatvessels. N Engl J Med 1997; 336:626–632<strong>Review</strong>s myocardial and great vessel injuries.Shaker KG, Hollingworth HM, Irwin RS, et al. Tracheobronchialinjuries caused by blunt trauma. J IntensiveCare Med 1995; 10:226–233Comprehensive review with illustrative chest radiographs.Wintermark M, Schnyder P. The Macklin effect: a frequentetiology for pneumomediastinum in severe bluntchest trauma. Chest 2001; 120:543–547The Macklin effect was identified by chest CT scans as the causeof pneumomediastinum in 39% of 51 patients, and was associatedwith prolonged ICU stays.Postcardiac Surgery <strong>Pulmonary</strong> ComplicationsBaisden CE, Greenwald LV, Symbas PN. Occult rib fracturesand brachial plexus injury following median sternotomyfor open-heart operation. Ann Thorac Surg 1984;38:192–194Two-thirds of the patients studied had positive bone scans.242 <strong>Pulmonary</strong> Complications of Cardiothoracic Surgery and Trauma (Krieger)ACC-PULM-09-0302-015.indd 2427/10/09 8:27:25 PM


Bardell T, Legare JF, Buth KC, et al. ICU readmission aftercardiac surgery. Eur J Cardiothorac Surg 2003; 23:354–359Preoperative renal failure and prolonged mechanical ventilation( 24 h) postoperatively predicted a higher incidence ofICU readmission.Branca P, McGaw P, Light RW, et al. Factors associatedwith prolonged mechanical ventilation following coronaryartery bypass surgery. Chest 2001: 119:537–546Fewer than 5% of patients required mechanical ventilation for 3 days after CABG. Preoperative cardiac or respiratoryinstability are predictive of prolonged MVS.Cohen AJ, Katz MG, Frenkel G, et al. Morbid results ofprolonged intubation after coronary artery bypass surgery.Chest 2000: 118:1724–1731Sixty-six of 1,112 patient undergoing CABG required MVSfor 48 h or reintubation. Multivariate regression analysisidentified three factors that were associated with prolongedintubation: elevated creatinine level, lower FEV 1, and apositive fluid balance at 24 h postoperatively. Prolongedintubation significantly increased hospital stay and mortalityrates.Dimopoulou I, Daganou M, Dafni U, et al. Phrenic nervedysfunction after cardiac operations: electrophysiologicevaluation of risk factors. Chest 1998; 113:8–14Thirteen of 63 surgery patients had delayed phrenic nerveconduction latency times postoperatively compared to preoperativetesting. The only risk factor that was significantlyassociated with this complication by logistic regression analysiswas the use of cardioplegic ice slush.El-Chemaly S, Abreu A, Krieger B. What are the risks ofpulmonary complications after cardiac surgery? J CritIllness 2003; 18:266–273General review with 39 references.Hoffman R, Yahr W, Krieger B. Diaphragmatic flutterresulting in failure to wean from mechanical ventilatorsupport after coronary artery bypass surgery. Crit CareMed 1990; 18:499–501Report of four patients who had diaphragmatic flutter aftercardiac surgery.Lee YC, Vaz MAC, Ely KA, et al. Symptomatic persistentpostcoronary artery bypass graft pleural effusionsrequiring operative treatment: clinical and histologicfeatures. Chest 2001; 119:795–800The effusions were lymphocytic ( 80% lymphocytes) and oftenresulted in fibrosis and occasional trapped lungs.Locke TJ, Griffith TL, Mould H, et al. Rib cage mechanicsafter median sternotomy. Thorax 1990; 45:465–468Thoracic wall discoordination was documented by magnetometersin 9 of 16 patients 1 week postoperatively.Massoudy P, Zahler S, Becker BF, et al. Evidence forinflammatory responses of the lungs during coronaryartery bypass grafting with cardiopulmonary bypass.Chest 2001; 119:31–36Proinflammatory cytokines (interleukin-6 and -8) increasedin pulmonary venous blood after bypass. In addition, adhesionmolecule counts on platelets and PMNs also becameelevated.Maillet JM, Besnerais P, Cantoni M, et al. Frequency, riskfactors and outcome of hyperlactatemia after cardiacsurgery. Chest 2003; 123:1361–1366A lactate level 3 mmol/L measured within 4 h after surgerywas associated with increased mortality.Milot J, Perron J, Lacasse Y, et al. Incidence and predictorsof ARDS after cardiac surgery. Chest 2001; 119:884–888The incidence of ARDS in a retrospective analysis of 3,278patients was 0.4% but carried a 15% mortality rate. Riskfactors identified by multivariate analysis included previouscardiac surgery, shock, and number of blood productsreceived.Ng CSH, Wan S, Yim APC, et al. <strong>Pulmonary</strong> dysfunctionafter cardiac surgery. Chest 2002; 121:1269–1277This article reviews the associated physiologic, biochemical,and histologic changes associated with the pulmonary dysfunctionfollowing cardiac surgery with particular attention toPMN activation, free radicals, and cytokines.Rady MY, Ryan T. Perioperative predictors of extubationfailure and the effect on clinical outcome after cardiacsurgery. Crit Care Med 1999; 27:340–347Clinical predictors of extubation failure included age 65 years;inpatient hospitalization; COPD; pulmonary hypertension;severe left ventricular dysfunction; redo operation; thoracicaorta procedures; and bypass time 120 min.Wynne R, Botti M. Postoperative pulmonary dysfunctionin adults after cardiac surgery with cardiopulmonarybypass: clinical significance and implications for practice.Am J Crit Care 2004; 13:384–393A nursing review that is worth reading with 159 references.Yende S, Quasney M, Tolley E, et al. Association of tumornecrosis factor gene polymorphisms and prolongedmechanical ventilation after coronary artery bypasssurgery. Crit Care Med 2003; 31:133–140Yende S, Quasney M, Tolley E, et al. Clinical relevance ofangiotensin-converting enzyme gene polymorphisms topredict risk of mechanical ventilation after coronaryartery bypass graft surgery. Crit Care Med 2004; 32:922–927The presence of a specific haplotype in the promoter region ofthe TNF gene site and angiotensin-converting enzyme gene<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 243ACC-PULM-09-0302-015.indd 2437/10/09 8:27:25 PM


polymorphisms were associated with a significantly higherrate of prolonged MVS following a CABG procedure.ThoracoscopyColt HG. Thoracoscopy: window to the pleural space.Chest 1999; 116:1409–1415<strong>Review</strong>s the indications for VATS and quotes a procedurerelatedmortality rate of 0.24% “in experienced hands.”Loddenkemper R. Thoracoscopy: state of the art. EurRespir J 1998; 11:218–221Comprehensive review of the diagnostic utility of “medical”thoracoscopy as performed by pulmonologists in Europe.244 <strong>Pulmonary</strong> Complications of Cardiothoracic Surgery and Trauma (Krieger)ACC-PULM-09-0302-015.indd 2447/10/09 8:27:26 PM


Sleep PhysiologyTeofilo L. Lee-Chiong, Jr., MD, FCCPObjectives:• Understand the various neural processes that control sleepand wakefulness• Describe the physiologic changes that occur duringsleep• Identify the consequences of sleep deprivation• <strong>Review</strong> the rules related to the scoring of sleep stages andrespiratory events in adultsKey words: circadian rhythm; polysomnography; sleep; sleepdeprivation; sleep homeostasis; sleep physiologySleep is a complex reversible state characterizedby both behavioral quiescence and diminishedresponsiveness to external stimuli. Sleep is generatedand maintained by CNS networks usingspecific neurotransmitters localized in specificareas of the brain.Neuroscience of SleepNeural systems generating wakefulnessinclude the ascending reticular formation in themedulla, pons, and midbrain; basal forebrain (neurotransmitter:acetylcholine); hypothalamus(hypocretin); locus ceruleus (norepinephrine);tuberomammillary nucleus (histamine); and ventraltegmental area and substantia nigra (dopamine).Neural systems generating non-rapid eyemovement (NREM) sleep consist of the hypothalamus(gamma aminobutyric acid [GABA]),basal forebrain (GABA and adenosine), and thalamus,which is where sleep spindles are generated.Finally, neural systems generating rapid eye movement(REM) sleep include the pons and basalforebrain. REM sleep is associated with activationof “REM-on” neurons (cholinergic) and inhibitionof “REM-off” neurons (noradrenergic, serotonergic,and histaminergic).Sleep and wake neurotransmitters consist ofthe following: (1) acetylcholine, the main REM sleepneurotransmitter; (2) adenosine, which is responsiblefor homeostatic sleep drive; (3) glutamate,the main CNS excitatory neurotransmitter;(4) GABA, the main CNS inhibitory neurotransmitterand the main NREM sleep neurotransmitter;(5) glycine, the main inhibitory neurotransmitter inthe spinal cord and is responsible for REM sleeprelatedmuscle atonia/hypotonia; and (6) hypocretin,the dysfunction of which is responsible fornarcolepsy.Sleep-Wake RegulationTwo basic intrinsic components interact toregulate the timing and consolidation of sleep andwake: sleep homeostasis, which is dependent onthe sleep-wake cycle, and circadian rhythm, whichis independent of the sleep-wake cycle. 1 These twoprocesses influence sleep latency, duration, andquality. In addition, the timing of sleep is alsodetermined by behavioral influences, such as socialactivities and work schedules.Sleep homeostasis is defined as increasingsleep pressure related to the duration of previouswakefulness: the longer a person is awake, thesleepier one becomes. This sleep pressure thendecreases after a sufficient duration of sleep.The circadian system consists of biologicalrhythms that are ubiquitous and genetically determined.Natural circadian rhythms free-run atslightly 24 h ( 24.2 h [referred to as “tau”]). Theprocess of entrainment adjusts the circadian rhythmto the external 24-h period by the use of environmentalcues called zeitgebers. Zeitgebers can bephotic or nonphotic (eg, mealtimes and social schedules)and cause either a phase advance, a shift of thecircadian period to an earlier time, or a phase delay,a shift of the circadian period to a later time.The suprachiasmatic nucleus (SCN) in theanterior hypothalamus is the master circadianrhythm generator in mammals. Other anatomicsites also harbor endogenous clocks. SCN activityis greater during the daytime than at night. Actionsof the SCN consist of promotion of wakefulnessduring the day as well as consolidation of sleepduring the night. Ablation of the SCN results in<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 245ACC-PULM-09-0302-016.1.indd 2457/10/09 10:17:33 PM


oth random distribution of sleep throughout theday and night and a reduction in the duration ofwaking periods (in some).There are two circadian peaks in wakefulness(wake-maintenance zones), namely in the latemorning and early evening; there are also twocircadian troughs in alertness (increased sleeppropensity) in the early morning and earlymidafternoon.Several afferent pathways provide inputs, bothphotic and nonphotic, to the SCN. The main afferentconnection involves the retinohypothalamictract that relays photic stimuli from the photosensitiveretina ganglion cells containing melanopsinto the SCN by the use of glutamate as its neurotransmitter.Retinal photoreceptors are mostsensitive to shorter-wavelength light (450 to 500nm [blue to blue-green]). An alternate afferent connectionexists that relays stimuli via the thalamiclateral geniculate nuclei and geniculohypothalamictract. Light entrainment is lost with disruption ofthe main pathway but not of the alternate pathway.The SCN, in turn, has widespread efferent projectionsto the basal forebrain, hypocretin neurons ofthe hypothalamus, locus ceruleus, thalamus, andpineal gland.Melatonin is synthesized and released by thepineal gland, with greatest secretion at night. Secretioninhibited by light exposure. Actions of melatoninon the sleep-wake rhythm consist of a phasedelay when administered in the morning andphase advance when given in the evening. Melatoninis less effective in phase shifting circadianrhythms than light exposure, but it possesses mildhypnotic properties.Physiologic Processes During SleepAutonomic Nervous SystemCompared with wakefulness, during NREMsleep there is a relative decrease in sympatheticactivity, accompanied by an increase in parasympatheticactivity. There is a further reduction insympathetic activity and enhanced parasympatheticactivity during REM sleep compared withNREM sleep. Transient increases in sympatheticactivity occur during phasic REM sleep.Respiratory SystemRespiration is controlled by both metabolic andbehavioral factors during the wake state. In contrast,only metabolic control is present duringsleep. Compared with levels during wakefulness,there is a decrease in both Pao 2and arterial oxygensaturation (Sao 2) and an increase in Paco 2duringsleep. Hypoxic and hypercapnic ventilatoryresponses, upper airway dilator muscle tone, andactivity of the accessory muscles of respiration alldiminish during NREM sleep, compared withwakefulness, and decrease further during REMsleep. The activity of the diaphragm remains intactduring REM sleep. NREM stage N1 sleep is characterizedby periodic breathing, with episodes ofhypopnea and hyperpnea. Respiratory patternsbecome stable and regular during stage N3 sleep.During REM sleep, respiration is irregular, withvariable respiratory rates and tidal volumes. Centralapneas or periodic breathing may emergeduring phasic REM sleep.Cardiovascular SystemDecreases in heart rate (HR), cardiac output(CO), and BP occur during NREM sleep comparedwith the wake state. Further reductions in HR, CO,and BP develop during tonic REM sleep comparedwith NREM sleep. Increases in HR, CO, and BPmay occur during phasic REM sleep compared wihNREM and tonic REM sleep, as well as duringawakenings. Nighttime systolic BP is commonlyapproximately 10% less than corresponding levelsduring the daytime (ie, “dipping” phenomenon).GI SystemSleep is associated with decreased swallowingrate, salivary production, and esophageal andintestinal motility. Basal gastric acid secretiondemonstrates a circadian pattern with peak levelsgenerally between 10:00 pm and 2:00 am and lowestlevels from 5:00 am to 11:00 am.Renal SystemUrine output diminishes during sleep as aresult of an increase in water reabsorption,246 Sleep Physiology (Lee-Chiong)ACC-PULM-09-0302-016.1.indd 2467/10/09 10:17:34 PM


decreased glomerular filtration, and increase inrelease of renin (during NREM sleep) and antidiuretichormone. Obstructive sleep apnea (OSA) isassociated with an increased risk of nocturia, withthe latter improving with optimal positive airwaypressure (PAP) therapy.Endocrine SystemCertain hormones exhibit increasing levelsduring sleep (eg, growth hormone, prolactin, parathyroidhormone, and testosterone), whereas levelsof others decrease during sleep (eg, cortisol, insulin,and thyroid-stimulating hormone). Acromegaly isassociated with a greater prevalence of OSA andcentral sleep apnea, and therapy of acromegalywith surgery or octreotide can improve sleeprelatedbreathing disorders (SRBDs). OSA canresult in lower levels of growth hormone and testosterone,both of which increase with PAP therapy.Finally, polycystic ovarian syndrome is associatedwith both increased testosterone levels and greaterrisk of OSA.OSA also increases the risk of insulin resistanceand type II diabetes mellitus, independent of bodymass index. There is an increase in both leptin andghrelin levels in OSA. Insulin sensitivity and levelsof leptin and ghrelin improve with PAP therapy.Musculoskeletal SystemSkeletal muscle relaxation (hypotonia or atonia)and inhibition of deep tendon reflexes developduring sleep.ImmunityProinflammatory cytokines, such as interleukin(IL)-1β and tumor necrosis factor-α, generallyenhance sleep, whereas antiinflammatory cytokines,including IL-4 and IL-10, suppress sleep.Viral and bacterial infections result in a reductionin NREM and increase in REM sleep. Changes inthe immune system also develop in persons withOSA, namely increases in C-reactive protein, IL-6,and tumor necrosis factor-α; levels improve withPAP therapy.ThermoregulationCore body temperature follows a circadianrhythmicity that peaks in the late afternoon andearly evening (from 6:00 pm to 8:00 pm) anddecreases at the onset of sleep, with lowest levelsat approximately 2 h before usual wake time (from4:00 am to 5:00 am). Changes in thermoregulationduring sleep include a decrease in core body temperature,decrease in thermal set point, reducedthermoregulatory responses to thermal challenges,and loss of heat production from shivering duringREM sleep. Nocturnal sleep typically occurs duringthe decreasing phase of the temperaturerhythm (after maximum core body temperature),whereas awakening occurs during the increasingphase of the temperature rhythm (after minimumcore body temperature).MetabolismMetabolism decreases during NREM sleepcompared with the wake state, and metabolic rateduring REM sleep is either similar or higher thanduring NREM sleep.Sleep DeprivationSeveral physiologic parameters becomeincreased during sleep deprivation, includingsubjective and objective sleepiness, sympatheticactivity, insulin resistance, and levels of cortisoland ghrelin. Medical errors and motor vehicleaccidents also increase. On the other hand, vigilance,cognition and attention, seizure threshold,resistance to infection, and levels of growth hormoneand leptin activity decrease during sleepdeprivation. 2PolysomnographyPolysomnography involves the continuous andsimultaneous recording of several physiologicvariables during sleep. These variables commonlyconsist of an electroencephalogram (EEG), electrooculogram(EOG), chin electromyogram (EMG),ECG, airflow, respiratory effort, oxygenation andventilation, snoring, and anterior tibialis EMG.Polysomnography is indicated for the following:(1) the diagnosis of SRBD; (2) PAP titration for<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 247ACC-PULM-09-0302-016.1.indd 2477/10/09 10:17:34 PM


SRBD; (3) follow-up after upper airway surgery ordental devices for OSA; and (4) evaluation of narcolepsy,idiopathic hypersomnia, periodic limbmovement disorder (PLMD), atypical or injuriousparasomnias, and nocturnal seizures.Diagnostic sleep studies can be classified intofour levels: level 1 (attended laboratory full polysomnography),level 2 (unattended full polysomnography),level 3 (cardiorespiratory sleep studiesincluding at least 4 bioparameters), and level 4 (limitedsleep studies with one to two bioparameters).A polygraph consists of a series of alternatingcurrent and direct current amplifiers and filters thatare used to record physiologic variables duringsleep. Alternating current amplifiers are commonlychosen for high-frequency (fast) physiologic variables,such as EEG, EOG, and EMG, whereas directcurrent amplifiers are used for low-frequency (slow)physiologic variables, such as oxygen saturation (ie,Sao 2). A derivation is defined as the difference involtage between two electrodes and can either bebipolar, ie, when two standard electrodes arematched to each other, or referential, ie, when a standardelectrode is matched to a reference electrode.In EEG, electrodes are placed based on theInternational 10-20 system, wherein each electrodeis provided with a letter, which represents a regionof the brain (F [frontal], C [central], O [occipital],and M [mastoid]), and a numerical subscript, withodd numbers representing left-sided electrodes,even numbers representing right-sided electrodes,and the designation Z used for midline electrodes.Recommended electrode placements for EEG areF4M1, C4M1, and O2M1. 3 Additional EEG electrodesare required when one is evaluating suspectednocturnal seizures.EEG voltage originates from the summedpotential activity of cortical neurons. EEG wavescan be differentiated based on frequency of EEGwaves into delta waves ( 4 Hz), theta waves(4–7 Hz), alpha waves (8–13 Hz), and beta waves( 13 Hz). Delta waves define the presence of stageN3 sleep. Theta waves are present in stages N1, N2,and R sleep. Alpha waves are seen during drowsinesswith eyes closed. Beta waves are presentduring alert wakefulness. Stage N2 sleep is characterizedby the presence of either K complexes orsleep spindles. A K complex consists of a highamplitude,biphasic wave with duration of 0.5 s andan initial sharp negative deflection followed bypositive high- voltage slow wave. Sleep spindles areoscillations with a frequency of 12 to 14 Hz lasting0.5 to 1.5 s; their amplitude is generally 50 μV.EOG is used to record the difference in potentials,referred to as a dipole, between the cornea,which is positively charged, and the negativelycharged retina. Recommended electrode placementsfor EOG are E1M2 and E2M2, with E1 representingan electrode below the left outer canthus,E2 an electrode above the right outer canthus, andM2 the right mastoid process. 3 The dipole changeswith eye movements. Two patterns of eye movementscan often be seen: slow rolling eye movementsthat occur during drowsiness when eyes areclosed and REMs that occur during wakefulness(ie, eye blinks) or REM sleep.Recording chin EMG involves the use of threeelectrodes that are located midline above the mandible,right of midline below the mandible, and leftof midline below the mandible. Derivation consistsof one electrode below and one electrode above themandible. ECG uses a modified lead II electrode.Airflow is measured by the use of either anoronasal thermal sensor, which is the recommendedtechnique for identifying apneas, or anasal air pressure transducer, which is recommendfor identifying hypopnea. With nasal air pressuremonitoring, inspiratory flow signals show a plateau(flattening) with obstructive events or reducedbut rounded signal with central events.Measuring respiratory effort involves esophagealmanometry or inductance plethysmography,which are important in distinguishing obstructive,central, and mixed apneas.Pulse oximetry is the recommended techniquefor measuring oxygenation, whereas transcutaneous(Ptcc0 2) or end-tidal (Petc0 2) monitoring isrecommended for the measurement of alveolarhypoventilation, especially in children. A microphoneis used to identify snoring.EMG of the anterior tibialis is important in thedetection of periodic limb movements duringsleep. Additional EMG electrodes can be placed inthe upper extremities to aid in the diagnosis ofREM sleep behavior disorder.Scoring Adult Sleep StagesPolysomnographic data are divided into 30-stime periods, or epochs. Standard sleep study248 Sleep Physiology (Lee-Chiong)ACC-PULM-09-0302-016.1.indd 2487/10/09 10:17:34 PM


paper speed is 10 mm/s (30 cm per epoch page).Each epoch is assigned a sleep stage that comprisesthe greatest percentage of the epoch. Sleep amongadults is typically composed of stage N1 (5%), N2(45%), N3 (25%) and R (25%).Stage wake is scored when 50% of the epochhas alpha EEG waves when eyes are closed. If alphawaves are absent, stage wake is defined by thepresence of vertical eye blinks, reading eye movements,or voluntary rapid open eye movements.Chin EMG tone is generally high. 3 Stage N1 sleepis present if alpha waves are replaced by low amplitude,mixed frequency (4–7 Hz) waves that occupy 50% of the epoch. Slow eye movements may bepresent. Sleep spindles and K complexes are absent.Tonic chin EMG levels are lower than stagewake. 3Stage N2 sleep is defined by low amplitude,mixed frequency waves with the presence of Kcomplexes (not associated with arousals) or sleepspindles during the first half of the epoch or duringthe last half of the previous epoch. 3 Stage N3 sleepis scored if at least 20% of the epoch is occupied byslow wave (0.5–2 Hz and 75 μV) EEG activityover the frontal regions. 3 Stage REM sleep is characterizedby low amplitude, mixed frequency EEGactivity, REMs, and low chin EMG tone. 3Scoring Respiratory and Movement EventsAn apnea is defined by a decrease in peakthermal sensor amplitude by at least 90% ofbaseline for a duration of at least 10 s. Apneas canbe either obstructive, when inspiratory effort ispresent throughout the entire event; central ifinspiratory effort is absent throughout the entireevent; or mixed when a central event is followedby an obstructive event. 3Hypopnea is a decrease in nasal pressure by atleast 30% of baseline for duration of at least 10 saccompanied by oxygen desaturation of at least4%. 3 A respiratory effort-related arousal is characterizedby breaths associated with increasingrespiratory efforts for at least 10 s and flattening ofthe nasal pressure waveform. Event precedes anarousal, and does not meet criteria for eitherapneas or hypopneas.Hypoventilation is present when there is anincrease in Paco 2of at least 10 mm Hg during sleepcompared to supine wake levels. 3 Cheyne-Stokesrespiration (CSR) is characterized by a crescendodecrescendoamplitude in respiration, with at least5 central apneas/hypopneas per hour of sleep. 3Periodic limb movements during sleep arescored if there are at least 4 consecutive leg movements,each 0.5–10 s in duration, and if the movementsare between 5 and 90 s apart. 3 REM sleepbehavior disorder is suggested by sustained chinEMG muscle activity or the presence of excessivetransient chin or limb EMG muscle activity duringREM sleep. 3Definitions of Sleep ParametersThe time when a sleep recording is started isdesignated lights out, and the time when sleeprecording ended is termed lights on. Time in bed(TIB) is the duration of monitoring between “lightsout” and “lights on.” Total sleep time (TST) is thesum of all sleep stages (ie, N1-3 sleep plus REMsleep). Time in bed and total sleep time are used tocalculate sleep efficiency (SE), which is the ratio ofTST to TIB [(TST 100)/TIB]. Wake time aftersleep onset (WASO) is the time spent awake fromsleep onset to final awakening.Sleep onset latency (SOL) is defined as the timefrom lights out to the onset of sleep (ie, the firstepoch of any sleep stage). REM sleep latency is thetime from sleep onset to the first epoch of REMsleep. REM sleep occurring within 10 to 15 min ofsleep onset is referred to as a sleep onset REM period(SOREMP).The apnea-hypopnea index (AHI) is the numberof apneas plus hypopneas per hour of sleep.The number of arousals per hour of sleep is referredto as an arousal index. An oxygen desaturationindex (ODI) refers to the number of oxygen desaturationevents per hour of sleep. Finally, theperiodic limb movement index (PLMI) is definedas the number of periodic limb movements perhour of sleep.Sleep ArchitecturePolysomnographic features of many primarysleep, medical, neurologic and psychiatric disorderscan be differentiated into two general patternsof sleep architecture: a high sleep input pattern(short sleep latency, high sleep efficiency, increasedtotal sleep time and reduced wake time after sleep<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 249ACC-PULM-09-0302-016.1.indd 2497/10/09 10:17:34 PM


onset) and a low sleep input pattern (prolongedsleep latency, reduced sleep efficiency andtotal sleep time, and increased wake time aftersleep onset). The high sleep input pattern can beseen following sleep deprivation, in disorderspresenting with excessive sleepiness, such asnarcolepsy, and during administration of sedatingmedications, whereas the low sleep input patternoften accompanies disorders presenting withinsomnia or use of stimulant medications.Epworth Sleepiness ScaleThe degree of sleepiness is often subjectivelydetermined using an eight-item questionnaire thatmeasures a person’s propensity to fall asleep inspecific situations in recent times, including thefollowing: (a) sitting and reading, (b) watchingtelevision, (c) sitting and inactive in a public place,(d) riding as a passenger in a car for an hour withouta break, (e) lying down to rest in the afternoon,(f) sitting and talking to someone, (g) sitting quietlyafter lunch without drinking alcohol, and(h) stopped in a car for a few minutes in traffic. Thechances of dozing are rated as one of the following:0 (never), 1 (slight chance), 2 (moderate chance) or3 (high chance). An aggregate score between 0 and9 is considered as representing a normal level ofalertness. Scores of 10 and above suggest the presenceof sleepiness that might require intervention.The correlation between sleepiness measuredsubjectively by the Epworth sleepiness scale andobjective measures using the multiple sleep latencytest (MSLT) and the maintenance of wakefulnesstest (MWT) is poor. Epworth sleepiness scale scoresare elevated in patients with OSA; scores improvefollowing effective therapy for OSA.Multiple Sleep Latency TestThe MSLT is an objective measure of a person’stendency to fall asleep in quiet situations,and is indicated for evaluation of unexplainedsleepiness and suspected narcolepsy, and to distinguishbetween narcolepsy and idiopathichypersomnia. 4 A regular sleep-wake schedule isrequired for at least 1 to 2 weeks prior to testing.Medications that affect sleep latency and REMsleep should be discontinued for at least 2 weeksbefore performing the test. Polysomnography(but not not a split-night study) should be performedimmediately before an MSLT to ensureat least 6 h of sleep and to exclude the presenceof either OSA or PLMD that can affect the interpretationof study results. In persons with OSA,PAP therapy should be used during both polysomnographyand MSLT. Smoking is not allowedprior to each nap trial, and persons should notdrink caffeine or engage in vigorous physicalactivity on the day of the study. A urine drugscreen may be performed on test day.The multiple sleep latency test consists of4 or 5 nap opportunities performed every 2 h,with each nap trial lasting 20 min in duration.The patient is asked to lie down in a comfortableposition in a dark, quiet room, close his/her eyesand try to fall asleep, while EEG, EOG, chinEMG, and ECG are continuously monitored.Sleep onset latency and SOREMPs, if present, aredetermined for each nap trial. Sleep onset latencyis recorded as 20 min if no sleep occurs duringa nap trial. Each nap trial is terminated after 20min if no sleep is recorded; if sleep is noted, thetest is continued for another 15 min to allowREM sleep to occur. The test is stopped after thefirst epoch of REM sleep. The patient is asked toget out of bed and to remain awake between naptrials.A short mean SOL, ie, 8 min, suggests thepresence of excessive sleepiness. A short SOL ispresent in persons with narcolepsy, idiopathichypersomnia, OSA and PLMD. It can also be seenduring sleep deprivation, following acutewithdrawal of stimulant agents, and in 15% to 30%of healthy individuals. Sleep onset REM periodsare present in persons with narcolepsy or OSA,during sleep deprivation, after withdrawal fromREM suppressants or alcohol, and in 1% to 3% ofhealthy individuals.Maintenance of Wakefulness TestThe MWT is an objective measure of a person’sability to remain awake in quiet situations for aspecified period of time. 4 An MWT is indicated toassess a person’s ability to maintain wakefulnessas well as to assess response to treatment for excessivesleepiness. This test is less sensitive than MSLTin measuring sleepiness.250 Sleep Physiology (Lee-Chiong)ACC-PULM-09-0302-016.1.indd 2507/10/09 10:17:35 PM


The MWT consists of four nap opportunities,each 40 min in duration and performed at 2-hintervals. The first nap trial is started about 1.5 to3 h after the person’s customary wake time. Standardmonitoring leads include EEG, EOG andchin EMG. The need for polysomnography priorto testing is not routinely indicated and shouldbe determined by the clinician for each specificindividual.During testing, the person is asked to sit in bedin a semi-reclined position in a dark, quiet room,and asked to try to stay awake. Measures to stayawake (eg, singing) are not allowed, and notobacco, caffeine, or stimulant agents arepermitted during test day. The nap trial is stoppedif either sleep occurs or after 40 min if no sleepis recorded.Sleep onset latency is determined for each nap;it is recorded as 40 min if no sleep occurs during anap trial. Mean SOL correlates with the ability tostay awake, with 8 min being abnormal; 8 ormore minutes but 40 min is considered indeterminate,and 40 min is considered normal.Sleep and AgingSleep requirements do not decline with aging.Older adults, however, have greater nocturnalsleep disturbance, more complaints of excessivesleepiness, and more frequent daytime napping.Older women are better able to maintainsatisfactory sleep compared with older men.Causes of sleep disturbance in older adults includeaging itself (rarely the sole cause); menopause;medical disorders, such as nocturia; neurologicdisorders, including dementia and Parkinsondisease; psychiatric disorders, such as depression;adverse effects of medications; and primary sleepdisorders. Aging is associated with a greaterprevalence of insomnia, OSA, central sleep apnea,PLMD, restless legs syndrome, REM sleep beha viordisorder, and advanced sleep phase disorder.Compared with younger adults, there is poorercorrelation between subjective sleep complaintsand objective polysomnography findings in thispopulation group. 5Among older adults, OSA remains more prevalentin men, but its risk increases with menopauseamong women. Hormonal replacement therapy(HRT) decreases the prevalence of OSA in postmenopausalwomen; nevertheless, data do notconclusively support the use of HRT as therapy forOSA among postmenopausal women. Comparedwith younger counterparts, OSA among olderadults is associated with relatively less risk of cardiopulmonarydiseases, and the AHI is less able topredict mortality risk.References1. Borbely AA, Achermann P. Sleep homeostasis andmodels of sleep regulation. J Biol Rhythms 1999;14:557−5682. Mullington JM, Haack M, Toth M, et al. Cardiovascular,inflammatory, and metabolic consequencesof sleep deprivation. Prog Cardiovasc Dis 2009;51:294−3023. Iber C, Ancoli-Israel S, Chesson A, et al, for theAmerican Academy of Sleep <strong>Medicine</strong>. The AASMmanual for the scoring of sleep and associatedevents: rules, terminology and technical specifications.1st ed. Westchester, IL: American Academyof Sleep <strong>Medicine</strong>, 20074. Standards of Practice Committee of the AmericanAcademy of Sleep <strong>Medicine</strong>. Practice parametersfor clinical use of the multiple sleep latency testand the maintenance of wakefulness test. Sleep2005; 28:113−1215. Ancoli-Israel S, Ayalon L, Salzman C. Sleep in theelderly: normal variations and common sleep disorders.Harv Rev Psychiatry 2008; 16:279−286<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 251ACC-PULM-09-0302-016.1.indd 2517/10/09 10:17:35 PM


Notes252 Sleep Physiology (Lee-Chiong)ACC-PULM-09-0302-016.1.indd 2527/10/09 10:17:35 PM


Sleep-Related Breathing DisordersTeofilo L. Lee-Chiong, Jr., MD, FCCPObjectives:• Differentiate the clinical and polysomnographic featuresof obstructive and central sleep apnea• Describe the various effective therapies for OSA• Distinguish the causes of hypercapnic vs nonhypercapnicCSA• Identify the key features of the hypoventilation syndromesKey words: central sleep apnea; hypoventilation; obstructivesleep apnea; positive airway pressure therapyObstructive Sleep ApneaObstructive sleep apnea (OSA) is characterizedby repetitive reduction or cessation of airflowdespite the presence of respiratory efforts causedby partial or complete upper-airway occlusionduring sleep. 1DefinitionsAn apnea is defined as a reduction in peakthermal sensor amplitude by at least 90% of baselinefor at least 10 s. Respiratory events can eitherbe central (no respiratory effort), obstructive (respiratoryefforts are present), or mixed (initial centralapnea followed by obstructive apnea). A hypopneais present if there is a decrease in nasal pressureby at least 30% of baseline for duration of at least10 s accompanied by at least 4% oxygen desaturation.Complex sleep apnea is characterized bycentral apneas that develop or become morefrequent during continuous positive airwaypressure (CPAP) titration for OSA.DemographicsOSA affects an estimated 24% of adult men and9% of adult women if OSA is defined by an apneahypopneaindex (AHI) of at least 5 events/ hour,or 4% of adult men and 2% of adult women if OSAis defined by an AHI of at least 5 events/h plus thepresence of excessive sleepiness. It affects menmore commonly than women, in whom prevalenceincreases with menopause.PathophysiologyThe upper airway can be conceptualized as acollapsible cylinder, with airflow determined byboth the difference in upstream (nasal) vs downstreampressure and airway resistance. Therefore,airflow is greater with the following: (1) greaterupstream pressure, (2) lesser downstream pressure,and (3) reduced airway resistance. Upper-airwaypatency is dependent on the balance of factors thatmaintain airway opening (activation of dilatormuscles) or promote airway closure (decrease inintraluminal pressure and Bernoulli forces). Airwaysize is also influenced by lung volume, whichdecreases during sleep. The critical closing pressure(Pcrit) is the intraluminal pressure at whichthe upper airway collapses; it becomes progressivelyless negative from nonsnorers and snorersto OSA. Activation of the upper-airway dilatormuscles lowers Pcrit. Ventilatory loop gain is alsoimportant in the pathogenesis of OSA. Personswith OSA are more likely to have intrinsicallyunstable negative feedback ventilatory controlsystem, with a high loop gain increasing the riskof periodic breathing.Compared with persons without the disorder,those with OSA tend to have narrower airways thatare more vulnerable to collapse. The most commonsites of upper-airway obstruction are behind thepalate (retropalatal), behind the tongue (retrolingual),or both. In OSA, repetitive upper-airwayobstruction is associated with episodic decreasesin arterial oxygen saturation (Sao 2); snoring thatmay alternate with periods of silence; arrhythmias,with a relative reduction in heart rate developingduring upper-airway obstruction followed by arelative increase in heart rate during apnea termination;arousal at apnea termination; and increasein BP in the immediate postapneic period.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 253ACC-PULM-09-0302-016.2.indd 2537/10/09 10:17:59 PM


Several factors might worsen hypoxemiaduring obstructive respiratory events, includinglow awake supine and baseline sleep Sao 2,increased percentage of total sleep time withapneas or hypopneas, greater duration of apnea orhypopneas, decrease in functional residual capacityand expiratory reserved volume, comorbid lungdisorders such as COPD, or stage of sleep (moresevere oxygen desaturation during rapid eyemovement [REM] sleep compared with non-rapideye movement [NREM] sleep).Classification Based on SeverityThe severity of OSA can be measured based onthe AHI as mild (5 to 15), moderate (16 to 30), orsevere ( 30). In addition to the AHI, other factorscan influence the clinical severity of OSA, such asthe degree of daytime sleepiness, nadir of Sao 2,severity of sleep fragmentation, presence of nocturnalarrhythmias, and the presence of comorbidcardiovascular or neurologic disorders.Risk FactorsRisk factors for OSA include a positive familyhistory of OSA, male gender (for adults), menopausalstate in women, aging, excess body weight,snoring, smoking and alcohol use, medication use(eg, muscle relaxants), and primary disorders, suchas untreated hypothyroidism (inconsistent data),acromegaly, neuromuscular disorders, and stroke.Androgen therapy has been shown to increase therisk of OSA. Central obesity, ie, waist-hip ratio, ismore important than general obesity.Specific craniofacial and oropharyngeal featurespredispose one to the development of OSA,including increased neck circumference, namely 17 inches in men and 16 inches in women; nasalnarrowing or congestion; macroglossia; low-lyingsoft palate; enlarged tonsils and adenoids; mid-facehypoplasia; retrognathia or micrognathia; andmandibular hypoplasia. The prevalence of OSA isgreater among African-Americans, Mexican Americans,and Asians and Pacific Islanders comparedwith white subjects. The apolipoprotein 4 gene isassociated with an increased risk and severity ofOSA, and polymorphism in the angiotensinconvertingenzyme gene is associated with the developmentof hypertension in persons with OSA.Clinical FeaturesPersons with OSA most commonly presentwith complaints of daytime sleepiness. It is importantto note that the severity of excessive sleepinessdoes not correlate closely with AHI. Other complaintsinclude repeated awakenings with gaspingor choking, snoring, witnessed apneas, attentiondeficit and/or hyperactivity (particularly in children),changes in mood such as treatment-resistantdepression, and decline in work or school performance.It is not unusual for OSA to be encounteredin persons complaining of a dry mouth/throatsensation on awakening, fatigue, gastroesophagealreflux, impaired memory and concentration,insomnia, morning headaches, nighttime diaphoresis,or nocturia.Physical ExaminationPhysical examination may reveal the presenceof excess body weight (ie, BMI 25 kg/m 2 ); largeneck circumference; nasal septal deviation or turbinatehypertrophy; crowded posterior pharyngealspace; enlarged tonsils and adenoids (especially inchildren); high, narrow hard palate; large uvula;low-lying soft palate; macroglossia; or retrognathiaor micrognathia. The physical examination resultsmay be entirely normal.ConsequencesIn addition to increased mortality rates, personswith OSA are more likely to have insulinresistance; parasomnias, including sleep-relatedeating disorder; ischemic heart disease; and stroke.Cardiac arrhythmias are frequently encountered,the most common of which is sinus arrhythmiathat consists of relative bradycardia at the onset ofthe apneic episode and relative tachycardia afterthe termination of the event. There is also anincreased likelihood of recurrence of atrial fibrillationafter successful cardioversion in persons withuntreated OSA.Systemic hypertension can develop when systemicBP fails to decrease during sleep (“nondipping”).OSA can further worsen heart function inpersons with congestive heart failure (CHF). <strong>Pulmonary</strong>hypertension and cor pulmonale candevelop in persons with severe OSA, for whom the254 Sleep-Related Breathing Disorders (Lee-Chiong)ACC-PULM-09-0302-016.2.indd 2547/10/09 10:18:00 PM


degree of oxygen desaturation may be predictiveof the risk of pulmonary hypertension. There is agreater likelihood of pulmonary hypertension inpersons with daytime hypoxemia and hypercapnia,morbid obesity, and comorbid COPD. <strong>Pulmonary</strong>hypertension secondary to OSA is generallymild. Other consequences of OSA include depressionand anxiety, impaired cognition (decreasedalertness, executive function, learning, and memory);erectile dysfunction; gastroesophageal reflux;nocturia; driving and work-related accidents;impaired school and work performance; andgreater health-care utilization.EvaluationEvaluation of persons suspected of having OSAshould consist of a thorough clinical history as wellas physical examination, with particular emphasison the upper airway and neck circumference.Laboratory testing should be individualized; routinescreening for hypothyroidism is not indicated.Polysomnography is required for the diagnosis ofOSA, the current standard of practice being anattended laboratory study. The latter can either bea full-night study with separate diagnostic andpositive airway pressure (PAP) titration studynights, or a split-night study consisting of an initialdiagnostic portion and a subsequent PAP titrationon the same night. During polysomnography,respiratory events are more frequent, last longer,and are associated with more profound oxygendesaturation during REM sleep. Paradoxical breathingmay be appreciated.TherapyTherapy for OSA requires a multidisciplinaryapproach, including general measures, PAP therapy,oral devices, and/or upper-airway surgery.The avoidance of alcohol, smoking, and musclerelaxants is important, as is proper sleep hygieneand safety counseling, such as not driving wheneverdrowsy. Optimal weight management is crucialfor persons who are overweight. Positionaltherapy, with the avoidance of a supine sleep position,may be considered if respiratory events occurexclusively or predominantly during supine sleepand if polysomnography demonstrates a normalAHI in the lateral or prone sleep position. Oxygentherapy is not indicated as sole therapy for OSAbut may be of benefit in persons with significanthypoxemia uncontrolled by PAP therapy alone.Topical nasal corticosteroids may be used as adjuncttherapy for persons with concurrent rhinitis. 2Modafinil or armodafinil is indicated for themanagement of residual excessive sleepinessdespite effective PAP therapy and if there is noother known cause for sleepiness. Hormonereplacementtherapy has been suggested for postmenopausalwomen; however, data regarding itsefficacy for this indication are inconsistent.PAP Therapy: PAP therapy is the treatment ofchoice for most persons with OSA. 3 PAP functionsas a pneumatic splint that maintains upper-airwaypatency and increases nasal pressure above Pcrit.Greater pressures may be required to control respiratoryevents during REM sleep and supine sleep.PAP therapy is indicated for persons with an AHIof at least 15 events per hour; or if the AHI is atleast 5 events/h but 14 events/h if there arecomplaints of excessive sleepiness, impaired cognition,mood disorder or insomnia, or documentedhypertension or coronary artery disease, or historyof stroke.There are several PAP modalities that are currentlyavailable, including CPAP, which providesa single constant pressure throughout the respiratorycycle. With the use of CPAP with expiratorypressure relief technology, a single pressure isprovided throughout the respiratory cycle, but atransient reduction in pressure during expirationand a subsequent return of pressure to baselinesetting before next inspiration is allowed. BilevelPAP provides two pressure levels during therespiratory cycle, namely: a higher level duringinspiration (ie, inspiration PAP), and a lower pressureduring expiration (ie, expiration PAP). Withautotitrating PAP (APAP), variable pressures areprovided based on device-specific diagnostic andtherapeutic algorithms; thus, this device is ableto automatically and continuously adjust deliveredPAP to maintain upper-airway patency. 4Adaptive servoventilation delivers pressure support(inspiration PAP minus expiration) thatincreases during periods of hypoventilation anddecreases during hyperventilation. Finally, nocturnalnoninvasive positive pressure ventilationprovides two pressure levels at a set rate to assistventilation.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 255ACC-PULM-09-0302-016.2.indd 2557/10/09 10:18:00 PM


Several methods have been developed to helpdetermine a single optimal CPAP pressure forsubsequent nightly use by the patient at home. Afull-night attended laboratory polysomnographyremains the preferred method, but a split-nightpolysomnography (ie, an initial diagnostic andsubsequent PAP titration portions on the samenight) is frequently used. Finally, titration using anAPAP device is increasing in popularity, but thereremains minimal data regarding its efficacyand applicability across different populationgroups.PAP therapy has been shown to decrease mortality,reduce sleepiness (both subjective and objective),decrease or eliminate snoring, normalizeAHI, and improve Sao 2. It also improves drivingsimulator steering performance, BP control, andleft ventricular function in persons with CHF.Health-care utilization declines after initiation ofPAP therapy. However, there have been inconsistentor inconclusive data on any benefits to sleepquality, quality of life, neurocognitive function, andmood. Persons initiating PAP therapy may experiencenasal congestion, dryness, epistaxis, or rhinorrhea;sinus discomfort or pain; facial skin irritation,rash or abrasion; eye irritation; gastric distentionas the result of aerophagia; or chest discomfort andtightness, many of which may result in the patientdiscontinuing therapy.Adherence to PAP therapy should be monitoredobjectively. Most studies report an objective compliance(ie, use for 4 h per night for 70% of nights)of 50 to 80%, with an average nightly use of just 5 h.Only patient education and the use of heatedhumidification have been consistently demonstratedto enhance PAP adherence. Factorspredicting the need for heated humidificationinclude the following: (1) age 60 years, (2) use ofdrying medications, (3) presence of chronic mucosaldisease, and (4) a history of uvulopalatopharyngoplasty.Other approaches that have beenrecom-mended to help improve a patient’s adherenceto PAP, such as the use of bilevel PAP, CPAPwith expiratory pressure relief technology andramping mechanism, and changing a poorly fittingnasal mask after therapy has started, have showninconsistent benefits.Bilevel PAP therapy may be considered forpersons who complain of excessively high pressureswith significant difficulty breathing outagainst these pressures, or of gastric distentioncaused by aerophagia. It may also be beneficial forpersons with comorbid obstructive or restrictivelung disease, hypoventilation syndrome, or withpersistent oxygen desaturation despite CPAPtherapy.APAP devices have been used for PAP titrationto identify a single fixed pressure for subsequenttreatment with a conventional CPAP device (thisis not recommended for split-night PAP titration)or PAP treatment when used in a self-adjustingmode for nightly therapy of OSA. APAP devicesare contraindicated for persons with CHF, significantrespiratory disease (eg, COPD), daytimehypoxemia and respiratory failure, and nocturnaloxygen desaturation unrelated to OSA, such as incases of obesity-hypoventilation syndrome.Compared with conventional CPAP, there areno significant differences with APAP in reductionsof AHI and arousal indexes, changes in sleep architecture,Sao 2levels, or subsequent CPAP acceptance.Mean airway pressure tends to be lower andpeak airway pressure tends to be greater (as theresult of air leaks) with APAP compared withCPAP. Proper mask fitting is crucial before thepatient uses APAP unattended. Finally, noninvasivepositive pressure ventilation is indicated forcases of persistent sleep-related hypoventilationand CO 2retention.Oral Devices: Oral devices are effective therapiesfor selected persons with snoring, mild-tomoderateOSA, and severe OSA (in some). Thereare two general types of oral devices: mandibularrepositioners, which displace the mandible andtongue anteriorly and are the most commonly usedoral devices; and tongue-retaining devices which,by securing the tongue in a soft bulb located anteriorto the teeth, hold the tongue in an anteriorposition. 5 Tongue-retaining devices are preferredfor edentulous persons or for those with compromiseddentition. As a class, oral devices have areported efficacy of 40 to 80%. Compliance isgreater than with PAP devices and ranges from50 to 80%. A follow-up polysomnography is indicatedafter optimal fit has been attained. In addition,periodic assessments by a dentist and sleepphysician are recommended to ensure continuedeffectiveness.Contraindications to the use of oral devicesinclude an inability to breathe nasally and sleep256 Sleep-Related Breathing Disorders (Lee-Chiong)ACC-PULM-09-0302-016.2.indd 2567/10/09 10:18:00 PM


apnea that is primarily central in nature. Thedevices generally are not recommended forgrowing children. In addition, mandibular repositionersshould not be used in persons with inadequateor compromised dentition or in those withsignificant temporomandibular joint dysfunction.Complications associated with the use of oraldevices consist of a dry mouth sensation or excessivesalivation, dental pain, undesirable dentalmovements with mandibular repositioners, andjaw or temporomandibular joint pain with mandibularrepositioners.Upper-Airway Surgery: Upper-airway surgerymay be considered for persons with definitive craniofacialor upper-airway abnormalities that areresponsible for OSA. Polysomnography is requiredafter upper-airway surgery to determine therapeuticefficacy, and patients require long-term follow-up.Several types of upper-airway surgicalprocedures have been developed to address specificupper-airway abnormalities, including tonsillectomyand adenoidectomy, which are commonlyused to treat childhood cases of OSA. Nasal septoplasty,polyp removal, and turbinectomy are usedto increase the dimensions of the nasal airway.Uvulopalatopharyngoplasty increases the size ofthe retropalatal airspace, whereas genioglossaladvancement, hyoid myotomy and suspension, andmandibular advancement enlarge the retrolingualairway. Uvulopalatopharyngoglossoplasty andmaxillomandibular advancement increase the retrolingual,retropalatal, and transpalatal airway.Finally, tracheotomy can be used to bypass thenarrow upper airway and is the only surgical procedurethat is consistently effective as a sole procedurefor OSA. In addition to tracheotomy, themost effective surgical procedures for the therapyof OSA are bariatric surgery for weight managementand maxillomandibular advancement.Medicolegal ConsiderationsSleepy persons with OSA are at increased riskof car accidents; the risk is increased further withconcurrent alcohol ingestion. A history of near-missaccidents predicts future risk of car accidents.Effective therapy of OSA decreases this risk.Patients with OSA should be instructed to avoiddriving whenever drowsy. Adherence to PAPtherapy must be assessed objectively in all patients,and periodic follow-up to identify adverse effectsis recommended. A multiple sleep latency testshould be considered whenever there is doubtabout a person’s degree of sleepiness. Modafiniland armodafinil, wake-promoting agents, maybenefit persons who are excessively sleepy despitePAP therapy. Consider reporting patients withexcessive sleepiness as the result of OSA to appropriateauthorities if (1) there is a history of severecar accidents related to untreated sleepiness; (2) ifprompt therapy for OSA cannot be provided; (3) ifthe patient refuses, or is consistently nonadherentwith, therapy for OSA; (4) if the patient fails torestrict driving until OSA has been adequatelycontrolled; or (5) if such situations are consideredreportable based on local legislation.Upper-Airway Resistance SyndromeThis syndrome is characterized by repetitivesleep-related episodes of decreased inspiratoryairflow attributable to increasing upper-airwayresistance accompanied by increased or constantrespiratory effort and arousals from sleep. 6 Upperairwayresistance syndrome should be excluded inpersons with unexplained sleepiness. Polysomnographydemonstrates an AHI that is 5 events/h.With esophageal pressure monitoring, increasinglynegative esophageal pressure excursions precedearousals and are followed by less negative esophagealpressure excursions as airflow increasesduring arousals. Nasal pressure monitoring demonstratesinspiratory flattening followed by arounded contour during arousals.SnoringRisk factors for snoring include obesity; supinesleep position; nasal obstruction; medications, suchas muscle relaxants, opioids or benzodiazepines;and alcohol use. Snoring is often loudest duringstage N3 sleep and diminishes during REMsleep. OSA is present in an estimated 25 to 95% ofsnorers.Central Sleep ApneaCentral sleep apnea (CSA) refers to repetitivecessation of airflow during sleep caused by theloss of ventilatory effort. Persons with CSA may<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 257ACC-PULM-09-0302-016.2.indd 2577/10/09 10:18:00 PM


present with insomnia or sleepiness or may beasymptomatic. 7PolysomnographyPolysomnography is necessary for the diagnosisof CSA and demonstrates cessation of respirationand ventilatory effort lasting at least 10 s, withat least five central apneas per hour of sleep. Respiratoryevents are most common during sleep onsetand N1/N2 sleep. There is an absence of chest andabdominal movement. No respiratory muscleactivity is present in diaphragmatic EMG, and nochanges in pressure are noted in esophageal pressuremonitoring.ClassificationCSA can be classified as either hypercapnicor nonhypercapnic. Hypercapnic CSA is characterizedby high sleep Paco 2and is often associatedwith high waking Paco 2. Persons with thehypercapnic form of CSA have decreased ventilatoryresponsiveness to hypercapnia. Causes ofhypercapnic CSA include neuromusculardisorders and long-term use of long-actingopioids. Persons with nonhypercapnic CSA havenormal or low waking Paco 2and an increa sedventilatory response to hypercapnia. In nonhypercapnicCSA, brief arousals during sleep triggera hyperventilatory “overshoot” that lowers Paco 2to less than its apneic threshold. Causes of nonhypercapnicCSA include idiopathic CSA, sleeponsetCSA, CSA caused by CHF, high-altitudeperiodic breathing, and complex sleep apnea (seethe subheading “Complex Sleep Apnea” tofollow).Idiopathic CSA: The etiology of this rare conditionis unknown. It affects men more commonlythan women, and its prevalence is greatest inmiddle-aged and older adults.Cheyne-Stokes Respiration (CSR): Periodicbreathing in CSR consists of recurring periods ofcrescendo--decrescendo ventilation separated bycentral apneas or hypopneas. Periodic breathingdevelops during NREM sleep and improves orresolves during REM sleep. CSR may develop inpersons with CHF or stroke. Risk factors for CSRin CHF include male gender, age 60 years,atrial fibrillation, and hypocapnia. There areseveral important differences between CSA andCSR. With CSA, the nadir of oxygen desaturationoften follows the termination of apnea, arousalsoccur at the termination of apnea, cycle time isshorter (generally 45 s), and the period ofhyperpnea is shorter. In contrast, CSR is associatedwith a more delayed nadir of oxygendesaturation, longer cycle time ( 45 s), and longerperiod of hyperpnea. In addition, arousalsassociated with CSR commonly occur at the peakof hyperpnea.High-Altitude Periodic Breathing: Cycles ofcentral apneas and hyperpneas can develop onascent to high altitude, usually at elevations 4,000 to 7,600 m. The likelihood of high-altitudeperiodic breathing is greater in persons withgreater hypoxic ventilatory drive, at greater elevations,after faster speed of ascent, and amongmen. Central apneas occur primarily duringNREM sleep, with respiratory patterns improvingduring REM sleep.Medication-Induced CSA: CSA can developduring administration of long-acting opioids,such as methadone, as the result of depression ofthe hypercapnic respiratory drive.CSA Caused by CHF: CSA or CSR can developin persons with CHF, the prevalence and severityof which are correlated with left ventricularfunction. The rate of mortality is greater in CHFpatients with CSR than in those without CSR.Sleep-Onset CSA: Central apneas may occurif Paco 2(greater during sleep and lower duringwakefulness) fluctuates above or below theapneic threshold. This sleep-onset CSA is generallytransient and resolves as sleep progresses.Complex Sleep Apnea: Complex sleep apnea isdefined as the development of CSA or CSR duringacute application of CPAP in patients with predominantlyOSA at baseline. This phenomenonoccurs in an estimated 15% of persons with OSAtitrated with CPAP.Therapy for CSAIt is important to treat any known underlyingcause of CSA, such as heart failure. Respiratorydepressants should be avoided in persons withhypercapnic CSA. High-altitude periodic breathingmay improve with oxygen therapy or after descentto lower altitudes. Effective pharmacologic258 Sleep-Related Breathing Disorders (Lee-Chiong)ACC-PULM-09-0302-016.2.indd 2587/10/09 10:18:00 PM


therapies for CSA include acetazolamide forhigh- altitude periodic breathing and hypnoticagents for sleep-onset central apneas. Many personswith CSA benefit from PAP therapy. CPAP orbilevel PAP may be tried for patients with CSA orCSR attributable to CHF; this type of therapy mightimprove cardiac function but may have no benefiton mortality, making close monitoring of therapeuticefficacy crucial. Adaptive servo ventilation hasbeen developed to manage persons with complexsleep apnea. Finally, nocturnal noninvasive ventilationmay be considered for those with hypercapnicforms of CSA.Hypoventilation SyndromesThe main features of sleep-related hypoventilationsyndromes are oxygen desaturation andincreased Paco 2during sleep, with Paco 2either 45 mm Hg or 45 mm Hg but abnormallyincreased relative to waking levels. Waking arterialblood gas levels may be normal or abnormal.Several mechanisms may be responsible forhypoventilation developing during sleep, includinga decrease in minute ventilation and/or tidalvolume, abnormal ventilation/perfusion relationships,or changes in ventilatory chemosensitivityand respiratory load responsiveness. Therapygenerally involves treating any underlyingdisorder/s and, if necessary, ventilatory assistanceduring sleep.Congenital Central Alveolar HypoventilationSyndromeFailure of automatic control of breathing in thissyndrome results in hypoxemia and hypercapnia.There is diminished responsiveness of central andperipheral chemoreceptors to hypoxemia andhypercapnia. Onset is usually in infancy. Hypoventilationis worse during sleep than wakefulness;during sleep, hypoventilation is more severeduring NREM sleep than REM sleep. Congenitalcentral alveolar hypoventilation syndrome maybe associated with autonomic dysfunction,Hirschsprung disease and neural crest tumors.Many cases of congenital central alveolar hypoventilationsyndrome involve de novo mutations in thePHOX2B gene. 8References1. Olson EJ, Park JG, Morgenthaler TI. Obstructivesleep apnea-hypopnea syndrome. Prim Care 2005;32:329−3592. Morgenthaler TI, Kapen S, Lee-Chiong T, et al.Practice parameters for the medical therapy ofobstructive sleep apnea. Sleep 2006; 29:1031−10353. Kushida CA, Littner MR, Hirshkowitz M, et al.Practice parameters for the use of continuous andbilevel positive airway pressure devices to treatadult patients with sleep-related breathing disorders.Sleep 2006; 29:375−3804. Morgenthaler TI, Aurora RN, Brown T, et al. Standardsof Practice Committee of the AASM. Practiceparameters for the use of autotitrating continuouspositive airway pressure devices for titrating pressuresand treating adult patients with obstructivesleep apnea syndrome: An update for 2007. Sleep2008; 31:141−1475. Kushida CA, Morgenthaler TI, Littner MR, et al.Practice parameters for the treatment of snoringand obstructive sleep apnea with oral appliances:an Update for 2005. Sleep 2006; 29:240−2436. Badr MS. Central sleep apnea. Prim Care 2005;32:361−747. Bao G, Guilleminault C. Upper airway resistancesyndrome—one decade later. Curr Opin PulmMed 2004; 10:461−4678. Amiel J, Laudier B, Attie-Bitach T, et al. Polyalanineexpansion and frameshift mutations of thepaired-like homeobox gene PHOX2B in congenitalcentral hypoventilation syndrome. Nat Genet2003; 33:459−461<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 259ACC-PULM-09-0302-016.2.indd 2597/10/09 10:18:01 PM


Notes260 Sleep-Related Breathing Disorders (Lee-Chiong)ACC-PULM-09-0302-016.2.indd 2607/10/09 10:18:01 PM


Nonrespiratory Sleep DisordersTeofilo L. Lee-Chiong, Jr., MD, FCCPObjectives:• Describe the different causes of insomnia and excessivesleepiness and the treatment options for each disorder• Identify the key features of the various parasomnias• Understand the clinical manifestations of restless legssyndrome• Distinguish the main presenting symptoms of the differentcircadian rhythm sleep disordersKey words: circadian rhythm sleep disorders; insomnia; narcolepsy;parasomnias; restless legs syndrome; sleepinessThe differential diagnoses of excessive sleepinessis extensive and include the following: (1) narcolepsy;(2) idiopathic hypersomnia; (3) circadianrhythm sleep disorders; (4) insufficient sleepsyndrome; (5) obstructive sleep apnea (OSA);(6) periodic limb movement disorder (PLMD);(7) posttraumatic hypersomnia (head injury);(8) recurrent hypersomnia, including Kleine Levinsyndrome; (9) medical, neurologic and psychiatricdisorders; and (10) medication or substance useor withdrawal. Transient insomnia can be a resultof the following: (1) acute stressors, (2) jet lag,(3) acute medication and substance use, and(4) shift work. Chronic insomnia can either presentas sleep-onset sleep disturbance, sleep maintenanceinsomnia, or early morning awakenings. Causesof sleep-onset insomnia consist of the following:(1) idiopathic insomnia; (2) psychophysiologicinsomnia; (3) restless legs syndrome (RLS);(4) delayed sleep phase disorder; (5) psychiatricdisorders, such as anxiety, mood disorder, or posttraumaticstress disorder; (6) paradoxical insomnia;or (7) use of stimulants. Sleep-maintenanceinsomnia can result from the following: (1) alcoholwithdrawal; (2) medical, neurologic and psychiatricdisorders, such as asthma, COPD, congestiveheart failure (CHF), and mood and panic disorders;(3) primary sleep disorders, including OSAand central sleep apnea (CSA), PLMD, psychophysiologicinsomnia, and parasomnias; and(4) medication or substance use. Early morningawakenings can be caused by advanced sleepphase disorder, depression, or withdrawal fromshort-acting hypnotic agents. There are manycauses of nocturnal oxygen desaturation, includingOSA, CSA, asthma, COPD, alveolar hypoventilationsyndromes, diaphragm paralysis,neuromuscular disorders, restrictive lung disease,CHF, and high altitude.InsomniaInsomnia is characterized by repeated difficultywith either falling or staying asleep that is associatedwith impairment of daytime function. 1 Sleepdisturbance can present as difficulty falling asleep(sleep-onset insomnia), frequent or prolongedawakenings (sleep-maintenance insomnia), finalmorning awakening that is earlier than desired(terminal insomnia), and unrefreshed feeling onawakening (nonrestorative sleep). Polysomnographically,insomnia can be defined as prolongedsleep-onset latency (SOL), 30 min; increasedwake time after sleep onset, 30 min; reducedsleep efficiency, 85%; and diminished total sleeptime, 6 to 6.5 h.DemographicsInsomnia is the most common sleep disorder,with 30 to 50% of adults complaining of occasionalinsomnia and 10 to 30% of adults describingchronic insomnia. Women are affected more frequentlythan men.Risk FactorsRisk factors for insomnia include the following:(1) female sex; (2) advancing age; (3) lower socioeconomicstatus or unemployment; (4) maritalstatus, either divorced or widowed; (5) poor healthstatus and physical disability; and (6) medical,neurologic, and psychiatric disorders (eg, respiratorydisorders).<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 261ACC-PULM-09-0302-016.3.indd 2617/10/09 10:18:21 PM


ConsequencesPersons with insomnia have an increased riskof psychiatric illness developing, such as majordepression. Other consequences of insomnia includefatigue, cognitive impairment, impaired academicand occupational performance, diminished qualityof life, and greater health-care utilization.ClassificationForms of insomnia can be classified, based onduration of sleep disturbance, as transient if theinsomnia lasts only a few days, or chronic if itpersists 1 to 3 months. Another useful distinctionclassifies the causes of sleep disturbance into primaryor comorbid insomnia. Primary (idiopathic)insomnia includes idiopathic insomnia, paradoxicalinsomnia, and psychophysiologic insomnia.Adjustment InsomniaIn this syndrome, sleep disturbance is a resultof an identifiable acute stressor, such as a momentouslife event, change in the sleep environment,or an acute illness. Sleep improves with resolutionof acute stressor or when adaptation to the stressordevelops.Idiopathic InsomniaThis type of insomnia is not associated withany identifiable etiology. It has its onset duringinfancy or early childhood, and its course is generallychronic and life-long.Inadequate Sleep HygienePoor sleep hygiene consists of activities orbehavior that increase arousal or decrease sleeppropensity and that are primarily responsible forsleep disturbance. These activities and behaviorare under a person’s control and can be eliminatedif the person desires to do so.Paradoxical InsomniaAlso known as sleep-state misperception, personsreport very minimal or no sleep during mostnights, but polysomnography generally demonstratesnormal sleep quality and architecture.Likewise, there is no daytime napping or impairmentof daytime functioning.Psychophysiologic InsomniaSleep disturbance is secondary to heightenedcognitive and somatic arousal at bedtime. Causesconsist of rumination and intrusive thoughts,increased agitation and muscle tone, and learnedmaladaptive sleep-preventing behavior, such asexcessive anxiety about the inability to sleep.Conditioned arousal is typically limited to aperson’s own bed and bedroom, with better sleepbeing described when attempted in anotherroom. Onset of sleep disturbance is generallyduring adolescence or early adulthood. Unlikegeneralized anxiety disorder, anxiety is limitedto issues related to sleep in psychophysiologicinsomnia.Common Medications That Can CauseInsomniaMany medications can cause insomnia; themost common include antidepressants such asfluoxetine or protriptyline, -adrenergic blockingantihypertensive agents, bronchodilators, decongestants,corticosteroids, and stimulant agents.EvaluationThe evaluation of insomnia relies primarily ona compatible history and sleep diary. Psychometrictests, including validated measures for depressionand anxiety, may be considered for selectedpatients. Polysomnography, actigraphy, andlaboratory tests are not routinely indicated; nevertheless,polysomnography may be helpful in personwith suspected OSA, PLMD, or paradoxicalinsomnia.Cognitive-Behavioral TherapyIn addition to instructions regarding propersleep hygiene, persons with insomnia often benefitfrom cognitive-behavioral treatments (CBTs);short-term benefits of these interventions are262 Nonrespiratory Sleep Disorders (Lee-Chiong)ACC-PULM-09-0302-016.3.indd 2627/10/09 10:18:21 PM


comparable with pharmacologic therapy and,unlike the latter, beneficial effects are sustainedover time. At long-term follow-up, CBT is moreeffective than pharmacotherapy. Effective CBTtechniques consist of cognitive therapy, paradoxicalintention, relaxation techniques, sleep restriction,and stimulus control. 2,3Cognitive therapy addresses dysfunctionalbeliefs accompanying insomnia, identifies irrationalexpectations and excessive worry related to theinability to sleep, challenges unrealistic concernsabout poor sleep, and provides a more appropriateunderstanding of insomnia and its accompanyingdaytime impairment.Paradoxical intention is designed to decreaseperformance anxieties related to efforts to fallasleep. Patients are instructed to go to bed and tryto stay awake as long as they can. Relaxation techniquesaddress both somatic and cognitive hyperarousaland reduce them by progressive musclerelaxation (ie, sequential tensing and relaxing ofvarious muscle groups), biofeedback, or guidedimagery.With sleep restriction, time spent awake in bedis reduced to enhance homeostatic sleep drive, andtime in bed is subsequently increased once sleepefficiency improves. In this technique, bedtime isadvanced or delayed based on calculated sleepefficiency. An earlier bedtime is allowed if sleepefficiency is 90%. In contrast, bedtime is delayedif sleep efficiency is 80%. No change in bedtimeis required if sleep efficiency is 80 to 90%. Patientsare instructed to wake up at the same time eachmorning, and no daytime napping is permitted.Stimulus control strengthens the association ofbedroom and bedtime to a conditioned responsefor sleep. Patients are instructed to use the bed onlyfor sleep or sex, lie down to sleep only when sleepy,get out of bed and go to another room if unable tofall asleep, engage in a restful activity, and returnto bed only when sleepy.PharmacotherapyHypnotic agents may enhance sleep but oftendo not improve daytime performance. In addition,there is minimal long-term beneficial effects onsleep after drug discontinuation. Benzodiazepinesand nonbenzodiazepine benzodiazepine receptoragonists are effective hypnotic agents. There isinsufficient evidence regarding the efficacy ofsedating antidepressants, sedating antipsychoticagents, antihistamines, and botanical compoundsfor the therapy of insomnia.Hypnotic agents are indicated for the treatmentof transient sleep disruption (eg, jet lag or adjustmentsleep disorder), chronic primary insomniathat fails to respond to cognitive behavioral therapy,and chronic comorbid insomnia that does notimprove with treatment of the underlying conditionand cognitive behavioral therapy. Hypnoticagents can be classified based on elimination halflivesinto short acting ( 1 h; ramelteon, zaleplon);intermediate acting (2 to 5 h); eszopiclone, triazolam,and zolpidem; long acting (5 to 24 h;estazolam and temazepam); and extra-long acting( 40 h; flurazepam and quazepam). Short-actingagents are usually used for sleep-onset insomnia,intermediate-acting agents for concurrent sleeponsetand sleep-maintenance insomnia, and longactingand extra-long-acting agents for earlymorning awakenings and daytime anxiety.Benzodiazepine receptor agonists bind to thegamma-aminobutyric acid-benzodiazepine (GABA-BZ) receptor complex. Various GABA-BZ receptorsubunits have different actions, namely BZ1 (hypnoticand amnesic actions) and BZ2 and BZ3(muscle relaxation, antiseizure and antianxietyactions). Benzodiazepines bind nonselectively tothe different GABA-BZ receptor subunits, BZ1, BZ2,and BZ3 and, thus, possess hypnotic, anxiolytic,myorelaxant, and anticonvulsant properties.Many adverse effects are associated with theuse of benzodiazepines, including (1) rebounddaytime anxiety, especially with short-actingagents; (2) residual daytime sleepiness with longactingagents; (3) cognitive and psychomotorimpairment; (4) development of tolerance definedas the need for increasingly higher dosages toattain similar therapeutic benefit during chronicuse; (5) withdrawal symptoms, such as anxiety,irritability and restlessness; (6) rebound insomnia;(7) respiratory depression and worsening of OSA;(8) increase risk of falls, particularly in some olderadults; and (9) dependency and abuse liability (lowrisk). Benzodiazepines are contraindicated duringpregnancy or lactation; in persons with significantrenal or hepatic impairment, in whom dosage<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 263ACC-PULM-09-0302-016.3.indd 2637/10/09 10:18:22 PM


adjustments are required; in cases of untreatedOSA; and in persons with severe obstructive andrestrictive ventilatory impairment.The nonbenzodiazepine benzodiazepine receptoragonists selectively bind to the BZ1 receptorsubunit. Duration of action varies, with zaleplonhaving the shortest, zolpidem having an intermediateeffect, and esopiclone having the longestduration of action. Compared with conventionalbenzodiazepines, this class of agents have a similarhypnotic action; possess no muscle relaxant, anticonvulsant,or anxiolytic properties; and areless likely to cause rebound insomnia, withdrawalsymptoms or tolerance, or to alter sleeparchitecture.Ramelteon is a melatonin receptor agonist withselectivity for the suprachiasmatic nucleus melatoninreceptor. Because of its short half-life, it isprimarily used for persons with sleep-onset insomnia.Ramelteon should not be used in personson fluvoxamine therapy or those with hepaticimpairment.Other hypnotic agents include trazodone, tricyclicantidepressants, first-generation histamineantagonists, and melatonin. Trazodone is widelypopular because of its lack of significant potentialfor tolerance or dependency; adverse effects associatedwith trazodone use consist of cardiacarrhythmias, orthostatic hypotension, and priapism.Tricyclic antidepressants also have numerousadverse effects, such as urinary retention,constipation, cardiac arrhythmias, orthostatichypotension, and exacerbation of restless legs andperiodic limb movements during sleep. The firstgenerationhistamine antagonists, includingdiphenhy dramine, constitute the majority of overthe-counterhypnotic agents. Adverse effects offirst-generation histamine antagonists consist ofrapid development of tolerance; residual daytimesedation as the result of long half-life; and anticholinergiceffects (eg, confusion, delirium, drymouth, and urinary retention). Melatonin is usedprimarily for insomnia associated with circadianrhythm sleep disorders. Melatonin has a shorthalf-life of 20 to 30 min.Excessive SleepinessExcessive daytime sleepiness (EDS) describesan inability to consistently achieve and sustainwakefulness and alertness to accomplish the tasksof daily living. Manifestations of EDS include frequentnapping, sleep attacks, microsleep episodes,hyperactivity in children, and automatic behavior.Consequences of EDS consist of greater risk ofaccidents, impaired work and academic performance,and mood disorder. Disorders that cancause excessive sleepiness include narcolepsy,insufficient sleep syndrome, idiopathic hypersomnia,and recurrent hypersomnia. Sleepiness canalso be caused by a variety of medical disorders orby drugs or substance use. 3,4NarcolepsyNarcolepsy is a neurologic disorder characterizedby the clinical tetrad of excessive sleepiness,cataplexy, sleep paralysis, and sleep hallucinations.However, only 10 to 15% of patients demonstratethis full tetrad. Excessive sleepiness is the first,primary, and most disabling symptom of narcolepsy.It can manifest as brief naps, each lastingabout 10 to 20 min, that occur repeatedly throughoutthe day. Sleepiness transiently improves afternaps. Sleep attacks can also occur.Cataplexy consists of transient episodes ofmuscle atonia/hypotonia that are precipitated byintense emotion, such as laughter, anger, or excitement.There is sparing of respiratory and oculomotormuscles, and memory and consciousness areunaffected. Cataplexy is the only pathognomonicsymptom of narcolepsy.Patients with narcolepsy may experience sleephallucinations that may be visual, auditory, tactileor kinetic, occurring at sleep onset (hypnagogic),or occurring on awakening (hypnopompic). Sleephallucinations may be accompanied by sleepparalysis, which could also be either hypnagogic,hypnopompic, or both. Sleep paralysis generallylasts a few seconds or minutes, affects voluntarymuscles, and spares the respiratory, oculomotorand sphincter muscles. Sensorium is unaffected.Sleep disturbance, manifesting as repetitive arousalsand awakenings, are common in persons withnarcolepsy who may complain of sleep-maintenanceinsomnia.Other clinical features of narcolepsy arememory impairment, automatic behavior, andhyperactivity and learning disability (in children).Patients with narcolepsy have an increased risk264 Nonrespiratory Sleep Disorders (Lee-Chiong)ACC-PULM-09-0302-016.3.indd 2647/10/09 10:18:22 PM


of OSA and CSA, periodic limb movements duringsleep, REM sleep behavior disorder, anddepression.Narcolepsy affects approximately 0.05% of theUS population. It may affect men more commonlythan women. Excessive sleepiness is usually thepresenting symptom, with onset generally from15 to 25 years of age. Clinical course is chronicwith persistent sleepiness; the severity of cataplexymay decrease over time. Narcolepsy isbelieved to result from loss of hypothalamic hypocretinneurons; other pathogenetic mechanismsinclude defective cholinergic system regulatingREM sleep and impairment of norepinephrineand dopamine systems.Narcolepsy can also present without cataplexy.In this syndrome, cataplexy-like symptoms, suchas prolonged episodes of tiredness or muscle weaknessassociated with atypical triggers, may bepresent. Narcolepsy without cataplexy accountsfor 10 to 50% of cases of narcolepsy.Narcolepsy with cataplexy can be diagnosedby history alone. Polysomnography followed bymultiple sleep latency testing is required if cataplexyis absent, atypical, or equivocal. Polysomnographygenerally demonstrates a shortened SOLof 10 min, sleep-onset REM periods (SOREMPs)of 20 min in 25 to 50% of cases, and repetitiveawakenings.The multiple sleep latency test typicallyshows a mean SOL of 8 min and at least twoSOREMPs. The combination of a shortened SOLand SOREMPs is present in 60 to 85% of cases ofnarcolepsy. The maintenance of wakefulness testmay be used to monitor treatment responseto stimulant medications used for excessivesleepiness.Levels of cerebrospinal fluid hypocretin-1 are 110 pg/mL or less than one third of mean normalcontrol values in persons with narcolepsy with cataplexy.Levels of cerebrospinal fluid hypocretin-1 arenormal in many persons with narcolepsy withoutcataplexy. Human leukocyte antigen typing haslimited diagnostic utility.Every person with narcolepsy should beinstructed on proper sleep hygiene, includingmaintaining regular sleep-wake schedules,obtaining sufficient nocturnal sleep duration,taking scheduled naps during the day, andavoiding driving until sleepiness is adequatelymanaged. Excessive sleepiness can be managedpharmacologically with modafinil, armodafinil,dextroamphetamine, or methylphenidate. Sleepdisturbance often improves with the use ofhypnotic agents or -hydroxybutyrate (sodiumoxybate). REM sleep-suppressant agents, such asserotonin reuptake inhibitors, tricyclic antidepressants,venlafaxine, and monoamine oxidaseinhibitors, as well as -hydroxybutyrate, can beused to reduce attacks of cataplexy.Insufficient Sleep SyndromeInsufficient sleep is the most common cause ofsleepiness in many industrialized countries,including the United States. In this syndrome,sleepiness is caused primarily by chronic voluntary,but unintentional, sleep deprivation or sleeprestriction. Sleepiness generally improves withsleep extension, as is frequently noted during“sleep-ins” on weekends or holidays. Diagnosis isbased on clinical history and sleep diaries, andpolysomnography is not indicated. Therapy consistsof sleep extension.Idiopathic HypersomniaPersons with idiopathic hypersomnia complainof constant sleepiness despite sufficient, or evenincreased, amounts of nighttime sleep. They oftendescribe frequent daytime napping which, unlikenarcolepsy, are longer and less refreshing. Etiologyfor the sleepiness is unknown. Cataplexy is absent.Diagnosis requires polysomnography and MultipleSleep Latency Test, the latter showing a shortenedSOL and less than two SOREMPs. Cerebrospinalfluid hypocretin-1 levels are normal. Therapyincludes sleep hygiene and stimulant agents.Recurrent HypersomniaBoth Kleine-Levin syndrome and menstrualrelatedsleep disorder can present as recurrentepisodes of hypersomnia occurring weeks ormonths apart, with sleep and alertness normalizingbetween episodes. Kleine-Levin syndrome is characterizedby sleepiness, hyperphagia, hypersexuality,aggressive behavior, and cognitive impairment.Cases mostly involve men. Etiology is unknown.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 265ACC-PULM-09-0302-016.3.indd 2657/10/09 10:18:22 PM


Hypersomnia Caused by Medical DisordersMany medical disorders can give rise to excessivesleepiness, including hepatic encephalopathy,hypothyroidism, Prader-Willi syndrome, renalfailure, CNS infections or tumors, head trauma,Parkinson disease, stroke, atypical depression,bipolar type II mood disorder, and seasonal affectivedisorder.Hypersomnia Caused by Drugs or SubstancesExcessive sleepiness may develop during useor abuse of sedative-hypnotic agents or after withdrawalfrom stimulants.Evaluation of SleepinessExcessive sleepiness is often diagnosed with acomprehensive sleep history aided by sleep diariesand/or actigraphy. Subjective tests of sleepiness,such as the Epworth sleepiness scale, are commonlyused. Polysomnography is necessary toexclude OSA and PLMD. Multiple sleep latencytesting generally demonstrates a mean SOL of 8 min, and the Maintenance of Wakefulness Testmay show a mean SOL of 40 min.Therapy for SleepinessSleep extension should be recommended forsuspected insufficient sleep syndrome. Schedulednapping may be beneficial for some patients, asis caffeine intake, which has additive effects tonapping. Many patients, especially those withmoderate-to-severe sleepiness, will require stimulantagents.ParasomniasParasomnias are physical or experiential phenomenathat occur during sleep. They consist ofactivation of skeletal muscles or the autonomicnervous system. Parasomnias can occur primarilyduring non-REM (NREM) sleep, during REMsleep, or occur randomly throughout the sleepperiod. Parasomnias occurring during NREM sleep(ie, disorders of arousal) include confusional arousals,sleep terrors, and sleepwalking. Parasomniasoccurring during REM sleep include nightmaresand REM sleep behavior disorder.NREM ParasomniasThe disorders of arousal occur predominantlyduring N3 sleep and mostly in the first third of thenight. They are most commonly encountered inchildren, and cases generally resolve by adolescence.Risk factors for NREM parasomnias includesleep deprivation, OSA, and PLMD. They oftenrespond to either proper sleep hygiene, such asavoidance of sleep deprivation, and scheduledawakenings. A trial of low-dose benzodiazepinesmay be considered for intractable cases. 5Confusional Arousals: In this disorder, episodesof confusion follow arousals from sleep, accompaniedby inappropriate behavior, amnesia, inconsolability,and diminished responsiveness to externalstimuli. There is minimal fear or autonomic hyperactivity.Most episodes last from 5 to 15 min.Sleepwalking: Ambulation during sleep maybe precipitated by sleep deprivation (which isthe most important risk factor), febrile states, orOSA. If ambulation is significantly disruptive ofthe individual’s sleep or if sleepwalking is potentiallyinjurious, therapy, such as avoidance ofsleep deprivation, scheduled awakenings, andmedications (eg, low-dose benzodiazepines) canbe tried.Sleep Terrors: Sleep terrors consist of abruptawakenings with profound fear and intense autonomicdischarge, such as tachycardia, tachypnea,sweating, and mydriasis. Vocalization, includingtalking or screaming, may accompany frenziedactivity and ambulation. Therapy is similar tothat for sleepwalking.REM ParasomniasREM parasomnias tend to occur during thesecond half of the sleep period. Episodes are characterizedby full alertness after awakening, gooddream recall, and minimal tachycardia or tachypnea.6Nightmare Disorder: Nightmares are unpleasantand frightening dreams that can suddenly awakena person from sleep, followed by a delayed returnto sleep. Nightmares can be precipitated by OSA,febrile illness, medications, or alcohol ingestion.266 Nonrespiratory Sleep Disorders (Lee-Chiong)ACC-PULM-09-0302-016.3.indd 2667/10/09 10:18:22 PM


Medications that can cause nightmares includeamphetamines, antidepressants, and β-blockers.REM Sleep Behavior Disorder: REM sleep behaviordisorder consists of abnormal “dream-enacting”behavior occurring during REM sleep accompaniedby loss of REM-related muscle atonia. Menare affected more commonly than women, andprevalence is greatest in adults 50 years of age.Its course is chronic and progressive. REM sleepbehavior disorder can result in injuries to thepatient or bed partner. It often responds favorablyto low-dose clonazepam at bedtime. Environ mentalprecautions are essential to minimize injury.Other ParasomniasCatathrenia is characterized by expiratorygroaning during sleep, particularly REM sleep.Hoarseness can be present on waking. Episodesare not associated with respiratory distress, oxygendesaturation, or cardiac arrhythmias, and anupper-airway examination is generally normal asis sleep architecture.Persons experiencing exploding head syndromemay describe an awakening with a loudsound or sensation of explosion in the head.Exploding head syndrome may be a variant ofsleep starts. Women are affected more commonlythan men. Events are not associated with pain orneurologic complications.In sleep enuresis, recurrent involuntary voidingmay occur during sleep. Enuresis can be classifiedas either primary or secondary. Primary sleepenuresis affects a child 5 years of age who hasnever been consistently dry during sleep for 6consecutive months. Secondary enuresis is presentwhen the child or adult, who had previously beendry, begins bedwetting again. Risk factors for secondarysleep enuresis include CHF, diabetes, OSA,seizures, use of diuretics, and urinary tract infectionor disease. Structural urinary tract pathologyshould be suspected if daytime enuresis is alsopresent, if there are abnormalities in the initiationof urination, or if urinary flow is abnormal. Evaluationof sleep enuresis commonly consists of aurinalysis and urine culture. Urologic evaluationmay be considered for intractable cases or if structuralabnormalities are suspected.Sleep-related eating disorder should be consideredif a person presents with complaints ofrepetitive bouts of eating or drinking during arousalsfrom sleep. There is often a lack of awarenessof the abnormal behavior, with total or partialamnesia. Risk factors include OSA, sleepwalking,and medication use (eg, zolpidem).RLSPersons with RLS may describe an urge tomove, or unpleasant sensations, in the legs (lesscommonly the arms) that (1) begin or worsen duringperiods of rest or inactivity; (2) are relievedtransiently by movement; and (3) are worse, oroccur only, at night. 7 RLS affects an estimated 3 to15% of the general population, and prevalence isgreater in those with anemia or uremia, duringpregnancy, and with aging. Women are more commonlyaffected, as are middle-aged and olderadults. Its course is chronic. An estimated 70 to 90%of persons have periodic limb movements duringsleep; conversely, one third of persons with periodiclimb movements during sleep have RLS.Risk factors for RLS include the following:(1) iron deficiency anemia; (2) uremia; (3) pregnancy;(4) Parkinson disease; (5) diabetes mellitus;(6) alcohol or caffeine ingestion; (7) smoking; and(8) medication use (eg, selective serotonin reuptakeinhibitors, tricyclic antidepressants, monoamineoxidase inhibitors, lithium, antihistamines, neuroleptics,and dopamine antagonists). RLS can give riseto either sleep-onset and sleep-maintenance insomniaor sleepiness caused by sleep fragmentation.The diagnosis of RLS is usually arrived at byclinical history. Laboratory evaluation (CBC count,serum iron, ferritin, folate, electrolytes, thyroidfunction tests, fasting glucose, and renal panel) arerecommended to exclude secondary causes of thedisorder. Polysomnography is not routinely indicated;however, if performed, it often demonstratesa delayed SOL, diminished sleep efficiency, andincrease in wake time after sleep onset. Periodiclimb movements during wakefulness 15/h maybe noted before sleep onset. As stated previously,periodic limb movements during sleep may bepresent. RLS has to be differentiated from akathisiarelated to the use of neuroleptic agents or dopaminereceptor antagonists or from peripheralneuropathy.Therapy for RLS should start with treatmentof known underlying causes or precipitating<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 267ACC-PULM-09-0302-016.3.indd 2677/10/09 10:18:22 PM


factors. Iron supplementation may be consideredif serum ferritin is 50 µg/L. Dopaminergicagents, such as levodopa, pramipexole, or ropinirole,are the preferred first-line agents. 8,9Adverse effects of dopaminergic agents includeaugmentation, or earlier onset or increased severityof symptoms, or involvement of other bodyparts such as the arms, which is more commonlyencountered with levodopa. Use of pramipexoleand ropinirole can give rise to nausea, sleepiness,orthostasis, and the development of compulsivedisorder. The administration of pergolide hasbeen associated with the development of pleuropulmonaryand cardiac valve fibrosis. Anotheruseful class of agents are the benzodiazepines,such as clonazepam, which provides symptomaticrelief. Opioid agents may be necessary for severesymptoms refractory to other therapy. Finally,gabapentin may be considered for RLS accompaniedby pain.PLMDIn this disorder, recurrent leg movements, consistingof partial flexion of the ankle, knee, and hipwith extension of the big toe, are accompanied bycomplaints of sleep disturbance or excessivesleepiness. Involvement of the upper extremityconsists of flexion at the elbow. PLMD shares manyof the risk factors of RLS. Polysomnography isrequired for diagnosis; periodic limb movementindex is abnormal if it is 5 in children or 15 inadults. Therapy of PLMD is similar to that of RLS.Specific therapy is not indicated for asymptomaticperiodic limb movements during sleep.Circadian Rhythm Sleep DisordersCircadian rhythm sleep disorders are causedby a recurrent or persistent misalignment betweenthe desired sleep schedule and the circadian sleepwakerhythm. They can present with either insomniaor sleepiness, or both. 10Advanced Sleep Phase DisorderPersons with advanced sleep phase disorder(ASPD) are often referred to a morning larks thatprefer an early bedtime, between 6:00 and 9:00 pm,and an equally early wake time, from 2:00 to5:00 am. 11 Sleep, itself, is normal for age. Sleep-relatedcomplaints include excessive sleepiness in the lateafternoon or early evening, or morning awakeningthat is earlier than desired. Onset of ASPD is commonlyduring middle age. It is diagnosed by historyand sleep diary or actigraphy. Therapy consists ofearly evening bright light therapy administeredbefore minimum core body temperature.Delayed Sleep Phase DisorderThis disorder is characterized by night owls thathave a preferred late bedtime, between 1 and 6 am,and a delayed wake time, from 10:00 am to 2:00 pm. 11There is generally no difficulty remaining asleepafter the onset of sleep. Onset of delayed sleepphase disorder is commonly during adolescence.Clinical course is chronic, but severity of nighttimeinsomnia and daytime sleepiness may diminishwith increasing age. Like ASPD, diagnosis if basedon history and sleep diary or actigraphy. Personswith delayed sleep phase disorder often respond totimed early morning light exposure provided afterthe minimum core body temperature accompaniedby evening avoidance of bright light. Chronotherapycan either involve a progressive phase delay orprogressive phase advancement. Melatonin administeredin the early evening is also helpful.Free-Running Circadian DisorderIn this circadian rhythm disorder, there is aprogressive daily delay in sleep-onset and waketimes, with the major sleep period “marching”progressively throughout the day, afternoon, andevening. 11 This pattern of the major sleep periodresults in periodically recurring complaints ofinsomnia or excessive sleepiness. Free-runningdisorder is rare in the general population, and mostaffected patients are totally blind and lack photicentrainment. It has a chronic course. This disorderis diagnosed by a compatible history, often aidedby sleep diaries or actigraphy. Polysomnographyis not routinely indicated for diagnosis. Therapyconsists of evening administration of melatonin. Aregular sleep-wake and daytime activity scheduleis important.268 Nonrespiratory Sleep Disorders (Lee-Chiong)ACC-PULM-09-0302-016.3.indd 2687/10/09 10:18:23 PM


Irregular Sleep-Wake RhythmThere are no stable circadian sleep-wake rhythmsin this disorder. Rather, variable, inconsistent, andmultiple sleep and wake periods occur throughoutthe day and from one day to another. 11 Nevertheless,aggregate sleep time during a 24-h period is normal.This disorder is most frequently seen in associationwith dementia or mental retardation. Diagnosisis arrived at by history and sleep diary or actigraphy.Evening administration of melatonin is usefulin establishing a more consistent bedtime.Jet LagJet lag consists of either transient insomnia orsleepiness after rapid eastward or westward airtravel across multiple time zones and is causedby the lack of synchrony to the new local timezone. 12 Thus, persons may complain of sleeponsetinsomnia and/or difficulty awakening thenext day after eastward flights. On the other hand,after westward flights, they may describe earlyevening sleepiness as well as early morningawakenings. Symptoms of jet lag remit spontaneouslywithin about a day for every time zonechange. If needed, therapy involves bright lighttherapy, with evening bright light exposure forwestward travel, and morning bright light exposureafter eastward travel. Short-acting hypnoticagents or melatonin may be used at bedtime forinsomnia.Shift Work Sleep DisorderShift work sleep disorder is defined by sleepdisturbance related to nonstandard work schedules,with persons either complaining of sleepinessand decreased alertness during nightshifts as well as insomnia during daytime sleepperiods. 12 Clinical history and sleep diaries arehelpful in the diagnosis of this disorder. Therapy,if required, consists of measures to increasenighttime alertness, such as bright light exposurein the workplace; napping before (or during)night work; psychostimulants (eg, caffeine,modafinil, or armodafinil); and measures toenhance daytime sleep, including hypnoticagents, melatonin, and restriction of daytimelight exposure.Medical DisordersHypertensionOSA is a risk factor for hypertension independentof known confounding factors; indeed, theodds of hypertension increase by approximately1% for each additional apneic event per hour ofsleep. There is often loss of the nocturnal decreasein BP (“dipping” phenomenon) in persons withOSA and improvement in BP control with positiveairway pressure therapy.Coronary Artery DiseaseRisk of coronary artery disease is increased inmiddle-aged persons with OSA. Possible mechanismsinclude endothelial dysfunction, hypercoagulability,insulin resistance, oxidative stress,and increased sympathetic activity duringsleep.Congestive Heart FailureThe prevalence of OSA, CSA, and Cheyne-Stokes respiration is increased in CHF. Alternatively,OSA may contribute to worsening leftventricular dysfunction, and the rate of mortalityis greater in CHF persons with worse apneahypopneaindexes.Cardiac ArrhythmiasThere is a decreased prevalence of prematureventricular contractions during sleep as the resultof greater parasympathetic tone and an increase inventricular arrhythmias during arousals fromsleep.Respiratory DisordersSeveral respiratory disorders, includingasthma, COPD, diaphragm paralysis, restrictivepulmonary diseases, and neuromuscular disordersare associated with significant sleep disturbance.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 269ACC-PULM-09-0302-016.3.indd 2697/10/09 10:18:23 PM


AsthmaPersons with asthma commonly complain ofinsomnia and sleep fragmentation. Nocturnalhypoxemia may develop as can nocturnal bronchoconstriction,the latter occurring as a result ofeither (1) circadian variability in airflow, which islowest in the early morning; (2) sleep-relatedchanges in autonomic nervous activity, ie, greaterparasympathetic tone and decrease in sympatheticactivity; (3) changes in lung capacity and inflammatorymediators; and (4) gastroesophageal reflux.Diagnosis is aided by showing a decrease in eveningpeak expiratory flow or FEV 1compared withdaytime values.COPDCOPD may present at night with repetitiveawakenings as the result of coughing or dyspnea.Nocturnal hypoxemia and hypercapnia maydevelop in advanced disease. Episodes of oxygendesaturation tend to be more frequent, of greaterduration, and more severe during REM sleep comparedwith NREM sleep. There are several mechanismsthat may be responsible for sleep-relatedoxygen desaturation, including hypoventilation(most important factor), ventilation/perfusionmismatching, and decrease in lung volumes. Overlapsyndrome is defined by the presence of bothCOPD and OSA. Compared with isolated COPD,persons with the overlap syndrome tend to havelower Pao 2, greater Paco 2, and greater mean pulmonaryartery pressures.Diaphragm ParalysisNocturnal hypoxemia can develop in personswith diaphragm paralysis and is worse duringREM sleep.Restrictive <strong>Pulmonary</strong> DiseasesRestrictive pulmonary diseases are commonlyassociated with sleep disturbance and frequentawakenings. Both OSA and CSA can develop.Nocturnal oxygen desaturation can either be transientor sustained and is worse during REMsleep.Neuromuscular DisordersNeuromuscular disorders are associated withan increased risk of OSA and CSA as well as nocturnaldyspnea. Nocturnal oxygen desaturationprimarily is caused by hypoventilation and occurspredominantly during REM sleep. Nocturnalhypoventilation can precede abnormalities duringwaking by months to years. The risk of nocturnaloxygen desaturation is greater in those with maximalinspiratory pressure 60 cm H 2O, or if FVCis 50% of predicted.References1. Schutte-Rodin S, Broch L, Buysse D, et al. Clinicalguideline for the evaluation and management ofchronic insomnia in adults. J Clin Sleep Med 2008;4:487–5042. Morin CM, Bootzin RR, Buysse DJ, et al. Psychologicaland behavioral treatment of insomnia:update of the recent evidence (1998–2004). Sleep2006; 29:1398–14143. Morgenthaler TI, Kapur VK, Brown TM, et al. Standardsof Practice Committee of the AASM. Practiceparameters for the treatment of narcolepsy andother hypersomnias of central origin. Sleep 2007;30:1705–17114. Wise MS, Arand DL, Auger RR, et al. Treatment ofnarcolepsy and other hypersomnias of central origin.Sleep 2007; 30:1712–17275. Mahowald MW, Schenck CH. Non-rapid eyemovement sleep parasomnias [review]. NeurolClin 2005; 23:1077–11066. Schenck CH, Mahowald MW. Rapid eye movementsleep parasomnias [review]. Neurol Clin2005; 23:1107–11267. Itin I, Comella CL. Restless legs syndrome. PrimCare 2005; 32:435–4488. Standards of Practice Committee of the AmericanAcademy of Sleep <strong>Medicine</strong>. Practice parametersfor the dopaminergic treatment of restless legssyndrome and periodic limb movement disorder.Sleep 2004; 27:557–5599. A <strong>Review</strong> by the Restless Legs Syndrome TaskForce of the Standards of Practice Committee ofthe American Academy of Sleep <strong>Medicine</strong>. Anupdate on the dopaminergic treatment of restlesslegs syndrome and periodic limb movement disorder.Sleep 2004; 27:560–583270 Nonrespiratory Sleep Disorders (Lee-Chiong)ACC-PULM-09-0302-016.3.indd 2707/10/09 10:18:23 PM


10. Morgenthaler TI, Lee-Chiong T, Alessi C, et al.Standards of Practice Committee of the AASM.Practice parameters for the clinical evaluation andtreatment of circadian rhythm sleep disorders.Sleep 2007; 30:1445–145911. Sack R, Auckley D, Auger RR, et al. Circadianrhythm sleep disorders: part II, advanced sleepphase disorder, delayed sleep phase disorder,free-running disorder, and irregular sleep-wakerhythm. Sleep 2007; 30:1484–150112. Sack RL, Auckley D, Auger RR, et al. Circadianrhythm sleep disorders: part I. Basic principles,shift work and jet lag disorders. Sleep 2007;30:1460–1483<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 271ACC-PULM-09-0302-016.3.indd 2717/10/09 10:18:23 PM


Notes272 Nonrespiratory Sleep Disorders (Lee-Chiong)ACC-PULM-09-0302-016.3.indd 2727/10/09 10:18:23 PM


<strong>Pulmonary</strong> Fungal InfectionsGeorge A. Sarosi, MD, FCCPObjectives:• Describe the epidemiology of the endemic mycosis• Describe the clinical syndromes of endemic mycosis• Familiarize participants with the various diagnostic testspertaining to pulmonary fungal infections• Show the various manifestations of aspergillosis andcandidiasis• Show different therapeutic strategies pertaining to fungaldiseasesKey words: aspergillosis; blastomycosis; candidiasis; coccidioidomycosis,cryptococcosis; histoplasmosisHistoplasmosisHistoplasmosis is the infection caused by thepathogenic, soil-dwelling fungus Histoplasma capsulatum.The organism is endemic throughout thesouth-central United States, especially along rivervalleys. Heavy concentrations of the fungus arefound in excrement of chickens, pigeons, starlings,and bats. A pulmonary infection will occur whenthe mycelium is disturbed and aerosolized fungalspores are inhaled.Clinical FeaturesPrimary <strong>Pulmonary</strong> Histoplasmosis: The primarydisease may have few or no symptoms. Even inthe asymptomatic patient, a chest radiograph mayshow patchy areas of pneumonitis with or withoutipsilateral hilar adenopathy. Patients become symptomaticaround 14 days after exposure. In the eventof a very large inoculum, the symptoms may beginsooner. The illness is an influenza-like disease withfever, chills, myalgia, and nonproductive cough.During the early nonimmune phase of the infection,the fungus disseminates widely throughout thebody. With the advent of specific T-cell−mediatedimmunity, the infection is brought under control.Patients who have been exposed to a very largeinfective inoculum will present with a potentiallylethal illness. The main difficulty is with gasexchange, and the chest radiograph shows a picturesimilar to that of ARDS. After recovery, the chestradiograph may return to normal or may be leftwith residual abnormalities. The initial patchy infiltratemay become denser and smaller, leading to theformation of pulmonary nodules. Central necrosiswithin these nodules can result in calcification, leadingto the development of the characteristic “target”lesion. Lymph node calcification alone or with apulmonary nodule is common. These calcificationsoccur more frequently in younger patients.Local complications may occur but, fortunately,not often. Extensive fibrosis (caused by exaggeratedresponse by the host) in the mediastinum maycause vascular compression and result in the superiorvena cava syndrome. Inflammatory massesadjacent to the esophagus may cause dysphagia.Calcified nodes may erode into various structures,leading to hemoptysis.Chronic Cavitary Histoplasmosis: This diseaseclosely resembles reinfection tuberculosis both insymptomatology and radiographic appearance.The mechanism of infection, unlike in tuberculosis,is that of a primary infection in a patient witha structurally abnormal lung, typically a middle-agedsmoker who has centrilobular emphysema.The acute infection in such patients usuallyresolves, but occasionally (in up to 20%) it will goon to produce progressive upper-lobe fibrocavitarydisease. Once the disease establishes itself and progressesunder observation, treatment is necessarybecause of continuing pulmonary destruction.Progressive Disseminated Histoplasmosis: Progressivedisseminated histoplasmosis (PDH) isa progressive extrapulmonary infection. It mayoccur as part of an overwhelming postprimaryspread (frequently seen in infants) with little orno tendency to granuloma formation despite thehuge burden of organisms that are present in reticuloendothelialtissues. Symptoms include highfever accompanied by diffuse lymphadenopathy,hepatosplenomegaly, and severe pancytopenia.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 273ACC-PULM-09-0302-017.indd 2737/10/09 8:29:14 PM


Death may occur within weeks and is usually secondaryto respiratory compromise.Dissemination also occurs in nonimmunocompromisedadults, in for whom the disease is asmoldering, subacute illness of many months induration. Fever is low grade at most, and most ofthe symptoms relate to weight loss. Physicalexamination shows skin and mucous membranelesions along with hepatosplenomegaly. Histopathologicexamination shows well-formedgranulomas and only rare organisms.In recent years, PDH has occurred primarily inpatients with significant T-cell dysfunction, includingorgan transplant recipients, patients receivingglucocorticoids and/or cytotoxic drug therapy formalignant neoplasms, and patients with HIV infection.In this latter group, PDH is frequently theAIDS-defining infection. The density of the organismsin biopsy material is greater than that in otherimmunosuppressed patients, helping materiallywith the laboratory diagnosis.It is important to note that perhaps as many asone half of patients with PDH have no symptomsrelated to the lung. Especially in patients withAIDS, PDH is primarily a debilitating, febrile illness.Eventually, chest radiographic abnormalitiesdevelop in all patients.The incidence of PDH complicating untreatedHIV infections has resulted in the acknowledgmentof parts of the world as being endemic for histoplasmosisthat were not known to harbor the fungus.These areas include the entire Caribbean basinextending all the way to the northern coast of SouthAmerica. In addition, many of the states of sub-SaharanAfrica have identified patients with histoplasmosis.The extent of the disease in Europe is not known.Although in Europe most patients who present withPDH come from endemic areas, a handful of casesappear to have had no exposure outside of Europe.The recent emergence of tumor necrosis factor(TNF)-α inhibitors has caused a clear-cut associationwith PDH. All anti-TNF-α agents may lead to PDH,but its occurrence is most frequent after the administrationof infliximab. In fact, PDH is the second mostcommon infectious complication after tuberculosis.Diagnostic TestsHistopathologic specimens reveal small yeaststhat are highly suggestive of histoplasmosis eitherwithin macrophages or lying free in areas of necrosis.Although the organisms occasionally may bevisualized in routine hematoxylin-eosin−stainedsections, as a general rule they are difficult to visualizeunless special stains are applied. Most experiencedpathology laboratories will use one of thesilver stains or the periodic acid-Schiff stain. Theorganism is best visualized from biopsies ofthe reticuloendothelial system, with the highestyield coming from bone marrow biopsy and fromblood cultures.Whenever skin lesions are present, biopsiesshould be performed. In patients with AIDS, thedensity of Histoplasma organisms is so high thatthey may be visible on Wright-stained smears ofthe buffy coat of the peripheral blood.The organism is readily recovered from biologicalmaterial, and whenever the disease is suspected,every attempt should be made to obtainadequate specimens for cultures. The availabilityof the lysis-centrifugation system has increased ourability to isolate the fungus from circulating blood,but standard radiometric tests also will grow thefungus readily. The main difficulty with cultureidentification is that it is time-consuming; wheneverpossible, histopathologic sections should beexamined because they are less time-consuming.Intradermal skin testing with histoplasmin is areliable epidemiologic tool, but it is useless for thediagnosis of individual cases. In addition, it is nolonger available commercially.Complement-fixing antibodies to mycelial andyeast phase antigens may be detected 4 to 6 weeksafter exposure (2 to 4 weeks after onset of fever).The titers usually decrease gradually during thenext 12 months, but in some patients they mayremain positive at a very high dilution for manyyears. Titers against the yeast antigen of 1:32 ina patient with an appropriate clinical historyshould suggest histoplasmosis. Similarly, a documentedfourfold increase of titer is diagnostic ofacute Histoplasma infection. Unfortunately,complement-fixing antibody test results are negativein as many as 50% of immunocompromisedpatients with PDH; thus, a negative complementfixingtest result cannot be used to rule out thedisease.The currently available immunodiffusion test,although highly specific, is not sensitive enoughfor routine clinical use. Its main use is in patients274 <strong>Pulmonary</strong> Fungal Infections (Sarosi)ACC-PULM-09-0302-017.indd 2747/10/09 8:29:14 PM


whose serum is anticomplementary. The availabilityof the Histoplasma polysaccharide antigen is amajor step forward. The test is extremely useful indiagnosing PDH, especially in immunocompromisedpatients. The test is able to detect excretionof this antigen in the urine and blood and can beused both for diagnosis and to follow the courseof the disease, including prediction of relapses.TreatmentPrimary pulmonary histoplasmosis requiresno treatment. In a rare incident of an individualwith massive inhalation and ARDS, amphotericinB or a lipid formulation of amphotericinB(L-amphotericin) is the treatment of choice whengas exchange abnormalities threaten his or her life.A total dose of 500 to 1,500 mg may be sufficient,although this dose is based on purely anecdotalinformation.Chronic cavitary histoplasmosis responds readilyto itraconazole 200 mg bid for at least 6 months.Although the response to amphotericin is quicker,the reduction in side effects that accompanies itraconazoleadministration makes it the agent ofchoice. Occasional treatment failures and delayedrelapses with itraconazole respond to a subsequentcourse of itraconazole or amphotericin. The newerazoles voriconazole and posaconazole both haveexcellent activity against the fungus, but their highcost makes them less attractive alternatives.Itraconazole can also be used to treat PDH inthe rare event of a patient whose illness is slow toprogress. In most other patients, the drug of choiceis amphotericin; the dose is rapidly accelerated to50 mg/d and continues until clinical stability isreached, which may take 500 to 1,000 mg of amphotericin.After clinical stability is reached, patientsshould be treated with itraconazole for 6 to 12months. Most nonimmunocompromised patientsshould respond well, with cure of the diseaseachieved. Immunocompromised patients whoseimmunosuppressive state cannot be improved andpatients with AIDS who continue to have CD4 counts < 200 μL require continued suppression.CoccidioidomycosisCoccidioidomycosis, or “cocci,” is the infectionby the fungus Coccidioides immitis. This organismis endemic in the desert southwestern UnitedStates and adjacent areas of Mexico and is responsiblefor approximately 100,000 new infections eachyear. Approximately 60% of infected patients haveno symptoms, whereas the remaining 40% usuallyhave only mild symptoms of a febrile, influenzalikeillness that lasts 1 to 3 weeks. Patients usuallyhave a nonproductive cough along with myalgiasand arthralgias, headaches, and pleuritic chestpain. The pleuritic chest pain may be very severe.A chest radiograph may show localized alveolarinfiltrate, a single nodular density, or a pleuraleffusion. Many patients with acute cocci have anillness that is indistinguishable from communityacquiredpneumonia. In a recent small series fromthe endemic area, acute cocci were seen in approximatelyone third of the patients. In rare instances,the patient may present with a pneumothoraxsecondary to the rupture of a subpleural cavity.Fortunately, most patients recover without furtherproblems. Approximately 5% may continue toshow persistent chest radiographic abnormalitiesbut remain in good health.Disseminated disease will be present in < 1%of patients, and it most commonly occurs amongblack, Filipino, and diabetic individuals, and inpatients who are receiving cytotoxic therapy orglucocorticoids for other disease entities. Disseminationis also likely to occur in organ transplantrecipients, patients receiving anti-TNF-αtherapy, as well as in patients infected withHIV.The so-called allergic manifestations of the illness,such as erythema multiforme or erythemanodosum, occur in up to 20% of patients, mostoften in young white women. These are thoughtto be favorable prognostic signs. In endemic areas,this manifestation of “valley fever” is readily recognizedby all practitioners. A persistent thinwalledcavity or a single nodule is the mostcommon radiographic residue of acute coccidioidomycosis.On occasion, a slowly progressivefibrocavitary disease may develop that resemblesreinfection tuberculosis.Disseminated CocciFortunately, this most-dreaded form of thedisease is uncommon. It may be a fulminant illnesswith development of meningitis along with bone<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 275ACC-PULM-09-0302-017.indd 2757/10/09 8:29:15 PM


and skin lesions. Other patients, especiallyseverely immunocompromised patients, willpresent with a rapidly progressive extensivepulmonary infiltrate that leads to air exchangeabnormalities and that may become fatal. For nonimmunocompromisedpatients, the onset of disseminationis usually more leisurely, and it mayfollow the primary infection by several weeks toseveral months. In many instances, the primaryinfection was asymptomatic, and thus cannot beaccurately dated.Cocci meningitis is a frequent manifestation ofextrapulmonary disease. Approximately one thirdof patients will have single organ dissemination tothe meninges. The inflammatory process involvesthe base of the brain, where it obstructs the flow ofcerebrospinal fluid (CSF) and entraps cranialnerves. Headache is common, and changes inmental acuity are frequently seen. Stiff neck andother meningeal signs are seldom seen. Examinationof CSF reveals a mononuclear cell pleocytosiswith depressed glucose and elevated protein levels.Other organs frequently involved in disseminateddisease include the bones and the skin. Eventually,most patients with disseminated cocci will havecutaneous involvement.Coccidioidomycosis is being recognized withincreased frequency in patients with HIV infection.Depending on immune status, the disease maylook no different from that found in non-HIV-infected patients; toward the end of the HIVinfection, when the CD4 count is 250 μL, coccidioidomycosisis usually a rapidly progressive,hematogenously spread, and rapidly fatal illness.The characteristic radiograph in this form of coccidioidomycosisis a micronodular infiltratethroughout all lung fields.Diagnostic TestsThe pathognomonic feature of cocci infectionis the giant spherule. This form of the organismmay range up to 100 μm and is filled with numeroussmall endospores. The spherule is sufficientlylarge and is readily seen on hematoxylin-eosinstainedsections. Giant spherules may also be seenby Papanicolaou stain-treated smears of sputumor pus, but less well as on potassium hydroxidetreatedsmears. The organism can readily be culturedin the laboratory, but great care must beexercised because this is a highly infectiousorganism.Most experts believe skin tests are helpful.Unfortunately, a single skin test cannot date theonset of the illness and is therefore better thoughtof as an epidemiologic tool than as a useful test forthe diagnosis of individual patients with cocci.Serologic diagnosis is extremely useful incoccidioidomycosis. Serum IgM precipitants maybe demonstrated by tube precipitations or bylatex agglutination. These are first detected 2 to3 weeks after the onset of symptoms of the primaryinfection and will eventually be noticed in75% of patients. Positive precipitins will diminishwithin 4 to 6 months. Serum IgG antibodies, theso-called complement-fixing antibodies, occur later;and by 3 months after the onset of symptoms,nearly 90% of patients have such antibodies.High titers or rising titers can also be used to alertthe clinician to the possibility of disseminatinginfection. Recently, immunodiffusion tests measuringboth IgM and IgG antibodies came intouse and are replacing other methods of measuringantibodies. Their sensitivity and specificityare very much the same, but they are much easierto perform.The height of the titer will vary from laboratoryto laboratory, and before a great deal of emphasisis placed on the relative height of a given titer, theorigin of the titer needs to be determined. Inapproximately 70% of patients with cocci meningitisthere are complement-fixing antibodies in theCSF. The presence of such antibodies is diagnosticof cocci meningitis. This is particularly importantbecause the organism is seldom cultured initiallyfrom CSF. Recently, an antigen-based diagnostictest became available that requires the specimensof either urine or blood. Initial results have beenpromising, but more data are needed.TreatmentPrimary pulmonary cocci require no treatment.However, treatment is advised for immunocompromisedpatients or other patients who have ahigh risk of dissemination (eg, diabetic, black, orpregnant patients). Treatment may also be recommendedfor those with severe, persistent infection.The usual dose of amphotericin for primary infectionunder these circumstances is between 500 and276 <strong>Pulmonary</strong> Fungal Infections (Sarosi)ACC-PULM-09-0302-017.indd 2767/10/09 8:29:15 PM


1,500 mg. Fluconazole, an orally administeredazole, is frequently used by practitioners in theendemic area for the treatment of acute coccidioidomycosisdespite the fact that the efficacy of thistreatment has never been established.Residual thin-walled cavities usually requireno treatment. Resection with or without amphotericinmay be needed if hemoptysis develops orif the cavities enlarge, move subpleurally, andthreaten rupture. The chronic fibrocavitary formof pulmonary coccidioidomycosis requires treatment.Although some patients respond to itraconazoleor fluconazole, more severely ill patientswill require amphotericin for effective treatment.Although the results with amphotericin are better,the treatment is seldom curative.Disseminated cocci have to be treated withamphotericin. An initial course of 2.5 g is recommended,but if a patient does not respond, the dosemay be increased to greater levels. In all patientswith disseminated cocci, a lumbar puncture must beperformed to rule out meningeal dissemination.Cocci meningitis always requires treatment.When amphotericin is chosen, both IV and intrathecalamphotericin therapy might be used. In manypatients, prolonged intrathecal therapy is necessaryfor the meningeal disease in addition to the standardcourse of IV amphotericin. It is administeredin doses ranging from 0.25 to 1 mg by cisternalpuncture, by lumbar puncture, or via an intraventricularreservoir. Treatment may be given up tothree times weekly and is usually continued untilboth the cell count and complement-fixing activitydiminish. At that time, the interval of administrationmay be extended to twice per week, then toonce per week, and eventually to biweekly ormonthly as the patient improves. This form of treatmentwill need to be continued for the remainderof the patient’s life. Recently, however, fluconazolereplaced amphotericin as primary therapy andamphotericin, is used in when fluconazole therapyhas failed. When fluconazole is used, it must becontinued for the patient’s life. The withdrawal offluconazole leads to an unacceptable relapse rate.Disseminated coccidioidomycosis involvingskin, bone, and noncavitating pulmonary lesionsmay be treated with itraconazole, 200 mg bid, orfluconazole, 400 mg/d. Ttreatment of coccidioidomycosisin patients with HIV infection is a dauntingchallenge. Response is poor, and the eventualprognosis is grim. Treatment should begin withamphotericin to the maximum tolerated dose, andonce the patient stabilizes, most investigators willswitch to fluconazole. The exact dose needed is notdetermined.BlastomycosisBlastomycosis refers to an infection by thefungus Blastomyces dermatitidis. The organism isendemic in the Midwest, southwestern UnitedStates, and the Canadian provinces Ontario andManitoba. Similar to histoplasmosis and coccidioidomycosis,blastomycosis is a soil fungus, existingin soil of high organic content in microfoci inthe endemic area.Clinical FeaturesAcute <strong>Pulmonary</strong> Blastomycosis: The infectionmay or may not be symptomatic. Radiographyreveals patchy alveolar to lobar infiltrates. Hilarenlargement is common, but pleural disease andcavitation are infrequent. The clinical symptomsinclude an influenza-like illness with high fever,and a cough productive of mucopurulent sputumalong with myalgias and arthralgias. Althoughsymptomatic disease is usually short-lived, radiographicabnormalities may persist for severalmonths. Progressive pulmonary infection withcontinued suppuration and eventual cavitation isan infrequent occurrence. An occasional patientmay present with rapidly progressive bilateralmacronodular illness, leading to ARDS.Most cases of blastomycosis present in achronic or subacute form. Patients usually have noantecedent history of an acute pulmonary illnessbut will complain of chronic respiratory symptomsof at least 2 months in duration. Symptoms includea chronic cough, low-grade fever, night sweats, andweight loss. The chest radiograph may revealeither a single mass or a lobar infiltrate. It is importantto realize that in certain patients, the primarypulmonary process may well have healed beforethe infection develops in distant sites.Diagnostic TestsThe fungus is readily visualized in expectoratedsputum or other purulent material after<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 277ACC-PULM-09-0302-017.indd 2777/10/09 8:29:15 PM


digestion of 10% potassium hydroxide. The large,single-budding organism is characterized by itsbroad neck of attachment and its double refractilecell wall. The organism is difficult to see in standardhematoxylin-eosin-stained sections but isreadily visualized with periodic acid-Schiff- orsilver-stained sections. The organism is readilyrecovered in cultures, although it is a somewhattime-consuming process. Depending on inoculumsize, cultures may take up to 3 weeks before theyare recognizable.There is no commercially available skin test,and the serologic diagnosis in blastomycosis isquite rudimentary. At present, even the presenceof a positive serologic test (measuring antibody)result using one of the more recent ultrasensitiveassays does not confirm the diagnosis ofblastomycosis—it merely points one in the directionof the infection. The recently available urineantigen has performed well in preliminary studies.TreatmentAcute pulmonary blastomycosis may notrequire treatment in some patients; most patientswith acute pulmonary disease should be treatedwith itraconazole, 200 mg bid, for at least 6 months.In severely ill patients in whom gas exchangeabnormality forces the issue, amphotericin is useduntil the patient reaches clinical, and this shouldbe followed by a course of oral itraconazole, 200mg bid, daily 6 months. All patients with blastomycoticmeningitis should receive amphotericin;there is no evidence that intrathecal amphotericinadministration improves prognosis. Because itraconazolereach high levels in CSF, high-dose fluconazoleis used as step-down therapy afterclinical stability is reached.CryptococcosisCryptococcosis is the infection caused by theencapsulated yeast Cryptococcus neoformans. In thenormal host, primary pulmonary infection seldomproduces a clinical illness. The organism existsworldwide and is frequently associated with birddroppings, especially from pigeons. Most patientswith significant cryptococcal disease have anunderlying immunocompromised state.Clinical FeaturesPrimary pulmonary cryptococcosis usually isdiscovered accidentally because most patientshave no symptoms. When symptoms are present,fever is modest and a chest radiograph frequentlyshows a large, localized lobar infiltrate. The usualnatural history of acute pulmonary cryptococcalinfection in immunocompetent hosts is that ofspontaneous recovery. Nevertheless, in somepatients with an apparently normal immune system,progressive pulmonary or disseminated diseasemay develop; the current recommendation isto treat all such patients with a course of fluconazole.If the decision is to treat a patient with pulmonarydisease, a lumbar puncture should be doneto make sure the patient does not have meningealinvolvement.The vast majority of the patients with cryptococcaldisease have cryptococcal meningitis. Thesymptoms frequently are trivial; change in mentalacuity and the presence of headaches frequentlyare the only clue to the involvement of the meningesby the fungus. More definite signs of meningealirritation are uncommon. CSF examination usuallyshows increased protein with a predominantmononuclear pleocytosis and depressed glucoselevels. It is important to remember that the CSFformula may be completely normal in patients withHIV coinfection.Diagnostic TestsThe organism is a single-budding yeast witha narrow-necked bud that may reach up to 12 μmin diameter. The character feature is the verylarge carbohydrate capsule. Histopathologicexamination of involved tissue will frequentlyshow the organism stained with mucicarmineand the various silver stains. Frequently, noinflammatory infiltrate surrounds the mass oforganisms. The large carbohydrate capsule is thebackground behind the India ink test, which willdemonstrate the organisms in up to 70% of theimmunocompromised patients’ lumbar puncturespecimens.The organism grows really readily in moststandard laboratory media and lends itself torapid laboratory diagnosis. Cultures findings in278 <strong>Pulmonary</strong> Fungal Infections (Sarosi)ACC-PULM-09-0302-017.indd 2787/10/09 8:29:15 PM


CSF are positive in 80 to 90% of patients. A bloodculture also should performed because it resultsfrequently are positive in immunocompromisedpatients. However, interpretation of sputum culturesis very difficult because the organism maybe a colonizer. There is no commercially availableskin test.Besides culture, serologic tests form the mainstayof diagnosis. Direct measure of the fungalantigen is possible with the widely available latexagglutination method. Results are positive in theCSF of 95% of patients with cryptococcal meningitis,and the method is more sensitive than theculture itself. It should also be determined in theblood since it frequently reaches very high titers,especially in patients with HIV infection.TreatmentAll immunocompetent patients with pulmonarycryptococcosis and a negative spinal tap resultshould be treated with a 6- to 12-week course offluconazole, 400 mg/d. <strong>Pulmonary</strong> cryptococcosisin immunocompromised patients is a progressiveillness, and all such patients should receive prompttreatment, even if the meninges are not involved.Most of the data dealing with treatment of cryptococcaldisease were derived from large groups ofpatients with cryptococcal meningitis, and treatmentof cryptococcal pulmonary disease is largelyderivative of these studies.All patients with cryptococcal meningitisshould receive amphotericin, 0.7 mg/kg/d, plusflucytosine, 100 mg/kg/d divided into four equaldoses, for 2 weeks. If at the end of 2 weeks theCSF is sterile, the patient should be switched tooral fluconazole, 400 mg/d, for 6 additionalweeks. The addition of flucytosine is clearly beneficial,but great care must be exercised to maintainflucytosine levels < 100 mg/mL to preventbone marrow suppression. The combination ofAIDS and cryptococcal meningitis is a particularproblem to treat. Although many patients respondto treatment, substantial numbers will not andproceed to die. Relapse is extremely common oncetreatment is stopped; thus, it is necessary to treatall responding patients with suppressive doses offluconazole, 200 mg/d, to decrease the frequencyof relapse.AspergillosisEpidemiologyAspergillosis sp are soil fungi found everywherethere is decaying vegetation. There are 300species of Aspergillus, but only a handful havebeen associated with human disease. By far themost common are Aspergillus fumigatus, followedby Aspergillus flavus, Aspergillus clavatus, and Aspergillusnidulans.PathogenesisAspergillus spores are inhaled regularly by allindividuals. Colonization of the respiratory treemay occur, especially after heavy exposure. Thesame Aspergillus spores may cause colonizationof any devitalized area of the lung. The most commonand certainly the most characteristic featureof aspergillosis is colonization of previouslyformed cavities in the lung. Aspergillus growingin preformed cavities produces a characteristiclesion, the intracavitary fungus ball or aspergilloma.Although occasional symptoms such ashemoptysis are distressing to the patient, as a generalrule the fungus does not invade surroundinglung tissue.Inhaled Aspergillus spores may also colonizethe mucus within the bronchi, especially when theyare associated with moderately severe asthma andthey are thick and tenacious in nature. In thisclinical setting, invasion of the mucus plug occurs,exacerbating the underlying airways disease andleading to increased inflammation and eventualfibrosis. This disease entity is allergic bronchopulmonaryaspergillosis.A recently described form of aspergillosisinvolves colonization of previously abnormallung substance. The involved patients usuallyhave a minimal degree of immunosuppressionand frequently have sarcoidosis or diabetes.Aspergillus spores may invade adjacent lung tissuesand produce a gradually progressive anddestructive process in lung containing centrilobularemphysema. Occasionally, fungus balls mayform in these cavities. Although distant spread israre, the disease may be locally invasive anddestructive.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 279ACC-PULM-09-0302-017.indd 2797/10/09 8:29:16 PM


The most lethal form of aspergillosis is disseminatedor pyemic aspergillosis. When sporesare inhaled by an abnormal host with reduced ornonfunctional neutrophils and macrophages, disseminationof the infection will occur. Withoutappropriate inflammatory cell activity, the fungiwill grow within the alveoli and invade adjacentvascular structures, leading to occlusion of thesevessels. Necrosis follows occlusion of the vessels,leading to wedge-shaped areas of infarction.Metastatic abscesses in brain, lung, liver, heart, andother organs are common. When lesions involvethe skin, they give rise to a characteristic skinlesion, with an area of central necrosis and a blackeschar. This skin lesion is referred to as ecthymagangrenosum. Occasionally, Aspergillus endocarditismay follow pyemic spread.Clinical ManifestationHemoptysis is the major clinical feature of thefungus ball. Although rare, occasional exsanguinatinghemorrhage has been noted. No drug therapyhas proven to be effective. In patients whose pulmonaryreserve warrants it, lung resection is aneffective mode of therapy. The principal symptomof allergic bronchopulmonary aspergillosis is episodicdyspnea, largely because of the underlyingreactive airways disease. The otherwise-severecourse of underlying asthma is punctuated byepisodes of worsening, when thick mucus plugsbecome inspissated in bronchi, causing an inflammatoryprocess distal to the obstruction. Thispropensity to cause bronchial obstruction gives riseto the characteristic radiographic pattern of thedisease, the so-called finger-in-glove appearance,in which multiple adjacent bronchi are distendedwith the mucus plug. Fever, cough, and pleuriticchest pain are common, and peripheral bloodeosinophilia is seen. Between clinical exacerbations,patients tend to return to their baselinesymptomatic state from their asthma. Eventually,however, because of the development of chronicfibrous changes, the restrictive lung function patternis overlaid on top of the reactive airwaysdisease.Disseminated or pyemic aspergillosis occursprimarily in patients with profound bone marrowsuppression. This danger is the greatest one inpatients with acute leukemia or bone marrowtransplantation. Characteristic clinical manifestationsare all related to the vascular occlusive natureof the disease, leading to a syndrome mimickingmultiple pulmonary emboli or stroke syndromesecondary to involvement of the intracranial bloodvessels. Aspergillosis also complicates AIDS in asmall number of patients. The same factors predisposingto aspergillosis are found in both patientswith and without HIV infection.DiagnosisAspergilloma is usually diagnosed by thecharacteristic radiographic pattern and confirmedby sputum culture. Diagnosis of allergic bronchopulmonaryaspergillosis is made with the recoveryof the fungus from sputum, the presence ofincreased IgE levels (especially specific IgE levels),peripheral blood eosinophilia, and the expectorationof bronchial plugs. Chronic necrotizing aspergillosisis diagnosed by the presence of a slowlyprogressive, usually upper-lobe disease that isalso accompanied by repeated isolations of Aspergillussp from sputum. The radiographic featuresof this disease frequently are indistinguishablefrom those of other chronic inflammatory lunginfections, such as tuberculosis or other fungalinfections.Disseminated aspergillosis is frequently a diagnosticchallenge. Blood culture results are almostalways negative; unless a highly characteristicclinical picture develops, such as the developmentof ecthyma gangrenosum, the diagnosis can bemade only by visualization in tissue or recovery inculture of the organism after either tissue biopsyor bronchoscopic specimen examination. Radiographicfindings are variable, but the so-called halosign surrounding an infiltrate is suggestive.Although frequently negative, a positive sputumculture result is extremely helpful in suggestingthe diagnosis. Measurement of serum galactomannanhas been shown to be the key in diagnosingearly disease in stem-cell transplant patients.TreatmentAspergilloma usually responds to surgery. Noantifungal agent has shown any consistent activityagainst the intracavitary fungus ball. Allergic bronchopulmonaryaspergillosis responds quite well to280 <strong>Pulmonary</strong> Fungal Infections (Sarosi)ACC-PULM-09-0302-017.indd 2807/10/09 8:29:16 PM


administration of high-dose glucocorticoids.Improvement is hastened and enlarged on by theaddition of itraconazole at a dose of 200 mg bid.Chronic necrotizing aspergillosis and disseminatedor pyemic aspergillosis are treated with systemicantifungal therapy.Amphotericin and L-amphotericin have beenused with variable success. Itraconazole has goodactivity against aspergillosis, but voriconazole isfavored. The new azole posaconazole is also active,but it exists only in oral form, precluding its use incritically ill patients. A new class of antifungalagent is the echinocandin. Although they all haveactivity against aspergillus, they are used primarilyas rescue therapy. Combination therapy withtwo different classes of antifungal drugs has beenused, but it is difficult to determine during whichcircumstances they should be used. Most reportsare case series only.CandidiasisCandida sp are normal inhabitants of themicrobiologic flora of people. Superficial Candidasp infections usually respond to simple measures.During the past 20 to 25 years, candidiasis becamea common problem in severely immunocompromisedpatients and more recently in nonimmunocompromisedcritically ill patients. Candida spbloodstream invasion is now the fourth-mostcommonnosocomial bloodstream invasion. Inaddition to Candida sepsis, other organs maybecome infected with the fungus. It is especiallycommon as a central line-associated sepsis incritically ill antibiotic-treated patients.Diagnosis requires recovery of the fungus inblood cultures or cultures from other sterile sites.Isolation of the fungus from other sites impliescolonization. Measurement of β (1-3) glucan detectscomponents of the fungal wall, and it may bepositive in candidiasis as well as other fungal infections.Of the three classes of antifungal agentsavailable, ie, polyenes, azoles, and echinocandins,the decision to pick one agent over another is leftto the discretion of the attending physicianAnnotated BibliographyBaddley JW, Perfect JR, Oster RA, et al. <strong>Pulmonary</strong>cryptococcosis in patients without HIV infection:factors associated with disseminated disease. Eur JClin Microbiol Infect Dis 2008; 27:937−943A useful discussion of the primary infection and the potentialfor extrapulmonary spread.Bergstom L, Yocum DE, Ampel NM, et al. Increasedrisk of coccidioidomycosis in patients treated withtumor necrosis factor alpha antagonists. ArthritisRheum 2004; 50:1959−1966Similar to histoplasmosis, the use of these agents often leadsto secondary infections.Bradsher RW Jr. <strong>Pulmonary</strong> blastomycosis. SeminRespir Crit Care Med 2008; 29:174−181An excellent and current review of the most common formof blastomycosis.Bradsher RW Jr., Chapman SW, Pappas PG. Blastomycosis.Infect Dis Clin North Am. 2003; 17:21−40Comprehensive presentation of the infection.Chapman SW, Dismukes WE, Proia LA, et al. Clinicalpractice guidelines for the management of blastomycosis:2008 update by the infectious Diseases Society ofAmerica. Clin Infect Dis 2008; 46:1801−1812The last word in treatment.Chow JK, Golan Y, Ruthazer R, et al. Factors associatedwith candidemia caused by non-albicans Candida speciesversus Candida albicans in the intensive care unit.Clin Infect Dis 2008; 46:1206−1213Discusses the fate on non-albicans Candida sp infections.Davies SF, Khan M, Sarosi GA. Disseminated histoplasmosisin immunologically suppressed patients.Am J Med 1978; 64:94−100Elaborates on the various immunosuppressive regimens thatappear to predispose to the development of progressive dissemination.Durkin M, Connolly P, Kuberski T. et al. Diagnosis ofcoccidioidomycosis with the use the Coccidiodes antigenenzyme immunoassay. Clin Infect Dis 2008; 47:e69−e73An exciting and promising new development.Eggiman P, Garbino J, Pittet D. Epidemiology of Candidaspecies infections in critically ill non-immunosuppressedpatients. Lancet Infect Dis 2003; 11:685−702A very careful review of the pertinent literature on thesubject.Fish DG, Ampel NM, Galgiani JN, et al. Coccidioidomycosisduring human immunodeficiency virus infection:a review of 77 patients. <strong>Medicine</strong> 1990; 69:384−391The first large article chronicling the impact of coccidioidomycosisas an opportunistic infection in AIDS.Galgiani JN, Ampel NM, Blair JE, et al. Coccidioidomycosis.Clin Infect Dis 2005; 41:1217−1223<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 281ACC-PULM-09-0302-017.indd 2817/10/09 8:29:16 PM


The most recent guidelines for the treatment of coccidioidomycosisby the Infectious Diseases Society of America.Goodwin RA Jr., Loyd JE, DesPrez RM. Histoplasmosisin normal hosts. <strong>Medicine</strong> 1981; 60:231−266This article expands on the many faces of acute histoplasmosis.For clinicians living in an endemic area, this article is a“must read.”Goodwin RA Jr., Owens FT, Snell JD, et al. Chronicpulmonary histoplasmosis. <strong>Medicine</strong> 1976; 55:413−452An excellent description of the chronic pulmonary form ofhistoplasmosis. Points out the seldom-appreciated fact thatthis form of histoplasmosis occurs only in individuals withabnormal lungs, usually smokers.Goodwin RA Jr., Shapiro JL, Thurman GH, et al. Disseminatedhistoplasmosis: clinical and pathologic correlations.<strong>Medicine</strong> 1980; 59:1−33The article elaborates in detail on the many varied clinicalmanifestations of PDH. It further enlarges on the observationthat progression of the initial dissemination occursmainly in abnormal hosts.Green RE, Schlamm HT, Oestmann JW, et al. Imagingfindings in acute invasive pulmonary aspergillosis:clinical significance of the halo sign. Clin Infect Dis2007; 44:373−379Discusses the prevalence and significance of the halo sign.Judson MA. Noninvasive Aspergillus pulmonarydisease. Semin Respir Crit Care Med 2004; 25:203−219An excellent and comprehensive treatise on the subject.Kerkering TM, Duma RD, Shadomy S. The evolutionof pulmonary cryptococcosis. Ann Intern Med 1981;94:611−616This study looks at the fate of pulmonary cryptococcosis inboth compromised and intact hosts. It documents that pulmonarycryptococcal disease in normal hosts is usually aself-limited disease, while in immunocompromised hosts itis only a single manifestation of disseminated cryptococcaldisease.Klein BS, Vergeront JM, Weeks RF, et al. Isolation ofBlastomyces dermatitidis in soil associated with a largeoutbreak of blastomycosis in Wisconsin. N Engl J Med1986; 314:529−534The first successful investigation of a point source epidemicof blastomycosis. The authors succeeded where othersrepeatedly failed: they isolated the fungus from naturein association with the largest outbreak of blastomycosis todate.Lortholary O, Fontanet A, Memain N, et al. Incidenceand risk factors of immune reconstruction inflammatorysyndrome complicating HIV-associated cryptococcosisin France. AIDS 2005; 19:1043−1049The immune reconstitution inflammatory syndrome mayoccur after successful treatment in all the endemic fungalinfections and in cryptococcosis.Maertens J, Theunissen K, Verhoef G, et al. Galactomannanand computed tomography-based preemptiveantifungal therapy in neutropenic patients at high riskfor invasive fungal infections: a prospective feasibilitystudy. Clin Infect Dis 2005; 41:1242−1250Discusses in detail the value of this approach.Marr KA. Fungal infections in hematopoietic stem celltransplant recipients. Med Mycol 2008; 46:293−302An excellent summary of the diseases discussed.Ostrosky-Zeichner L, Alexander BD, Kett DH, et al.multicenter clinical evaluation of the (1-3)beta-D-glucanassay as an aid to diagnosis of fungal infections inhumans. Clin Infect Dis 2005; 41:654−669An other diagnostic test of great promise.Pappas PG, Kauffman CA, Andes D, et al. Clinicalpractice guidelines for the management of candidiasis:2009 update by the Infectious Diseases Society ofAmerica. Clin Infect Dis 2009; 48:503−535Comprehensive recommendations for the treatment of bothalbicans and non-albicans Candida.Regnard JF, Icard P, Nicolosi M, et al. Aspergilloma:a series of 89 surgical cases. Ann Thorac Surg 2000;69:898−903A large contemporary series exploring the surgical optionfor an aspergilloma.Saag MS, Graybill RJ, Larsen RA, et al. Practice guidelinesfor the management of cryptococcal disease InfectiousDiseases Society of America. Clin Infect Dis 2000;20:711−717Although somewhat dated, this is still the official treatmentguideline.Saraceno JL, Phelps DTR, Ferro TJ, et al. Chronic necrotizingpulmonary aspergillosis: approach to management.1997; 131:1435−1441An excellent expose on this uncommon lung infection.Saubolle MA. Laboratory aspects in the diagnosis of coccidioidomycosis.Ann N Y Acad Sci 2007; 1111:301−314A comprehensive reference for the intricacies of diagnosis.Valdivia L, Nix D, Lindberg E, et al. Coccidioidomycosisas a common cause of community-acquired pneumonia.Emerg Infect Dis 2006; 12:958−962An important consideration for practitioners in the endemicarea.Walsh TY, Anaissie EJ, Denning DW, et al. Treatmentof aspergillosis: clinical practice guidelines of theInfectious Diseases Society of America. Clin Infect Dis2008; 46:327−360282 <strong>Pulmonary</strong> Fungal Infections (Sarosi)ACC-PULM-09-0302-017.indd 2827/10/09 8:29:16 PM


The official treatment guide of a very complicated topic.Wheat LJ. Histoplasmosis. Experience during outbreaksin Indianapolis and review of the literature.<strong>Medicine</strong> 1997; 76:339−354A careful description of the largest outbreak ever studied.Wheat LJ, Connolly-Stringfield P, Baker RL, et al. Disseminatedhistoplasmosis in the acquired immunodeficiencysyndrome: clinical findings, diagnosis andtreatment, and review of the literature. <strong>Medicine</strong> 1990;69:361−374An excellent review of PDH complicating AIDS. Since theadvent of the AIDS epidemic, this form of histoplasmosismade its appearance in parts of the country where histoplasmosispreviously was never recognized.Wheat LJ, Freifeld AG, Kleiman MB, et al. Clinical practiceguidelines for the management of patients withhistoplasmosis: 2007 update by the Infectious Societyof America. Clin Infect Dis 2007; 45:807−825This report is the most up-to-date treatment guide.Wheat LJ, Goldman M, Sarosi G. State-of-the-art reviewof pulmonary fungal infections. Semin Respir Infect2002; 17:158−181Description of the epidemiology and ecology of the majorendemic mycoses.Wood KL, Hage CE, Knox KS, et al. Histoplasmosis aftertreatment with anti-tumor necrosis factor-alpha therapy.Am J Respir Crit Care Med 2003; 167:1279−1282Introduction of new agents have led to complicating infectionsby fungi, tuberculosis and other agents.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 283ACC-PULM-09-0302-017.indd 2837/10/09 8:29:16 PM


Notes284 <strong>Pulmonary</strong> Fungal Infections (Sarosi)ACC-PULM-09-0302-017.indd 2847/10/09 8:29:17 PM


Other <strong>Pulmonary</strong> DisordersMark J. Rosen, MD, FCCPObjectives:• Discuss the pathogenesis, clinical features, and treatment ofpulmonary disorders that occur in patients with sickle-cellhemoglobinopathies• <strong>Review</strong> the pathogenesis, clinical features, and treatmentof lung disease in patients with liver disease• <strong>Review</strong> the pathogenesis and clinical features of pulmonaryoxygen toxicity• Discuss pulmonary injury caused by radiation therapy ofmalignancy• Outline the pulmonary disorders that occur after thermalinjury and smoke inhalation• <strong>Review</strong> the causes and effects of methemoglobinemiaKey words: acute chest syndrome; carbon monoxide; hepatopulmonarysyndrome; oxygen toxicity; pulmonary radiationinjury; smoke inhalation<strong>Pulmonary</strong> Disorders in Persons WithSickle-Cell HemoglobinopathiesThe sickle-cell (SC) hemoglobinopathies arecharacterized by a predominance of hemoglobinS. The most common is SC anemia (hemoglobinSS), which affects approximately 1 in 600 AfricanAmericans. Other disorders occur when hemoglobinS is associated with another abnormal hemoglobin.SC disease occurs predominantly in AfricanAmericans in the United States and in Hispanicsfrom the Caribbean, Central America, and SouthAmerica.PathogenesisWhen deoxygenated hemoglobin SS polymerizes,the RBCs assume a rigid configuration andocclude the microvasculature, leading to ischemiaor infarction of tissues. Compared with hemoglobinAA, hemoglobin SS has reduced affinity foroxygen, facilitating polymerization. Altered propertiesof these RBCs lead to hemolysis and vasoocclusiveepisodes. SS RBCs also interact withthe vascular endothelium, and the increasedadherence of these cells to endothelial surfaces isan important feature of vasoocclusive crises. Inaddition, the release of arginase and free hemoglobinfrom hemolyzing RBCs leads to nitric oxide(NO) dysregulation, endothelial dysfunction, andpulmonary hypertension. SC disease also is associatedwith a hypercoagulable state as the resultof thrombocytosis and the procoagulant effects ofRBC membrane lipids, but the incidence of venousthromboembolism does not appear to be increasedin these patients. The following three overlappingpulmonary syndromes are described in patientswith SC hemoglobinopathies: acute chest syndrome(ACS), fat embolism syndrome, andchronic restrictive lung disease with pulmonaryhypertension.Acute Chest SyndromePatients with SC disease frequently presentwith a syndrome of chest pain, fever, and cough,usually shortly after the onset of a painful crisis.In adults, the radiograph typically shows multilobeor lower lobe pulmonary opacities; pleural effusionis present in approximately 15% of cases. Whensevere, ACS resembles ARDS. ACS is the leadingcause of death in patients with SC disease, mostcommonly because of respiratory failure or vasoocclusiveneurologic complications. The pathogenesisof ACS often is multifactorial, and it is usuallyimpossible to demonstrate one specific cause foran episode. Disorders leading to ACS include infection,fat embolism from infarcted bone marrow,pulmonary microvascular occlusion, and in situthrombosis. The most common pathogens identifiedin ACS are Chlamydia pneumoniae, Mycoplasmapneumoniae, and respiratory viruses. Other bacteriaare less common causes of ACS but include Streptococcuspneumoniae, Haemophilus influenzae, andStaphylococcus aureus. Thoracic bone infarctionoccurs commonly during an acute painful crisisand may lead to atelectasis and pneumonia bycausing splinting from pain.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 285ACC-PULM-09-0302-018.indd 2857/11/09 1:40:23 AM


Because the pathogenesis of ACS is multifactorial,each treatment is of uncertain benefit in aspecific patient. Patients almost always are treatedwith analgesics, are hydrated to prevent hemoconcentration,and are given oxygen supplementationto reduce sickling. Most patients receive antibioticsempirically because it usually is impossible todetermine whether or not they have pneumonia.Antibiotics should be directed against both “typical”and “atypical” pathogens. In severe cases, thetransfusion of RBCs may improve ACS rapidly.Incentive spirometry was shown to prevent atelectasisand pulmonary infiltrates in patients withSC disease who were hospitalized with acute chestpain. The use of inhaled NO to inhibit microvascularadhesion of RBCs appears promising butcannot be recommended until large clinical trialsdetermine its efficacy, safety, and proper dose.Fat Embolization SyndromeBesides being a common cause of ACS, fatembolization may cause a systemic disorder withneurologic changes, renal failure, petechiae, andmultilobe opacities or ARDS. The syndrome usuallyoccurs during an active painful crisis; the circulationof free fatty acids from necrotic bonemarrow is implicated in the pathogenesis of diffusevascular injury that may lead to ARDS and multiorganfailure. Fat globules may be detected insputum and urine, and bone scan findings usuallyare positive. The systemic fat embolization syndromeappears to be especially common in patientswith hemoglobin SC and in the postpartum periodof pregnant women. Treatment is supportive.Chronic Lung Disease and <strong>Pulmonary</strong>HypertensionProgressive, irreversible, and profound lungdisease may develop in patients with SC disease,especially in those with recurrent ACS. Undoubtedly,recurrent fat embolization, infarction, andinfection all contribute to this syndrome. As thedisease progresses, increasing hypoxemia, radiographicevidence of pulmonary fibrosis, and apredominantly restrictive functional impairmentdevelop in patients.<strong>Pulmonary</strong> hypertension (PH) is common inpatients with SC disease. Mild PH, defined as apulmonary arterial systolic pressure 35 mm Hg,occurs in approximately 20% of adult patients andsevere disease (pulmonary arterial systolic pressure 45 mm Hg) in 10%. The presence of PH is also isassociated with a high risk of death. Hemolysis andNO dysregulation play central roles in the developmentof PH. NO is a soluble gas molecule that isproduced by the conversion of l-arginine to citrullineand mediates pulmonary vascular tone throughthe relaxation of smooth muscle and reduction ofthe expression of endothelin-1 and endothelinreceptor. NO also reduces the endothelial expressionof adhesion molecules, decreases plateletactivation, and inactivates reactive oxygen species.Hemolysis releases arginase from RBCs, reducingthe plasma availability of arginine for NO synthesis;free hemoglobin itself is also an NO scavenger.Increasing severity of hemolysis, as determinedby serum free hemoglobin, lactate dehydrogenase,indirect bilirubin, and blood reticulocyte count,correlates with the increasing risk of developingPH. PH is most often diagnosed by echocardiographicestimation of systolic pressure derivedfrom measurements of the velocity of the tricuspidvalvular regurgitant jet. In general, patients withPH should undergo cardiac catheterization both toconfirm the diagnosis and to evaluate left ventricularpressures. New therapies under investigationinclude arginine supplementation and drugsused in the treatment of idiopathic pulmonaryarterial hypertension (IPAH).Liver-Lung SyndromesBecause venous blood flows from the liver tothe lungs, these organs are in some ways interdependent.PH may cause hepatic congestion andascites. Conversely, decreased hepatic synthetic andmetabolic function may lead to deranged productionand metabolism of vasoactive substances andother mediators that profoundly affect the lung.The major pulmonary complications of liver diseaseare the so-called hepatopulmonary syndrome (HPS),portopulmonary hypertension, α 1-antitrypsin deficiency,and hepatic hydrothorax.HPSLiver disease, hypoxemia, and pulmonaryvascular abnormalities referred to as intrapulmonary286 Other <strong>Pulmonary</strong> Disorders (Rosen)ACC-PULM-09-0302-018.indd 2867/11/09 1:40:24 AM


vascular dilations constitute the hallmark of HPS.Hypoxemia is caused by precapillary and capillarydilation and pulmonary and pleural arteriovenousanastomoses. The most common presenting symptomis dyspnea, but the presence of stigmata ofchronic liver disease (spider nevi, clubbing) andsevere hypoxemia strongly suggest the diagnosis.Patients may note platypnea (ie, shortness of breathwhile upright) because of increased shunting ofpulmonary blood to the systemic circulationthrough spider nevi in the lung bases. Arteriovenousshunts are the most important determinantof hypoxemia in patients with severe HPS; thesepatients may not respond well to the administrationof supplemental oxygen. Vascular dilationsprobably cause hypoxemia by diffusion-perfusionimpairment, in which oxygen may not diffuse allthe way to the center of a dilated pulmonary capillary.Serum that streams through the center of thatcapillary is therefore unoxygenated. Intrapulmonaryvascular dilations may be caused by dysregulatedvasoactive mediators leaving the liverand entering the lungs, causing remodeling of thepulmonary vessels. NO is also implicated in thepathogenesis of this disorder. The lungs of patientswith cirrhosis generate excessive NO, which in turnleads to vasodilation.An estimated 15 to 20% of patients with cirrhosishave HPS. It is diagnosed by use of the followingcriteria: portal hypertension (with orwithout cirrhosis); arterial hypoxemia (alveolararterialoxygen difference, 15 mm Hg); andpulmonary vascular dilation demonstrated by theechocardiographic appearance of microbubbles inthe left atrium within three to six cycles after theinjection of hand-agitated circulation into a peripheralvein or abnormal brain uptake seen on radionuclidescanning. The most effective treatment forHPS is liver transplantation, which may lead toresolution of hypoxemia after several months.Long-term oxygen therapy is recommendedfor patients with HPS, but its usefulness may belimited in the presence of significant shunts. Pharmacologicapproaches have been disappointing,and liver transplantation is the only known effectivetreatment. However, HPS is associated withincreased postoperative mortality. A Pao 2 60 mmHg is considered to be an indication for transplantation,and patients with this disorder are givenhigh priority on waiting lists.Portopulmonary HypertensionUp to 20% of patients with cirrhosis and portalhypertension have modestly increased pulmonaryarterial systolic pressures ( 25 mm Hg at rest or 30 mm Hg with exercise) that are caused byincreased cardiac output despite reducedpulmonary vascular resistance. However, portopulmonaryhypertension, with pulmonaryvasoconstriction and pathologic changes indistinguishablefrom IPAH, develops in 1 to 2% ofpatients with cirrhosis. Criteria for this diagnosisinclude the presence of liver disease and portalhypertension, pulmonary artery systolic pressure 25 mm Hg, mean pulmonary arterial occlusionpressure 15 mm Hg, and pulmonary vascularresistance 250 dynes ⋅ s ⋅ cm -5 . The diagnostic andtreatment strategies are similar to those for IPAH,but less well studied, although hepatotoxicity ofsome agents (such as bosentan) are of special concern.Unfortunately, and for unknown reasons, theresponse of patients with portopulmonary hypertensionto liver transplantation is unpredictable,and many centers consider severe portopulmonaryhypertension to be a contraindication for transplantation.α 1-Antitrypsin DeficiencyThe pulmonary manifestations of this geneticdefect are actually the result of a primary hepaticabnormality. With a gene mutation, abnormal α 1-protein levels accumulate in hepatocytes and arenot released from the liver. The resulting low circulatingconcentration of this protective proteaseinhibitor leads to excessive neutrophil elastaseactivity in the lungs and the destruction of pulmonaryelastic tissue. The section on COPD discussesthis disorder in detail.Hepatic HydrothoraxSome patients have congenital anatomicdefects in the diaphragm. If ascites develops, thepositive infradiaphragmatic pressure moves fluidthrough these defects into the negative-pressurepleural space, causing the accumulation of pleuralfluid with the same characteristics as the ascites.This condition is referred to as hepatic hydrothorax.Pleural fluid is typically a transudate and occurs<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 287ACC-PULM-09-0302-018.indd 2877/11/09 1:40:24 AM


more often on the right side than the left. Spontaneousbacterial empyema may occur with orwithout spontaneous bacterial peritonitis. Theinitial treatment is directed at reducing the volumeof ascites by the restriction if salt and theadministration of diuretics. Thoracentesis may behelpful to relieve dyspnea and hypoxemia acutely;however, the fluid usually reaccumulates. Chesttube drainage is often hazardous because asciticfluid translocates rapidly into the pleural space,leading to volume depletion and electrolyteabnormalities. Pleurodesis is generally unsuccessfulbecause the fluid usually reaccumulates toorapidly for the pleural surfaces to come togetherand adhere. Surgical repair of diaphragmaticdefects by videothoracoscopy can be attempted,but few centers are experienced in performingthe procedure, and the results often are disappointing.Peritoneovenous shunts usually are not successfulin patients with hepatic hydrothoraxbecause the pleural space has a lower pressure thanthe venous system and because fluid moves preferentiallyto the pleural space. The transjugularintrahepatic portosystemic shunt is an option inpatients who have recurrent hydrothorax despitereceiving diuretic therapy and undergoing repeatedthoracenteses. The best treatment for refractoryhepatic hydrothorax is liver transplantation.Oxygen ToxicityOxygen is necessary for life, but it also has thepotential to be toxic through the generation of freeradicalintermediates. A balance of oxidative andantioxidant processes maintains life while maintainingorgan function. Our knowledge of oxygentoxicity is largely inferred from experimental dataobtained from other mammalian species and fromclinical observations.Mechanisms of Oxygen-Induced InjuryThe stepwise reduction of oxygen to waterproduces free radicals that in turn are injurious totissues. These include the superoxide anion (-O 2−),hydrogen peroxide (H 2O 2), and hydroxyl radicalspecies (-OH − ). The latter is among the most reactivebiological molecules known and is capable ofdamaging almost any component of living cells.Oxygen radicals have a beneficial role because theyare used by neutrophils during phagocytosis andthe killing of bacteria and may be importantmediators of vascular tone by interacting with NO.However, unopposed hyperoxia promotes theexcessive production of these molecules (termedoxidative stress), leading to damage to cellular andlysosomal membranes and in turn increasing cellpermeability, releasing lysosomal proteolyticenzymes, inactivating cellular enzymes, and damagingDNA. Thus, many cells that are exposed tosufficient oxidative stress stop dividing and die,some by entering apoptotic pathways and committing“suicide,” others by oxidation-induced necrosis.This may be the case with ischemia-reperfusioninjury in patients with myocardial infarction andstroke. The recruitment of neutrophils augmentstissue injury by releasing radicals when activated.Oxidant injury is opposed by antioxidantmechanisms. Intracellular enzymes that reduceoxidant activity include superoxide dismutase(eliminates superoxide anion), catalase (metabolizesH 2O 2), and glutathione peroxidases (reduceslevels of H 2O 2and lipid peroxides). Vitamin Einhibits lipid peroxidation in membranes, anddeficiencies of vitamin E result in the increasedsusceptibility of cells to oxidant injury.The lung and capillary endothelium are theonly tissues that can be exposed to high concentrationsof oxygen because other tissues consumemore oxygen than is dissolved in the blood. Thus,lung cells may be exposed to a Po 2of 600 mmHg in a pure oxygen atmosphere under normobaricconditions, but other tissues are rarely exposed tooxygen tensions that are 10 mm Hg above normalbecause dissolved O 2accounts for 10% oftotal O 2content.Other factors may influence the degree ofinjury on exposure to O 2. Previous exposurestimulates the production of antioxidant enzymes,conferring protection at the next exposure. Inexperimental models, endotoxin also induces antioxidantenzyme activity and increases tolerance tohigh concentrations of O 2. Conversely, drugs andtoxins enhance lung injury caused by oxygen.Hyperoxia and bleomycin exert synergistic toxiceffects on the lung, probably because bleomycinincreases the generation of free O 2radicals. Amiodarone,nitrofurantoin, mitomycin, and paraquat288 Other <strong>Pulmonary</strong> Disorders (Rosen)ACC-PULM-09-0302-018.indd 2887/11/09 1:40:24 AM


also potentiate hyperoxic injury. NO combines withO 2−to form the peroxynitrite anion ONOO, whichis potentially extremely toxic.Clinical Manifestations of <strong>Pulmonary</strong> OxygenToxicityPatients may notice retrosternal pain afterinhaling pure oxygen for a few hours, and tracheobronchitisis found in biopsy specimens. Mucociliaryvelocity is impaired, possibly as the resultof damage to ciliated airway epithelium. It is difficultto diagnose parenchymal oxygen toxicity inhumans because the clinical picture is alwaysobscured by the disease that prompted the needfor oxygen therapy. In experimental models, damageto type I pneumocytes increases permeabilityand leads to the recruitment of inflammatory cellsinto capillaries and alveolar septae. The alveolarepithelium may undergo extensive necrosis withhyaline membrane formation and be replaced withproliferating type II cells. With ongoing injury,fibrosis ensues. These changes are morphologicallyindistinguishable from ARDS as the result of othercauses.What Is a “Safe” Concentration of Oxygen?There is no treatment for oxygen-induced lunginjury; therefore, the best strategy should be toprevent it by minimizing exposure. However, nosafe threshold has been established in humans. Thequestion is in some ways misguided because anyconcentration is potentially harmful in the absenceof adequate antioxidant defenses. In studies ofexperimental lung injury, even 50% O 2enhanceslung damage, and that concentration is the thresholdat which the replication of human pulmonaryepithelial cells is inhibited. In patients who receivedbleomycin, any supplemental O 2is deemed tobe potentially hazardous. For general clinical management,most clinicians assume that a fractionof inspired oxygen of 0.5 is safe, even for prolongedperiods, but this has not been establishedconvincingly.<strong>Pulmonary</strong> Radiation InjuryRadiation is intentionally given to the lung inthe treatment of lung cancer and cannot be avoidedin radiation therapy of cancers of the esophagusand breast. The lungs are exposed in mantle radiationof the mediastinum for lymphoma, and thelungs are the dose-limiting tissue for whole-bodyirradiation in preparation for bone marrow transplantation.Killing tumor cells is accompanied bythe destruction of surrounding normal lung tissue,leading to the following two clinical syndromes:radiation pneumonitis and radiation fibrosis.Although many patients who receive radiotherapyhave radiographic changes, symptoms developonly in approximately 5%. With advances in combinedchemotherapy and radiation therapy forcancers, this disorder should become even lesscommon.PathogenesisRadiation causes the death of cells in mitosis.Type II pneumocytes and capillary endothelial cellshave the greatest turnover rates and are, therefore,the most susceptible to radiation injury. Histopathologicchanges include edema and increasedpermeability of small vessels and capillaries, followedby the obstruction by platelets, fibrin, andcollagen. Damage to type II pneumocytes leads tothe formation of hyaline membranes, followed byhyperplasia and interstitial fibrosis. Tissue injuryalso causes a cytokine response that recruitsimmune cells to perpetuate the inflammatoryresponse and subsequent reparative fibrosingprocess.Predisposing FactorsThe magnitude of lung injury depends on severalfactors:1. Irradiated volume of lung tissue: The risk ofpneumonitis increases with the volume ofexposed lung.2. Total dose: The total dose of radiation predictsthe development of radiation fibrosis better thanpneumonitis. There is a steep dose−responserelationship; 30 Gy is usually well tolerated, 40 Gy almost always causes radiographicchanges, and 50 Gy almost always causessymptomatic lung injury.3. Fraction size: Dividing the total dose over timereduces the risk of lung injury.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 289ACC-PULM-09-0302-018.indd 2897/11/09 1:40:25 AM


4. Previous irradiation increases the risk.5. Chemotherapy may potentiate radiation damage.The classic example is bleomycin.6. Withdrawal of therapy with corticosteroidsmay lead to overt pneumonitis when subclinicallung injury becomes overt.Radiation PneumonitisAfter a latent period of up to 6 months, dyspnea(the most common symptom), dry cough,pleuritic chest pain, and fever may develop. Thephysical examination findings usually are normal.A progression of chest radiographic findings istypical, manifesting as ground-glass opacities thatmay progress to consolidation, coalescing to forma sharply demarcated opacity corresponding to thefield of radiation. Radiographic changes may occuroutside the radiation field, perhaps attributable toa lymphocyte-mediated hypersensitivity reaction.Pleural effusions are observed occasionally.Most patients with radiation pneumonitis havemild symptoms and require no therapy. Althoughcontrolled studies in humans are lacking, animaldata and clinical experience indicate that therapywith corticosteroids is an effective treatment inpatients with more severe disease. The optimaldose is unknown; prednisone, approximately60 mg/d, is recommended, followed by a graduallytapered dose. The disease may flare during a steroidtaper. There is no effective treatment forradiation-induced fibrosis.Radiation FibrosisWhen lung injury is chronic, fibrosis may occur.This fibrosis usually is not apparent until at least6 months after exposure, may take up to 2 years toevolve, and generally remains stable thereafter. Thesymptoms depend on the extent and severity offibrosis, ranging from no symptoms to severe dyspnea,hypoxemia, and death. Mediastinal fibrosismay cause superior vena cava syndrome or constrictivepericarditis, and exposure of the coronaryarteries to the beam may cause obstruction. Severefibrosis may occur in the absence of clinicallyapparent pneumonitis. The chest radiographtypically shows volume loss in the affected areas.Bronchiectasis and pleural thickening are commonfindings on CT scans. At times, confluent fibrosismay be difficult to distinguish from the recurrenceof a tumor.Smoke InhalationSmoke generated in a fire is a suspension ofparticles and toxic gases in hot air. Fire victimsmay experience a variety of inhalation injuries,including thermal injury to the airways and theeffects of inhaling carbon monoxide (CO) and otherproducts of combustion.Thermal and Toxic InjuryThe same high temperatures that burn the skinmay injure the upper airways, and inhalationinjury is an important predictor of mortality inburn victims. The peripheral airways and alveoliare not burned unless steam is inhaled because theheat is dissipated centrally. Along with hot air, sootand toxic gases also are inhaled. Upper airwayobstruction caused by laryngeal edema or spasmmay be severe, especially in patients with facialburns. Smoke inhalation is associated with bronchospasm,impaired mucociliary function, mucushypersecretion, inflammation, and edema. Theexpectoration of soot with sputum reflects smokeinhalation but not necessarily airway injury.The chest radiograph findings usually are normalat the time of presentation. Therefore, fiberopticlaryngoscopy and bronchoscopy should beperformed to assess the degree and extent of airwaydamage in patients with suspected airwayinjury. Even if the airway is not grossly compromisedat the time of the initial examination, inflammationand edema usually progress during thefirst 24 to 48 h. The presence of edema andblistering should prompt intubation, but if theairway appears uninjured, then the patient shouldbe observed closely. In the absence of a significantairway burn, intubated patients can usually beextubated safely after a few days, when the edemasubsides. Corticosteroid administration does notattenuate the course of pulmonary injury andincreases the risk of infection and death.With improvements in burn wound management,pneumonia is now the most common causeof infection in burn units and is a very commoncause of death. Pathogens associated with nosocomialpneumonia, especially Pseudomonas aeruginosa,290 Other <strong>Pulmonary</strong> Disorders (Rosen)ACC-PULM-09-0302-018.indd 2907/11/09 1:40:25 AM


are also common in burn patients. Herpes simplextracheobronchitis is common, perhaps due to directextension from an oral source in patients who areimmunocompromised due to severe burn injury.CO PoisoningThe inhalation of CO is the most common causeof unintentional poisoning death in the UnitedStates. CO is produced by the incomplete combustionof carbonaceous materials in the presence ofa decrease in ambient oxygen. Fires, automobileexhaust, and unvented coal, kerosene, and woodburningstoves and fireplaces are the most commoncauses of CO poisoning.CO competes with O 2for hemoglobin bindingsites. Its affinity is 200 times greater than that ofO 2, and CO shifts on the oxyhemoglobin bindingcurve to the left. The result is decreased O 2contentof arterial blood and decreased O 2release to thetissues, leading to tissue hypoxia. In addition, CObinds to proteins and increases production of freeradicals that injure tissues and impair mitochondrialfunction and oxygen utilization. Pao 2may benormal, and pulse oximetry overestimates arterialoxygen saturation because most devices do notreliably distinguish oxyhemoglobin from carboxyhemoglobin.However, direct measurement of arterialoxygen saturation and content are reduced. Themajor signs of CO poisoning are CNS and cardiacdisturbances, and patients with severe CO poisoning( 40%) may have severe lactic acidosis.CO measurements may be used to guide treatment,but they do not necessarily correlate withclinical abnormalities, especially when measuredhours after exposure. In asymptomatic persons,levels up to 20% may require no treatment, butbecause normobaric oxygen is inexpensive, easilyavailable, and generally considered to be safe,some authorities recommend its administrationwith a goal of lower CO levels to 5%. The use of100% oxygen reduces the half-life of CO in theblood from around 5 hours to 80 minutes. The roleof hyperbaric oxygen is controversial, but mostexperts recommend therapy with hyperbaricoxygen in severe cases (ie, CO 40%, coma or otherCNS dysfunction, arrhythmia, or cardiac ischemia).Hyperbaric oxygen treatment hastens the resolutionof acute symptoms, and there are some studiesthat support the notion that it protects the patientagainst the development of delayed neuropsychiatricdysfunction, although this issue remains acontentious one.After apparent recovery from the initial intoxication,cognitive and personality changes, Parkinsonism,incontinence, dementia, or psychosisdevelop in approximately 25 to 50% of patientswithin the next few weeks. The delayed neuropsychiatricsyndrome appears not to be directly relatedto cerebral ischemia alone; postischemic reperfusioninjury, the effects of CO on vascular endothelium,and oxygen radical-mediated brain lipidperoxidation also may play a role. The lack of availabilityof hyperbaric chambers, concerns about therisks of transporting critically ill patients to hyperbaricfacilities, and the lack of optimal patientselection criteria and treatment regimens still limitthe use of this therapeutic modality.MethemoglobinemiaAnother effect of oxygen on metabolism comesthrough its interaction with iron; oxidizing agentsconvert the ferrous iron (Fe 2+ ) in hemoglobin to theferric form (Fe 3+ ), creating methemoglobin. Whenoxidizing agents are administered in excess or topatients with deficiency in enzyme systems thatconvert methemoglobin to hemoglobin, toxic levelsof methemoglobin may develop. This moleculehas a much greater affinity for oxygen than oxyhemoglobin,reducing the oxygen content of arterialblood. The presence of methemoglobin alsoshifts the oxyhemoglobin curve to the left. Thus,both the decreased oxygen content of arterial bloodand the increased affinity of oxygen for hemoglobinlead to reduced tissue oxygenation and thesymptoms associated with methemoglobinemia.Cyanosis generally develops in patients at methemoglobinlevels of approximately 15%, symptomsdevelop at levels of approximately 30%, and mentalstatus changes at levels of approximately 50%.Methemoglobin levels 70% usually are fatal.Commonly used drugs are oxidants, whichinclude metoclopramide, chloroquine, dapsone,local anesthetics (eg, benzocaine and lidocaine),nitrates (eg, nitroglycerin, nitroprusside, and NO),and sulfonamides; however, their use may causemethemoglobinemia in susceptible patients. Highlevels of methemoglobin turn blood brown, andthe blood does not turn red when exposed to air.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 291ACC-PULM-09-0302-018.indd 2917/11/09 1:40:26 AM


Pulse oximetry is inaccurate in the diagnosisbecause the peak absorbance of methemoglobin isat 631 nm, whereas pulse oximetry estimates oxygensaturations at wavelengths of 660 and 940 nm.Pulse oximeters may read falsely high oxyhemoglobinlevels, and when methemoglobin levels ofapproximately 30% are detected, the oximetertends to read the oxygen saturation at a value ofapproximately 85%. A clue to the diagnosis in thiscase is the finding of a high Pao 2at the same timethat the pulse oximeter reads 85%. The diagnosisis established using cooximetry, which measuresoxyhemoglobin, deoxyhemoglobin, carboxyhemoglobin,and methemoglobin as percentages of totalhemoglobin.Treatment usually is indicated for patients withmethemoglobin levels of 30%. Methylene blueis a cofactor of nicotinamide adenine dinucleotidephosphate-methemoglobin-reductase and increasesthe capacity of that enzyme to reduce to ferric ironlevels. The usual dose is 1 to 2 mg/kg for over5 min; methylene blue may increase methemoglobinlevels in doses of 15 mg/kg and in patientswith glucose-6-diphosphate deficiency.Annotated BibliographySC <strong>Pulmonary</strong> SyndromesBellet PS, Kalinyak KA, Shukla R, et al. Incentive spirometryto prevent acute pulmonary complications insickle cell diseases. N Engl J Med 1995; 333:699−703Forty percent of patients with SC disease and chest painhad thoracic bone infarction. <strong>Pulmonary</strong> complications(atelectasis and infiltrates) were much less frequent in agroup randomized to be treated with incentive spirometrycompared with those who were not. This finding suggeststhat atelectasis is an important factor in the developmentof ACS.Gladwin MT, Sachdev V, Jison ML, et al. <strong>Pulmonary</strong>hypertension as a risk factor for death in patients withsickle cell disease. N Engl J Med 2004; 350:886−895Describes the incidence and poor prognosis of PH complicatingSC disease and its relationship to the severity ofhemolysis.Gladwin MT, Vichinsky E. <strong>Pulmonary</strong> complicationsof sickle cell disease. N Engl J Med 2008; 359:2254−2265Recent and thorough review of mechanisms and clinicalfeatures of ACS and PH. The authors propose a model ofpulmonary complications in SC disease as driven by eithervaso-occlusion or the effects of intravascular hemolysis onendothelial cell and vascular function.Morris CR, Kato GJ, Poljakovic M. Dysregulated argininemetabolism, hemolysis-associated pulmonaryhypertension and mortality in sickle cell disease. JAMA2005; 294:81−90This investigation supports the hypothesis that hemolysisdrives the development of PH in patients with SC disease byrelease of arginase and hemoglobin from RBCs.Powars DR, Weidman JA, Odon-Maryon T, et al. Sicklecell chronic lung disease: prior morbidity and therisk of pulmonary failure. <strong>Medicine</strong> (Baltimore) 1988;67:66−76Describes a progressive syndrome of irreversible pulmonarydisease that may lead to profound hypoxemia, PH, cor pulmonale,and death. The risk of the development of chroniclung disease increases with the number of episodes of ACSand painful crises.Vichinsky EP, Neumayr LD, Earles AN, et al. Causesand outcomes of the acute chest syndrome in sickle celldisease. N Engl J Med 2000; 342:1855−1865Multicenter study of 671 episodes of ACS in 583 patientsthat demonstrates the importance of fat embolism and pulmonaryinfection with atypical pathogens and shows thatrespiratory failure and neurologic disorders are commoncomplications. Packed RBC transfusions and bronchodilatorsimprove oxygenation.Vichinsky EP, Styles LA, Colangelo LH, et al. Acutechest syndrome in sickle cell disease: Clinical presentationand course. Blood 1997; 89:1787−1792Prospective study of 1,700 episodes of ACS. Adults areoften afebrile and report dyspnea, chills, and severe pain,usually after a vaso-occlusive event. Cases are more commonin winter; respiratory failure generally rapidly develops inpatients with fatal cases of ACS, and one-third of fatal casesare bacteremic.Liver–Lung Syndromesde Campos JRM, Filho LOA, Werebe EC, et al. Thoracoscopyand talc poudrage in the management ofhepatic hydrothorax. Chest 2000; 118:13−17This procedure was effective in approximately 50% of casesand associated with a high 3-month mortality.Deibert P, Allgaier HP, Loesch D, et al. Hepatopulmonarysyndrome in patients with chronic liver disease:role of pulse oximetry. BMC Gastroenterol 2006; 6:15An arterial oxygen saturation of 92% in the supineposition and/or a decrease of 4% after a change to the292 Other <strong>Pulmonary</strong> Disorders (Rosen)ACC-PULM-09-0302-018.indd 2927/11/09 1:40:26 AM


upright position correlated strongly with the presence ofHPS.Garcia N, Mihas AA. Hepatic hydrothorax: Pathophysiology,diagnosis and management. J Clin Gastroenterol2004; 38:52−58Useful review of current concepts.Lazaridis KN, Frank JW, Krowka MJ, et al. Hepatichydrothorax: Pathogenesis, diagnosis and management.Am J Med 1999; 107:262−267A thorough review with useful algorithms for diagnosis andtreatment.Rodriguez-Roisin R, Krowka MJ. Hepatopulmonarysyndrome—a liver-induced lung vascular disorder. NEngl J Med 2008; 358:2378−2387Thorough review of the mechanisms, clinical features, diagnosis,and treatment.Rodriguez-Roisin R, Krowka MJ. Herve P, et al. <strong>Pulmonary</strong>-hepaticvascular disorders. Eur Respir J 2004;24:861−880<strong>Review</strong>s hepatopulmonary syndrome and portopulmonaryhypertension, with recommendations for diagnosis andtreatment by an expert task force.Oxygen ToxicityCrapo JD. Morphologic changes in pulmonary oxygentoxicity. Annu Rev Physiol 1986; 48:721−731<strong>Review</strong> of the histopathology associated with oxygen pulmonaryinjury.Gaston B, Drazen JM, Loscalzo J, et al. The biology ofnitrogen oxides in the airways. Am J Respir Crit CareMed 1994; 149:538−551NO may combine with oxygen radicals to form highly toxiccompounds. Enthusiasm for NO therapy should be tempereduntil its efficacy and safety are established.Heffner JE, Repine JE. State of the art: <strong>Pulmonary</strong> strategiesof antioxidant defense. Am Rev Respir Dis 1989;140:531−554<strong>Review</strong> of antioxidant mechanisms that prevent hyperoxicinjury.Ingrassia TS 3rd, Ryu JH, Trastek VF, et al. Oxygenexacerbatedbleomycin pulmonary toxicity. Mayo ClinProc 1991; 66:173−178Even “acceptable” levels of O 2supplementation potentiatebleomycin lung toxicity.Pryor WA, Houk KN, Foote CS, et al. Free radical biologyand medicine: It’s a gas, man! Am J Physiol RegulIntegr Comp Physiol 2006; 291:R491−R511A thorough review of oxygen and the mechanisms of radicalgeneration and oxidative stress.Repine JE, Bast A, Lankhorst I, et al. Oxidative stressin chronic obstructive pulmonary disease. Am J RespirCrit Care Med 1997; 156:341−357This “state of the art” article develops the hypothesis thatoxidant-antioxidant disturbances are pivotal to the developmentof lung injury in patients with COPD.Radiation InjuryArbetter KR, Prakash UBS, Tazelaar HD, et al. Radiation-inducedpneumonitis in the “nonirradiated” lung.Mayo Clin Proc 1999; 74:27−36Six cases, with excellent radiographs. Patients had histologyof bronchiolitis obliterans organizing pneumonia, suggestingan immunologic mechanism.Choi YW, Munden RF, Erasmus JJ, et al. Effects ofradiation therapy on the lung: radiographic appearancesand differential diagnosis. Radiographics 2004;24: 985−997<strong>Review</strong> of radiation techniques and types of radiationinducedinjury, with excellent images.Roach M 3rd, Gandara DR, You HS, et al. Radiationpneumonitis following combined modality therapyfor lung cancer: Analysis of prognostic factors. J ClinOncol 1995; 13:2606−2612Results of a retrospective analysis of almost 2,000 patientsfrom 24 series. The daily fraction size, number of daily fractions,and total dose were associated with the risk of radiationpneumonitis.Roberts CM, Foulcher E, Zaunders JJ, et al. Radiationpneumonitis: a possible lymphocyte-mediated hypersensitivityreaction. Ann Intern Med 1993; 118:696−700After unilateral radiotherapy for breast carcinoma, patientswith pneumonitis had similar increased BAL fluid lymphocytecounts and gallium uptake in both lungs. Treatmentof a hypersensitivity reaction may account for theeffectiveness of therapy with corticosteroids in radiationpneumonitis.Smoke Inhalation and COCobb N, Etzel RA. Unintentional carbon monoxiderelateddeaths in the United States, 1979 through 1988.JAMA 1991; 266:659−663Improvements in the insulation of homes often lead to poorventilation. Coupled with the use of coal, kerosene, and woodin stoves and fireplaces, the risk of CO poisoning increases.Ernst A, Zibrak JD. Carbon monoxide poisoning.N Engl J Med 1998; 39:1603−1608Excellent, concise review.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 293ACC-PULM-09-0302-018.indd 2937/11/09 1:40:26 AM


Hampson NB. Pulse oximetry in severe carbon monoxidepoisoning. Chest 1998; 114:1036−1041Pulse oximetry overestimates oxygen saturation in CO poisoning.Hantson P, Butera R, Clemessy J, et al. Early complicationsand value of initial and paraclinical observationsin victims of smoke inhalation without burns. Chest1997; 111:671−675Emphasizes the value of laryngoscopy in patients with smokeinhalation.Masanes MJ, Legendre C, Lioret N, et al. Fiberopticbronchoscopy for the early diagnosis of subglottalinhalation injury: comparative value in the assessmentof prognosis. J Trauma 1994; 36:59−67Bronchoscopy was more reliable in the immediate diagnosisof inhalational airway disorders than clinical, laboratory,and radiographic findings. It was also the strongest predictorof the development of ARDS and death.Ryan CM, Schoenfeld DA, Thorpe WP, et al. Objectiveestimates of the probability of death from burn injuries.N Engl J Med 1998; 338:362−366In a retrospective review of 1,665 patients with acute burninjuries, three risk factors for death were identified: age 60years, 40% of body surface area burned, and inhalationinjury. When all three risk factors are present, the mortalityrate was 90%.Shirani KZ, Pruitt BA, Mason AD. The influence ofinhalational injury and pneumonia on burn mortality.Ann Surg 1987; 205:82−87In a US Army Institute of Surgical Research study, pneumoniawas present in 55% of burn patients who died andwas a major predictor of excess mortality.Weaver LK. Carbon monoxide poisoning. N Engl JMed 2009; 360:1217−1225<strong>Review</strong> of mechanisms, clinical features, and treatment, withhelpful illustrations.Weaver LK, Hopkins RO, Chan KJ, et al. Hyperbaricoxygen for acute carbon monoxide poisoning. N Engl JMed 2002; 347:1057−1067Three hyperbaric oxygen treatments reduced the incidenceof the delayed emergence of cognitive dysfunction by almost50% compared with a control group. Optimal patient selectioncriteria and treatment regimens need more investigation.MethemoglobinemiaKarim A, Ahmed S, Siddiqui R, et al. Methemoglobinemiacomplicating topical lidocaine used during endoscopicprocedures. Am J Med 2001; 111:150−153König MW, Dolinski SY. A 74-year-old womanwith desaturation following surgery. Chest 2003;123:613−616Novaro GM, Aronow HD, Militello MA, et al. Benzocaine-inducedmethemoglobinemia: experience from ahigh-volume transesophageal echocardiography laboratory.J Am Soc Echocardiogr 2003; 16:170−175These articles discuss the mechanisms of methemoglobinemiaand common clinical situations in which they mayoccur.294 Other <strong>Pulmonary</strong> Disorders (Rosen)ACC-PULM-09-0302-018.indd 2947/11/09 1:40:27 AM


Medical Statistics/Test-Taking StrategiesDavid W. Kamp, MD, FCCPObjectives:• Understand and analyze how well a clinical test “performs”in diagnosis• Comprehend the “how” and “why” one test is better thananother• <strong>Review</strong> how we should use test results to help make medicaldecisions• Summarize recommendations for successful test-takingstrategiesKey words: medical statistics; probability; receiver operatorcurvesIntroductionA firm understanding of medical statistics isvitally important for interpreting the medical literature,yet it often is perceived as a mysteriousscience. This was nicely captured in Mark Twain’sinfamous remark regarding statistics: “There arelies; damn lies and statistics.” Because many physicianslack the necessary skills in statisticalanalysis to adequately assess the quality and conclusionsof studies, medical decisions may beimpaired. 1 For example, only 18% of medical students,residents, and attending physicians at severalteaching hospitals correctly answered a simplehypothetical problem that involved decisionanalysis. Berwick et al 2 administered a simplestatistics test to medical students, residents, privateattending physicians, and academic attendingphysicians and found that the percentages of correctanswers were 73%, 70%, 55%, and 74%, respectively.A more recent study 3 found that 59% of277 medical residents had trouble performingbasic numerical tasks that translated into difficultyinterpreting medical literature. Thus, physiciansrequire a better understanding of the fundamentalsof medical statistics. There are many detailedresources, including Web-based materials, whichare available to enhance the understanding ofmedical statistics. 4,5 This chapter reviews the basicprinciples underlying medical statistics that areclinically relevant as well as important for takingpulmonary boards.An important concept about statistics is that itdoes not define “truth” but rather, when used correctly,it allows a sense of what the truth is basedon (ie, the data and assumptions that have beenused). Every statistical test has important assumptionsunderlying its validity. Although some statisticaltests are fairly rigorous (ie, insensitive to theviolation of assumptions), the results will be questionableif an incorrect statistical test is used. Studydesign is the most important aspect of the scienceof statistics. Minor study flaws (eg, randomization,use of appropriate controls, and double blinding)will limit confidence in the results that have beenanalyzed, even those that use the proper statisticaltests.For example, an analysis 6 of 250 controlled trialsfound that studies with inadequate concealmenthad a 41% greater treatment effect comparedwith those in trials with adequately concealedtreatment allocation and that studies that were notdouble-blinded had a 17% greater treatment effectcompared with those in double-blinded studies. Inan analysis of 32 studies, van Nieuwenhoven et al 7found an inverse relationship between the qualityof the methods and the benefit of the selectivedecontamination of the digestive tract on the incidenceof ventilator-associated pneumonia. Thus,studies with the most appropriate methodologicdesign demonstrate the least benefit of selectivedecontamination of the gut on the incidence ofventilator-associated pneumonia and, as such,strongly argue against such a policy.Another important example is illustrated instudies published in the 1980s and 1990s suggestingthat there is a pathologic supply dependenceof oxygen utilization in which, unlike the normalsituation, oxygen uptake (V . o 2) increased as oxygendelivery (Do 2) increased. This finding led toa decade in which critical care physicians regularlyused hemodynamic monitoring in patients<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 295ACC-PULM-09-0302-019.indd 2957/10/09 8:35:10 PM


with sepsis in an attempt to increase Do 2to supranormallevels so as to improve V . o 2. However,these conclusions turned out to be incorrect basedon an artifact derived from improper methodologyin which Do 2and V . o 2were each calculatedfrom measurements containing a common variable,cardiac output (CO) (ie, Do 2 CO Cao 2;V . o 2 CO(Cao 2 CV . o 2), where Cao 2is the oxygencontent or the arterial blood whereas CV . o 2is theoxygen content or the venous blood.Basic TerminologyThe purpose of medical statistics is to understandthe “true” effect in the general populationon the basis of the results of a single experiment ina limited population sample (eg, the effect on FEV 1of methacholine). The basic assumption of all studiesis that the underlying truth can be “estimated”from a single, properly designed experiment.Because we cannot possibly measure the FEV 1response to methacholine in the entire population,we must use statistics. Expressed another way,there would be no need for statistics if we couldmeasure the end point of interest in the entirepopulation. Because statistics must be used toevaluate study populations of interest, manyimportant issues regarding study samples (eg, sizeand selection/exclusion criteria) must be carefullyconsidered in determining how relevant the dataare to the general population.Any measured parameter has some distributionwithin a study population (eg, Gaussian, uniform,skewed right, skewed left, bimodal, u-shaped,exponential, and others). A Gaussian distributionis defined as a symmetric, unimodal, bell-shapeddistribution with a size and shape in patients withand without disease that may be completely distinctor may nearly totally overlap (Fig 1). Thereare statistical tests for assessing whether a sampledvariable has been drawn from a population withan underlying Gaussian distribution. Occasionally,a non-Gaussian distribution can be transformed tobe Gaussian (ie, a logarithmic transformation ofdata skewed to the right, such as length of stay inthe ICU), which facilitates parametric statisticaltests on the data. The assumption that data have aparticular distribution is important and prone toerror, but when it is true, it greatly simplifies statisticaltesting.Figure 1. Defining “abnormal” values of a test.The primary aim of most statistical analyses isthe comparison of the central tendency betweendifferent groups. To do this, there are some fundamentalstatistical terms that physicians mustunderstand, including the following:• Variable: defined as the things that are measuredor manipulated in research. Most researchbelongs to one of two general categories, correlationalor experimental. In correlationalresearch, the variables are not manipulated butare only measured to determine relations (correlations)between some set of variables (eg,FEV 1and age). Correlational research does notenable causal relationships to be established.In experimental research, some variables aremanipulated, and then the effects of this manipulationon other variables are assessed (eg,studies examining the effects of varying dosesof methacholine on FEV 1). Experimental dataare the only conclusive way to demonstratecausal relations between variables. For example,we can infer that “A influences B” if alterationsin variable A result in predictable changesin variable B. Independent variables are thoseitems that are manipulated in the experimentalresearch, whereas dependent variables arethose items that are only measured and arenever manipulated experimentally.• Mean: defined as the average (∑X 1)/N, whereX 1is a symbol referring to the N individual valuesof your distribution. Thus, the first value inthe distribution would be X 1, the second wouldbe X 2, and so on until the last value, X N.296 Medical Statistics/Test-Taking Strategies (Kamp)ACC-PULM-09-0302-019.indd 2967/10/09 8:35:10 PM


• Median: defined as the value of X such that 50%of the values are higher and 50% are lower.The advantage of using the median is that itis not as influenced by data outliers. For symmetricdistributions (including Gaussian),mean equals median. For distributions that areskewed to the high side (eg, length of stay in theICU) the mean is greater than the median (eg, afew patients with a very long length of stay willsubstantially move the mean upward but willhave very little if any impact on the median,which gives us a more accurate indication ofthe length of stay for most patients).• Percentile: defined such that of all the valuesunder consideration, Z% are less than that, sothat (100 – Z)% are greater than that. For example,the fact that your 2006 pulmonary boardscores are in the 95th percentile indicates that95% of all other scores are worse than yourscores. In contrast, an extramural grant scoreof 150 that is in the 5th percentile indicatesthat only 5% of grants have lower scores thatare better (similar to golf, in which low scoresare better in grading grants). The median is the50th percentile, whereas we often also look atthe <strong>25th</strong> and 75th percentiles.• Variance: defined as a measure of the “spread” ofthe data distribution. It is determined numericallyby (∑(X i X ) 2 / (N 1), in which X is defined as the population mean.• SD: defined as (variance) 1/2 . The SD is a commonlyused measure of data variation. Onlywhen a data set has a Gaussian distributionis it true that 95% of values in the sample fallwith in 2 SDs of the mean. The SEM, whichalso often is used to express data variance, isthe theoretical SD of all sample means of size nthat are derived from a population.• Sensitivity: defined as the fraction of those withthe disease whose test result is positive (ie, greaterthan the cut point). It is also known as the truepositivefraction, which is defined numerically as(TP/ [TP FN]), where TP is the true-positiveresult and FN is the false-negative result (Fig 1). Itis also defined as (1 – FN fraction). For example,if 95 of 100 patients with asthma have a positivemethacholine challenge result, then the sensitivityof the methacholine challenge in detectingpatients with asthma is 95%. A high sensitivitymeans few FN results.• Specificity: defined as the fraction of thosepersons without the disease who have a negativetest result (ie, less than the cut point). It isalso known as the true negative fraction, whichis defined numerically as (TN/[TN FP]) inwhich TN is the true-negative result and FPis the false-positive (Fig 1). It is also definedas (1 – FP fraction). For example, if 25 of100 patients without asthma have a positivemethacholine challenge result, then the specificityof the methacholine challenge is 75%(75/[75 25]). A high specificity means few FPresults.• The null hypothesis: the hypothesis states thatthe means of two data sets in the underlyingpopulation are identical (ie, difference betweenmeans 0). Statistical testing is performed toeither support or refute the null hypothesis.The null hypothesis accepted is the same as sayingthat there is “no statistically significantdifference” between the two groups. The nullhypothesis rejected is the same as saying that ishighly likely that there is a “statistically significantdifference” between the two groups.In summary, when interpreting data sets, it isimportant to identify the central tendency as wellas the spread of the distribution. For Gaussiandistributions, the data typically are expressed asthe mean SD. For most other distributions, thedata are better expressed as the median and interquartilerange (ie, <strong>25th</strong>/50th/75th percentiles). Forvery small samples, the minimum and maximumvalues (range) should be listed. Sensitivity is usedfor ruling out disease, especially as a screeningtool, whereas specificity is used to rule in disease.Some useful mnemonics include the following:“snout” for sensitivity (rules out disease), and“spin” for specificity (rules in disease). For example,the d-dimer test often is used to rule out thepresence of pulmonary embolism (PE) in patientswho have not undergone recent surgery or whohave cancer or liver disease because of its highsensitivity, whereas, when the d-dimer test resultis positive, the PE protocol chest CT scan or, lesscommonly pulmonary angiogram, is used becauseof the high specificity of the test. Finally, bewareof data outliers since these grossly distort distributionsand statistical test results that are based onthe mean.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 297ACC-PULM-09-0302-019.indd 2977/10/09 8:35:11 PM


Hypothesis TestingStatistics often are performed to assess samplesor a set of samples that are taken from the studypopulation to make assumptions about the entirepopulation. For measuring a continuous variable(eg, FEV 1), one should determine whether thepopulation has a Gaussian distribution. Statisticalcomparisons of means within samples is based onthe principle that if you took the mean value of FEV 1in a random sample of N members of the wholepopulation and do this many times (N 50 – 100),then the distribution of sampling means will beGaussian regardless of the distribution in theunderlying population (ie, the central limit theorem).Because we do not typically generate numeroussamples but rather use one such sample foreach arm of the study (eg, the FEV 1response tomethacholine in healthy patients vs patients withasthma) to compare the means of these samples anduse statistical methods to make inferences aboutthe truth in the underlying populations (ie, to identifypatients who are healthy or who haveasthma).Hypothesis testing begins with the nullhypothesis that the means of the two groups in theunderlying population are identical. Statisticaltesting will either refute or confirm this hypothesis.Because these are independent samples from thepopulation, even if the null hypothesis is true, it isnot likely that the two samples will have identicalmeans. Therefore, we use the sampling distributionto tell us how unlikely it is that two samples thatare drawn from the same populations will have adifference in the size observed. This is done bypicking some arbitrary threshold for believing thatthe likelihood that the two samples are from thesame population is sufficiently small that we thinkit is acceptable to conclude that they are probablynot. An arbitrary threshold of 5% (p 0.05) is usuallyused (called the type I error or level error).However, there may be circumstances in whichone accepts as a relevant statistically significantdifference a p value of 0.05 and others in whicha clinically relevant effect may not be acceptedunless the p value were much 0.05 (ie, p 0.01,in which there is an arbitrary threshold of a 1%chance that the samples are identical). A study thatfails to achieve statistical significance does notprove the null hypothesis but rather simply saysthat the data do not provide sufficient evidence toreject it.Student t Test for Comparing MeansThe Student t test is a parametric test that isused to compare a continuous, metric variable (eg,FEV 1) within two groups or the effect of two alternativetreatments in a single group. This testassumes that either the underlying populationdistribution is Gaussian or that N is large enoughthat the distribution of sampling means is Gaussian.An unpaired t test should be used if there aredifferent persons in the two groups (eg, separategroups of normal patients and asthma patients) toexamine FEV 1normal vs FEV 1asthma, withthe null hypothesis being FEV 1normal FEV 1asthma. However, a paired t test shouldbe used if instead the experiment gives placeboand albuterol to the same patients at different times(preferably in random order) to determine whetherFEV 1placebo – FEV 1albuterol is zero (nullhypothesis accepted). The specific question ofinterest determines whether differences betweengroups should be assessed in both directions (twotailedtests) or only in one direction (one-tailedtests).Power Analysis, Type II Errors, and Sample SizeTo avoid erroneously concluding that there isnot a true difference between two sample groupsrequires an adequately sized N value to ensure thatthe two group means are far enough apart tominimize overlap between the groups. An insufficientN value to address this possibility is calleda type II error. If the chance of a type II error is ,then 1 – defines the power of the study to detecta true difference between the groups. For example,if the chance of a type II error is 10%, then thepower of the study to detect a real difference is90%. Similar to type I errors, the threshold set fortype II errors is arbitrary, but typically is 10% or20% (90% or 80% power, respectively). An 80%power indicates that the sample size is largeenough so that there is a 20% probability thatthere really is a difference but random fluctuationprevented detecting statistical significance (eg,p 0.05). This highlights the importance of calculatingsample size in the study design. Sample size298 Medical Statistics/Test-Taking Strategies (Kamp)ACC-PULM-09-0302-019.indd 2987/10/09 8:35:12 PM


is calculated by choosing a numerical value for and and by using some reasonable estimatesfrom the available literature about the estimateddifference one expects to find between the meansof the two groups using standard published formulas.Sample size determination should alwaysbe performed before the study and not after thestudy to validate the results.Nonparametric TestsNonparametric testing is used when the distributionof the variable of interest in the underlyingpopulation is non-Gaussian (ie, not bell-shaped).Nonparametric statistical testing uses ranks ratherthan the actual numerical values so that the averageranks rather than the means for the groups arecompared. For example, FEV 1values for the placebogroup are 2.4 L, 2.8 L, and 3.0 L and those forthe methacholine group are 2.1 L, 2.3 L, and 2.7 L.If we put these values in ascending rank order, itwould look as follows: rank 3, 2.4 L; rank 5, 2.8 L;rank 6, 3.0 L; rank 1, 2.1 L; rank 2, 2.3 L; and rank4, 2.7 L. Therefore, the average rank for placebois 14/3 4.6, whereas that for methacholine is7/3 2.3. The data are then analyzed with the testequivalent of t tests. For example, with unpaireddata sets the use of the Wilcoxon rank sum test orMann-Whitney U test would be appropriate,whereas for paired data use of the Wilcoxon signedrank test is appropriate. There are at least tworeasons for not using nonparametric tests when thedata set is Gaussian, including the following: nonparametrictests are a bit more inflexible for detectingdifferences and, thus are more prone to type IIerrors, and the confidence intervals (CIs) cannotbe easily calculated.Multiple ComparisonsMultiple statistical comparisons to a singlecontrol group can lead to spurious differenceswhen none really exist. For comparisons, thechance of finding one spurious “statistically significant”result is 1 – 0.95 (ie, for 10 comparisons,there is a 40.1% chance of getting at least one suchtype I error).There are several approaches to prevent thisproblem, including the following: (1) make anadjustment for multiple comparisons by using aBonferroni correction, which is a very rigorous wayto maintain protection against type I errors and iscalculated by dividing the chosen p value (0.05) bythe number of comparisons, and to use thisadjusted p value to reject any of the various nullhypotheses (ie, if we do 10 pairwise comparisons,the adjusted p value would be calculated as 0.05/10 or 0.005 to be considered statisticallysignificant). (2) Use a smaller p value (eg, 0.01 or0.005). (3) Identify a primary outcome before thestudy if multiple outcomes will be assessed, withthe understanding that the decision to accept orreject the null hypothesis will hinge on this singleoutcome. Although many studies identify a primaryoutcome, secondary outcomes with an unadjustedp 0.05 often are reported, especially whenthe primary outcome is not statistically significant.Because the utility of such an approach is fraughtwith error, ideally the significance of any secondaryend points should be addressed by an appropriatelydesigned study with that new finding as theprimary end point (ie, a second interferon for idiopathicpulmonary fibrosis trial that is underway todetermine whether there is a survival benefit withactive treatment, as suggested in the initial studyas a secondary end point). (4) Combine the multipleend points into a single overall end point.(5) Use analysis of variance (ANOVA) with post-hoctesting, which is one of the most common approaches,as described in the section to follow.ANOVAThe one-way ANOVA is the appropriate statisticalapproach to use when comparing two or moreindependent groups (ie, to assess whether the FEV 1is different among groups that were randomizedto receive five different treatment regimens comparedwith a single control group). ANOVA, whichis subject to the same data distribution assumptionsas is t testing (ie, Gaussian distribution),provides an overall measure of whether there arestatistically significant differences among thegroups overall (p 0.05). Once a true difference isnoted among the groups, there are a number ofdifferent post-hoc tests (eg, Dunnett, Duncan,Scheffé, least significant difference, Tukey, Student-Newman-Keuls, and others) one can perform thatprevent type I errors that might result from performingall the separate pairwise comparisons that<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 299ACC-PULM-09-0302-019.indd 2997/10/09 8:35:12 PM


are possible in the five groups in our example. Fornonparametric data, the Kruskal-Wallis test is aversion of a one-way ANOVA but without post-hoctests. A Bonferroni correction on the pairs of groupsstudied is often used in this situation.Proportions or CountsWhen comparing data that are categoric, ratherthan numerical, the number, fraction, or proportionof patients in each category is analyzed. Similar toa numerical continuous variable analysis, theobserved proportions in the sample are used toapproximate those in the entire population of interest,with inferences about the entire populationmade from the single sample. When the N is large,the distribution approaches a Gaussian curve.These studies use n m tables, with the value ineach cell being either the number or the fraction ofpatients, and then are analyzed with the use of aχ 2 or Fisher exact test. For example, patients in theICU who experienced stress gastritis bleedingwhen they received prophylaxis with sucralfate,cimetidine, omeprazole, or placebo yield a 2 4table such as that shown in Table 1.Confidence Intervals: An Alternative toHypothesis TestingHypothesis testing using p values is limited bysimply addressing the null hypothesis based on ap value threshold with an arbitrary cutoff. A statisticallysignificant difference may be technicallycorrect but of limited clinical significance. Forexample, a large study may show that a bronchodilatorraises the mean FEV 1by 50 mL (p 0.01)compared with placebo, but the magnitude of thatincrease probably will not be likely to generate anybenefit for patients and, as such, is not clinicallyTable 1. Hypothetical Relationship Between Various GI ProphylacticRegimens and Stress Gastritis Bleeding in the ICUType of ProphylaxisDid Bleed,No. (%)Did Not Bleed,No.Sucralfate 8 (6.6) 114Cimetidine 13 (10.8) 107Omeprazole 16 (14.8) 92Placebo 21 (14.9) 120significant. In contrast, a small study that is notstatistically significant may indeed provide a clinicallyrelevant benefit of the new drug. CIs providea way to more easily judge the clinical relevance ofthe findings while reporting the observed effect (ie,the point estimate) in the study along with the 95%CI. The CI provides the following two useful piecesof information: the point estimate, which is thesingle value, based on the study, that is most likelyto represent the true difference between the drugand placebo, and the 95% CI, which is the plausiblerange of differences within which lies the truedifference. No value within this 95% CI can berejected as representing the true effect within a 5%significance level. The null hypothesis is accepted(ie, there is no difference between groups) if the95% CI includes the value representing no effect(ie, 0% difference in FEV 1or 1.0 for an odds ratio[OR; defined below] for death in a study withmortality as the end point), whereas the nullhypothesis is rejected if the 95% CI does not crossthis value.The CI provides information beyond the “noeffect”value. With the use of a hypotheticalexample, the authors of a small study of survivaltime that used a new drug for the treatment ofadvanced small cell lung cancer (median survivaltime with standard therapy, 6 months) found thatthe 95% CI for median survival when using thenew drug was 0.50 to 1.68. Although this rangeincludes 1.00 (ie, it is a negative study because itcrosses the no-effect line), the CI is very useful.The study shows that, at best, we can expect thenew drug to increase the median survival timeby only 4 months (0.68 6 months) and mayreduce survival time to 3 months (0.5 6 months),which questions the utility of any future largerstudy.Relative Risks and ORsPhysicians who read the medical literatureshould be familiar with the fundamentals behindterms such as absolute risk (AR), relative risk (RR),AR reduction (ARR), RR reduction (RRR), and OR.The authors of a study reporting that a new druggave a RRR for death of 25% sounds great, but ifthe risk of death in the control group is 4%, thenthe new drug reduced the death rate by just 1%.Physicians must then decide whether this difference300 Medical Statistics/Test-Taking Strategies (Kamp)ACC-PULM-09-0302-019.indd 3007/10/09 8:35:12 PM


is a clinically relevant one when factoring in otherissues, such as the cost and side effects of the drug.Table 2 shows how AR is calculated from a hypotheticalstudy in which the authors examinedcough in patients who were randomized to receiveenalapril or valsartan.RR (also called the risk ratio) is defined as theratio of the rates of cough occurrence. In Table 2,the risks for cough in the two groups are 44.6% and28.1%, respectively, so the RR 0.281/0.446 0.63,which translates into a 37% lower risk of the developmentof a cough while one is receiving therapywith valsartan compared with therapy with enalapril.Of note, RR cannot be calculated for casecontrolstudies because the case group is chosenbased on having the disease under study whereasthe control subjects are selected because they donot have the disease.ARR (also called risk difference) is defined asthe absolute difference in rates of occurrencebetween the experimental and control patients ina trial. Using the data in Table 2, ARR is defined as44.6% – 28.1% 16.5% ARR for the developmentof a cough while one is receiving therapy withvalsartan.RRR is defined as the percentage difference inrates of occurrence of events (good or bad)between the experimental and control patients ina trial. RRR 1 – RR, which in our hypotheticalTable 2 would be 1 – 0.63 0.37. It is also calculatedas follows: ARR/AR among the appropriatereference group 0.165/0.446 0.37 or 37% RRRfor the development of a cough while one is receivingtherapy with valsartan compared with therapywith enalapril. A major limitation when using theRR and RRR is that they can be misleading. Forexample, if the RRR is 33%, this could result fromthe actual risk decreasing from 3 to 1% or bydecreasing from 66 to 22%. Without knowing theARs for each condition, the RR and RRR can havevery different implications for the efficacy of atreatment or the incidence of a side effect.LR (also called the likelihood ratio) summarizesinformation about the diagnostic test or treatmentresponse by combining data about the sensitivityand specificity. 8,9 The LR informs how much apositive or negative result alters the likelihood thatthe patient would have the disease or treatmentresponse of interest. The LR of a positive test result(LR) is defined as follows: sensitivity / (1 – specificity).The LR of a negative test result (LR–) isdefined as follows: (1 – sensitivity) / specificity. Auseless test has a LR 1.0, whereas a test is betterthe more it is further from 1.0; 1.0 for LR andlower than 1.0 for LR–. The LR can then be used tocompare various test results to ascertain which hasbetter discriminative power. An example of howLRs are calculated is illustrated in Figure 2OR is defined as the ratio of the odds of aparticular end point for each treatment. In ouraforementioned example, the odds of a coughdeveloping while one is receiving therapy withenalapril a/b 29/36 0.81, whereas the oddsof a cough developing while one is receivingtherapy with valsartan c/d 18/46 0.39. TheOR of a cough developing while one is receivingtherapy with enalapril compared with therapywith valsartan would be calculated as 0.81/0.39 2.07. The utility of ORs is that they are the naturalmeasure arising from certain statistical approaches(eg, logistic regression, case-control studies, andmeta-analyses). ORs are not an intuitive measureof risk. For example, the OR of 2.07 in this exampledoes not mean that enalapril-treated patients hadtwice as much chance as valsartan-treated patientsof having a cough develop (that is the RR, whichis 1.59 0.446/0.281). As illustrated in Table 3, theRR approximates the OR only if the event (eg, outcomeor complication) is rare, as may occur withthe effect of an active drug (vs placebo).The number needed to treat (NNT) is definedas the ratio of 1/ARR. The NNT represents thenumber of patients who would need to be treatedwith the better therapy to avoid a single case of theadverse outcome. For example, the ARR in the rateTable 2. Hypothetical Relationship Between Two Different Angiotensin-Converting Enzyme Inhibitors and CoughDrug Cough No Cough Total No. Patients with Cough, % (AR)Enalapril 29 a 36 b a b 65 a/(a b) 44.6%Valsartan 18 c 46 d c d 64 c/(c d) 28.1%<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 301ACC-PULM-09-0302-019.indd 3017/10/09 8:35:12 PM


Table 3. Hypothetical Relationship Between Placebo and a Test Drug in Causing an Adverse EventVariables No. Events, No. Without Event, No. Odds AR RR RRR, % ARR, % ORPlacebo 100 2 98 2/98 2/100 0.5 50 1 0.495Drug 100 1 99 1/99 1/100Placebo 100 90 10 90/10 90/100 0.5 50 45 0.091Drug 100 45 55 45/55 45/100of cough while one is receiving therapy with valsartanover therapy with enalapril is 0.165 (16.5%),and therefore, treating 1/0.165, or 6.1 patients, withvalsartan instead of with enalapril will eliminatea single case of drug-induced cough. The best indexto use in describing the data depends on that whichyou are examining. Typically, the one that mosttransparently reflects the actual difference in eventrates is the ARR or the reciprocal of ARR, the NNT.The RR and RRR obscure the actual change in ratesand may have widely different implications whenthe actual data are assessed. The OR reflects theactual change in rates but in a manner that is notalways intuitive.Medical Decision Making: ReceiverOperator Characteristic CurvesPhysicians must use tests that have some arbitrarycut point between normal and abnormalbased on studies in patients with and without thedisease of interest. Choosing a cut point involvesinevitable trade-offs between sensitivity andspecificity. The usual way this trade-off is examinedis via the receiver operating characteristic(ROC) curve (Fig 3). Note that the ROC curve livesin a box with lengths and widths both equal to 1.0,with sensitivity on the y-axis and 1-specificity(otherwise known as the FP rate) on the x-axis. TheFigure 2. Calculating measures of discrimination of a new serum test (T) to detect mesothelioma (D) in a group of asbestosworkers with a mesothelioma prevalence of 20%.302 Medical Statistics/Test-Taking Strategies (Kamp)ACC-PULM-09-0302-019.indd 3027/10/09 8:35:13 PM


Figure 3. ROC curve.ROC curve analysis provides a global sense of howwell a test distinguishes patients with disease fromdisease-free individuals. This distinction is particularlyhelpful when there are multiple tests thatexist for a given disease to decide which is superior.Inevitably, there is a trade-off between sensitivityand specificity at all possible values of the cutpoint. If a particular test result is no better than aflip of the coin (ie, a useless test), then the graphwill show a diagonal line from the bottom left-handcorner to the upper right-hand corner. On the otherhand, when the distributions of test results betweenthose with and without the disease are widelyseparated (ie, a helpful test with little overlap), thenthe area under the curve is high (ie, approaches1.0). In this instance, the best cut point for a testresult is one that is located in the upper-most lefthandcorner of the graph.To illustrate how ROC curves are used to determinethe best test cut point, the role of d-dimertesting is helpful. Clinicians use d-dimer testing asa strategy to rule out PE and avoid the use ofexpensive testing. For this purpose, the ideald-dimer test cut point requires virtually everyonewith PE to have a positive test result (ie, very highsensitivity and very small FN rate), but we canaccept FP d-dimer test results (ie, low specificity)because we will confirm the presence of PE byimaging. Thus, a low value for the cut point withthe highest specificity while maintaining a verylow FN rate (ie, a d-dimer enzyme-linked immunosorbentassay value of 500 ng/mL) would bechosen based on data from the medical literature.The cut point chosen should maximize “utility” forthe entire population under consideration. Thisrequires assigning some measure of the costs thatoccur as a consequence of both FP and FN results(eg, money, morbidity, and mortality). The optimaloperating point is where the slope of the ROCcurve is defined as:{Cost of FP misclassification} (1 – probability ofdisease) –––––––––––––––– –––––––––––––––––––{Cost of FN misclassification} (Probability ofdisease)Thus, when the costs of FN misclassifications arelarge, you need a point at which the ROC slope issmall (ie, the upper right of the ROC curve wheresensitivity is sure to be high and FN results areminimized; Fig 2). However, when the costs associatedwith FP results are large, you need an operatingpoint at which the ROC curve slope is large (ie,on the left side of the ROC curve where specificityis large and FP results are thus minimized). In arare disease, there is typically a large ROC curveslope (ie, left side of the ROC curve where specificityis large and FP results are minimized). This ispreferred because, in patients with rare diseases,even good test results have more FP results thanTP results, and as a consequence, the cumulativeadverse costs of FP results are magnified.Bayes TheoremSensitivity and specificity alone are inadequateto meaningfully understand the implications of atest result in a real clinical practice. Cliniciansrequire an understanding of Bayes theorem thatinvolves determining the positive predictive value(PPV) and negative predictive value (NPV) of a testby calculating the posttest probability of diseasebased on a pretest probability of disease, sensitivity,and specificity, as follows: (1) the PPV is definedas the percentage of patients with a positive testresult (ie, greater than the cut point) who reallyhave the disease in question and is determined bythe equation (TP)/(TP FP); (2) the NPV is definedas the percentage of patients with a negative testresult who are actually disease-free and is determinedby the equation (TN)/(TN FN); (3) posttestprobability of disease is defined as 100% – NPV(this provides an estimate of the probability of thedisease after calculating the PPV and NPV fromthe sensitivity and specificity); and (4) the likelihoodratio (LR) is defined as the ratio of the<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 303ACC-PULM-09-0302-019.indd 3037/10/09 8:35:14 PM


probability of a test result in patients with the particulardisease in question to the probability of thatsame test in patients who are free of that disease.The LR for a positive test result (ie, LR ) is determinedfrom the ratio of sensitivity/(1 – specificity).The LR of a negative test (ie, LR ) is determinedfrom the ratio of (1 – sensitivity)/specificity. Theposttest odds reflect the odds that the patient hasthe disease of interest after the test is conducted(ie, pretest odds LR).The best way to determine the PPV and NPVof a particular test is to set up a 2 2 table, asillustrated in Table 4. For hypothetical purposes,we will determine the PPV and NPV of d-dimertesting f or PE in which the pretest probability forPE in our patient population is 25%. The PPVreflects the posttest probability of disease after apositive test result. In the hypothetical case, thePPV of a positive d-dimer test result is only 62%(ie, not much better than the flip of a coin) and isof little clinical value in the diagnostic evaluationof PE. However, the value of d-dimer testing is inits NPV, which in our illustrative example is 98%,as calculated in Table 4. Thus, the posttest probabilityof disease after a negative d-dimer test resultwould be 2% (ie, 100 to 98%). The high NPV of d-dimer testing (98% in our hypothetical model) andthe low posttest probability of PE (2%) is veryattractive in excluding PE when the d-dimer testresult is low in this patient population (PE estimate,25%). Accordingly, this allows clinicians to confidentlyfocus on other causes for the patient’s complaints.Notably, however, if the PE diseaseprevalence is greater (ie, 75%), then the NPVdecreases and the posttest probability of PEincrease. For example, when the same sensitivityand specificity is used, the FN test result incidenceis 4 of 24 negative test results; therefore, the NPVdecreases to approximately 83% and the posttestprobability of PE is 17%, which are values thatTable 4. Calculating the PPV and NPV of d-Dimer Testing in a Population With an Estimated PE Prevalence of 25%Steps Disease Present Disease Absent Total1*Positive test result TP FPNegative test result FN TNPE prevalence 25 752 † Positive test result 24Negative test result 1PE prevalence 25 753 ‡ Positive test result 24 15Negative test result 1 60PE prevalence 25 754 § Positive test result 24 15 39Negative test result 1 60 61PE prevalence 25 75 1005 || Positive test result 24 15 39PPV 24/39 62%Negative test result 1 60 61NPV 60/61 98%PE prevalence 25 75 100*Presume a theoretical group of 100 patients with possible PE and fill in the columns.†Sensitivity 95%; therefore, of the 25 patients with disease, 24 have a positive test result, and 1 does not.‡Specificity 20%; therefore, of the 75 patients without disease, 60 have a negative test result, and 15 have positive test results.§Add up across.||Compute the PPV and NPV from the rows.304 Medical Statistics/Test-Taking Strategies (Kamp)ACC-PULM-09-0302-019.indd 3047/10/09 8:35:14 PM


may not be clinically reassuring in stopping theevaluation for PE. Indeed, this is precisely whatlimits the predictive value of d-dimer testingin patients who are deemed to be at very high riskfor PE.It is important to remember that no test is perfect(ie, 100% sensitivity and 100% specificity).Furthermore, even the “gold standard” test can bewrong in a small percentage of cases (ie, pulmonaryangiogram to diagnose PE has both FP results[erroneous reading by a radiologist] and FN results[sampling error or erroneous reading by a radiologist]).Thus, the sensitivity and specificity are insufficientfor informing clinicians about the utility ofa particular test in medical decision making. It isessential to know or to make an estimate about thepretest probability of disease in the patient populationof interest.As illustrated previously in our example ofd-dimer testing in patients with PE, in patientswith a low pretest probability, even a test with highsensitivity will result in a very high percentage ofthe positive test results being false-positive ones,especially if the specificity is not very high. Incontrast, in patients with a high pretest probability(ie, 75%), even a test with high sensitivity willresult in enough FN results that one will be preventedfrom confidently ruling out PE even if thed-dimer test result is negative. Clinicians shoulduse the Bayes theorem to determine the PPV, NPV,and posttest probability of disease to avoid importantassumptions and conclusions about the testresults. Calculation of the PPV and NPV beforethe test is performed, especially if it is expensive,dangerous, or invasive, will help in the assessmentof whether the test result will be of much clinicalbenefit.MetaanalysisMetaanalysis involves a statistical analysis oftwo or more primary data sets that have been publishedor otherwise communicated to address aspecific question or hypothesis. The methods,strengths, and weaknesses of this approach haverecently been reviewed. 10 Compared with a singlestudy, the major potential strength of a properlyperformed metaanalysis is that it enables betterdetermination of whether a particular theoryor finding is correct based on the accumulatedfindings of multiple data sets. The major limitationis that it is unable to correct for underpoweredprimary data sets that introduce sample bias aswell as insensitivity to confounders.In summary, there are some fundamental statisticalprinciples that clinicians must be intimatelyfamiliar with to properly interpret the medicalliterature as well as to make logical medical decisions.In particular, an understanding of how basicstatistical terms are calculated, how hypothesistesting occurs, as well as the concepts behind ROCsin medical decisions are all vitally important. Table5 summarizes the appropriate statistical tests thatare available to analyze data in a wide variety ofsituations, some of which are well beyond thescope of this review. The best statistical test is basedon the nature of the dependent (Y) variable and theindependent (X) variable.Test-Taking StrategiesTest Preparation TipsThere are a number of common sense recommendationsto optimize success on any test. First,confidence in your test taking begins with adequatetest preparation beforehand. The principlesbehind successful test preparation include the following:(1) budget your time to be sure that youhave sufficient time to study so that you are wellprepared for the test; (2) go to a review course ifpossible (it is important to pay attention to hintsthat the instructor may give, as well as to takecareful notes and to ask questions about areas thatyou find confusing; (3) review any material from,for example, old tests, sample problems, reviewmaterial, or textbooks that might be on the test; (4)put the main ideas/information/formulas onto asheet that can be quickly reviewed many timesbecause this makes it easier to retain the key conceptsthat will be on the test; (5) review for severalshort periods rather than one long period, becauseyou will find that you retain information betterand be less fatigued; (6) avoid pulling an “allnighter”before the test, and plan on getting at least7 to 8 h of sleep the night before taking the test; (7)eat a healthy meal before a test to ensure that youwill have the energy and focus necessary; (8) arriveat the testing site well before the start time (youmay wish to set your alarm as well as a backup<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 305ACC-PULM-09-0302-019.indd 3057/10/09 8:35:15 PM


Table 5. Summary of the Most Appropriate Statistical Test Based on the Dependent and Independent Variables*Nature of the Dependent Variable Nature of the Independent Variables Tests To UseContinuousTwo independent groupsThat is GaussianUnpaired t testWith a large NUnpaired t testNeither Gaussian orlarge NWilcoxon rank-sum testMann-Whitney U testCategoric or ordinalχ 2 testFisher exact test (2 2 tables only)ContinuousMore than two independent groupsThat is GaussianANOVAThat is not GaussianKruskal-Wallis testCategoric or ordinalχ 2 testContinuous, Gaussian with repeated/Repeated-measures ANOVAserial measurements for each individualContinuousPaired (one population measuredThat is Gaussiantwice or two populations that are Paired t testWith a large Nmatched up)Paired t testNeither Gaussian or large NWilcoxon signed rank testCategoric or ordinalMcNemar testContinuous and Gaussian A single continuous variable CorrelationUnivariate linear regressionBinaryUnivariate logistic regressionContinuous and GaussianTwo or more variables, continuousand/or categoricMultivariate linear regressionANCOVABinaryMultivariate logistic regression*ANCOVA analysis of covariance. From http://bama.ua.edu/~jleeper/627/choosestat.html.alarm); and (9) go to the bathroom before walkinginto the examination room to avoid wasting anytime worrying about your bodily needs during thetest.Test-Taking TipsOnce a test is underway, there are also a numberof helpful recommendations to optimize success.(1) Read the directions carefully! Do not rush,but rather pace yourself by reading the entire questionand finding the key words. Always read thewhole question carefully and do not make assumptionsabout what the question might be. Simply notfollowing directions can result is a loss of manyeasy points on questions to which you actuallyknew the answers. (2) When you first receive yourtest, do a quick survey of the entire test so that youknow how to efficiently budget your time. (3) Bringa watch to the test with you so that you can betterpace yourself. (4) Keep a positive attitude throughoutthe whole test and try to stay relaxed. If youfeel nervous or overwhelmed, stop and take a fewdeep breaths to relax, then move to a portion of thetest with which you are more comfortable. If youdo not know an answer, skip it, go on with the restof the test, and come back to it later. Perhaps onanother part of the test there will be something thatwill help you out with that question. (5) Do theeasiest problems first; do not stay on a problem onwhich you are stuck, especially when time is a factor.(6) Do not worry if others finish before you butrather focus on the test in front of you. If you havetime left, look over your test, make sure that youhave answered all the questions, and only changean answer if you misread or misinterpreted thequestion because the first answer that you putdown is usually the correct one. Look for carelessmistakes and make sure that you have not left outany answers or parts of answers.Multiple Choice Test-Taking StrategiesThere are a variety of useful recommendationsto consider when answering multiple choice questions.(1) Read the question before you choose the306 Medical Statistics/Test-Taking Strategies (Kamp)ACC-PULM-09-0302-019.indd 3067/10/09 8:35:15 PM


answer after reading all the answers. Look for thecentral idea of each question. What is the mainpoint? (2) Come up with the answer in your headbefore looking at the possible answers. In thisway, the choices given on the test will not throwyou off or trick you. (3) Eliminate the answers thatyou know are not right. (4) Avoid answers thatbegin with always, never, none, except, most, orleast. Underline these or other key words on thetest paper. (5) Because there is no guessing penaltyon the pulmonary boards, always make aneducated guess, and select an answer. If you haveto guess, the following guidelines are helpful:(A) because the length of choices can be a clue,choose the longest; (B) if two choices are similar,choose neither; (C) if two choices are opposites,choose one of them; and the most generalalternative is usually the right answer. (6) Do notkeep on changing your answers because your firstchoice is usually the right one, unless you haveincorrectly read the question. (7) In “all of theabove” and “none of the above” choices, if youare certain one of the statements is true, do notchoose “none of the above.” Likewise, if one ofthe statements is false, do not choose “all of theabove.” (8) In a question with an “all of the above”choice, if you see at least two correct statements,then “all of the above” is probably the answer.(9) A positive choice is more likely to be true thana negative one. (10) If there is an “all of the above”option, and you know that at least two ofthe choices are correct, select the “all of the above”choice. (11) The correct answer is usuallythe choice with the most information.True/False Test TipsThere are a number of common sense recommendationsto consider when answering true/falsequestions. The basic ground rule for answeringtrue/false items is that if any part of the statementis not true, then the student should select false asthe answer. (1) Usually, there are more true answersthan false answers because false statements oftenare more difficult to construct as looking correct.Thus, if you are guessing on a true/false question,there is a 50% chance of being right with a trueanswer. (2) As with multiple choice questions, readthrough each statement carefully and pay attentionto the qualifiers and keywords. Qualifiers like“never,” “always,” and “every” mean that thestatement must be true all of the time. Usually,these types of qualifiers lead to a false answer.Qualifiers like “usually,” “sometimes,” and “generally”mean that the statement can be consideredtrue or false depending on the circumstances. Usually,these type of qualifiers lead to an answer oftrue. (3) If any part of the question is false, then theentire statement is false, but just because part of astatement is true does not necessarily make theentire statement true. (4) Every part of a true sentencemust be true. (5) Negatives can be confusing.If the question contains negatives such as “no,”“not,” or “cannot,” drop the negative and readwhat remains to decide whether that sentence istrue or false. If it is then true, then its opposite, ornegative, is usually false. Remember, double negativesmake a positive (ie, not uncommon actuallymeans common). (6) Questions that state a reasontend to be false. Words in the statement that causejustification or reason (ie, since, because, when, orif) tend to make the statement false. Pay close attention.The reason that is given maybe incorrect orincomplete.AcknowledgmentThe author is very appreciative of the materialsand comments provided by Dr. Allan Garland(Case Western Reserve University School of <strong>Medicine</strong>,Cleveland, OH). This work was supported inpart by a Merit <strong>Review</strong> grant from the Departmentof Veterans Affairs.References1. Casscells W, Schoenberger A, Graboys TB. Interpretationby physicians of clinical laboratoryresults. N Engl J Med 1978; 299:999−10012. Berwick DM, Fineberg HV, Weinstein MC.When doctors meet numbers. Am J Med 1981;71:991−9983. Windish DM, Huot SJ, Green ML. <strong>Medicine</strong> residents’understanding of the biostatistics andresults of the medical literature. JAMA 2007;298:1010−10224. Windish DM, Diener-West M. A clinician-educator’sroadmap to choosing and interpreting statisticaltests. J Gen Intern Med 2006; 21:656−660<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 307ACC-PULM-09-0302-019.indd 3077/10/09 8:35:15 PM


A succinct overview of statistical terminology that alsoprovides a logical approach for selecting the appropriatestatistical test.5. Guyatt GH, Haynes RB, Jaeschke RZ, et al. Users’guides to the medical literature: XXV. Evidencebasedmedicine: principles for applying the users’guides to patient care. JAMA 2000; 284:1290−1296This is the <strong>25th</strong> installment of a superb multipart seriesreviewing various topics that practically illustrate howphysicians should use medical statistics and evidencedbasedmedicine in their practice.6. Schulz KF, Chalmers I, Hayes RJ, et al. Empiricalevidence of bias: dimensions of methodologicalquality associated with estimates of treatmenteffects in controlled trials. JAMA 1995;273:408−4127. van Nieuwenhoven CA, Buskens E, van Tiel FH,et al. Relationship between methodological trialquality and the effects of selective digestive decontaminationon pneumonia and mortality in criticallyill patients. JAMA 2001; 286:335−3408. Knottnerus JA, van Weel C, Muris JWM. Evalua -tion of diagnostic procedures. BMJ 2002; 324:477−4819. Perera R. Making sense of diagnostic test likelihoodratios. ACP Journal Club 2007; 146:A8−A910. Noble JH. Meta-analysis: methods, strengths, weaknesses, and political uses. J Lab Clin Med 2006;147:7−20This is an excellent overview of a potentially powerfulstatistical tool that can be misused if not well understood.It contains 106 references.308 Medical Statistics/Test-Taking Strategies (Kamp)ACC-PULM-09-0302-019.indd 3087/10/09 8:35:15 PM


Acid-Base DisordersJanice L. Zimmerman, MD, FCCPObjectives:• Define terminology used in describing acid-basedisturbances• Diagnose simple and complex acid-base disorders• Differentiate anion gap and non-anion gap metabolic acidoses• <strong>Review</strong> appropriate treatment for major causes of acidosisand alkalosisKey words: acid; acidemia; acidosis; alkalemia; alkalosis;anion gap; base; strong ion differenceAcid-base disturbances are common in the criticallyill and result from a variety of clinical disorders.Appropriate diagnosis and management of acidbasedisorders require accurate interpretation ofsimultaneous measurements of electrolytes, albumin(alb), and arterial blood gases, as well as knowledgeof compensatory physiologic responses.Acid-Base PhysiologyThe acidity of body fluids is measured interms of hydrogen ion concentration, which isexpressed as pH (the negative log of the hydrogenion concentration). Hydrogen ion concentrationand pH are inversely related. Changes in aciditycan be stabilized although not totally correctedby endogenous buffers, molecules that accept ordonate hydrogen ions. Important buffers include thefollowing: bicarbonates (HCO 3 H ↔ H 2CO 3↔H 2O CO 2); proteins (protein H ↔ H-protein);hemoglobin (hemoglobin H ↔ H-hemoglobin);and phosphates (HPO 42 H ↔ H 2PO 4).The major buffering system in the body is thecarbonic acid–bicarbonate pair. The relation ofpH to this buffering system is expressed by theHenderson-Hasselbalch equation: pH pK logHCO 3/H 2CO 3 6.1 log HCO 3/0.03 Paco 2.The interrelation of H , Paco 2, and HCO 3canalso be illustrated by the Henderson equation. Thisequation is helpful as a bedside tool to predictor evaluate the accuracy of the three acid-baseparameters: H 24 Paco 2/HCO 3. An estimationof H can be made from pH. Between pH 7.2and 7.5, H changes by 1 mmol/L for each 0.01-Uchange in pH (Table 1).Although acid-base status is typically characterizedby H and HCO 3for clinical problemsolving, Stewart’s physicochemical analysis moreaccurately describes the determinants of acid-basestatus in biological systems. Three independentvariables determine acid-base changes: (1) Paco 2.(2) The strong ion difference (SID). This is the netelectric charge on strong electrolytes. It is the sumof all strong cation concentrations minus the sumof all strong anion concentrations, or SID (Na K Ca 2 Mg 2 ) (Cl other strong anions).The normal value is 38 to 42 mmol/L. The SID isdependent primarily on renal function but mayalso be impacted by GI function and tissue metabolism.(3) The total concentrations of nonvolatileweak acids. The main weak acids are inorganicphosphate (Pi) and serum proteins, particularlyalbumin.Other variables such as H and HCO 3aredependent variables that change based on primaryalterations in the independent variablesabove. Evaluation of acid-base status by Stewart’sTable 1. Relationship Between pH and H pHH , mmol/L7.65 227.60 257.55 287.50 327.45 357.40 407.35 457.30 507.25 567.20 637.15 717.10 797.05 897.00 100<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 309ACC-PULM-09-0302-020.indd 3097/10/09 8:35:37 PM


approach requires assessment of all the independentvariables.This chapter will primarily review the moretraditional analysis of acid-base problems usingformulas based on H and HCO 3. However, somecorrelates of Stewart’s approach will be mentionedand may help in understanding the limitations oftraditional acid-base analysis.DefinitionsThe following definitions are used: acidemia,an increase in H and a decrease in arterial pH;alkalemia, a decrease in H and elevation of arterialpH; acidosis, a process that, if unopposed, will leadto a decrease in pH; and alkalosis, a process that, ifunopposed, will lead to an increase in pH.Acid-Base DisturbancesThe development of an acid-base disorder istraditionally identified by changes in the bicarbonateconcentration (the metabolic or renal component)or by changes in Paco 2(the respiratorycomponent). A primary metabolic disturbanceresults from a primary alteration in bicarbonateconcentration, whereas a primary respiratory disturbanceresults from a primary alteration in Paco 2.In Stewart’s approach, metabolic disturbancesresult from changes in the strong ion differenceand/or concentrations of weak acids.Features of acid-base disturbances are noted inTable 2. Primary acid-base disturbances also initiatecompensatory responses that are predictable. Compensatoryprocesses help normalize the arterial pHbut usually never return the pH fully to normal.Appropriate compensatory responses requireTable 2. Features of Acid-Base DisturbancesDisorders Primary Problem pH CompensationMetabolicacidosisMetabolicalkalosisRespiratoryacidosisRespiratoryalkalosis↓HCO 3, ↓SID, ↑alb, ↑Pi ↓ ↓Paco 2↑HCO 3, ↑SID, ↓ alb ↑ ↑Paco 2↑Paco 2↓ ↑HCO 3↓Paco 2↑ ↓HCO 3normal functioning of lungs and kidneys, andfailure to develop a compensatory response definesthe presence of a second primary disorder.Metabolic AcidosisIn metabolic acidosis, the following are seen:(1) defect: increased acid accumulation or decreasedextracellular HCO 3; decreased SID, increasedconcentration of alb, or increased Pi; (2) laboratorymanifestation: decreased pH, decreased plasmaHCO 3; and (3) compensation: increased ventilationleading to decreased Paco 2; exchange of intracellularNa and K for extracellular H ; increasedrenal H excretion with urinary buffers and regenerationof new bicarbonate (slower process).The degree of respiratory compensation can beestimated by the following formulas: Paco 2 1.5 HCO 3 8 2; and Paco 2 1.2 HCO 3.During prolonged acidosis, when respiratorycompensation is complete, the last two digits of pHequal Paco 2(if pH 7). The lower limit of normalcompensation is Paco 2of 10 mm Hg.EtiologiesMetabolic acidosis can result from an increasein endogenous acid production that overwhelmsrenal excretion (eg, ketoacidosis), exogenous acidinput, excessive loss of bicarbonate (eg, diarrhea),or decreased renal excretion of endogenous acids(eg, chronic renal failure). Metabolic acidoses areusually further characterized by the anion gap (AG).Unmeasured anions (proteins, phosphates, sulfates,and organic acids) in healthy persons exceed theunmeasured cations (potassium, calcium, andmagnesium), and this difference results in the AG.The AG can be estimated by the following formula:AG Na (Cl HCO 3) 10 4.The AG can be increased because of a decreasein unmeasured cations, an increase in unmeasuredanions, or laboratory error in measurement. Theclassification of metabolic acidoses by the AGis shown in Table 3. In addition, lactic acidosishas been associated with drugs such as metformin,nucleoside reverse transcriptase inhibitors,linezolid, lorazepam, and propofol. Base deficitdetermined by measured plasma pH and Paco 2,and hemoglobin concentration using algorithmsand nomograms of Siggard-Andersen, has also been310 Acid-Base Disorders (Zimmerman)ACC-PULM-09-0302-020.indd 3107/10/09 8:35:38 PM


Table 3. Classification of Metabolic Acidosis by AGIncreased AGLactic acidosisType A (tissue ischemia)Type B (altered cell metabolism)KetoacidosisDiabetesAlcohol inducedStarvationRenal failureToxin ingestionSalicylatesMethanolEthylene glycolParaldehydeNormal AG (hyperchloremic)GI loss of HCO 3–Ureteral diversionDiarrheaIleostomyProximal colostomyRenal loss of HCO 3–Proximal renal tubular acidosisCarbonic anhydrase inhibitorRenal tubular diseaseAcute tubular necrosisChronic tubulointerstitial diseaseDistal rental tubular acidosis (type 1 and 4)Hypoaldosteronism, aldosterone inhibitorsPharmacologicAmmonium chlorideHyperalimentationDilutional acidosisused to evaluate metabolic acidosis. However, thisapproach is less reliable in critically ill patients.The AG has several limitations as the soleindicator of a metabolic acidosis. An AG acidosiscan exist even with a normal AG in patients whoare severely hypoalbuminemic or have pathologicparaproteinemias. In such patients, the normal AGmay be as low as 4 to 5 mmol/L. For every decreaseof 1 g/dL in alb, a decrease of 2.5 to 3 mmol in AGwill occur. Pathologic paraproteinemias lower theAG because Igs are largely cationic. Another exceptioncan occur when an elevated AG does not reflectan underlying acidosis. In patients with significantalkalemia (usually pH 7.5), alb is more negativelycharged, which increases unmeasured anions.Normal AG acidoses can also be categorized bythe potassium level. Hypokalemia ( 3.5 mmol/L)is associated with ureteral diversion, diarrhea,proximal colostomy, ileostomy, proximal renaltubular acidosis (RTA), type 1 distal RTA, andhyperalimentation. Hyperkalemia ( 4.5 mmol/L)can be found in hypoaldosterone states, ammoniumchloride administration, and type 4 distal RTA.A decreased AG implies an increase in unmeasuredcations or a decrease in unmeasured anions.Possible etiologies include paraproteinemias,hypoalbuminemia, hyponatremia, lithium toxicity,profound hyperkalemia, hypercalcemia, orhypermagnesemia, and halide poisoning (bromideand iodide).In an uncomplicated AG metabolic acidosis,every increase of 1 mmol/L in the AG should resultin a concomitant decrease of 1 mmol/L in HCO 3.Deviation from this relation suggests a mixed acidbasedisorder. The difference between these two valueshas been termed the gap, and can be expressedas follows: gap (deviation of AG from normal) (deviation of HCO 3from normal). If the normal AGis assumed to be 12 mmol/L and the normal HCO 3is 24 mmol/L, then the following equation results:gap (AG 12) (24 HCO 3).The normal value for gap should be zero.However, variance in measurements can result ina gap of 06. If the gap is positive, then eithera simultaneous metabolic alkalosis or a respiratoryacidosis exists. If the decrease in HCO 3isgreater than the increase in AG, which results ina negative gap, then a concomitant normal AGacidosis (hyperchloremic) or chronic respiratoryalkalosis with compensating hyperchloremicacidosis may exist. Small deviations of the gapmay not indicate mixed acid-base disorders, andclinical information must always be considered inthe evaluation.In Stewart’s approach, a decreased SID inmetabolic acidosis results from an elevated Cl oraccumulation of unidentified anions analogous toa normal AG or increased AG acidosis describedabove. A simplified formula that can be used toestimate the SID takes into account changes in alband phosphate: SID HCO 3 0.28 alb g/L 1.8Pi mmol/L.Clinical ManifestationsThe clinical features of metabolic acidosisdepend primarily on the underlying disorder.Some effects of acidosis can be related to the changein pH. The effects of acidemia on the cardiovascularsystem have been questioned and may not besignificant.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 311ACC-PULM-09-0302-020.indd 3117/10/09 8:35:38 PM


CNS: Increased cerebral blood flow; decreasedmental status.<strong>Pulmonary</strong>: Hyperventilation with rapid and/ordeep respiration (Kussmaul respiration); increasein pulmonary vascular resistance.Cardiovascular: Potential decreased myocardialcontractility; arrhythmias; potential decreasedresponsiveness to catecholamines; arterial vasodilation.Metabolic: Insulin resistance; hyperkalemia;hypercalcemia; increased protein catabolism;catecholamine stimulation. Metabolic acidosisshifts the oxyhemoglobin dissociation curve tothe right, which promotes oxygen release at thetissue level.TreatmentTreatment requires appropriate diagnosis ofthe etiology of metabolic acidosis and correctionof the underlying disorder. Therapy for a non-AG metabolic acidosis requires replacing volumelosses with a bicarbonate-containing solution thathas a low chloride concentration. Treatment ofmost AG metabolic acidoses is well established (ie,insulin for ketoacidosis, dialysis for renal failure).Adjunctive therapy with sodium bicarbonate inthis setting is usually not recommended. Treatmentof lactic acidosis with bicarbonate does notimprove survival. In circumstances when pH 7.1is associated with severe hemodynamic compromise,some studies show that bicarbonate mayimprove cardiovascular responsiveness. However,myocardial performance is often normal in the settingof metabolic acidosis. Potential complicationsof bicarbonate administration frequently outweighpossible benefits (Table 4). Other nonbicarbonatebuffers, such as tris-hydroxymethyl aminomethane,dichloroacetate, and carbicarb, may avoidTable 4. Potential Complications of Bicarbonate AdministrationVolume overloadParadoxical CSF/intracellular acidosisRespiratory acidosisImpaired oxygen delivery (tissue hypoxia)HypokalemiaHypocalcemiaHypernatremiaHyperosmolalityOvershoot alkalemiasome of the complications, but clinical data do notsupport their use.Metabolic AlkalosisIn metabolic alkalosis, the following are seen:(1) defect: extracellular fluid increase in base or lossof acid; increased SID or decreased concentration ofalb; (2) laboratory manifestation: increase in pH andincrease in HCO 3; and (3) compensation: hypoventilationleading to increased Paco 2. Paco 2may rise6 to 7 mm Hg for every increase of 10 mmol/L inHCO 3. This response is limited by hypoxemia to amaximum Paco 2of 50 to 55 mm Hg. A characteristicfeature of metabolic alkalosis is that the kidneyscontribute to the process by adding HCO 3to thecirculation and hyperabsorbing filtered HCO 3rather than enhancing HCO 3excretion.EtiologiesThe etiologies of metabolic alkalosis can bedivided into those characterized by chloride depletion(hypovolemic) and those characterized by chlorideexpansion (hypervolemic). These etiologies arealso referred to as chloride or saline responsive andchloride or saline resistant, respectively (Table 5). InStewart’s approach, an increased SID in metabolicalkalosis is due to decreased Cl . HypokalemiaTable 5. Etiologies of Metabolic AlkalosisHypovolemic, Cl − depletedGI loss of H +VomitingGastric suctionCl − rich diarrheaVillous adenomaRenal loss of H +DiureticsPosthypercapniaHypervolemic, Cl − expandedRenal loss of H +Primary hyperaldosteronismPrimary hypercortisolismAdrenocorticotropic hormone excessPharmacologic hydrocortisone/mineralocorticoid excessRenal artery stenosis with right ventricular hypertensionRenin secreting tumorHypokalemiaBicarbonate overdosePharmacologic overdose of NaHCO 3Milk-alkali syndromeMassive blood transfusion312 Acid-Base Disorders (Zimmerman)ACC-PULM-09-0302-020.indd 3127/10/09 8:35:38 PM


is common to both types of metabolic alkalosis.Measurement of urine chloride (UCl) is helpful indistinguishing the two categories: UCl 20 mmol/Lin chloride depletion, and UCl 20 mmol/L inchloride expansion. In metabolic alkalosis, UClis a more accurate reflection of intravascular volumethan urine sodium because sodium must beexcreted with excess HCO 3. Diuretic use alters theutility of UCl for assessing volume status.Clinical ManifestationsSevere metabolic alkalosis is associated with ahigh mortality rate in hospitalized patients.Cardiovascular: Tachycardia; arrhythmias; increasedcardiac contractility.CNS: Decreased cerebral blood flow; seizures;altered mental status.Metabolic: Ionized hypocalcemia; hypokalemia;hypophosphatemia; impaired enzyme function. Incontrast to metabolic acidosis, metabolic alkalosisshifts the oxyhemoglobin dissociation curve to theleft, which increases oxygen affinity for hemoglobinand may impair oxygen release at the tissuelevel.TreatmentThe clinical setting along with measurementof UCl usually allows appropriate classification ofthe metabolic alkalosis. Etiologies associated withchloride depletion respond to volume replacement(saline responsive). Normal saline solutionoffers more chloride per liter (154 mmol/L) thanlactated Ringer solution (104 to 109 mmol/L).Etiologies associated with chloride expansionusually indicate imbalances in the renal-adrenalaxis. Potassium deficiencies should be correctedas appropriate. It is rarely necessary to considerhydrochloric acid (0.1 N solution) administrationor deliberate hypoventilation to correct severemetabolic alkalosis. Acetazolamide, 250 to 375 mgqd or bid, will increase excretion of HCO 3butmay result in acidemia and worsening hypokalemiabecause of kaliuresis. A single 500-mg dose ofacetazolamide may be sufficient.Respiratory AcidosisIn respiratory acidosis, the following are seen:(1) defect: ineffective alveolar ventilation; increasein carbon dioxide production; (2) laboratorymanifestation: increased Paco 2; decreased pH;and (3) compensation: buffers (primarily intracellularproteins); increased respiratory rate; slowrenal response ( 24 to 36 h) with increasedbicarbonate reabsorption. The normal respiratoryresponse to hypercapnia is an increase in alveolarventilation. The stimulus to increase ventilation ismediated by changes in the H of the cerebrospinalfluid (CSF), which affects chemoreceptors in themedulla. Because CSF is essentially devoid of nonbicarbonatebuffers, carbon dioxide that diffusesreadily across the blood-brain barrier results in amarked increase in CSF H . The CSF pH is correctedtoward normal by a slower rise in CSF HCO 3thatresults from transfer of cerebral or blood HCO 3.A very small increase in plasma HCO 3can beseen acutely because of titration of intracellularnonbicarbonate buffers. For each increase of 10 mmHg in Paco 2, the HCO 3increases by 1 mmol/L.Acute increases in Paco 2change the plasmaHCO 3only 4 to 5 mmol/L to a maximum of 30 to32 mmol/L. If the respiratory acidosis is acute, thepH should decrease by 0.08 for each increase of10 mm Hg in Paco 2.The renal reabsorption of HCO 3is slower andplays a minimal role in acute respiratory acidosis.Over hours to days, the kidneys compensate withproximal reabsorption of filtered bicarbonate andsecretion of H in the distal tubule, which is trappedby ammonia and excreted. Bicarbonate reabsorptionis accompanied by loss of Cl . This mechanismprovides adequate, although incomplete pHcompensation for chronic respiratory acidosis. Thetransition from acute to chronic respiratory acidosisis defined by HCO 3retention due to renal compensation.In chronic respiratory acidosis, the HCO 3increases 3.5 mmol/L for each increase of 10 mm Hgin Paco 2. The limit of renal compensation in chronicrespiratory acidosis is HCO 3of approximately45 mmol/L. Higher values suggest an associatedsecondary metabolic alkalosis. In chronic respiratoryacidosis, the pH decreases by 0.03 for eachincrease of 10 mm Hg in Paco 2.EtiologiesCommon causes of acute respiratory acidosisinclude airway obstruction, respiratory centerdepression, neuromuscular disorders, and pulmonarydisorders (Table 6).<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 313ACC-PULM-09-0302-020.indd 3137/10/09 8:35:38 PM


Table 6. Etiologies of Respiratory AcidosisAirway obstructionForeign bodyTongue displacementLaryngospasmObstructed endotracheal tubeSevere bronchospasmObstructive sleep apneaRespiratory center depressionGeneral anesthesiaSedative, narcotic drugsCerebral injury, ischemiaIncreased CO 2productionMalignant hyperthermiaShiveringHypermetabolismHigh carbohydrate dietNeuromuscular disordersDrugs or toxinsElectrolyte disordersSpinal cord injuryGuillain-Barré syndromeMyasthenia gravisPolymyositisLung conditionsRestrictive diseaseObstructive diseaseHemothorax or pneumothoraxFlail chestAcute lung injuryObesity-hypoventilation syndromeInappropriate ventilator settingsClinical ManifestationsThe clinical manifestations of acute respiratoryacidosis depend on the absolute increase inPaco 2, the rate of Paco 2increase, and severity ofassociated hypoxemia. Intentional hypoventilation(permissive hypercapnia) is well tolerated and haslittle impact on hemodynamics.Nervous System: Somnolence; anxiety or confusion;psychosis; tremors, myoclonus, or asterixis;headache; papilledema. (Cerebral vasodilationleads to increased intracranial pressure.)Cardiovascular System: Tachycardia; hypertension;supraventricular/ventricular arrhythmias;peripheral vasodilation.TreatmentTreatment involves the rapid identificationof the etiology of respiratory acidosis andimplementation of corrective action. In manycircumstances, intubation and mechanical ventilationare necessary to support alveolar ventilation.Administration of NaHCO 3for acute respiratoryacidosis is not indicated. Remember that NaHCO 3is metabolized to carbon dioxide, which mayworsen acidosis. In addition, normally functioningkidneys excrete most of the bicarbonate in theacute setting.A therapeutic pitfall to avoid in the treatmentof respiratory acidosis is creation of a posthypercapnicalkalosis. This condition most commonlyoccurs when a patient with compensated chronicrespiratory acidosis is overventilated to a normal ornear-normal Paco 2. A high plasma HCO 3resultsin alkalemia. Appropriate ventilator managementand goals should prevent this situation.Respiratory AlkalosisThe following are seen in respiratory alkalosis:(1) defect: primary hyperventilation; (2) laboratorymanifestation: decreased Paco 2; increased pH; and(3) compensation: protein or hemoglobin release ofhydrogen ion; slow renal response with bicarbonateloss in urine. Alveolar ventilation is regulatedby several factors: chemoreceptors in the medulla(sensitive to H ) and great vessels (sensitive tooxygen), cortical input (voluntary control), andpulmonary chemoreceptors and stretch receptors.Any of these factors or a combination may lead tohyperventilation. The HCO 3decreases 2 mmol/Lfor each decrease of 10 mm Hg in Paco 2in acuterespiratory alkalosis.Similar to changes noted in respiratory acidosis,pH increases 0.08 for every decrease of 10 mmHg in Paco 2in acute respiratory alkalosis, and pHincreases 0.03 for each decrease of 10 mm Hg inPaco 2in chronic respiratory alkalosis. However,chronic respiratory alkalosis is unique among acidbasedisorders in that pH may return to normal ifthe condition is prolonged. Persistent hypocapniaresults in decreased renal H secretion. PlasmaHCO 3decreases 5 mmol/L for each decrease of10 mm Hg in Paco 2.EtiologiesThe etiologies of respiratory alkalosis are listedin Table 7.314 Acid-Base Disorders (Zimmerman)ACC-PULM-09-0302-020.indd 3147/10/09 8:35:39 PM


Table 7. Etiologies of Respiratory AlkalosisHypoxemic drive<strong>Pulmonary</strong> disease with arterial-alveolar gradientCardiac disease with right-to-left shuntCardiac disease with pulmonary edemaHigh altitudeAcute and chronic pulmonary diseaseEmphysema<strong>Pulmonary</strong> embolism<strong>Pulmonary</strong> edemaMechanical overventilationStimulation of respiratory centerNeurologic disordersPainPsychogenicLiver failure with encephalopathySepsis/infectionSalicylatesProgesteronePregnancyFeverClinical ManifestationsAcute respiratory alkalosis results in neurologic,cardiovascular, muscular, and metabolicabnormalities.Nervous System: Confusion or dizziness; seizures;paresthesias; circumoral numbness.Cardiovascular System: Tachycardia; arrhythmias(especially at pH 7.6).Muscular System: Cramps; carpopedal spasms.Metabolic Changes. Hypokalemia; hypophosphatemia;ionized hypocalcemia. In chronicrespiratory alkalosis, mild hypokalemia (fromintracellular shift) and hyperchloremia (from renalretention) result.TreatmentSevere alkalemia is associated with high mortalityand requires aggressive treatment. Therapy isdirected at the underlying cause because pharmacologicagents are not available for the alkalemia.Complex Acid-Base DisordersWhen a single process, such as metabolicalkalosis, results in the acid-base disturbance,it is classified as simple. In many patients, multipleacid-base disturbances exist concurrentlyand are labeled complex acid-base disorders(Table 8). Triple acid-base disorders involve twometabolic disturbances with a superimposedprimary respiratory disturbance. Such conditionsmay be seen in patients with a chronic acid-basedisturbance who have a superimposed mixeddisturbance. Clues to the presence of a complexdisorder include a normal pH (with the exceptionof respiratory alkalosis), Paco 2and HCO 3deviating in opposite directions, and a pH changein the opposite direction for a known primarydisorder.Complex disorders require an organizedapproach to interpretation of acid-base parametersfor accurate diagnosis. The followingTable 8. Mixed Acid-Base DisordersAcid-Base DisordersRespiratory acidosis with metabolic acidosisRespiratory alkalosis with metabolic alkalosisRespiratory acidosis with metabolic alkalosisRespiratory alkalosis with metabolic acidosisMetabolic acidosis with metabolic alkalosisClinical SyndromesCardiopulmonary arrestIntoxication with methanol, ethylene glycolCOPD with lactic acidosis, diabetic ketoacidosis, etc.Severe hypophosphatemiaRespiratory failure with renal failure, diarrhea, etc.Cirrhosis with diuretic use, vomitingPregnancy with hyperemesisOverventilation in COPD patientCOPD with diuretic use, vomiting, gastric suctionSevere hypokalemiaSepsisSalicylate intoxicationRenal insufficiency with congestive heart failure, pneumoniaAdvanced liver disease with lactic acidosisUremia or ketoacidosis with vomiting, gastric suction, diuretics, etc.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 315ACC-PULM-09-0302-020.indd 3157/10/09 8:35:39 PM


Table 9. Acid-Base Calculations*Acid-Base DisorderFormulaRespiratory acidosisAcute Decrease in pH 0.08 (Paco _________ 40) 210Increase in [HCO – 3] ______Paco 2 310Chronic Decrease in pH = 0.03 (Paco _________ 40) 210Increase in [HCO – 3] 3.5 _______PaCO 210Respiratory alkalosisAcute Increase in pH 0.08 _________(40Paco ) 210Decrease in [HCO – 3] 2 ______Paco 210Chronic Increase in pH 0.03 _________(40Paco ) 210Decrease in [HCO – 3] 5 7 _______PaCO 210Metabolic acidosis AG [Na]([Cl ][HCO – 3]Paco 2 1.5 [HCO – 3] 8 2gap (AG12)(24[HCO – 3])Metabolic alkalosis Increase in Paco 2 0.60.7 [HCO – 3]*AG = anion gap.considerations are one system that may be usefulin determining the disturbances present:1. Determine the overall acid-base condition by measuringpH. Is acidemia or alkalemia present?2. Determine if the primary process is meta -bolic (HCO 3deviation) or respiratory (Paco 2deviation).3. If a respiratory disturbance is present, determineif it is acute or chronic.4. If a metabolic disturbance is present, determineif respiratory compensation is appropriate.5. If a metabolic acidosis is present, calculate the AG.6. If an AG metabolic acidosis is present, calculatethe gap to determine if other metabolic disturbancesare present.A summary of useful calculations for acid-baseevaluation is found in Table 9.Annotated BibliographyAdrogue HJ, Madias NE. Management of life- threateningacid-base disorders: first of two parts. N Engl J Med 1998;338:26–34Adrogue HJ, Madias NE. Management of life-threateningacid-base disorders: second of two parts. N Engl JMed 1998; 338:107–111<strong>Review</strong> of adverse consequences and management of severeacid-base disorders.Dubose TD. Acid-base disorders. In: Brenner BM, ed.Brenner and Rector’s the kidney. 6th ed. Philadelphia,PA: WB Saunders, 2000; 925–997A detailed review of all acid-base disorders.Fencl V, Jabor A, Kazda A, et al. Diagnosis of metabolicacid-base disturbances in critically ill patients. Am JRespir Crit Care Med 2000; 162:2246–2251Use of the SID compared to the AG and base excess in analysisof acid-base status.Forsythe SM, Schmidt GA. Sodium bicarbonate forthe treatment of lactic acidosis. Chest 2000; 117:260–267A review of the literature on the use of bicarbonate to correctlactic acidosis and its lack of benefit.Kaufman D, Kitching AJ, Kellum JA. Acid-base balance.In: Hall JB, Schmidt GA, Wood LDH, eds. Principles ofcritical care. 3rd ed. New York: McGraw-Hill, 2005;1201–1208This chapter outlines an approach to acid-base disorders usingSID.316 Acid-Base Disorders (Zimmerman)ACC-PULM-09-0302-020.indd 3167/10/09 8:35:39 PM


Kellum JA. Disorders of acid-base balance. Crit Care Med2007; 35:2630–2636A review of quantitative and traditional approaches to acidbaseanalysis.Levraut J, Grimaud D. Treatment of metabolic acidosis.Curr Opin Crit Care 2003; 9:260–265A review of the effects of acidosis and treatmentapproaches.Narins RG, Emmett M. Simple and mixed acid-basedisorders: a practical approach. <strong>Medicine</strong> (Baltimore)1980; 59:161–187An old but classic review of all acid-base disorders.Narins RG, Kupin W, Faber MD, et al. Pathophysiology,classification and therapy of acid-base disturbances. In:Arieff AI, DeFronzo RA, eds. Fluid, electrolyte, and acidbasedisorders. New York: Churchill Livingstone, 1995;105–198A detailed review of pathophysiology and management of allacid-base disorders.Paulson WD. How to interpret the anion gap. J Crit Illness1997; 12:96–99Brief review of the use of the AG and its limitations.Salem MM, Mujais SK. Gaps in the anion gap. Arch InternMed 1992; 152:1625–1629A detailed review of limitations of the use of the AG in clinicalpractice.Stewart PA. Modern quantitative acid-base chemistry.Can J Physiol Pharmacol 1983; 61:1444–1461The original explanation of physicochemical analysis and useof SID in acid-base problems.Wrenn K. The delta gap: an approach to mixed acid-basedisorders. Ann Emerg Med 1990; 19:1310–1313A concise review of the derivation, interpretation, and clinicalapplication of the gap.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 317ACC-PULM-09-0302-020.indd 3177/10/09 8:35:39 PM


Notes318 Acid-Base Disorders (Zimmerman)ACC-PULM-09-0302-020.indd 3187/10/09 8:35:40 PM


Drug-Induced Lung DiseasesDavid W. Kamp, MD, FCCPObjectives:• Appreciate the diverse clinical syndromes of drug-inducedpulmonary diseases• Understand the general approach to the patient with suspecteddrug toxicity• <strong>Review</strong> the common abnormalities associated with specificchemotherapeutic agents• Comprehend the typical manifestations of pulmonarytoxicity caused by nonchemotherapeutic agentsKey words: acute lung injury; drugs; hypersensitivity pneumonitis;pulmonary fibrosisDrug-induced lung diseases have challengedclinicians since the dawn of modern medicine. In1880, William Osler described a patient with pulmonaryedema associated with opiate exposureand suggested that there was a pathophysiologicrelationship. In 1972, Edward Rosenow 1 extensivelyreviewed this topic and identified 20 drugsthat clearly caused pulmonary toxicity. A decadelater, Cooper and colleagues 2 expanded the listto 37 drugs. Since then, 350 drugs have beenimplicated in causing a wide array of pulmonarymanifestations that involve virtually all componentsof the respiratory system including theairways, parenchyma, pleura, pulmonary circulation,mediastinum, vocal cords, and respiratorymuscles (Table 1). The number of drugs that causelung disease will undoubtedly continue to growas new agents and biological response modifiersare developed.Although the clinical syndromes associatedwith certain drugs are well defined, the full scopeof the epidemiology of drug-induced lung diseaseis not firmly established, partly because of the factthat the diagnosis is one of exclusion (eg, infectionand tumor). Notably, there are no pathognomonicclinical, laboratory, physiologic, radiographic, orhistologic findings. Also, most reactions are idiosyncratic,without any clear relationship to dose ortime of exposure. Indeed, some drugs can causetoxicity years after exposure (eg, cyclophosphamide).With few exceptions, the risk factors for druginducedlung diseases are poorly defined. Moreover,confounding variables such as the use ofother drugs, oxygen, or radiation therapy, each ofwhich can cause pulmonary injury or have interactiveeffects (eg, bleomycin and oxygen), oftenhamper the diagnosis. Rechallenge with the implicateddrug is rarely performed because effectivealternative agents are nearly always available.Thus, clinicians evaluating patients with possibledrug-induced pulmonary symptoms must obtaina thorough drug exposure history, maintain a highindex of suspicion, and use a systematic diagnosticapproach that is reviewed herein.The management of patients with drug-inducedpulmonary side effects is largely supportive. Typically,therapy with the implicated drug is withdrawn,and a trial of corticosteroids is considered,particularly in the setting of significant symptomsand/or gas-exchange abnormalities. The scientificbasis for using corticosteroids is, unfortunately,Table 1. Major Clinical Syndromes of Drug-Induced<strong>Pulmonary</strong> Disease1. Chronic pneumonitis/fibrosis*2. Hypersensitivity-type lung disease*3. Acute noncardiogenic pulmonary edema*4. Bronchiolitis obliterans with organizing pneumonia5. Alveolar hypoventilaton6. Bronchospasm7. Cough8. Concentric bronchiolitis obliterans9. Pleural effusions10. Venous thromboembolism11. <strong>Pulmonary</strong> vasculitis12. <strong>Pulmonary</strong> hypertension13. Drug-induced SLE14. Alveolar hemorrhage15. <strong>Pulmonary</strong> renal syndrome16. Alveolar proteinosis17. Mediastinal abnormalities (eg, adenopathy, lipomatosis,or mediastinitis)18. Panlobular emphysema19. <strong>Pulmonary</strong> calcinosis20. Pseudosepsis syndrome*Most common pulmonary manifestations.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 319ACC-PULM-09-0302-021.indd 3197/10/09 10:21:00 PM


supported by anecdotal reports rather than welldesignedcontrolled studies.Major Clinical Syndromes of Drug-Induced <strong>Pulmonary</strong> DiseaseChronic Pneumonitis/FibrosisA wide array of drugs has been implicated incausing chronic interstitial pneumonitis and/orfibrosis, making them the most common manifestationsof drug-induced lung disease. Some of theagents implicated are listed in Table 2. There is asmall increase in antidepressant use, especiallytricyclic-related agents and selective serotoninreceptor inhibitors, in patients with idiopathicpulmonary fibrosis (IPF; odds ratio [OR], 1.52; 95%confidence interval 1.24 to 1.86). 3 These patientscommonly present with insidious onset of coughand dyspnea. Weight loss and clubbing may alsobe present, raising the possibility of an underlyingmalignancy or IPF, respectively. Chest radiographyand high-resolution chest CT (HRCT) generallyreveal reticular infiltrates beginning in the basilarsubpleural regions and progressing to diffusedisease. The results of pulmonary function tests(PFTs) typically show reduced lung volumes(eg, restrictive physiology), a reduced diffusingcapacity of the lung for carbon monoxide (Dlco),and arterial hypoxemia at rest or with exercise.Table 2. Some Drugs That Cause Chronic Pneumonitis/FibrosisChemotherapeuticAgentsNonchemotherapeuticAgentsHypersensitivity-Type Lung DiseaseVirtually any drug can cause a generalizedhypersensitivity-type reaction with respiratorysymptoms that are associated with pulmonaryinfiltrates and eosinophilia (PIE). The agents commonlyimplicated are listed in Table 3. Patients canpresent with Loeffler syndrome, consisting of anacute onset over several days of cough, dyspnea,fever, rash, myalgias, peripheral eosinophilia, andfleeting infiltrates. Alternatively, patients maypresent with chronic eosinophilic pneumoniaconsisting of a subacute onset over several monthsof low-grade fever, night sweats, nonproductivecough, and weight loss. Establishing a diagnosismay be challenging because not all findings, especiallyperipheral eosinophilia, are present in eachpatient. The diagnosis is often secured by bronchoscopywith BAL and biopsy and, in someinstances, after a prompt response to therapy withcorticosteroids. The prognosis is generally favorablewith a mortality rate of 1%.Acute Noncardiogenic <strong>Pulmonary</strong> EdemaNoncardiogenic pulmonary edema can occurafter exposure to a variety of drugs (Table 4). Thesepatients generally present with acute dyspnea anda nonproductive cough that develops during aperiod of hours. A physical examination of thechest reveals inspiratory crackles, whereas arterialblood gas levels show hypoxemia. The chest radiographgenerally demonstrates diffuse acinar and/or ground-glass infiltrates. The histopathology can• Azathioprine• BCNU• Bleomycin*• Busulfan• Chlorambucil• Cyclophospamide• Fludarabine• Gemcitabine• 6-Mercaptopurine• Methotrexate• Mitomycin C• Taxanes (paclitaxel/docetaxel)• Tyrosine kinase inhibitors(imatinib)*Most commonly implicated.• Amiodarone*• Anti-TNF--targetedtherapy• Cocaine• Gold• Heroin• Methysergide• Nitrofurantoin• Penicillamine• Phenytoin• Sirolimus• Statins• Sulfasalazine• TocainideTable 3. Some Drugs That Cause Hypersensitivity-TypeLung DiseaseChemotherapeuticAgents• Azathioprine• Bleomycin• Busulfan• Imatinib• Methotrexate*• Nitrogen mustards• Procarbazine• Taxanes*Most commonly implicated.NonchemotherapeuticAgents• Antibiotics (-lactam andsulfa-containing)*• Anti-TNF--targeted therapy• Isoniazide• Nitrofurantoin• NSAIDsPenicillamine• Phenytoin• Statins320 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3207/10/09 10:21:00 PM


Table 4. Some Drugs That Cause Acute Noncardiogenic<strong>Pulmonary</strong> EdemaChemotherapeuticAgents• Azathioprine• Cytosine arabinoside• Gemcitabine• G-CSF• IL-2• Methotrexate• Mitomycin C• Nitrogen mustards• Retinoic Acid• TNF• Vinblastine (withmitomycin)NonchemotherapeuticAgents• Amiodarone• Aspirin and NSAID overdose• Cocaine• Opiate overdose (eg, heroin)• Sedative/hypnotic drugoverdose• Sulfasalazine• Tocolytic therapy (eg,terbutaline, ritodrine)• IFN-1bTable 5. Some Drugs That Cause Bronchilolitis ObliteransWith Organizing Pneumonia (BOOP) / Cryptogenic OrganizingPneumonia (COP)ChemotherapeuticAgents• Bleomycin• Cyclophosphamide• Methotrexate• Mitomycin CNonchemotherapeuticAgents• Amiodarone• Amphotericin B• Carbamazepine• Cocaine• Nitrofurantoin• Penicillamine• Phenytoin• Sirolimus• Sulfasalazine• Ticlodipinebe similar to ARDS. Several mechanisms have beenimplicated in causing drug-induced noncardiogenicpulmonary edema. First, some drugs increasethe filtration coefficient of the respiratory membrane,making it more permeable (eg, overdoseof narcotics/sedatives). Second, certain agentsdepress the CNS, resulting in neurogenic pulmonaryedema. Finally, some drugs cause an idiosyncraticreaction resulting in noncardiogenicpulmonary edema within hours of absorption. Theprognosis varies depending on the offendingagent. For example, pulmonary edema associatedwith an overdose of salicylates is potentiallyreversible with appropriate management, whereaspatients with carmustine-induced pulmonaryedema generally have a poor prognosis.Cryptogenic Organizing PneumoniaA variety of drugs cause bronchiolitis obliteranswith organizing pneumonia (BOOP), whichis preferentially now called cryptogenic organizingpneumonia (COP; Table 5). Thus, any new diagnosisof COP requires a careful review of their drugexposure before concluding that COP is “cryptogenic.”These patients typically present withcough, dyspnea, crackles on physical examination,and patchy airspace infiltrates seen on thechest radiograph. PFT results reveal a mixedobstructive and restrictive defect. Gold andpenicillamine, the two best-characterized etiologicagents, are often used in the management ofrheumatoid arthritis (RA), a disease that can alsocause COP. Thus, it may be difficult to distinguishdrug-induced COP from the patient’s underlyingcollagen vascular disorder. Management requiresa high clinical suspicion, lung biopsy, promptwithdrawal of therapy with the implicated drug,and the administration of corticosteroids. Theoutcome is generally favorable, as it is with idiopathicCOP.Alveolar HypoventilationAlveolar hypoventilation is caused by drugsthat induce respiratory depression or block respiratorymuscle function. Patients with underlyingpulmonary or neuromuscular disorders are particularlyprone to the development of acute hypercarbicrespiratory failure. Some of the agentsimplicated in causing neuromuscular blockade ormotor neuropathies/myopathies are listed in Table6. Aminoglycoside-induced neuromuscular blockadeis a rare but potentially life-threatening adverseeffect that has been described in patients who havebeen exposed to neomycin, streptomycin, tobramycin,gentamicin, amikacin, kanamycin, andnetilmicin. 4 The risk of aminoglycoside-inducedneuromuscular blockade is increased in the presenceof a disease or drug that promotes neuromuscularblockade, increasing aminoglycoside druglevels, hypomagnesemia, and hypocalcemia. Themanagement of these disorders requires a highclinical suspicion to identify the offending drugand withdrawal of therapy with the agent to avoidfurther respiratory failure.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 321ACC-PULM-09-0302-021.indd 3217/10/09 10:21:01 PM


Table 6. Some Drugs That Cause Alveolar HypoventilationIsolated CoughNeuromuscular blockade• Aminoglycosides• Cocaine• Gamma-hydroxybutyrate(GHB)Motor neuropathies or myopathies• Amiodarone• Captopril• Corticosteroids• DiureticsBronchospasm• Opiates (eg,heroin)• Polymixins• Sedative/hypnoticdrugs• Isoniazid• Phenytoin• ProcainamideCough, which is one of the most commonmanifestations of drug-induced lung disease, is avagus nerve-mediated reflex caused by variouschemical and mechanical stimuli virtually anywherewithin the upper and lower respiratorytracts. Cough often occurs in patients with druginducedbronchospasm or drug-induced interstitiallung disease. However, angiotensin-convertingenzyme (ACE) inhibitors induce an isolated nonproductivecough without associated bronchospasmor parenchymal lung disease in nearly 10%of patients receiving any type of ACE inhibitor. 5Drug-induced bronchospasm is caused byvarious agents (Table 7). Patients generally presentwith wheezing, cough, and/or dyspnea. Spirometrytypically shows airways obstruction. The mechanismof drug-induced bronchospasm varies withthe particular agent. In patients with asthma, unlikehealthy individuals, -adrenergic blockers inducebronchospasm within minutes by mechanismsinvolving the inhibition of adrenergic bronchodilatortone. Virtually any route of administration (eg,oral, IV, or ophthalmic) can induce bronchospasm.Aspirin-induced bronchospasm is mediated byan enhanced 5-lipoxygenase pathway, resulting inthe production of bronchoconstricting cysteinylleukotrienesand, to a lesser extent, by a reductionin bronchodilating prostaglandins (eg, prostaglandinE 2). Dipyridamole causes bronchospasm byaugmenting the levels of adenosine, a bronchoconstrictingagent. Gold and penicillamine cause irreversibleairways obstruction due to concentricbronchiolitis obliterans.Pleural EffusionsAlthough a number of drugs can cause a pleuraleffusion (Table 8; see also Dr. Sahn’s chapter onpleural disease in this book), the number is muchless than those that involve the lung parenchyma. 6Pleural effusions occur with acute onset as part ofa hypersensitivity reaction after exposure to amiodarone,methotrexate, and nitrofurantoin, as wellas a systemic lupus erythematosus (SLE)-like reactionthat is described below. Anticoagulants caninduce an acute bloody effusion. Chronic pleuraleffusion may occur after long-term exposure todrugs that induce a delayed hypersensitivity-typeresponse (eg, methotrexate or procarbazine) or inassociation with the development of interstitialpulmonary inflammation/fibrosis (eg, busulfanand methotrexate). A variety of chemotherapeuticagents, including some of the newer antineoplasticagents, are associated with an increased risk ofpleural fluid formation. 7Table 7. Some Drugs That Cause BronchospasmChemotherapeuticAgents• IL-2• Methotrexate• Vinblastine• Vinca alkaloids plusmitomycinNonchemotherapeuticAgents• Aspirin and NSAIDs• -blockers• Contrast media• Corticosteroids• Dipyridamole• Gold• Nitrofurantoin (acute)• Opiates (cocaine, heroin)• Penicillamine• ProtamineTable 8. Some Drugs That Cause Pleural EffusionsChemotherapeuticAgentsBleomycinBusulfanEtoposideGemcitabineMethotrexateMitomycinProcarbazineTaxanesTyrosine kinase inhibitorsNonchemotherapeuticAgentsAmiodaroneAnticoagulantsBromocriptineDantroleneEsophageal sclerosing agentsMethysergideNitrofurantoin322 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3227/10/09 10:21:01 PM


<strong>Pulmonary</strong> Vascular DiseaseDrugs that affect the pulmonary vascular circulationby causing venous thromboembolism(VTE), pulmonary hypertension, vasculitis, orpulmonary venoocclusive disease are listed inTable 9. Oral contraceptives and other estrogencontainingagents can induce VTE. The VTE risksare relatively small in patients using second-generationand third-generation oral contraceptivescompared with nonusers (15, 30, and 5 per 100,000patients treated, respectively). 8 There is an importantsynergistic interaction between hypercoagulableconditions (eg, factor V Leiden deficiency)and the risks of VTE from hormonal replacementtherapy. 9 Thus, the risk from hormonal replacementin the absence of other VTE risk factorsappears too small to recommend against usingthese agents.Drug-induced pulmonary hypertension canoccur after exposure to a variety of drugs by mechanismsthat are unclear. A case-control study 10showed that appetite suppressants (eg, amphetamines,fenfluramine, and dexfenfluramine) areassociated with an increased risk of primary pulmonaryhypertension (OR: overall, 6.3; if the drugswere used for 3 months, 20). Althoughfenfluramine and dexfenfluramine have beenwithdrawn from the market worldwide, cliniciansshould remain vigilant for drug-induced pulmonaryhypertension because an estimated 5 millionadults in the United States continue to use appetitesuppressants. Oral contraceptives and otherestrogen-containing agents have also been associatedwith an increased risk of pulmonaryhypertension.An inflammatory vasculitis has been noted inconjunction with a generalized hypersensitivityreaction in patients exposed to busulfan, nitrofurantoin,or illicit drugs such as cocaine and heroin.However, it is uncertain whether the vascularresponse represents a primary or secondary effectof drug exposure. Leukotriene receptor antagonists(LTRAs), such as zafirlukast and montelukast,rarely (ie, 1 case per 15,000 to 20,000 patient-yearstreated) induce an idiosyncratic syndrome similarto Churg-Strauss vasculitis with peripheral eosinophilia,eosinophilic vasculitis, peripheral neuropathy,and cardiac failure. 11 Because most of thepatients were receiving either high-dose inhaledor systemic corticosteroids, a preexisting vasculitismay have accounted for their symptoms. However,a review of the literature through 2007 identified62 cases from 40 publications, 7 of whom wereentirely steroid naïve (eg, not receiving any formof inhaled or oral steroids when symptoms developed),suggesting that there may be a causal relationship.12 An accompanying editorial reviewingthis study as well as another case-control studyinvolving 78 patients concluded that Churg-Straussvasculitis can occur in association with exposureto LTRA in three settings: (1) use for worseningasthma, (2) use resulting in clinical improvementleading to a reduction in steroids and an unmaskingTable 9. Some Drugs That Cause <strong>Pulmonary</strong> Vascular DiseaseComplication Chemotherapeutic Agents Nonchemotherapeutic AgentsThromboembolic diseaseEstrogens/hormonal treatmentPhenytoinSteroids<strong>Pulmonary</strong> hypertension Mitomycin Aminorex (recalled)IL- 2AmphetaminesDexfenfluramine (recalled)Fenfluramine (recalled)L-tryptophan (recalled)Oral contraceptivesVasculitis Busulfan Cocaine/heroinNitrofurantoinZafirlukast/montelukastVeno-occlusive disease Bleomycin Oral contraceptivesBusulfanBCNUMitomycin<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 323ACC-PULM-09-0302-021.indd 3237/10/09 10:21:01 PM


of the vasculitis, and (3) use in steroid-naïvepatients. 13 <strong>Pulmonary</strong> venoocclusive disease is avery rare adverse effect reported after exposure tovarious agents (Table 9).Miscellaneous Drug-Induced <strong>Pulmonary</strong>ReactionsThere are a variety of less common druginducedadverse pulmonary effects (Table 10). SLEis associated with the use of a wide variety ofdrugs, although five agents are involved in 90%of the cases (Table 10). More recently, anti-tumornecrosis factor (TNF)-–targeted therapy (eg, etanercept,infliximab, and adalimumab) has also beenassociated with drug-induced SLE reactions. Fiveto twelve percent of all cases of SLE are caused bydrugs. It is unclear whether these agents cause orunmask a latent case of SLE. Hydralazine-inducedand isoniazid-induced SLE occur more frequentlyin patients who are slow acetylators of these drugs.Hydralazine-induced SLE occurs in up to 20% ofpatients receiving doses of 400 mg/d but can befound in patients receiving low-dose therapy.Procainamide-induced SLE is time related ratherthan dose related in that positivity for antinuclearantibodies (ANAs) develops in approximately50% of patients after 3 months of therapy and innearly all patients by 1 year. Notably, patients withdrug-induced SLE, unlike other patients with SLE,are usually negative for double-stranded DNAANA. Typically, pleurisy and dyspnea, in conjunctionwith systemic complaints of fever and arthralgias,develop in an insidious manner after usingthe drug for several months or even years. Chestradiograph abnormalities include pleural effusions,atelectasis, diffuse interstitial infiltrates, andalveolar infiltrates. The withdrawal of therapy withthe offending agent generally results in the promptresolution of symptoms within days, but occasionally,corticosteroids are required for symptomaticrelief.Alveolar hemorrhage and hemoptysis canoccur after exposure to certain drugs (Table 10).Drug-induced hemoptysis is most commonlycaused by pulmonary embolism with infarction.Penicillamine can cause a pulmonary-renal syndromesimilar to Goodpasture syndrome. Therapywith oral anticoagulants can induce spontaneouspulmonary hemorrhage days to years after theonset of therapy. Abciximab, a chimeric monoclonalantibody directed against platelet glycoproteinIIb/IIIa receptor that reduces restenosis afterangioplasty and stent placement, can cause severealveolar hemorrhage. 14 This relatively rarecomplication (7 of 2,533 patients [0.3%] vs 0 of 5,412control subjects) should be suspected in patientspresenting with hemoptysis, hypoxemic respiratoryTable 10. Miscellaneous Drug-Induced <strong>Pulmonary</strong> ReactionsConditionImplicated DrugsDrug-Induced SLE Hydralazine ProcainamideIsoniazidPenicillamineAnti-TNF--targeted therapyQuinidineAlveolar hemorrhageAbciximabBevacizumabAlemtuzumabAmiodaroneAnticoagulantsBusulfanCocaineAmphotericin BMitomycin CNitrofurantoinPenicillamineAlveolar proteinosisChurg-Strauss vasculitis Zafirlukast MontelukastMediastinal abnormalities Methotrexate PhenytoinLipomatosisCorticosteroidsMediastinitisEsophageal sclerosing agentsPanlobular emphysemaMethylphenidate<strong>Pulmonary</strong> calcification Antacids High-dose vitamin DCalciumPseudosepsis syndromeChronic salicylate intoxication324 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3247/10/09 10:21:01 PM


failure, diffuse alveolar infiltrates, and progressivedecline in hemoglobin levels within hours to 2 daysafter the first dose of abciximab. Bevacizumab, amonoclonal antibody against vascular endothelialgrowth factor that is used to treat a variety of cancers,including lung and colon cancer, can result infatal pulmonary hemorrhage. 15Mediastinal abnormalities can be a manifestationof an adverse drug effect (Table 10). Phenytoincan induce a pseudolymphoma syndrome thatis associated with peripheral adenopathy and,rarely, mediastinal adenopathy. Methotrexatemay cause transient hilar adenopathy during ahypersensitivity-type response. Typically, theadenopathy regresses 1 to 2 weeks after drug withdrawal.Mediastinal fullness caused by lipomatosisis an unusual manifestation of an adverse effect ofcorticosteroids. Although mediastinal widening inpatients receiving corticosteroids may raise thesuspicion of adenopathy, the diagnosis of lipomatosisis suggested by the characteristic chest radiographappearance of a straight mediastinal borderwithout the lumpy features of adenopathy and thetypical CT scan findings of lipid-containing tissue.Mediastinitis associated with fever and chest painis a relatively rare adverse effect of esophagealvariceal sclerotherapy.Other rare pulmonary adverse drug effectsreported in the literature include busulfan-inducedalveolar proteinosis, methylphenidate-inducedpanlobular emphysema, and pulmonary parenchymalcalcium deposition associated with hypercalcemicconditions or drugs such as antacids, calcium,and high-dose vitamin D. Long-term salicylateingestion can cause a pseudosepsis syndrome(reviewed in the section “AntiinflammatoryDrugs”).Approach to the Patient WithSuspected Drug-Induced Lung Diseasebased on the presenting clinical syndrome review edabove as well as by the presence of localized ordiffuse infiltrates (Table 11). In some instances, thepatient may have minimal or absent symptoms andfindings but demonstrate abnormalities on thechest radiograph or PFT results. A high level ofclinical suspicion is required because drug-induceddisease may occur years after initial exposure andthe radiographic abnormalities may be subtle.Diagnostic ApproachBecause drug-induced pulmonary toxicity hasa profound impact on patient management, it isimportant to establish a firm diagnosis as soon aspossible. This begins with an understanding of theclinical syndromes caused by various agents combinedwith the prudent use of noninvasive studiesand invasive procedures. Evaluation of the chestradiograph, PFT results and, on occasion, chest CTscan help narrow the differential diagnosis. Noninvasivestudies that may be useful include thefollowing: (1) an echocardiogram to assess cardiacfunction; (2) sputum studies to identify an infectiouspathogen (eg, Gram stain, culture, directfluorescent antibody for Pneumocystis carinii orLegionella, and acid-fast bacteria stain and culture);and (3) immunologic studies to exclude collagenvascular disorders and vasculitis.If a firm diagnosis is not established, then thejudicious use of invasive diagnostic procedures iswarranted. Typically, fiberoptic bronchoscopy withBAL and transbronchial lung biopsy is the nextstep in the evaluation of patients with localized ordiffuse pulmonary disease of unclear etiology.Table 11. Differential Diagnosis of Radiographic AbnormalitiesDiffuse DiseaseLocalized DiseaseDifferential DiagnosisAs mentioned previously, the diagnosis ofdrug-induced lung disease is one of exclusionbecause there are no pathognomonic criteria. Thedifferential diagnosis is broad, especially in cancerpatients receiving therapy with multiple drugs,often including immunosuppressive agents. However,the differential diagnosis can be narrowedInfectionMalignancyLymphangitic metastasis<strong>Pulmonary</strong> edema<strong>Pulmonary</strong> fibrosisRadiation pneumonitis/fibrosisLeukoagglutinin reactionARDSHypersensitivity pneumonitis<strong>Pulmonary</strong> hemorrhageDrug toxicityInfectionMalignancy<strong>Pulmonary</strong> emboliRadiation pneumonitisDrug toxicity<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 325ACC-PULM-09-0302-021.indd 3257/10/09 10:21:02 PM


Although this procedure has an excellent diagnosticyield for infections and malignant lesions(approximately 70 to 90%), it has a lower diagnosticyield for interstitial inflammatory lesions,including those caused by drug-induced toxicity.In the proper clinical setting, a nondiagnostic bronchoscopymay be sufficient to suggest the diagnosisof drug-induced injury. A thoracoscopic biopsyor, less commonly, an open-lung biopsy is recommendedin patients for whom the diagnosisremains unclear and the differential diagnosisincludes other pulmonary abnormalities, especiallyinterstitial lung disease.These procedures have the greatest diagnosticyield and a relatively low complication rate, evenin critically ill patients. The histopathologicchanges associated with drug toxicity are nonspecific,revealing elements of diffuse alveolar damage,fibrinous exudate, atypical or reactive alveolartype II cells, inflammatory cells, and fibrotic foci.One study 14 showed that up to 20% of patients withdiffuse infiltrates undergoing an open-lung biopsyhad histologic changes that could be attributed toa drug reaction.PFTsThe role of PFTs for detecting adverse druginducedpulmonary effects has been extensivelyinvestigated. The majority of the studies havefocused on bleomycin, busulfan, and amiodarone.The most common PFT result abnormalities arediminished lung volumes and a reduced Dlco.However, it is controversial whether these abnormalitiesaccurately predict that clinically overtdisease will develop. 17,18 Overt drug-induced diseasecan develop in patients in the absence of PFTresult abnormalities and, conversely, patients canhave abnormal PFT results in the absence of druginducedlung disease. Patient effort and anemiaare important confounding variables that affectPFT results. For most drugs, well-designed clinicalstudies validating the role of PFTs to detect adversepulmonary effects are not available. Until definitivedata are available, it is reasonable for clinicians tocontinue using PFTs in select settings with the goalof identifying subclinical disease as long as thephysiologic interpretation is made in the contextof the entire clinical situation.Chemotherapy-Associated <strong>Pulmonary</strong>ToxicityAzathioprine/MercaptopurineAzathioprine, a purine analog that inhibitsDNA synthesis, is an immunosuppressive agentused in the treatment of nonmalignant disease(eg, IPF and organ transplantation). Mercaptopurine,the active metabolite of azathioprine, is anantineoplastic agent. Azathioprine rarely ( 1%)causes pulmonary fibrosis, hypersensitivity-typereactions/PIE, or diffuse alveolar damage.BCNUCarmustine, or bis-chloronitrosourea (BCNU),is the best-known and extensively studied memberof the nitrosourea family of drugs that includeslomustine (chloroethylnitrosourea), semustine(methyl-chloroethylnitrosourea), and chlorozotocin.These cytotoxic agents, which are unique intheir ability to cross the blood-brain barrier, areactive against a variety of neoplasms, includingCNS malignancies. BCNU causes interstitial pulmonaryfibrosis and granulomatous inflammationthat can progress after drug withdrawal. Thepathogenesis of BCNU-induced pulmonary toxicityis unclear. BCNU promotes oxidant-inducedlung injury by inhibiting glutathione reductase,thereby reducing glutathione stores, which is animportant antioxidant defense. BCNU causes cytotoxicchanges characterized by alveolar type II cellhyperplasia and dysplasia, fibroblastic foci of proliferation,and interstitial fibrosis.Patients presenting with BCNU-induced pulmonarytoxicity will typically note an insidiousonset of a nonproductive cough and dyspnea associatedwith reticular nodular interstitial infiltrateson the chest radiograph. However, the onset ofsymptoms is highly variable, appearing withindays to as many as 17 years after beginning therapyBCNU. 19 Risk factors include the total dose, otheragents, and preexisting lung disease. In terms ofdose, the incidence of pulmonary toxicity withhigh-dose BCNU ( 1,500 mg/m 2 ) varies from 20to 50%, whereas with low-dose BCNU the incidenceis on the order of 1 to 5%, primarily whenadministered concurrently with other agents.Other agents such as cyclophosphamide and326 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3267/10/09 10:21:02 PM


adiation increase the risk of BCNU-induced lunginjury; however, a synergistic interaction has notbeen documented. In regard to preexisting lungdisease, patients with preexisting symptomaticpulmonary disorders, especially with a reducedvital capacity or Dlco, are typically excluded fromreceiving BCNU. A reduced Dlco without chestradiographic abnormalities occurs in patientsreceiving BCNU.Although there are no prospective studies toguide the timing of PFT monitoring, PFTs aretypically performed during BCNU therapy tomonitor for subclinical disease as well as intermittentlyfor the long term, given the extended latencyperiod between treatment and pulmonary toxicity.This approach is supported in part by the pooroutcomes of patients in whom BCNU-inducedpulmonary fibrosis develops. BCNU-induced lunginjury is associated with a mortality rate of nearly90%. Corticosteroid therapy has little impact inpreventing or treating BCNU-induced lung injury.The primary treatment is to withdraw BCNUpromptly and provide supportive care. Althoughthere is less information about the frequency ofpulmonary injury caused by the other nitrosoureas,each has been reported to have effects similar tothose of BCNU.BleomycinBleomycin, which is a cytotoxic antibiotic thatis isolated from Streptomyces verticillus, is very usefulin treating patients with head and neck carcinomas,germ-cell tumors, and Hodgkin andnon-Hodgkin lymphomas. Bleomycin accumulatesin the skin and lung, resulting in skin ulcerationsand pulmonary fibrosis. Bleomycin-induced pulmonarydamage, which is the major dose-limitingside effect, was first recognized in 1972. 20 The overallincidence of pulmonary toxicity is 10% (range,3 to 40%) and is fatal in 1 to 2% of patients.The pathogenesis of bleomycin-induced lunginjury has not been firmly established. 18 Becausebleomycin reproducibly causes interstitial pneumonitisand fibrosis in animal models, it hasbecome a paradigm for the study of the pathogenicmechanisms underlying pulmonary fibrosis. Earlybleomycin-induced pulmonary injury shows acuteand organizing diffuse alveolar damage that issimilar to the fibroproliferative phase of ARDS.Bleomycin binds to intracellular iron in alveolarepithelial and vascular endothelial cells and, in thepresence of oxygen, generates highly reactive oxygenspecies (ROS), such as hydroxyl radicals.ROS alter important cellular components (eg,DNA, lipids, and proteins), resulting in damage tothe respiratory membrane. Intratracheal or IVbleomycin increases cytokine levels within the lungin animal models. The important cytokines implicatedin the pathogenesis of bleomycin-inducedpulmonary toxicity include transforming growthfactor (TGF)-, TNF-, interleukin (IL)-1, and others.Pathogenic roles for ROS and certain cytokinesare supported by the protective effects of variousagents, including antioxidants, iron chelators, IL-1receptor antagonists, IL-12, keratinocyte growthfactor, CD36 peptides (binds TGF- 1), or antibodiesagainst TGF-, TNF-, or CD3 receptors. 18A crucial role for alveolar epithelial cell injuryis supported by the finding that transgenic micedeficient in the epithelial cell-restricted integrinv6 do not exhibit bleomycin-induced pulmonaryinflammation or fibrosis because they are unableto activate latent TGF-. 21 Bleomycin lung toxicitycan also be prevented by blocking excessive alveolarepithelial cell apoptosis or DNA damage, byenhancing DNA repair, and by augmenting fibrinolysis.22–25 Notably, the expression of the bleomycinresistantStreptoalloteichus hindustanus gene, whichencodes a protein that binds bleomycin and blocksDNA damage, can prevent bleomycin-inducedfibrosis. 24 Recently, increased expression of epithelialcell-derived integrin v6 was noted 26 in thedistal alveoli of patients with IPF and antibodiesagainst v6 blocked bleomycin-induced fibrosisin mice, even when administered after the developmentof fibrosis. A disease-specific gene expression-profilingstudy in which the authors usedmultiple murine models of lung disease identified12 unique transcripts characteristic of bleomycininducedpulmonary fibrosis that were not evidentin the other models. 27 Although the pathogenicmechanisms underlying bleomycin toxicity arebetter understood and lead to potentially novelantifibrotic therapies, it is unclear why certainpatients are predisposed to pulmonary fibrosis.There are three major clinical manifestations ofbleomycin-induced lung toxicity: (1) chronic interstitialfibrosis, (2) hypersensitivity-type disease,and (3) COP. Interstitial fibrosis is by far the most<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 327ACC-PULM-09-0302-021.indd 3277/10/09 10:21:02 PM


common, occurring in approximately 11% of thosepatients treated. Patients typically present with asubacute onset of a nonproductive cough anddyspnea within a few weeks to 6 months aftertreatment. 15 Pleuritic substernal chest pain occursin approximately 3% of patients. However, nearly20% of patients with bleomycin-induced lunginjury will be asymptomatic. Physical examinationfindings may show tachypnea, bibasilar crackles,and hyperpigmented skin lesions. Chest radiographstypically reveal basilar, predominantreticular, or fine nodular infiltrates that oftenoriginate from the costophrenic angles. Other lesscommon radiographic patterns include patchyalveolar infiltrates, lobar consolidation, and lungnodules. PFT results generally show reduced lungvolumes and Dlco. Patients with bleomycininducedhypersensitivity-type disease may havesimilar symptoms but often with an associatedfever, malaise, and chest radiographic changes ofchronic pneumonitis and ground glass-appearingparenchyma with little fibrosis. The radiographicand pathologic distinctions between fibrosis andpneumonitis are not always clear.Risk factors for the development of bleomycininducedpulmonary toxicity include the following.(1) Total dose: a dose relationship has beendescribed with an incidence of 3 to 5% when thetotal dose is 300 U but 20% when the total doseis 500 U 18 ; however, pulmonary toxicity can occurafter receiving only 20 U. (2) Oxygen: bleomycin,more than any other drug, has a well-known synergistictoxic interaction with oxygen that can occuryears after bleomycin exposure. Although thisreaction does not develop in all patients, theinspired oxygen concentration for patients whohave ever received bleomycin should be kept 25%, if possible. (3) Radiation: radiation therapy,even years after bleomycin exposure, increases therisk. (4) Age: persons 70 years of age are at anincreased risk of lung injury. (5) Abnormal renalfunction: patients with reduced renal function aremore susceptible to bleomycin-induced pulmonarytoxicity because of the fact that the kidneys primarilyexcrete bleomycin. (6) Concurrent use of othercytotoxic agents: some drugs may increase therisk of bleomycin-related injury, including cyclophosphamide,doxorubicin, granulocyte colonystimulatingfactor (G-CSF), methotrexate, andvincristine.The management of patients with bleomycininducedpulmonary toxicity centers on withdrawingthe drug and limiting exposure to exogenous oxygen,as noted previously. Although the overallmortality rate is approximately 1 to 2%, the mortalityrate varies from 10 to 83% in patients withpulmonary disease. Corticosteroids (60 to 100 mg/d)are generally administered to all patients withclinically significant toxicity and then slowlytapered according to the patient’s clinical response.Clinical improvement typically occurs withinweeks and may take 2 years to completely resolve.Select patients will be left with residual radiographicand abnormalities on the PFT results.BusulfanBusulfan, an alkylating agent that is used in themanagement of chronic myeloproliferative disorders,was the first chemotherapeutic agent implicatedin causing chronic pneumonitis/pulmonaryfibrosis. In general, alkylating agents (eg, busulfan,cyclophosphamide, chlorambucil, and melphalan)have a lower frequency of inducing lung injurythan other cytotoxic agents. Like bleomycin, theseagents can induce synergistic pulmonary damagewhen a patient is exposed to oxygen, radiation, orother cytotoxic chemotherapeutic agents. The incidenceof symptomatic busulfan-induced pulmonaryfibrosis is approximately 4 to 5%. However,histologically evident fibrosis and cytotoxicchanges (hyperplastic and dysplastic type II cells)occur in up to 46% of busulfan-treated patients,most of whom are asymptomatic. 28 A thresholddose has not been established.The clinical presentation is an insidious onset,often 3 years after initiating therapy, of cough,dyspnea, fever, malaise, and weight loss. The PFTresults show restrictive lung volumes, a reducedDlco, and hypoxemia. The chest radiograph typicallyshows diffuse interstitial and alveolar infiltrateswith a basilar predominance. Occasionally,pleural effusions, nodular densities, or normalchest radiograph findings are noted. Managementcenters on drug withdrawal and the administrationof corticosteroids. The prognosis is poor, with themortality rate ranging from 50 to 80%. Alveolarproteinosis has also been reported after exposureto busulfan. Unlike primary alveolar proteinosis,it does not respond to therapeutic BAL.328 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3287/10/09 10:21:02 PM


CyclophosphamideCyclophosphamide, an alkylating agent thatcauses pulmonary toxicity less frequently thanbusulfan, is widely used to treat malignant diseases(eg, lymphomas, breast carcinoma, and ovariancarcinoma) and nonmalignant diseases (eg, collagenvascular disorders, IPF, and Wegener granulomatosis).The incidence of adverse pulmonaryeffects is 1%. The extensive clinical indicationsfor cyclophosphamide and its frequent use in conjunctionwith other cytotoxic drugs increases thelikelihood that lung toxicity will be seen by pulmonologists.Although cyclophosphamide aloneinduces pulmonary toxicity in humans relativelyinfrequently, 29 it clearly does so in animals. 30 Thepathogenesis of pulmonary toxicity has not beenestablished but likely involves oxidant-mediatedmechanisms. Cyclophosphamide is metabolizedinto two active agents, phosphoramide mustardand acrolein, both of which reduce hepatic glutathionestores.Chronic pneumonitis and/or fibrosis are themost common clinical manifestations of pulmonarytoxicity. Symptoms develop as soon as2 weeks to as long as 13 years after the initiationof therapy, without any clear dose relationship.Patients present with cough, dyspnea, fever, andbibasilar interstitial infiltrates seen on chest radiographs.Synergistic toxicity has been described inpatients receiving radiation therapy or other cytotoxicagents. 31 Cyclophosphamide-induced pulmonaryfibrosis, unlike IPF, may present with bilateralpleural thickening typically without clubbing and“Velcro” crackles. 29 The prognosis is generally poor,with a mortality rate approaching 50%. Mostpatients are treated with corticosteroids based onanecdotal reports of benefit.MethotrexateMethotrexate is a folic acid antagonist that isused in the management of malignant and nonmalignantinflammatory diseases. The incidenceof pulmonary toxicity is nearly 7% for high-dosemethotrexate, as used to treat malignancies, but2 to 3% with low-dose methotrexate regimensused to treat chronic inflammatory conditionssuch as RA. Methotrexate-induced pulmonarytoxicity does not have a clear dose relationshipbecause it occurs over a broad range of doses (40to 6,500 mg).Clinical manifestations include the following:(1) hypersensitivity-type disease, the most commonform; (2) chronic pneumonitis/fibrosis;(3) COP; (4) acute chest pain; (5) noncardiogenicpulmonary edema; (6) acute pleurisy/pleural effusions;and (7) bronchospasm. 32 Methotrexateinducedhypersensitivity-type reactions typicallyoccur within 10 days to 4 months after initiatingtherapy but can occur up to 1 month after therapywith methotrexate ceases. 32 Patients may complainof fever, cough, dyspnea, arthralgias and, less commonly,skin rash (approximately 17%). Chest radiographsreveal diffuse interstitial infiltrates withground-glass changes seen on HRCT scans. Otherradiographic abnormalities that are occasionallyseen include nodular infiltrates, hilar/ mediastinaladenopathy, and pleural effusions. Blood eosinophiliais present in nearly 40% of patients. BALfluid findings typically reveal a predominance oflymphocytes, especially T- suppressor cells, whichare characteristic of a hypersensitivity reaction. 33The diagnosis of methotrexate-induced pulmonarytoxicity requires the following three majorcriteria: (1) hypersensitivity pneumonitis by histopathology,(2) radiographic evidence of interstitialand/or alveolar infiltrates, and (3) blood cultures(if febrile) and sputum cultures (if available) thatare negative for microbes. Patients must also havethree of the following five minor criteria: (1) dyspneaof 8 weeks in duration, (2) nonproductivecough, (3) room air oxygen saturation of 90%,(4) Dlco of 70% of predicted, and (5) leukocytecount of 15,000 cells/L. 34Risk factors for the development of methotrexateinducedpulmonary toxicity include the following:(1) appearance of symptoms within the first 32weeks of therapy, (2) multidrug regimens (synergywith cyclophosphamide has been reported), (3)diabetes mellitus, (4) age 50 years, (5) rheumatoidpleuropulmonary disease, and (6) hypoalbuminemia.35 Dose, frequency, smoking status,previous lung disease, and route of administrationroute (eg, oral, IV, intrathecal, and IM) do not consistentlyaffect the frequency of methotrexateinducedadverse pulmonary effects. Notably, PFTresults have not been helpful in identifying patientswith RA who are at risk for the development ofpulmonary toxicity while receiving long-term<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 329ACC-PULM-09-0302-021.indd 3297/10/09 10:21:03 PM


low-dose methotrexate. 36 Chronic pneumonitis/fibrosis, which occurs less commonly as the resultof methotrexate than other chemotherapeuticagents, typically presents 4 months to 3 yearsafter the initiation of therapy. Management centerson the withdrawal of therapy with the drug andthe administration of corticosteroids. Chronicfibrosis will develop in approximately 7% ofpatients with methotrexate-induced hypersensitivityreactions, and 8% will die of progressive respiratoryfailure.MitomycinMitomycin is an alkylating cytotoxic antibioticthat is used in the treatment of breast, GI, gynecologic,and lung carcinomas. The incidence of pulmonarycytotoxicity is approximately 5% (range,3 to 39%). 37 However, the full scope is difficult toassess because mitomycin is typically administeredconcurrently with other agents and is also a cotoxinwith oxygen and radiation. The presenting featuresas well as the radiographic and physiologicchanges are similar to those seen with bleomycininducedinterstitial pneumonitis/fibrosis. Of note,this reaction seems to occur most frequently afterthe third cycle of treatment. Although serial monitoringof the Dlco to detect clinically occult diseaseis unproven, it is generally recommended andwidely used for this purpose. Prednisone can rapidlyresolve the symptoms and interstitial infiltrates.A unique reaction of mitomycin is theinduction of microangiopathic hemolytic anemiaconcurrently with noncardiogenic pulmonaryedema and renal failure. The mortality rate associatedwith this uncommon reaction is 90%.The combination of mitomycin with vinca alkaloids(eg, vinblastine and vincristine) induces anacute onset ( 3 h after receiving the vinca alkaloid)of bronchospasm that is associated with focal or diffuseinterstitial infiltrates seen on chest radiographsand hypoxemia. 38 Vinca alkaloids alone do not causethis adverse side effect. Chronic respiratory symptomsthat respond to corticosteroids develop inapproximately two thirds of the patients.Retinoic AcidAll-trans-retinoic acid (ATRA) is a highly effectivebiological response modifier that inducesclinical remission in patients with acute promyelocyticleukemia by augmenting the differentiationof malignant stem cells into mature neutrophils.An ARDS-like picture, termed retinoic acid syndrome,developed in nearly 25% of patients whowere initially treated with ATRA. Prednisolonetherapy (75 mg/d) reduces the incidence of ATRAto approximately 8%. 39 The syndrome is characterizedby the sudden onset of fever, dyspnea, effusions(pleural and pericardial), diffuse alveolarinfiltrates seen on chest radiographs, and hypoxemicrespiratory failure. Before the efficacy ofsteroid therapy was recognized, more than half ofthe patients required mechanical ventilation, andthe mortality rate was 33%.The pathogenesis has not been established butlikely relates partly to the rapid increase in matureleukocytes. Autopsy studies 39 reveal diffuse alveolardamage with edema and hemorrhage as wellas leukemic cells in the interstitium. Mechanismsimplicated in ATRA-induced capillary leak syndromeinclude the release of vasoactive cytokines,oxidants, and lipid mediators from inflammatorycells, as well as enhanced expression of leukocyteadhesion molecules that impair leukocyte transitthrough the pulmonary microcirculation. Somepatients in whom retinoic acid syndrome developswhile they are receiving prednisolone haveresponded to therapy with dexamethasone. Managementis otherwise largely supportive, similarto patients with ARDS.Other Chemotherapeutic Agents/NewerAntineoplastic DrugsOther chemotherapeutic drugs are less commonlyassociated with adverse pulmonary effects,which are seen more often than with the agentsreviewed earlier. Chlorambucil, which is an alkylatingagent similar to cyclophosphamide, is used inthe treatment of lymphoproliferative disorders aswell as nonmalignant diseases. <strong>Pulmonary</strong> toxicityis relatively rare, but chronic pneumonitis/fibrosishas been reported. Cytosine arabinoside, which isan antimetabolite used to treat acute leukemia,causes noncardiogenic pulmonary edema in 13 to20% of patients. 40 The mortality rate varies from2 to 50%. Corticosteroids may improve outcome,but this finding remains uncertain. Fludarabine, anantimetabolite that is used in the management of330 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3307/10/09 10:21:03 PM


patients with chronic lymphoproliferative disorders,can cause chronic pneumonitis/fibrosis aswell as a hypersensitivity-type reaction in nearly9% of patients that is often responsive to therapywith corticosteroids. 41Although older chemotherapeutic agentsinduce a wide array of pulmonary toxicity in nearly10% of patients, there is less information concerningthe frequency as well as the clinical manifestationsresulting from exposure to the neweragents. 7,42 Gemcitabine, an agent that is increasinglybeing used in the management of a variety of solidtumors, can induce an ARDS-like condition that ispotentially fatal. 42 The estimated frequency ofgemcitabine-induced pulmonary toxicity rangesfrom 1 to 1.4% and, in addition to ARDS, includesnonspecific interstitial pneumonitis (NSIP), pulmonaryfibrosis, and pleural effusions. 7,42Taxanes (eg, paclitaxel and docetaxel) are usedin the management of solid tumors (eg, breast,lung, gastric, and ovarian). Paclitaxel inducesbronchospasm and a type 1 hypersensitivity reactionin up to 30% of patients. 7,42 Other much lesscommon taxane-induced pulmonary manifestationsinclude NSIP, hypersensitivity pneumonitis,pulmonary fibrosis, ARDS, and pleural effusion.Topoisomerase I inhibitors (eg, irinotecan andtopotecan), which are used in the management ofvarious solid tumors, cause pulmonary toxicitythat includes NSIP and bronchiolitis obliterans. 7,42Synergy between irinotecan and paclitaxel orradiation increases the frequency of pulmonarytoxicity from 1.8 to 13% or 56%, respectively. Tyrosinekinase inhibitors (eg, gefitinib and imatinib)uncommonly cause pulmonary toxicity (approximately2%), consisting of NSIP, hypersensitivitypneumonitis, pulmonary fibrosis, COP, alveolarhemorrhage, ARDS, and pleural effusions. 7,42Biological response modifiers, such as IL-2and TNF-, are used in experimental protocols totreat various malignancies. Each can induce noncardiogenicpulmonary edema which, in the caseof IL-2, is also associated with massive fluid retention.Granulocyte-macrophage CSF and G-CSFcan cause a hypersensitivity-type pneumonitiswhen administered in conjunction with othercytotoxic agents. Also, G-CSF, especially in thepresence of other cytotoxic agents, can induceARDS. 43 Although it is beyond the scope of thisreview, immunosuppressive agents can promotethe development of opportunistic pulmonaryinfections.Nonchemotherapy-Associated<strong>Pulmonary</strong> ToxicityAntiinflammatory DrugsAspirin: Aspirin (acetylsalicylic acid), which isthe most commonly prescribed drug worldwide,causes the following two forms of pulmonarytoxicity: bronchospasm and noncardiogenic pulmonaryedema. Aspirin-induced asthma (AIA)occurs in 1% of healthy individuals and in upto 20% of asthmatic individuals. However, aspirinsensitivity occurs in a larger proportion of asthmaticpatients with nasal polyps, which is a clinicaltriad that was first identified by Samter andBeers. 44Symptoms of AIA occur within minutes tohours after ingestion and may be associated withfacial flushing, rhinorrhea, angioedema, and conjunctivitis.The pathogenesis of AIA is mediatedby the enhanced production of cysteinyl leukotrienesvia the 5-lipoxygenase pathway and, to alesser extent, by a reduction in bronchodilatingprostaglandins secondary to cyclooxygenase inhibition.Aspirin-sensitive asthmatic patients whoare challenged with aspirin have increased levelsof urinary leukotriene E 4, which is a general markerof serum leukotriene levels, because of theincreased activity of leukotriene C synthase, therate-limiting step in leukotriene synthesis. 45 Asexpected, they exhibit symptomatic and spirometricimprovement in the presence of 5-lipoxygenaseinhibition and cysteinyl leukotriene receptorantagonism. 45 An increased risk of AIA is alsoassociated with several single nucleotide polymorphismsin the promoters of prostaglandin-E 2receptorsubtype 2, cysteinyl leukotriene receptor 1,cysteinyl leukotriene receptor 2, and T-cell T box. 46A similar syndrome occurs with other nonsteroidalantiinflammatory drugs (NSAIDs). Agents thatdo not block the cyclooxygenase pathway, suchas acetaminophen and salicylate, can be safelyused in patients with aspirin-induced or NSAIDinducedbronchospasm. 45Aspirin-induced noncardiogenic pulmonaryedema occurs in 10 to 15% of patients with a severesalicylate overdose. This can occur inadvertently<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 331ACC-PULM-09-0302-021.indd 3317/10/09 10:21:03 PM


in long-term aspirin users, typically elderly patientswith multiple medical problems, or intentionallyin individuals attempting suicide. Thesepatients will often present with dyspnea, tachypnea,altered mental status, and a chest radiographrevealing diffuse alveolar infiltrates. Most patientswill manifest with a simple respiratory alkalosisor a mixed anion gap metabolic acidosis plus arespiratory alkalosis. The constellation of findingscan mimic a pseudo-sepsis syndrome. 47 The diagnosisis made by first considering aspirin ingestionin the differential diagnosis of patients presentingwith noncardiogenic pulmonary edema or a sepsislikepicture and then determining the serumsalicylate level. Respiratory alkalosis is typicallyseen with serum salicylate levels of 35 mg/dL(therapeutic range, 10 to 20 mg/dL), whereas noncardiogenicpulmonary edema occurs at serumlevels of 45 mg/dL. Although the pathogenesisof aspirin-induced noncardiogenic pulmonaryedema is unclear, it likely results from increasedcapillary permeability resulting from toxic levelsof salicylates. Management is based on drug withdrawaland supportive care, often in the ICU withthe patient receiving mechanical ventilation. Alkalinediuresis should be performed as soon as thediagnosis is suspected to enhance renal clearanceand thereby reduce the serum salicylate level.Hemodialysis is reserved for patients with aspirinassociatedseizures, refractory acidosis, coma, orvery high salicylate levels (80 to 100 mg/dL). Outcomeis generally favorable in young patients withan acute salicylate overdose. However, there is ahigh mortality rate in older patients who havemultiple medical problems and a long-term historyof the ingestion of salicylates, and they have a moresubtle presentation.Gold: Gold has been widely used for decadesin the management of chronic inflammatoryconditions, most notably RA but also pemphigus,psoriatic arthritis, bronchial asthma, andankylosing spondylitis. Both the oral (auranofin)and the IM (gold sodium thiomalate) preparationscan induce chronic pneumonitis/interstitialfibrosis. Bronchiolitis obliterans, in the presenceor absence of organizing pneumonia, occursless frequently. Interstitial lung disease causedby gold therapy can be distinguished from thepatient’s underlying rheumatoid lung disease bythe following: (1) an acute onset of dyspnea anda nonproductive cough; (2) fever (ie, temperature 38C); (3) skin rash; (4) crackles on chestexamination; (5) absence of finger clubbing orsubcutaneous nodules; (6) the presence of bloodeosinophilia, proteinuria, or liver dysfunction; (7)BAL fluid lymphocytosis with a predominanceof CD8 (suppressor) cells typical of a hypersensitivity-typereaction; (8) alveolar opacities in anonbasilar distribution; and (9) an HRCT scanthat shows bronchovascular infiltrates, ratherthan peripheral infiltrates. 48 A favorable clinicalresponse typically occurs after drug withdrawaland a course of corticosteroids. Recurrent diseasehas been documented following retreatment,which is not warranted.NSAIDs: NSAIDs are among the most commonlyprescribed drugs because they are widelyused in the management of an array of rheumatologicdisorders as well as for minor musculoskeletalpain in healthy individuals. Similar to salicylates,NSAIDs can precipitate asthma as well as noncardiogenicpulmonary edema in aspirin-sensitiveindividuals. Ophthalmic NSAIDs also can triggerbronchospasm. Nearly every type of NSAID caninduce a hypersensitivity-type reaction that maybe associated with PIE. In general, prompt resolutionoccurs with drug withdrawal, and rarely, acourse of corticosteroids is required.Methotrexate: See the section “Chemotherapy-Associated <strong>Pulmonary</strong> Toxicity ” for a discussion.Penicillamine: Penicillamine is an antiinflammatory,antifibrotic, and copper-chelating agentthat is used in the treatment of RA, scleroderma,primary biliary cirrhosis, and Wilson disease. Themanifestations of penicillamine-induced pulmonarytoxicity include the following: (1) interstitialpneumonitis/fibrosis; (2) bronchiolitis obliterans,in the presence or absence of organizing pneumonia;(3) drug-induced SLE; and (4) alveolar hemorrhagedue to a pulmonary-renal syndrome.Patients with penicillamine-induced interstitialpneumonitis/fibrosis generally present with theinsidious onset of dyspnea and cough associatedwith restricted lung volumes and a reduced Dlco.A subset of patients present with a hypersensitivity-typereaction associated with peripheral bloodeosinophilia and increased serum IgE levels. Aswith gold therapy, it can be challenging to distinguishdrug-induced pulmonary toxicity from thepatient’s underlying collagen vascular disorder.332 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3327/10/09 10:21:04 PM


The incidence of penicillamine-induced bronchiolitisobliterans is unknown but is estimated tooccur in 1% of patients with RA receiving thistherapy. There are no clear risk factors. Patientspresent with a subacute onset of dyspnea, cough,and wheeze. The chest radiograph reveals hyperinflationin the absence of infiltrates, whereas PFTresults are used to confirm the presence of increasedlung volumes as well as airflow limitation withouta bronchodilator effect. Lung biopsy specimensfrom these patients reveal bronchiolar constrictioncaused by mononuclear inflammation and fibrosis.Management centers on drug withdrawal, supportivetherapy, and the consideration of a trial ofcorticosteroids, azathioprine, or cyclophosphamide.However, there are no data demonstratingthe value of immunosuppressive therapy. Theprognosis for patients with penicillamine-inducedbronchiolitis obliterans is poor; nearly 50% ofpatients die, whereas the remainder have very poorquality of life due to a severe permanent residualobstructive impairment.Penicillamine rarely induces Goodpasturesyndrome, which consists of diffuse alveolar hemorrhageand rapidly progressive glomerulonephritiscaused by autoantibodies directed againstcomponents of the alveolar and glomerular basementmembranes. There are case reports of thissyndrome occurring after penicillamine was usedto treat RA as well as Wilson disease, demonstratingthat it is not simply attributable to the patient’sunderlying collagen vascular disease. Typically,the serum ANA level is increased, but the doublestrandedDNA antibodies seen in classic SLE arenot apparent. However, it is important to considerSLE or an antineutrophil cytoplasmic antibodyrelatedvasculitis in the differential diagnosisbecause a number of these patients will not haveserum and tissue antiglomerular basementmembrane antibodies. Despite drug withdrawaland corticosteroids, the mortality rate is nearly50%, and there is a high incidence of progressionto chronic renal failure necessitating long-termdialysis.TNF-–Targeted Therapy: Anti–TNF- antibodies(eg, etanercept, infliximab, and adalimumab)are increasingly being used to block theeffects of TNF- in autoimmune diseases suchas RA and Crohn disease. Because TNF- has acritical role in host defense against infections, notsurprisingly its utility is complicated by pulmonaryinfections, most prominently the reactivationof tuberculosis as well as fungal infections. 49,50It is recommended that all patients receiving anti-TNF--targeted therapy be screened and treatedfor latent tuberculosis before the drug is used.Further, these agents, especially etanercept andinfliximab, can also cause various noninfectiouspulmonary manifestations such as NSIP, pulmonaryfibrosis, loosely formed granulomatousinflammation, pleural effusi ons, increased size ofRA nodules, and SLE-like reactions. 50,51Antimicrobial DrugsAntibiotic-Induced Hypersensitivity-Type Disease:The most common antibiotics causing a hypersensitivity-typelung reaction in the presence of PIEinclude -lactam and sulfa antibiotics. Less commonly,fluoroquinolones, tetracycline, erythromycin,nitrofurantoin, isoniazid, paraaminosalicylicacid, and ethambutol cause this type of a reaction.There are no known risk factors, and most reactionsare idiosyncratic. Typically, the patients presentduring a period of 1 to 4 weeks (can be delayed upto 1 year) with minimal, nonspecific respiratorycomplaints. Prompt resolution occurs with drugwithdrawal perhaps with a short course of corticosteroidsand antihistamines.Nitrofurantoin: Nitrofurantoin has been usedfor decades in the management of acute urinarytract infections and as long-term suppressivetherapy for patients with asymptomatic bacteruria.Nitrofurantoin has the following twomain distinct adverse pulmonary manifestations:(1) acute hypersensitivity-type reaction and(2) chronic pneumonitis/fibrosis that mimics IPF.Alveolar hemorrhage and noncardiogenic pulmonaryedema have also been described but are lesscommon. The incidence of nitrofurantoin-inducedadverse pulmonary effects is not firmly established.One study 52 has shown that the nitrofurantoin-inducedacute hypersensitivity-type reactionthat was severe enough to warrant hospitalizationoccurred in 1 in 5,000 new drug administrations,whereas the chronic form occurred in 1 in 50,000new drug administrations.Acute nitrofurantoin pulmonary toxicity is avery rare ( 0.1%) side effect that occurs within1 month of the first dose in 86% of patients.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 333ACC-PULM-09-0302-021.indd 3337/10/09 10:21:04 PM


Symptoms consist of dyspnea, cough, fever, chestpain, and a macular/papular skin rash . There isan elevated erythrocyte sedimentation rate (ESR)and peripheral blood eosinophilia in most patients.The chest radiograph shows a mixed alveolar/interstitial infiltrative pattern but is normal in 18%of patients. One third of patients have a smallpleural effusion. PFT results typically reveal arestrictive pattern with a reduced Dlco. Althoughthe diagnosis is generally made clinically, a lungbiopsy specimen and BAL fluid should be obtainedin any patient in whom infection, metastatic tumor,and/or an inflammatory condition are in the differentialdiagnosis. Histopathologic findingsinclude interstitial inflammation with lymphocytesand plasma cells and, in the more severe cases,alveolar edema, hyaline membranes, and hemorrhage.The prognosis is generally favorable withdrug withdrawal and, occasionally, therapy withcorticosteroids. Although ARDS develops in somesituations, the overall mortality rate is 1%. 53Chronic nitrofurantoin pulmonary toxicityoccurs in elderly patients who have been treatedwith long-term suppressive therapy for asymptomaticbacteruria. Although much less commonthan the acute form, it is similar to the acute formin that patients present with dyspnea and cough,but, in contrast, fever and rash are uncommon.Unlike the acute form, patients with the chronicform of the disease have systemic complaints offatigue and weight loss. Infrequently, they have anincreased ESR and peripheral blood eosinophilia.Low-level elevations of the serum ANA and rheumatoidfactor can be seen that are similar to thosein patients with IPF. The chest radiograph showsa bilateral interstitial infiltrative pattern typicallywithout pleural effusions. PFT results generallyreveal a restrictive pattern with a reduced Dlco.As with the acute form, the diagnosis is usuallymade on clinical grounds. However, lung biopsyand BAL fluid are indicated in any patient whomight have an infection, a metastatic tumor, and/or an inflammatory condition. The characteristichistopathologic findings include lymphocyticinterstitial inflammation along with interstitial andalveolar fibrosis that can be indistinguishable fromthe usual interstitial pneumonitis form of IPF.Unlike the generally reversible acute form of thedisease, three fourth of patients with chronic nitrofurantoinpulmonary toxicity do not improve at allor are left with substantial residual disease despitedrug withdrawal and a trial of corticosteroids. Theoverall mortality rate is nearly 8%. 53Sulfasalazine/Mesalamine: Sulfasalazine, whichis an antibiotic used in the management of inflammatorybowel disease, can cause COP, PIE, pulmonaryfibrosis, and bronchospasm. Sulfapyridine,which is the carrier component of sulfasalazine,is responsible for the majority of the side effects.However, mesalamine, the clinically active component,can also induce pulmonary adverse sideeffects.Pentamidine: Pentamidine can cause bronchospasmwhen administered IV or via nebulizer.The adverse effect can be prevented by pretreatmentwith -agonists or ipratropium.Cardiovascular DrugsAmiodarone: Amiodarone is an effective agentfor treating ventricular and supraventriculararrhythmias that are refractory to other drugs.However, the utility of this drug is limited byadverse pulmonary effects that can occur in 5 to10% of patients, as well as toxicity to the eyes, thyroidgland, liver, GI tract, and nervous system. Theadverse effects of amiodarone are typically associatedwith a high daily dose ( 400 mg/d) and aprolonged duration of therapy ( 12 months). Ametaanalysis 54 of randomized, placebo-controlled,low-dose amiodarone efficacy trials involving1,465 patients receiving 400 mg/d amiodaronefor a minimum of 12 months demonstrated a trendtoward increased pulmonary toxicity comparedwith those patients receiving placebo (OR, 2.0; 95%confidence interval, 0.9 to 5.3; p 0.07), but thisdifference did not reach statistical significance. Incontrast, compared with placebo, the use of lowdoseamiodarone did increase pulmonary toxicity(1.2% vs 3.8%, respectively) in the Canadian MyocardialInfarction Amiodarone Trial. 55 An analysisof 237 patients with amiodarone-induced pulmonarytoxicity demonstrated that only patient age( 60 years; OR, 3; 95% confidence interval, 1.3 to7) and duration of therapy ( 6 at 12 months; OR,18; 95% confidence interval, 6 to 52) significantlyimpacted the risk. 56The mechanisms underlying amiodaroneinducedpulmonary toxicity are not firmly established.As reviewed elsewhere, 57 several theories334 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3347/10/09 10:21:04 PM


that are not mutually exclusive have been implicated,including the following: (1) direct cellulardamage in part due to alveolar epithelial-cell apoptosis,(2) enhanced phospholipidosis, (3) immunologicinjury, (4) ROS-induced damage, (5) alterationsin membrane properties, (6) increased intracellularcalcium in vascular endothelial cells, (7) activationof G proteins, and (8) inhibition of pulmonaryepithelial cell sodium-potassium adenosine triphosphatase.More recent studies show that amiodaronecauses lung epithelial-cell mitochondrialdysfunction, ROS production, and mitochondriaregulatedapoptosis that can be blocked with theadministration of angiotensin system antagonists.58,59 Amiodarone has a large volume of distributionwith a very long half-life of 30 to 60 days.Thus, the antiarrhythmic effects as well as theadverse effects of amiodarone will persist forweeks to months after the drug is withdrawn fromtherapy. Amiodarone is an iodine-containing phospholipaseinhibitor that causes lipid accumulationin nearly all tissues, especially the lungs, skin, andliver. This blockade results in the accumulation ofundigested surfactant phospholipids in the lung,a feature that is seen in virtually all patients receivingthe drug.The histologic features of amiodarone-inducedpulmonary toxicity include the following: (1) accumulationof foamy macrophages with characteristiclamellated inclusions in the interstitium andalveolar spaces; (2) hyperplasia of alveolar type IIcells; and (3) widening of the alveolar septa withinfiltration of lymphocytes, plasma cells, eosinophils,and neutrophils. Edema, intraalveolar fibrinexudation, lamellar inclusions in pulmonary endothelialcells, bronchiolar epithelial cells and type IIcells, and fibrosis can also occur.The clinical and laboratory manifestations ofamiodarone-induced pulmonary toxicity includethe following: (1) interstitial pneumonitis/fibrosis,(2) ARDS, (3) COP, (4) mass lesions that cancavitate, (5) eosinophilic pneumonia, (6) diffusealveolar hemorrhage, and (7) pleural effusions.Uncommon reactions include hypersensitivitypneumonitis, alveolar hypoventilation, and bronchospasm.The patients typically present with aninsidious onset of a nonproductive cough anddyspnea with radiographic evidence of asymmetricchronic pneumonitis/fibrosis and an elevatedESR. Occasionally, ill-defined alveolar infiltrates,a lung mass, or pleural effusions may be seen.Risk factors for amiodarone-induced adversepulmonary side effects are not firmly established,but some that have been implicated include thefollowing: (1) maintenance dose of 400 mg/d,(2) age 60 years, (3) duration 6 to 12 months,(4) angiography/acute lung injury, and (5) cardiothoracicsurgery/ARDS. The total cumulative doseor serum levels of amiodarone are not useful. Further,a reduction in the Dlco and increased galliumuptake may support a clinical diagnosis, but theyare not reliable predictors of pulmonary toxicity inthe absence of clinical and radiographic abnormalities.The diagnosis of amiodarone-induced pulmonarytoxicity is one of exclusion, especially heartfailure, pulmonary embolism, infection, and otherinflammatory interstitial diseases (eg, COP andchronic eosinophilic pneumonia). Chest CT scansmay reveal high-attenuation areas caused bythe iodine present in amiodarone. However, thisfinding is a nonspecific one that may be present inhealthy subjects. The typical PFT abnormalitiesinclude a reduction in the Dlco and reduced lungvolumes. Although a decreased Dlco is a nonspecificfinding that is seen in all patients withamiodarone-induced disease, a normal Dlco suggeststhat heart failure, rather than amiodarone,may account for the respiratory symptoms. Bronchoscopymay be necessary to exclude infection.KL-6, a high-molecular-weight glycoproteinsecreted by alveolar type II cells, is a potentiallyuseful serum marker that is increased ( 500 U/mL)in patients with interstitial lung disease causedby IPF, radiation, and amiodarone. 60,61 Serum KL-6levels are typically 500 U/mL in patients withcongestive heart failure and pneumonia. Althoughfurther studies are necessary to corroborate thefindings of these small observational studies, thesensitivity and specificity of an increased serumKL-6 level in identifying interstitial diseaseapproaches 94% and 96%, respectively. 60,61The management of amiodarone-induced pulmonarytoxicity includes drug withdrawal and theinitiation of a new antiarrhythmic agent or implantationof an automatic cardioverter/defibrillator ifrequired. A trial of corticosteroids is often used insymptomatic patients, but the efficacy of steroidshas not been established. Radiographic resolutiongenerally occurs over 2 months, and patients are<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 335ACC-PULM-09-0302-021.indd 3357/10/09 10:21:04 PM


treated for at least 6 months to reduce the likelihoodof relapse. Recurrent amiodarone pulmonarytoxicity can also occur. If amiodarone is the onlyeffective agent in a patient with life-threateningarrhythmias, the dose must be reduced to theminimum that is effective (ideally 400 mg/d)and corticosteroids added to therapy.ACE Inhibitors: Cough occurs in up to 5 to 15%of patients receiving any type of ACE inhibitor. 5The cough generally appears from 1 to 2 monthsup to 1 year after initiating therapy. Patientswithout asthma in whom ACE inhibitor-inducedcough develops, compared with those patientswithout cough, are slightly more sensitive to methacholine.However, ACE inhibitors do not consistentlycause airways obstruction, nor are patientswith asthma at an increased risk. 62 Although themechanism of ACE inhibitor-induced cough hasnot been firmly established, it likely involvesblocking the metabolism of cough-inducing neuropeptidessuch as substance P and bradykinin.Before extensive testing is performed to evaluatethe source of a cough in a patient receivingACE inhibitors, therapy with the drug shouldbe discontinued. Patients with an ACE inhibitorinducedcough will generally show resolutionwithin 1 to 4 days. Management is simple: avoidall ACE inhibitors. Patients can be switched to anangiotensin II receptor antagonist (eg, losartan),which rarely induces cough. 63 However, losartancan induce angioedema in a manner similar to thatof ACE inhibitors. 64 If ACE inhibitors are requiredfor the management of a patient’s cardiovasculardisease and there are no alternative agents, thereare reports that the cough can be partly controlledwith oral theophylline or nebulized cromolyn orlidocaine. 65,66Angioneurotic edema is a rare adverse effect ofACE inhibitors (0.1 to 0.2% of patients) that ispotentially life threatening. 67 It is manifested byswelling of the tongue, lips, and mucousmembranes within hours or at most 1 week afterinitiating treatment, and can rapidly evolve intoacute respiratory distress, upper airway obstruction,and death. Patients with a history of idiopathicangioneurotic edema should avoid using ACEinhibitors because they may be at increased risk.The mechanism underlying this serious adverseeffect is unknown, but immune pathways, increasedtissue levels of bradykinin or histamine, and/or adeficiency of complement 1-esterase inactivatorhave all been implicated. The management of ACEinhibitor-induced angioedema includes drugwithdrawal, attention to airway patency and, ifsevere, the subcutaneous injection of epinephrine(0.01 mL/kg body weight of a 1:1,000 solution) every15 to 20 min along with IV saline solution to correcthypotension. Diphenhydramine (1 to 2 mg/kg upto 50 mg IV or IM) and steroids are also useful.-Blockers: -Adrenergic receptor blockersadministered by any route (eg, oral, IV, or ophthalmicsolutions) can precipitate bronchospasm,especially in patients with asthma or COPD. Propranolol,a nonselective -blocker, causes dosedependentreductions in airflows and shouldbe avoided in patients with asthma or COPD. 68Timolol, a nonselective -blocker used in ophthalmicsolutions, has predictable adverse effects onthese patients, similar to propranolol, and shouldalso be avoided. Severe bronchospasm and sporadicdeaths have been reported with nonselective-blocker ophthalmic solutions. Physiciansshould specifically inquire about ophthalmicsolu tions in patients with glaucoma presentingwith respiratory complaints since most patientsdo not consider these to be “drugs.” Selective 1-blockers (eg, atenolol and betaxolol ophthalmicsolution) are better tolerated in patients withairflow obstruction.A 2002 metaanalysis 69 showed that selective 1-blockers are well tolerated in patients withasthma and COPD and result in minimal reductionsin FEV 1when used over the long term. A recentstudy showed that cardioselective -blockersreduce the rate of mortality in patients with COPDundergoing vascular surgery. 70 Thus, in carefullyselected patients with COPD, the use of a cardioselective-blocker can be safely used. However,caution is warranted: 1-selectivity is relativebecause 1-selective inhibitors, especially in highdoses, can block 2-adrenergic receptors and triggerbronchospasm. Esmolol is the drug of choicein critically ill patients with asthma or COPDwho require a -blocker (eg, unstable angina)because it is an IV selective 1-blocker in doses of 300 g/kg/min and has an extremely short halflife(approximately 9 min).Statins: Statins are the most widely pre scribedlipid-lowering drug primarily because of their welldocumentedbeneficial effect on cardiovascular336 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3367/10/09 10:21:05 PM


morbidity and mortality. Statins inhibit cholesterolformation by blocking 3-hydroxy-3-methylglutamylreductase. There have been several case reports 71,72of statin-induced hypersensitivity pneumonitisand NSIP resulting from exposure to pravastatin,lovastatin, or simvastatin. The chest radiographicabnormalities include bilateral patchy ground-glassopacities and interstitial infiltrates. Although themechanism of lung injury is unknown, increasedalveolar CD8 suppressor T lymphocytes, hyperplastictype II pneumocytes, peripheral bloodeosinophilia, and clinical deterioration accompanyingstatin reintroduction support hypersensitivityin the pathogenesis. Statin-induced pneumonitismay become more commonly observed when usedat the higher doses advocated in some reports forthe prevention of cardiovascular diseases.Illicit DrugsThe use of illicit drugs continues to be a significanthealth-care problem of epidemic proportionsworldwide. Respiratory symptoms in patientsusing illicit drugs may be caused by a wide arrayof pulmonary disorders, including the following:(1) alveolar hypoventilation (hypercarbic respiratoryfailure); (2) aspiration; (3) noncardiogenicpulmonary edema; (4) barotrauma (eg, pneumothorax,pneumomediastinum, and pneumoperitoneum); (5) endocarditis/septic emboli;(6) foreign-body granulomatosis; (7) PIE; (8) COP;(9) alveolar hemorrhage; (10) bronchospasm;(11) interstitial pneumonitis/fibrosis; and (12) HIVassociatedinfections. Knowledge about the drugsused, the route of administration, and the presenting clinical syndrome helps to narrow the broaddifferential diagnosis in these patients.Noncardiac <strong>Pulmonary</strong> Edema: Noncardiac pulmonaryedema is an infrequent complication ofheroin, cocaine, or methadone abuse, but heroinabuse is by far the most common cause. 73 Naloxone,an opiate antagonist, also causes a similarsyndrome. There are no known risk factors foropiate-induced pulmonary edema. Although thepathogenesis is not established, several mechanismshave been proposed, including thefollowing: (1) altered alveolar/capillary permeability;(2) neurogenic pulmonary edema; (3) directopiate cytotoxicity; (4) drug hypersensitivity; and(5) hypoxemic alveolar injury. However, none ofthese mechanisms have been adequately testedin animal models or human studies. Typically,patients present within minutes to hours (at most24 h) after use with dyspnea, reduced respirationsand mental status, miotic pupils, and a chest radiographdemonstrating perihilar alveolar infiltratesin a “batwing” distribution. Pathologic studiesfrom autopsies have revealed changes similar toARDS, such as the following: (1) alveolar edemaand hemorrhage, (2) diffuse alveolar damage,(3) hyaline membranes, (4) interstitial inflammatorycell infiltrates, and (5) type II cell hyperplasia.Management centers on supportive care includingmechanical ventilation, which is needed in 40%. 73Unlike other causes of ARDS, the prognosis in opiate-inducednoncardiogenic pulmonary edema isgood, with resolution of pulmonary edema 48to 72 h, and nearly all patients survive.Cocaine: Cocaine, which is an alkaloid extractedfrom the leaves of Erythroxylon coca, is a sympathomimeticagent that causes topical anesthesiaand CNS stimulation. In 1995, an estimated 1.5million people in the United States regularly usedcocaine. 74 Cocaine hydrochloride, the salt formthat can be injected IV or rapidly absorbed acrossthe nasal mucosa, causes pulmonary manifestationssimilar to those of IV opiates (eg, heroin). Aunique feature of cocaine is that it causes nasalmucosal vasoconstriction that can result in ischemicnecrosis and the perforation of the nasal septum.Free-base cocaine, which is derived from thealkaline crystallization of the salt, vaporizes whenheated and is rapidly absorbed across the respiratorymembranes. The term crack refers to thepopping sound that occurs when cocaine crystalsare heated.Free-base crack cocaine smoking causes a distinctset of pulmonary abnormalities that are termedcrack lung. 75 These patients typically present withcough, chest pain, dyspnea, hemoptysis, andwheezing. Chest pain from either a pulmonary orcardiac origin occurs in approximately 20 to 40% ofcocaine users. Cocaine can induce direct coronaryartery vasoconstriction. Unlike IV crack users,patients with crack lung from inhalation may have acough productive of black, soot-like material (approximately10 to 33% of patients), bronchospasm(approximately 50% of patients), and thermal burnsto the upper and lower airways. Other pulmonary<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 337ACC-PULM-09-0302-021.indd 3377/10/09 10:21:05 PM


complications of crack cocaine use include PIE;COP; pulmonary hypertension; and alveolarhemorrhage/hemoptysis, and barotrauma(eg, pneumothorax, pneumomediastinum, andpneumoperitoneum). Barotrauma is likely causedby the Valsalva maneuver that is performed toenhance alveolar-capillary cocaine absorption.The Dlco is either normal or mildly reducedwithin weeks to months after using crack cocaine. 75,76The reason for the reduction in Dlco is unclear,but it may reflect a loss of alveolar-capillary surfacearea. 75–77 Some evidence 78 has demonstratedthat crack users have increased lower respiratorytract iron and ferritin levels that may contributeto lung injury via oxidant-induced mechanisms.Foreign-Body Granulomatosis: Foreign-body granulomatosis is caused by the IV injection of insolubleparticulate contaminants used to “cut” streetheroin or from talc or cellulose in crushed tablets(eg, amphetamines, methadone, or hydromorphine).Initially, patients may be asymptomaticdespite abnormal chest radiograph findingsshowing reticular nodular interstitial infiltrates.However, relentless progression of the infiltratesgenerally occurs accompanied by increasing dyspnea.The nodules may coalesce in a manner similarto that seen in silicosis, resulting in progressivemassive pulmonary fibrosis with surroundingcysts and bullae. Notably, PFT results reveal airwaysobstruction, rather than restriction, that progressesto severe disease with a reduced Dlco. 79The lungs demonstrate vascular and interstitialnoncaseating giant-cell granulomatous infiltratescontaining birefringent talc crystals. As the diseaseprogresses, parenchymal destruction andpulmonary hypertension occur, leading to respiratoryfailure/cor pulmonale. The poor outcomenoted in most patients is generally not altered bytherapy with corticosteroids.Miscellaneous AgentsContrast Media: Both the ionic and nonionicforms of contrast media can induce bronchospasmand a reduction in airflow. Typically, this occurswithin 5 min of infusion and resolves within 30 minafter infusion. Contrast media rarely induces potentiallyfatal leukostasis in the pulmonary arteriolesand capillaries. 80 This occurs within minutes to anhour after injection, resulting in dyspnea andhypoxemic respiratory failure caused by noncardiogenicpulmonary edema. Although the mechanismis unclear, a complement-mediated pathwayhas been implicated. Management centers on supportivecare, and therapy with high-dose corticosteroids,IV diphenhydramine, and heparin.Interferon-1b: Although there are no proventreatments for IPF, some studies 81 have suggesteda possible role for interferon (IFN)-1b based onits antifibrotic actions (eg, decreases fibroblast proliferationand collagen synthesis, augments fibroblastapoptosis, and inhibits profibrotic cytokineproduction) and its protective effect in animalmodels of pulmonary fibrosis. Acute respiratoryfailure developed in four patients with severeIPF (total lung capacity, 38 to 59% of predicted;Dlco, 20 to 37% of predicted) after 2 to 35 doses ofIFN-lb. 81 In two patients who underwent a lungbiopsy, pathologic studies revealed diffuse alveolardamage. To date, there is no clear beneficial rolefor IFN-lb in the management of IPF, yet seriousadverse pulmonary toxicity can occur, especiallyin patients with severe disease.Tocolytic Agents: Tocolytic agents, such as albuterol,terbutaline, ritodrine, isoxuprine, and salbutamol,are -adrenergic 3agents that are usedto inhibit uterine contractions during prematurelabor. Although acute pulmonary edema reportedlyoccurs in 0.5 to 5% of patients, in one study 82of 8,700 patients there was an incidence of only0.32%. The mechanism of pulmonary edema hasnot been established but is believed to be causedby -adrenergic–induced peripheral vasodilationand increased intravascular volume. Afterthe tocolytic drug is discontinued, the vasculartone normalizes, thereby promoting fluid movementinto the extravascular spaces, including thealveoli. Left ventricular function and pulmonarywedge pressure are usually normal. Patients typicallypresent during or within 12 h of delivery(rarely beyond 12 h postpartum) with an acuteonset of dyspnea, cough productive of pinktingedsputum, chest pain, tachycardia, tachypnea,hypoxemic respiratory failure, and diffusealveolar infiltrates seen on a chest radiograph.The differential diagnosis includes the following:gastric acid aspiration, cardiac pulmonaryedema, pulmonary embolism, amniotic fluidembolism, and peripartum cardiomyopathy. The338 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3387/10/09 10:21:05 PM


transthoracic echocardiogram findings generallyare normal. Management involves supportivecare and diuresis. The prognosis is generallyfavorable, with intubation/mechanical ventilationrequired in 10% of patients and a mortalityrate of nearly 3%. 83Sirolimus (Rapamycin): Sirolimus is an immunosuppressiveagent that is used in organ-transplantpatients in the presence or absence of cyclosporine.The frequency of pulmonary toxicity varies from“rare” to approximately 11%. 84,85 Of 24 patientswith sirolimus-associated pneumonitis, symptomsoccurred within 5 months of initiating therapyand consisted of cough (96%), fatigue (83%), fever(67%), and dyspnea (33%). 84,85 The HRCT scan/pathologic patterns include COP (79%), groundglassattenuation (17%), and, less commonly, interstitialfibrosis, alveolar proteinosis, necrotizingvasculitis, and pleural effusions. 84,85 BAL findingsshow a lymphocyte predominance (CD4 CD8)with occasional eosinophils and rarely (8%) diffusealveolar hemorrhage. Sirolimus withdrawal resultsin complete recovery within weeks to 6 months.References1. Rosenow EC III. The spectrum of drug-inducedpulmonary disease. Ann Intern Med 1972; 77:977–9912. Cooper JA Jr, White DA, Matthay RA. Druginducedpulmonary disease: part 1. Cytotoxicdrugs. Am Rev Respir Dis 1986; 133:321–3403. Hubbard R, Venn A, Britton J. Exposure to antidepressantsand the risk of cryptogenic fibrosingalveolitis: a case-control study. Eur Respir J 2000;16:409–4134. Snavely SR, Hodges GR. The neurotoxicity of antibacterialagents. Ann Intern Med 1984; 101:92–1045. Howard PA, Dunn MI. Is your patient’s coughcaused by an ACE inhibitor? J Respir Dis 1997;18:762–7686. Morelock SY, Sahn SA. Drugs and the pleura. Chest1999; 116:212–2217. Dimopoulou I, Bamias A, Lyberopoulos P, et al.<strong>Pulmonary</strong> toxicity from novel antineoplasticagents. Ann Oncol 2006; 17:372–3798. Poulter NR, Chang CL, Farley TMM, et al. Venousthromboembolic disease and combined oral contraceptives:results of international multicentrecase control study. Lancet 1995; 346:1575–15829. Vandenbroucke JP, Rosing J, Bloemenkamp KWM,et al. Oral contraceptives and the risk of venousthrombosis. N Engl J Med 2001; 344:1527–153510. Abenhaim L, Moride Y, Brenot F, et al. Appetitesuppressantdrugs and the risk of primary pulmonaryhypertension. N Engl J Med 1996; 335:609–61611. Drazen JM, Israel E, O’Byrne PM. Drug therapy:treatment of asthma with drugs modifying the leukotrienepathway. N Engl J Med 1999; 340:197–20612. Nathani N, Little MA, Kunst H, et al. Churg-Strauss syndrome and leukotriene antagonists use:a respiratory perspective. Thorax 2008; 63:883–88813. Beasley R, Bibby S, Weatherall M. Leukotrienereceptor antagonist therapy and Churg--Strausssyndrome: culprit or innocent bystander. Thorax2008; 63:847–84814. Kalra S, Bell MR, Rihal CS. Alveolar hemorrhageas a complication of treatment with abciximab.Chest 2001; 120:126–13115. Sandler A, Gray R, Perry MC, et al. Paclitaxel carboplatinalone or with bevacizumab for non-smallcell lung cancer. N Engl J Med 2006; 355:2542–255016. Cockerill FR III, Wilson WR, Carpenter HA, et al.Open lung biopsy in immunocompromisedpatients. Arch Intern Med 1985; 145:1398–140417. Mason JW. Prediction of amiodarone-induced pulmonarytoxicity [editorial]. Am J Med 1989; 86:2–318. Sleifer S. Bleomycin-induced pneumonitis. Chest2001; 120:617–62419. O’Driscoll BR, Hasleton PS, Taylor PM, et al. Activelung fibrosis up to 17 years after chemotherapywith carmustine (BCNU) in childhood. N Engl JMed 1990; 323. 378–38220. Yagoda A, Mukherji B, Young C, et al. Bleomycin,an antitumor antibiotic: clinical experience in 274patients. Ann Intern Med 1972; 77:861–87021. Munger JS, Huang X, Kawakatsu H, et al. Theintegrin v6 binds and activates latent TGF 1amechanism for regulating pulmonary inflammationand fibrosis. Cell 1999; 96:319–32822. Eitzman DT, McCoy RD, Zheng X, et al. Bleomycininduced pulmonary fibrosis in transgenic micethat either lack or overexpress the murine plasminogenactivator inhibitor-1 gene. J Clin Invest1996; 97:232–23723. Wang R, Ibarra-Sunga O, Pick R, et al. Abrogationof bleomycin-induced epithelial cell apoptosis andlung fibrosis by captopril or by a caspase inhibitor.Am J Physiol Lung Cell Mol Physiol 2000; 279:L143–L151<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 339ACC-PULM-09-0302-021.indd 3397/10/09 10:21:05 PM


24. Tran PL, Weinbach J, Opolon P, et al. Preventionof bleomycin-induced pulmonary fibrosis afteradenovirus-mediated transfer of the bacterial bleomycinresistance gene. J Clin Invest 1997; 99:608–61725. He Y-H, Wu M, Kobune M, et al. Expression ofyeast apurinic/apyrimidinic endonuclease (APN1)protects lung epithelial cells from bleomycintoxicity. Am J Respir Cell Mol Biol 2001; 25:692–69826. Horan GS, Wood S, Ona V, et al. Partial inhibitionof integrin v6 prevents pulmonary fibrosiswithout exacerbating inflammation. Am J RespirCrit Care Med 2008; 177:56–6527. Lewis CC, Yang JTH, Huang X, et al. Diseasespecificgene expression profiling in multiple modelsof lung disease. Am J Respir Crit Care 2008;177:376–38728. Heard BE, Cooke RA. Busulphan lung. Thorax1958; 23:187–19329. Malik SW, Myers JL, DeRemee RA, et al. Lung toxicityassociated with cyclophosphamide use. twodistinct patterns. Am J Respir Crit Care Med 1996;154:1851–185630. Cooper JAJr, Merrill WW, Reynolds HY. Cyclophosphamidemodulation of bronchoalveolarcellular populations and macrophage oxidativemetabolism: possible mechanisms of pulmonarypharmacotoxicity. Am Rev Respir Dis 1986;134:108–11431. Trask CW, <strong>Joan</strong>nides T, Harper PG, et al. Radiation-inducedlung fibrosis after treatment of smallcell carcinoma of the lung with very high-dosecyclophosphamide. Cancer 1985; 55:57–6032. Imokawa S, Colby TV, Leslie KO, et al. Methotrexatepneumonitis: review of the literature and histopathologicalfindings in 9 patients. Eur Respir J2000; 15:373–38133. White DA, Rankin JA, Stover DE, et al. Methotrexatepneumonitis: bronchoalveolar lavage findingssuggest an immunologic disorder. Am Rev RespirDis 1989; 139:18–2134. Searles G, McKendry RJ. Methotrexate pneumonitisin rheumatoid arthritis: potential risk factors;four case reports and a review of the literature. JRheumatol 1987; 14:1164–117135. Alarcon GS, Kremer JM, Macaluso M, et al. Riskfactors for methotrexate-induced lung injury inpatients with rheumatoid arthritis: A multicenter,case-control study; methotrexate-Lung StudyGroup. Ann Intern Med 1997; 127:356–36436. Cottin V, Tebib J, Massonnet B, et al. <strong>Pulmonary</strong>function in patients receiving long-term low-dosemethotrexate. Chest 1996; 109:933–93837. Linette DC, McGee KH, McFarland JA. Mitomycin-inducedpulmonary toxicity: case report andreview of the literature. Ann Pharmacother 1992;26:481–48438. Raderer CM, Kornek M, Hejna F, et al. Acute pulmonarytoxicity associated with high-dose vinorelbineand mitomycin. Ann Oncol 1996; 7:973–97539. Wiley JS, Firkin FC. Reduction of pulmonary toxicityby prednisolone prophylaxis during all-transretinoic acid treatment of acute promyelocytic leukemia:Australian Leukaemia Study Group. Leukemia1995; 9:774–77840. Andersson BS, Cogan BM, Keating MJ, et al. Subacutepulmonary failure complicating therapywith high-dose Ara-C in acute leukemia. Cancer1985; 56:2181–218441. Helman DL, Byrd JC, Ales NC, et al. Fludarabinerelated pulmonary toxicity. Chest 2002; 122:785–79042. Vahid B, Marik PE. <strong>Pulmonary</strong> complications ofnovel antineoplastic agents for solid tumors. Chest2008; 133:528–53843. Azoulay E, Attalah H, Hauf A, et al. Granulocytecolony stimulating factor or neutrophil-inducedpulmonary toxicity: myth or reality? Chest 2001;120:1695–170144. Samter M, Beers RF Jr. Concerning the nature ofintolerance to aspirin. J Allergy 1967; 40:281–29345. Babu KS, Salvi SS. Aspirin and asthma. Chest 2000;118:1470–147646. Kim S-H, Park H-S. Pathogenesis of nonsteroidalanti-inflammatory drug-induced asthma. CurrOpin Allergy Clin Immunol 2006; 6:617–62247. Leatherman JW, Schmitz PG. Fever, hyperdynamicshock, and multiple-system organ failure: A pseudosepsissyndrome associated with chronic salicylateintoxication. Chest 1991; 100:1391–139648. Tomioka R, King TE Jr. Gold-induced pulmonarydisease: clinical features, outcome, and differentiationfrom rheumatoid lung disease. Am J RespirCrit Care Med 1997; 155:1011–102049. Keane J, Gershon S, Wise RP, et al. Tuberculosis associatedwith infliximab, a tumor necrosis factor neutralizingagent. N Engl J Med 2001; 345:1098–110450. Mutlu GM, Mutlu EA, Bellmeyer A, et al. <strong>Pulmonary</strong>adverse events of anti TNF- antibody therapy.Am J Med 2006; 119:639–646340 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3407/10/09 10:21:05 PM


51. Thavarajah K, Wu P, Rhew EJ, et al. <strong>Pulmonary</strong>complications of tumor necrosis factor-targetedtherapy. Respir Med 2009; 103:661–66952. Jick SS, Jick H, Walker AM, et al. Hospitalizationsfor pulmonary reactions following nitrofurantoinuse. Chest 1989; 96:512–51553. Holmberg L, Boman G. <strong>Pulmonary</strong> reactions tonitrofurantoin: 447 cases reported to the SwedishAdverse Drug Reaction Committee 1966–1976. EurJ Respir Dis 1981; 62:180–18954. Vorperian VR, Havighurst TC, Miller S, et al.Adverse effects of low-dose amiodarone: a metaanalysis.J Am Coll Cardiol 1997; 30:791–79855. Cairns JA, Connolly SJ, Roberts R, et al. Randomizedtrial of outcome after myocardial infarctionin patients with frequent or repetitive ventricularpremature depolarisations: CAMIAT. Lancet 1997;349:675–68256. Ernawati DK, Stafford L, Hughes JD. Amiodaroneinducedpulmonary toxicity. Br J Clin Pharmacol2008; 66:82–8757. Reasor MJ, Kacew S. An evaluation of possiblemechanisms underlying amiodarone-inducedpulmonary toxicity. Proc Soc Exp Biol Med 1996;212:297–30458. Nicolescu AC, Ji Y, Comeau JL, et al. Direct mitochondrialdysfunction precedes reactive oxygenspecies production in amiodarone-induced toxicityin human peripheral lung epithelial HPL1Acells. Toxicol Appl Pharmacol 2008; 227:370–37959. Nikaido A, Tada T, Nakamura K, et al. Clinical featuresof and effects of angiotensin system antagonistson amiodarone-induced pulmonary toxicity.Int J Cardiol 2008 Dec 22. [Epub ahead of print]60. Yasuhiro E, Ritsuko H, Kenta U, et al. KL-6 as apotential new marker for amiodarone-inducedpulmonary toxicity. Am J Cardiol 2000; 86:229–23161. Ohnishi H, Yokoyama A, Kondo K, et al. Comparativestudy of KL-6, surfactant protein A, surfactantprotein D, and monocyte chemoattractantprotein-1 as serum markers for interstitial lungdisease. Am J Respir Crit Care Med 2002; 165:378–38162. Kaufman J, Casanova JE, Riendl P, et al. Bronchialhyperreactivity and cough due to angiotensin convertingenzyme inhibitors. Chest 1989; 95:544–54863. Conigliaro RL. Gleason PP. Losartan-inducedcough after lisinopril therapy [letter]. Am J HealthSyst Pharm 1999; 56:914–91564. Acker CG, Greenberg A. Angioedema induced bythe angiotensin II blocker losartan [letter]. N Engl JMed 1995; 333:157265. Aldis WL. Cromolyn for cough due to angiotensinconvertingenzyme inhibitor therapy: preliminaryobservations [letter]. Chest 1991; 100:1741–174266. Fogari R, Zoppi A, Tettamanti F, et al. Effects ofnifedipine and indomethacin on cough inducedby angiotensin-converting enzyme inhibitors: adouble-blind, randomized, cross-over study. J CardiovascPharmacol 1992; 19:670–67367. Zafar H, Hall WD. Cough and angioneuroticedema associated with angiotensin-convertingenzyme inhibitor therapy. Ann Intern Med 1992;117:234–24268. Chester EH, Schwartz HJ, Fleming GM. Adverseeffect of propranolol on airway function in nonasthmaticchronic obstructive lung disease. Chest1981; 79:540–54469. Salpeter SR, Ormiston TM, Salpeter EE. Cardioselective3-blockers in patients with reactive airwaydisease: a meta-analysis. Ann Intern Med 2002;137:715–72570. van Gestel YRBM, Hoeks SE, Sin DD, et al.Impact of cardioselective 3-blockers on mortalityin patients with chronic obstructive pulmonarydisease and atherosclerosis. Am J Respir Crit CareMed 2008; 178:695–70071. Liebhaber MI, Wright RS, Gelberg HJ, et al. Polymyalgia,hypersensitivity pneumonitis and otherreactions in patients receiving HMG-CoA reductaseinhibitors: a report of 10 cases. Chest 1999;115:886–88972. Lantuejoul S, Brambilla E, Brambilla C, et al. Statininducedfibrotic nonspecific interstitial pneumonia.Eur Respir J 2002; 19:577–58073. Sporer KA, Dorn E. Heroin-related noncardiogenicpulmonary edema. Chest 2001; 120:1628–163274. Albertson TE, Walby WF. Respiratory toxicitiesfrom stimulant use. Clin Rev Allergy Immunol1997; 15:221–24175. Haim DY, Lippmann ML, Goldberg SK, et al. Thepulmonary complications of crack cocaine: a comprehensivereview. Chest 1995; 107:233–24076. Itkonen J, Schnoll S, Glassroth J. <strong>Pulmonary</strong> dysfunctionin “freebase” cocaine users. Arch InternMed 1984; 144:2195–219777. Kleerup EC, Koyal SN, Marques-Magallanes JA,et al. Chronic and acute effects of “crack” cocaineon diffusing capacity: membrane diffusion, and<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 341ACC-PULM-09-0302-021.indd 3417/10/09 10:21:06 PM


pulmonary capillary blood volume in the lung.Chest 2002; 122:629–63878. Janjua TM, Bohan AE, Wesselius LJ. Increasedlower respiratory tract iron concentrations in alkaloidal(“crack”) cocaine users. Chest 2001; 119:422–42779. Pare JP, Cote G, Fraser RS. Long-term follow-up ofdrug abusers with intravenous talcosis. Am RevRespir Dis 1989; 139:233–24180. Hauggaard A. Non-cardiogenic pulmonaryoedema after intravenous administration of nonioniccontrast media. Acta Radiol 1996; 37:823–82581. Honore I, Nunes H, Groussard O, et al. Acuterespiratory failure after interferon 1b therapyof end-stag pulmonary fibrosis. Am J Respir CritCare Med 2003; 167:953–95782. Perry KG Jr., Morrison JC, Rust OA, et al. Incidenceof adverse cardiopulmonary effects withlow-dose continuous terbutaline infusion. Am JObstet Gynecol 1995; 173:1273–127783. Pisani RJ, Rosenow EC III. <strong>Pulmonary</strong> edema associatedwith tocolytic therapy. Ann Intern Med 1989;110:714–71884. Champion L, Stern M, Israel-Biet D, et al. Sirolimusassociated pneumonitis: 24 cases in renal transplant recipients. Ann Intern Med 2006; 144:505–50985. Roberts RJ, Wells AC, Unitt E, et al. Sirolimusinduced pneumonitis following liver transplantation.Liver Transpl 2007; 13:853–855Annotated BibliographyInternet SiteFoucher P, Camus P, the GEPPI. Pneumotox on line.Available at: http://www.pneumotox.com. AccessedMarch 28, 2009This is an excellent resource for a complete up-to-date listingof all the drugs reported to cause adverse pulmonary effects.It includes a detailed reference list, starting with the initialdescription and including the most current reports, for 350 agents organized alphabetically and by chest radiographicpattern.General <strong>Review</strong>sCleverley JR, Screaton NJ, Hiorns MP, et al. Druginducedlung disease: high resolution CT and histologicfindings. Clin Radiol 2002; 57:292–299A study comparing HRCT scan appearances with histologicfindings in 20 patients with drug-induced lung disease thatfound concordance in only nine patients (45%). The mostcommon HRCT scan findings included ground-glass opacities(85%), consolidation (70%), interlobar septal thickening(75%), and central lobular nodules (40%). The HRCTscan findings are of limited value in determining the histologicpicture and prognosis in such patients.Cooper JAD Jr. Drug-induced lung disease. Adv InternMed 1997; 42:231–268An excellent overview of the area emphasizing the pulmonaryclinical syndromes and diagnostic approach (149 references).Limper AH. Drug-induced pulmonary disease. In:Mason RJ, Murray JF, Broaddus VC, et al, eds. Murray& Nadel’s textbook of respiratory medicine. 4th ed.Philadelphia, PA: WB Saunders Co., 2005; 1888–1908An excellent overview of the topic that includes 167 references.Meyers JL. Diagnosis of drug reactions in the lung.Monogr Pathol 1993; 32–53A detailed analysis of the characteristic pathologic changesassociated with drug-induced adverse pulmonary effects thatincludes 121 references.Chemotherapeutic AgentsAlarcon GS, Kremer JM, Macaluso M, et al. Risk factorsfor methotrexate-induced lung injury in patients withrheumatoid arthritis. Ann Intern Med 1997; 127:356–364A multicenter, case-control study demonstrating the strongestclinical predictors for lung injury.Castro M, Veeder MH, Mailliard JA, et al. A prospectivestudy of pulmonary function in patients receivingmitomycin. Chest 1996; 109:939–944A prospective study of the role of PFTs in 133 patients showingthat serial changes in the Dlco did not predict clinicaltoxicity.Cherniak RM, Abrams J, Kalica AR. NHLBI Workshopsummary: pulmonary disease associated withbreast cancer therapy. Am J Respir Crit Care Med 1994;150:1169–1173A concise summary of lung toxicity due to chemotherapeuticagents used to treat breast cancer.Cottin V, Tebib J, Massonnet B, et al. <strong>Pulmonary</strong> functionin patients receiving long-term low-dose methotrexate.Chest 1996; 109:933–938This prospective study of 124 patients receiving low-dosemethotrexate shows that pneumonitis developed in 3.2% of342 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3427/10/09 10:21:06 PM


patients and that PFTs did not detect abnormalities prior toclinical presentation.Dimopoulou I, Bamias A, Lyberopoulos P, et al. <strong>Pulmonary</strong>toxicity from novel antineoplastic agents. AnnOncol 2006; 17:372–379A thorough review of adverse pulmonary effects from novelchemotherapeutic agents that includes 127 references.Limper AH. Chemotherapy-induced lung disease. ClinChest Med 2004; 25:53–64A succinct review of adverse pulmonary effects from new andold chemotherapeutic agents that includes 79 references.Sleifer S. Bleomycin-induced pneumonitis. Chest 2001;120:617–624A nice overview of the area with 96 references.Nonchemotherapeutic AgentsEvans RB, Ettensohn DB, Fawaz-Estrup F, et al. Goldlung: recent developments in the pathogenesis, diagnosis,and therapy. Semin Arthritis Rheum 1987;16:196–205A review of 60 patients with gold-induced pulmonarytoxicity and an extensive review of the literature.Everitt DE, Avorn J. Systemic effects of medicationsused to treat glaucoma. Ann Intern Med 1990; 112:120–125A concise summary of the adverse effects from commonlyused ophthalmic solutions.Hunt LW, Rosenow EC III. Asthma-producing drugs.Ann Allergy 1992; 68:453–462An extensive review of all the drugs that induce bronchospasmas well as the mechanisms implicated.Israili ZH, Hall WD. Cough and angioneurotic edemaassociated with angiotensin-converting enzyme inhibitortherapy: a review of the literature and pathophysiology.Ann Intern Med 1992; 117:234–242A concise summary of ACE inhibitor-induced cough andangioneurotic pulmonary edema clinical presentation,pathogenesis, and management.Kennedy JI Jr. Clinical aspects of amiodarone pulmonarytoxicity. Clin Chest Med 1990; 11:119–129A good overview of the clinical, radiographic, and histologicfeatures as well as management of amiodarone-induced lungdisease.Leatherman JW, Scmitz PG. Fever, hemodynamic shockand multiple organ failure: a pseudosepsis syndromeassociated with chronic salicylate intoxication. Chest1991; 100:1391–1396A concise review of acute and chronic salicylate toxicity.Pisani RJ, Rosenow EC III. <strong>Pulmonary</strong> edema associatedwith tocolytic therapy. Ann Intern Med 1989;110:714–718An overview of the clinical presentation and managementof 58 patients with tocolytic therapy-induced pulmonaryedema.Reasor MJ, Kacew S. An evaluation of possible mechanismsunderlying amiodarone-induced pulmonarytoxicity. Proc Soc Exp Biol Med 1996; 212:297–304A detailed review of how amiodarone may injure the lungswith 89 references.Tomioka H, King TE Jr. Gold-induced pulmonary diseases:clinical features, outcome, and differentiationfrom rheumatoid arthritis. Am J Respir Crit Care Med1997; 155:1011–1020A thorough review of the topic with 137 references.Illicit DrugsAlbertson TE, Walby WF. Respiratory toxicities fromstimulant use. Clin Rev Allergy Immunol 1997; 15:221–241A concise update of the topic with 178 references.Haim DY, Lippmann ML, Goldberg SK, et al. The pulmonarycomplications of crack cocaine: a comprehensivereview. Chest 1995; 107:233–240A succinct review of the pulmonary adverse effects of crackcocaine.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 343ACC-PULM-09-0302-021.indd 3437/10/09 10:21:06 PM


Notes344 Drug-Induced Lung Diseases (Kamp)ACC-PULM-09-0302-021.indd 3447/10/09 10:21:06 PM


Hemodynamic Monitoring and ShockJanice L. Zimmerman, MD, FCCPObjectives:• Understand the indications, complications, and use of pulmonaryartery catheterization• Apply hemodynamic information to determine appropriateclinical management• <strong>Review</strong> techniques of estimating cardiac output• Outline types, characteristic hemodynamic patterns, andmanagement of shockKey words: cardiac output; hemodynamics; monitoring;pulmonary artery catheter; sepsis; shock<strong>Pulmonary</strong> artery catheterization (PAC) with aflow-directed, balloon flotation catheter was introducedinto clinical practice in 1970. Many physiciansbelieve that the ability to assess cardiac functionand left ventricular filling pressure with PAC ishelpful in diagnosis and therapy. Because PAC is amonitor, its clinical utility depends on appropriateinterpretation of data and institution of effectivetherapeutic measures. Observational studies foundthat use of PAC was associated with either no reductionin complication rates or increased mortalityand use of resources, in comparison with patientswho did not undergo PAC. Randomized studiesof PAC use in high-risk surgical patients, patientswith shock and ARDS, patients with heart failure,and patients admitted to the ICU have found nodifference in mortality between groups managedwith and without PAC. The first part of this chapterwill concentrate on reviewing the principles ofPAC and correct interpretation of data rather thanon the definitive benefit of the PAC. More detailedinformation on aspects of PAC use, including waveformanalysis, is available through the <strong>Pulmonary</strong>Artery Catheterization Education Project, availableat www.pacep.org.Indications and ComplicationsPAC is without therapeutic purpose with the rareexception of emergency cardiac pacing. PAC shouldbe considered for use when hemodynamic informationis needed for optimum care of the patient thatthat cannot be supplied by clinical evaluation orother noninvasive modalities. Studies have demonstratedthat PAC data are more accurate than clinicalassessment in determining hemodynamic variablesin complicated cases. Common indications for PACmonitoring are listed in Table 1. Many of the diagnosticuses of PAC have been supplanted by the use ofechocardiography. The most common diagnostic usesof PAC are to determine intravascular volume statusand cardiac function, which allow assessment of thetype of shock and cause of pulmonary edema.Various complications are associated with PAC,but dysrhythmias such as premature ventriculardepolarizations or conduction abnormalities thatoccur during passage of the catheter are most common(Table 2). Dysrhythmias have been describedin 13 to 78% of patients, and the risk is increasedin the presence of cardiac ischemia, hypotension,hypokalemia, hypocalcemia, hypoxemia, andacidosis. When possible, metabolic abnormalitiesshould be corrected before PAC. In patients withpreexisting left bundle-branch block, the potentialfor development of complete heart block exists.This rare occurrence is more likely when the leftbundle-branch block is new and associated witha recent myocardial infarction. Other complicationsoccur at an overall rate of 4 to 7%, but thefrequency of data misinterpretation or misapplicationis unknown.Use of the PACInsertionThe basic multilumen, 110-cm pulmonary artery(PA) catheter is depicted in Figure 1. The distallumen opens at the catheter tip, and the proximallumen (right atrial lumen) opens approximately30 cm from the tip. A thermistor is located 3to 5 cm from the tip. Adaptations that allow<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 345ACC-PULM-09-0302-022.indd 3457/10/09 11:10:50 PM


Table 1. Indications for PAC MonitoringDiagnosticEvaluate volume statusEvaluate cardiac functionDetermine type of shockDetermine cause of pulmonary edemaDiagnose cardiac tamponadeDiagnose ventricular septal ruptureDiagnose acute mitral regurgitationDiagnose constrictive pericarditisTherapeutic/preventiveGuide therapyMonitor response to fluids, diuretics, vasoactive drugs, andpositive pressure ventilationMonitor high-risk patients perioperativelycontinuous venous oximetry and right ventricularcardiac output measurements are available. A PAcatheter (usually 7F or 7.5F) is preferably insertedaseptically into the subclavian vein or internaljugular vein using a modified Seldinger technique.However, clinical circumstances of an individualpatient dictate the appropriate selection of the site.Insertion through the femoral vein may requirefluoroscopic assistance. Fluoroscopic guidancemay also be necessary in patients with low cardiacoutput states, dilated right-heart chambers, severetricuspid regurgitation, or PA hypertension. Ideally,PAC should be performed by a physician whois skilled in the insertion technique, knowledgeableof complications, and capable of appropriatelyinterpreting and utilizing the data obtained.The catheter should be advanced with continuousECG and pressure monitoring. Once thecatheter has passed into an intrathoracic vein asnoted by respiratory fluctuation in pressures, theballoon should be inflated with 1.3 to 1.5 mL of airto facilitate passage through the right heart andinto the PA. In the average patient, the distanceTable 2. Complications of PACEstablishment of central venous accessArterial puncture/damagePneumothoraxBleedingAir embolismNeuropathyCatheterization procedureDysrhythmiasRight bundle-branch blockCatheter knottingCatheter presenceThrombosisInfectionPA ruptureBalloon ruptureEndocardial damage<strong>Pulmonary</strong> embolism/infarctionMisinterpretation of datafrom insertion to the right atrium is approximately10 cm from a subclavian approach, 10 to 15 cm fromthe right internal jugular, and 35 to 45 cm from thefemoral site. The PA is usually reached within 50to 55 cm from the internal jugular or subclavianvein and within 65 to 70 cm from the femoral vein.Greater insertion lengths may indicate coiling inthe right atrium or right ventricle. Characteristicwaveforms and pressures are obtained duringpassage through the right atrium, right ventricle,and PA (Table 3, Fig 2). Atrial pressure waves canusually be recorded from the right atrium and thePA occlusion position. The a-wave is absent inatrial fibrillation.If particular waveforms are not seen, placementof the catheter may be incorrect. The clinicianmust always be alert for the possibility ofdysrhythmias and abnormal waveform patterns.The balloon should be deflated after a PA occlusionpressure (PAOP) waveform is verified. WithFigure 1. PAC (Swan-Ganz catheter).346 Hemodynamic Monitoring and Shock (Zimmerman)ACC-PULM-09-0302-022.indd 3467/10/09 11:10:50 PM


Table 3. Components of Hemodynamic WaveformsWaveformAtrial pressure and PAOPa-wavec-wavev-wavex-descenty-descentRight ventricularPADescriptionAtrial contraction; follows P-wave (within PR interval) on ECG in right atrial tracing; nearend of or after QRS in PAOP tracing.Sudden motion of mitral and tricuspid valve rings toward right and left atria at onset ofventricular systole; follows a-wave but not always present in right atrial tracing andnot visible in PAOP tracing.Venous filling of left and right atria during ventricular systole when mitral and tricuspidvalves closed; near end of T-wave on ECG in right atrial tracing, after T-wave in PAOPtracing.Atrial relaxation and sudden downward motion of the atrioventricular junction duringearly ventricular systole; follows a-wave.Rapid atrial emptying following opening of mitral and tricuspid valves; follows v-waveSystolic peak with rapid downstroke to end-diastolic pressure; peak systolic wave occursafter the QRS on ECG but before the T-wave with end-diastole in the T-QRS period.Systolic peak and diastolic trough with dicrotic notch due to closure of pulmonic valve;peak systolic wave occurs after the QRS but before the peak of T-wave on ECG;end-diastolic trough near end of QRS.deflation, a characteristic PA pressure waveformshould be apparent. If a PA pressure tracing is notpresent, continued occlusion of the pulmonaryvessel (“overwedging”) is possible and requiresrepositioning of the PAC to prevent pulmonaryinfarction. A chest radiograph should be obtainedafter the procedure to evaluate proper positioning.Daily chest radiographs should be inspected fordistal migration of the catheter tip.MeasurementsThe PAC allows measurement of right atrialpressure, right ventricular pressure (during passagethrough the chamber), PA pressure, and PAOP(“wedge” pressure). In addition, cardiac output andmixed venous oxyhemoglobin saturation (Svo 2)can be directly measured. Normal values for thesemeasurements are listed in Table 4. In general,trends in these measurements over time are moreuseful clinically than single values. From these primarymeasurements, other physiologic variablesthat may be useful in evaluation and managementof patients can be calculated. These equations andvalues are listed in Table 5.Large variations in intrathoracic pressure dueto labored spontaneous respiration (eg, severe bronchospasm)or mechanical ventilation, especiallywith positive end-expiratory pressure (PEEP),can cause difficulty in interpreting measurementof vascular pressures. To minimize these effects,vascular pressures are measured at end-expiration.Digital readouts of mean PAOP are not adequate inpatients with rapid or labored breathing becauseFigure 2. Characteristic waveforms and pressures seen during insertion of a PAC. A valid wedge pressure is not a straight line buthas a-waves and v-waves indicative of transmitted left atrial pressure.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 347ACC-PULM-09-0302-022.indd 3477/10/09 11:10:51 PM


Table 4. Normal Values Obtained From PACVariablesRight atrial pressure, mm Hg 2–8Right ventricularpressure, mm HgSystolic 20–30DiastolicPA pressure, mm HgSystolic 20–30Diastolic 5–15Normal RangeSame as right atrialpressurePAOP, mm Hg6–12 (must be less thandiastolic PA pressure)Svo 2, % 65–75more than end-expiration will be averaged. Visualassessment of the pressure tracing allows moreaccurate determination of PAOP. Disconnection ofpatients from the ventilator or decrease in PEEP inorder to measure hemodynamic variables are notrecommended. Either of these actions alters hemodynamicconditions and may result in deteriorationof clinical status.PAOPThe measurement of PAOP is a major reasonfor PAC. Accurate measurement is essential forcorrect physiologic interpretation of data. The correctposition of the catheter for determination ofPAOP can be confirmed using the following criteria:(1) Mean PAOP should be less than or equal to thePA diastolic pressure, and lower than the mean PApressure. (2) The waveform should be characteristicof left atrial pressures with an a-wave and v-wave.A damped waveform, straight line, or waveformdeflections only related to respiratory changesare not acceptable. Balloon inflation and deflationshould consistently result in disappearance andreappearance of the PA pressure tracing. (3) Afree-flowing column of fluid or continuous flushexcludes catheter obstruction. (4) Blood gas analysisof a sample withdrawn from the presumed PAOPposition should usually have an oxygen saturationgreater than or equal to systemic arterial blood.The correct catheter position is necessary tointerpret the PAOP as a reflection of left atrialTable 5. Physiologic Data Derived From Invasive Monitoring*Cardiac index:CO/BSA (normal range, 2.4 to 4.4 L/min/m 2 )Systemic vascular resistance:(MAP − CVP)/CO 79.9 (normal range, 900 to 1,400 dyne•s•cm -5 )<strong>Pulmonary</strong> vascular resistance(MPAP − PAOP)/CO 79.9 (normal range, 150 to 250 dyne•s•cm -5 )Stroke volume indexCI/HR 1,000 (normal range, 36 to 48 mL/beat/m 2 )Stroke volume indexSV/BSA CI/HR (normal range, 30 to 65 mL/min/m 2 )Left ventricular stroke work indexSVI (MAP − PAOP) 0.0136 (normal range, 43 to 61 g•m/m 2 )Right ventricular stroke work indexSVI (MPAP − CVP) 0.0136 (normal range, 7 to 12 g•m/m 2 )O 2contentHb arterial O 2saturation 1.36 (Po 2 0.003) [normal approximately 19.5 mL/dL]Arteriovenous oxygen content differenceCao 2 Cvo 2(normal range, 3 to 5 mL/dL)Oxygen deliveryCO Cao 2 10 (normal range, 800 to 1,200 mL/min)Oxygen consumptionCO (Cao 2− Cvo 2) 10 (normal range, 180 to 280 mL/min)<strong>Pulmonary</strong> shunt (veno-arterial admixture)(Cco 2− Cao 2)/(Cco 2− Cvo 2) [normal range, 3 to 5%]*BSA body surface area; Cao 2 arterial O 2content; Cco 2 pulmonary capillary O 2content (assumed equal to alveolar Po 2);CI cardiac index; CO cardiac output; Cvo 2 mixed venous oxygen content; CVP central venous pressure; Hb hemoglobinconcentration; HR heart rate; MPAP mean pulmonary artery pressure; SV stroke volume; SVI stroke volume index.348 Hemodynamic Monitoring and Shock (Zimmerman)ACC-PULM-09-0302-022.indd 3487/10/09 11:10:51 PM


pressure. The pulmonary vascular system distal tothe catheter tip must be patent and provide a bloodfilledconnection with the left atrium. Left atrialpressures cannot be estimated in the upper lung(zone 1), where alveolar pressure usually exceedsPA and venous pressures (Fig 3). These conditionsclose pulmonary capillaries, and blood flow doesnot occur. In the central lung area (zone 2), flow isdetermined by the PA pressure and alveolar pressurewith alveolar pressure exceeding pulmonaryvenous pressure. Balloon inflation converts zone2 to zone 1 conditions by stopping blood flow andallowing collapse of the vessel distal to the balloon.Thus, PAOP reflects alveolar pressure rather thanleft atrial pressure in these zones. In zone 3, PAand venous pressures exceed alveolar pressuresand capillaries remain open providing the necessaryconnection between the catheter tip and leftatrium. With supine positioning, most of the lungis in zone 3, and flow-directed catheters usuallyenter these areas. Placement in non-zone 3 areasor conversion of zone 3 to other zones can occurwith hypovolemia, change in patient position, orapplication of positive pressure ventilation. It hasbeen estimated that if the PAOP changes 50% ofthe change in PEEP, the catheter tip is likely to bein a non-zone 3 position.The PA diastolic pressure is sometimes usedto estimate the PAOP if wedging is unsuccessful.The PAOP is usually only 1 to 3 mm Hg less thanthe PA diastolic pressure in individuals with a normalheart rate and pulmonary vasculature. Withtachycardia ( 120 to 130 beats/min) or pulmonaryhypertension (eg, chronic lung disease or hypoxemia),the difference in the two measurements maybe large and vary over time.Appropriate interpretation of the measurementsobtained with a PAC requires an understanding ofthe physiologic assumptions. The PAOP is a commonmeasurement that is obtained to reflect left ventricularfilling pressure (left ventricular end-diastolicpressure). Balloon occlusion of a branch of the PAinterrupts flow and allows transmission of pressureback from the pulmonary veins and, ultimately, theleft atrium. In general, the PAOP approximates leftatrial pressure in the absence of any mechanicaldisruption of the large pulmonary veins. Pressureswithin the left atrium are a reflection of the pressurewithin the left ventricle at end-diastole. Left ventricularend-diastolic pressure is a major determinantof left ventricular end-diastolic volume, or preload.However, preload is also determined by complianceof the ventricle (pressure-volume relationship) andtransmural ventricular distending pressure (intracavitarypressure less juxtacardiac pressure). Figure4 illustrates these principles. Changes in juxtacardiacpressure, as with PEEP therapy (or auto-PEEP), orchanges in ventricular compliance, as with ischemiaor vasoactive agents, alter the interpretation of thePAOP as a reflection of myocardial function. Theeffect of PEEP on juxtacardiac pressure dependsFigure 3. Catheter tip placement can occur in various westzones of the lung. Pa alveolar pressure; Pa arterial pressure;Pv venous pressure.Figure 4. Wedge pressure (PW) may be associated with differentventricular volumes, or preload, depending on whetherjuxtacardiac pressure is elevated (top, A), as may occur withPEEP therapy, or whether ventricular compliance is decreased(bottom, C), as may occur with ischemia. B illustrates a normalheart and juxtacardiac pressure.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 349ACC-PULM-09-0302-022.indd 3497/10/09 11:10:51 PM


Table 6. Conditions That Alter Pressure and VolumeRelationships*PAOP overestimates LVEDPMitral valve obstructionAtrial myxomaMitral valve regurgitation<strong>Pulmonary</strong> venous obstruction/vasoconstrictionIntracardiac shunt (left to right)Non-zone 3 catheter positionTachycardiaPAOP underestimates LVEDPAortic regurgitationDecreased left ventricular compliancePAOP and LVEDP do not reflect preload (LVEDV)Positive pressure ventilationAbnormal left ventricular complianceVasoactive drugsDilated right ventricle*LVEDP left ventricular end-diastolic pressure;LVEDV left ventricular end-diastolic volume.on lung and chest wall compliance. Usually, PEEPcauses a decrease in left ventricular transmuralpressure despite an increase in measured PAOP.Overall, measured PAOP will overestimate leftventricular end-diastolic pressure with PEEP,but the extent is difficult to determine precisely.In addition, mitral or aortic valve dysfunctionalter the association of PAOP with left ventricularend-diastolic pressure (Table 6). The PAOP as ameasure of preload does not reliably predict fluidresponsiveness.The PAOP is an indirect assessment of pulmonarycapillary pressure but does not equal pulmonarycapillary pressure. The pulmonary capillarypressure exceeds left atrial pressure and PAOP bya variable degree depending on vascular resistancein the pulmonary venous system. A normal PAOPmay be present despite high pulmonary capillarypressures if marked resistance exists. Physiologic(eg, pulmonary embolism and hypoxia) and pharmacologicalterations (eg, prostaglandin administrationand sympathetic stimulation) affect thepulmonary venous resistance. In such cases, thePAOP may underestimate the degree to whichelevated pulmonary capillary pressure contributesto the development of pulmonary edema.Thermodilution Cardiac OutputDetermination of cardiac output by thermodilutioninvolves the injection of an indicator throughthe proximal port (right atrium) with detection ofa temperature change by a thermistor at the distalend of the catheter. The volume and temperatureof injectate and the area under the thermodilutioncurve enable the cardiac output to be calculated.Thermodilution cardiac output generally has a goodcorrelation with the Fick or dye dilution technique.However, even under the best circumstances, thismeasurement has a reliability of 10%. To ensure thebest accuracy, the following steps should be taken:(1) temperature and volume of the injectate mustbe carefully measured; (2) average of three separatemeasurements obtained at least 1 min apart;(3) inject the solution over 2 s at the same phase ofrespiration, which preferentially is end-expiration;(4) use of room temperature injectate except forpatients who are severely hypothermic ( 30C)who require iced injectate; a temperature differenceof 10C is necessary for accurate measurements; and(5) inspect graphic display of cardiac output thermodilutioncurve for rapid upstroke and gradual,smooth return to baseline.Falsely low measurements may be obtained inthe presence of tricuspid regurgitation and falselyelevated results in the presence of intracardicshunts or distal catheter position. Measurementsmay be unreliable in the presence of atrial fibrillationbecause of poor reproducibility. Other noninvasiveand semiinvasive measurements of cardiacoutput that are used include the following:Transesophageal Doppler UltrasonographyThis technique determines the velocity of bloodin the descending thoracic aorta during ventricularsystole. Stroke volume is estimated from thisvelocity using a nomogram and measurement ofaortic diameter. Proper positioning is necessaryfor accurate measurements, and requires trainingand experience.CO 2Partial RebreathingMonitors using this technique are based on theFick principle applied to CO 2. Determinations aremade of CO 2production and arterial CO 2contentduring normal and rebreathing conditions. Patientsmust be receiving controlled mechanical ventilationbecause an increase in minute ventilationduring the rebreathing period in spontaneously350 Hemodynamic Monitoring and Shock (Zimmerman)ACC-PULM-09-0302-022.indd 3507/10/09 11:10:52 PM


eathing patients reduces the accuracy of the cardiacoutput estimation. Inaccurate measurementsalso result with low minute ventilation, high shuntfraction, and high cardiac output.Pulse Contour AnalysisAnalysis of the arterial pressure waveform isused to compute stroke volume with or withoutinitial calibration of cardiac output by transpulmonarythermodilution or indicator dilution (eg,lithium chloride). The pulse contour methodologyis based on changes in the pulse pressure waveform.The pulse pressure varies with the strokevolume and changes in vascular compliance. Anelevated pulse pressure variation may predictfluid responsiveness in critically ill patients.These techniques are not reliable when the arterialwaveform is poorly defined or arrhythmiasare present.Studies of the reliability, validity, and clinicalutility of the less invasive techniques reportvariable results. The advantages, disadvantages,and limitations of the various techniques suggestthat a single method may not be applicable to allpatients. Further clinical investigations are neededto determine the optimal method of assessing thehemodynamic status of critically ill patients.Mixed Venous Blood MeasurementsMeasurements of Svo 2are obtained by withdrawingblood from the distal port of the catheterwith the balloon deflated. At least 2.5 to 3 mL ofblood must be eliminated to allow for catheter deadspace before removal of blood for analysis. Bloodshould be removed slowly in patients receivinghigh fraction of inspired oxygen levels or PEEP toavoid contamination with oxygenated capillaryblood. Inaccurate measurements also occur in thepresence of severe mitral regurgitation, left-torightcardiac shunts, and distal tip placement. PAcatheters and special central venous catheters thatprovide continuous measurement of Svo 2make useof fiberoptic reflectance oximetry.The Svo 2reflects the balance between tissue O 2delivery and O 2consumption. The four variablesthat determine Svo 2are cardiac output, arterialoxygen saturation, hemoglobin, and tissue oxygenconsumption.Although this measurement is often used toreflect changes in cardiac output, decreases in arterialoxygen saturation (hypoxemic lung disease)or hemoglobin as well as increased tissue oxygenconsumption (fever, hypermetabolism, shivering,seizures, pain, or work of breathing) can result indecreases in Svo 2. A normal Svo 2in a stable patientusually indicates that tissue oxygen needs arebeing met. An exception to this rule is in sepsis andcertain poisonings (eg, cyanide), in which tissuehypoxia may exist with normal or elevated Svo 2.In many critically ill patients without PAC, centralvenous catheterization allows measurement ofcentral venous oxyhemoglobin saturation (Scvo 2).The Scvo 2is usually 2 to 3% less than Svo 2becausethe lower body extracts less oxygen. However, inpatients with shock, this relationship is altered dueto increased oxygen consumption by the GI tract.Under such conditions, the Scvo 2is 5 to 7% higherthan Svo 2, but both values change in parallel.Clinical Indications for HemodynamicMonitoringMyocardial Infarction and Heart FailureThe first use of PAC was in patients withmyocardial infarction complicated by shock. Estimationof preload by PAOP, measurement of cardiacoutput, and calculation of systemic vascularresistance (SVR) allow institution and titration offluids and inotropic or vasopressor agents. Newerless invasive techniques may also provide similarinformation. The recognition of ischemic right ventriculardysfunction has implications for treatmentdifferent from that of left ventricular dysfunction.The characteristic hemodynamic findings in acuteright ventricular dysfunction include an elevatedright atrial pressure that equals or exceeds PAOP,a steep y descent, and a right ventricular pressuretracing with a diastolic dip and plateau (squareroot sign). The right atrial pressure may increaseduring inspiration (Kussmaul sign) or with thehepatojugular reflux maneuver.PAC has been utilized in patients with heartfailure to optimize inotropic and vasodilatortherapy. The randomized trial of PAC use inpatients with severe symptomatic heart failure(ejection fraction 30%) failed to show any benefitbut was associated with a greater incidence<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 351ACC-PULM-09-0302-022.indd 3517/10/09 11:10:53 PM


of in-hospital adverse events. The benefit of PACin cardiogenic shock has not been evaluated inclinical trials.ShockPAC or less invasive techniques may be helpfulin the diagnosis of shock not due to myocardialinfarction. Characteristic hemodynamic patternsthat use filling pressures, cardiac output, and/orSVR are usually sufficient to distinguish cardiogenic,distributive, hypovolemic, and obstructiveshock.Mitral RegurgitationChronic mitral regurgitation is best diagnosedby noninvasive echocardiography. In acute decompensation,as may occur in myocardial infarction,PAC information will often distinguish mitralregurgitation from other complications if echocardiographyis not readily available. Elevatedleft atrial pressure and PAOP are noted, and tall,peaked v-waves are seen in the PAOP tracing. Abifid PA waveform composed of the PA systolicwave and the v-wave may be seen. Regurgitantflow makes accurate assessment of the PAOPdifficult if not impossible. In such circumstances,the a-wave on the PAOP tracing can be used toestimate left ventricular end-diastolic pressure.The pressure tracing should be examined togetherwith a simultaneous ECG tracing. The v-waveoccurs at the end of the T-wave on ECG, with thePA waveform occurring earlier in the cardiac cycle.Prominent v-waves can also be noted when the leftatrium is distended and noncompliant due to leftventricular failure from any cause.Tricuspid RegurgitationTricuspid regurgitation usually occurs in thesetting of pulmonary hypertension and right ventriculardilatation. Tricuspid regurgitation is associatedwith elevated right atrial and right ventricularend-diastolic pressures. A prominent early v-waveis seen in the right atrial pressure tracing with PAC,and the x-descent may be absent or diminished.A steep y-descent is present. In severe tricuspidregurgitation, the right atrial pressure tracing mayresemble the right ventricular pressure tracing.Intracardiac ShuntThe presence of an intracardiac shunt may bedetermined using PAC to obtain blood samplesfor oxygen content in the right atrium, right ventricle,and PA, or by using PAC with oximetry.This assessment may be helpful in distinguishingacute ventricular septal defect from acute mitralregurgitation in myocardial infarction. A step-upin oxygen content ( 10%) between the right ventricleand right atrium indicates a ventricular septaldefect. With development of an acute ventricularseptal defect, the right atrial pressure, PA pressure,and PAOP are elevated.<strong>Pulmonary</strong> EdemaThe clinical differentiation of noncardiac andcardiac pulmonary edema may be difficult, particularlyin critically ill patients. The differentiationof the two types of pulmonary edema is importantbecause therapeutic interventions and prognosisdiffer significantly. Patients with noncardiacpulmonary edema have normal or near-normalPAOP in contrast to the elevated PAOP noted inmyocardial dysfunction.Acute Massive <strong>Pulmonary</strong> EmbolismMassive pulmonary embolism can produce aclinical presentation similar to cardiogenic shock.Although rarely needed as a diagnostic tool, PACwill demonstrate elevated systolic and diastolicPA pressures in massive pulmonary embolismbut no elevation in PAOP. The mean PA pressurerarely exceeds 40 mm Hg unless there is a chroniccomponent of pulmonary hypertension.Acute Cardiac TamponadeEchocardiography is most commonly used todiagnose pericardial effusion and its hemodynamiceffects on myocardial function. In those instances inwhich echocardiography is not available, or whenthere is a question of hemodynamic impact of pericardialeffusion, PAC may be helpful particularlywith focal effusions. The characteristic hemodynamicpattern of acute cardiac tamponade includes elevationand equalization of right-sided and left-sidedfilling pressures (right atrium and PAOP) and a352 Hemodynamic Monitoring and Shock (Zimmerman)ACC-PULM-09-0302-022.indd 3527/10/09 11:10:53 PM


prominent systolic x-descent and blunted y-descenton the right atrial pressure tracing. The right atrialpressure declines in inspiration in contrast to rightventricular infarction and constrictive pericarditis.Constrictive PericarditisConstrictive pericarditis produces equalizationand elevation of right-sided and left-sided fillingpressures similar to cardiac tamponade. In addition,constrictive pericarditis usually produces anearly diastolic dip followed by a pressure plateauin the right and left ventricular pressure tracings.An inspiratory increase in pressure (Kussmaulsign) can often be seen in the right atrial tracingalong with a systolic x-descent and prominent earlydiastolic y-descent.Restrictive CardiomyopathyThe hemodynamic pattern in restrictive cardiomyopathyis similar to constrictive pericarditis.Differentiation of the two conditions may be verydifficult. Left-sided filling pressures tend to behigher than right-sided filling pressures in restrictivecardiomyopathy. In addition, the diastolicpressure plateau is usually less than one third ofthe right ventricular systolic pressure.ShockShock is a syndrome of impaired tissue oxygenationand perfusion that results from one ofthe following mechanisms: an absolute or relativedecrease in oxygen delivery; ineffective tissueperfusion; or impaired utilization of deliveredoxygen. Shock results when the oxygen balanceis disturbed and oxygen demands exceed supply.Although hypotension is often present in shockstates, BP may be normal or even slightly increaseddue to compensatory vasoconstriction. Additionalclinical manifestations in shock are related to tissuehypoperfusion, the compensatory responsesinitiated by shock, and the underlying etiologyof shock.Shock is usually classified into the followingfour categories: hypovolemic; distributive;cardiogenic; and obstructive (Table 7). However,most patients with shock demonstrate featuresof more than one type of shock (ie, mixed shock).For example, a patient who is in septic shock willlikely have a hypovolemic component beforeresuscitation in addition to the decreased SVR thatis characteristic of distributive shock.The different types of shock are characterizedby hemodynamic patterns (Table 8) that assistTable 7. Classification of ShockHypovolemicHemorrhagic (trauma and GI bleed)Nonhemorrhagic (eg, external loss, interstitial fluidredistribution)DistributiveSepsisAdrenal crisisNeurogenic (spinal shock)AnaphylaxisThyroid stormCardiogenicMyopathic (eg, ischemia, cardiomyopathy)Mechanical (eg, valvular lesions, septal defects)ArrhythmiasObstructiveMassive pulmonary embolismTension pneumothoraxCardiac tamponadeConstrictive pericarditisTable 8. Hemodynamic Patterns of Shock*Type of Shock SVR CO CVP PAOP SvO 2or ScvO 2DistributiveSeptic ↓ ↑ or N* (rarely↓) ↓ or N ↓ or N ↓ or NNeurogenic ↓ ↑ or ↓ ↓ ↓ ↑ or NHypovolemic ↑ ↓ ↓ ↓ ↓Cardiogenic ↑ ↓ ↑ ↑ ↓ObstructiveCardiac tamponade ↑ ↓ ↑ ↑ ↓<strong>Pulmonary</strong> embolism ↑ ↓ ↑ ↓ or N ↓*N Normal; CO cardiac output; CVP central venous pressure; ↑ increase; ↓ decrease.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 353ACC-PULM-09-0302-022.indd 3537/10/09 11:10:53 PM


with diagnosis and the determination of appropriatemanagement. Specific measurements maynot always be available, and clinical findings in aspecific patient may be variable due to the specificetiology of shock, the underlying cardiac function,the duration of shock, and degree of resuscitation.Management Principles of ShockThe management of shock requires treatmentof the underlying etiology, and the restoration ofadequate oxygen delivery and tissue perfusion.Oxygen delivery is optimized with interventionsto achieve an appropriate BP, cardiac output, andoxygen content of arterial blood (see Table 5). Thefirst goal is a minimum BP to maintain blood flowto the heart and vital organs while other componentsof oxygen delivery are addressed. An initialmean arterial pressure (MAP) of 65 mm Hg isusually recommended, but the specific goal for anindividual patient should aim to optimize tissueperfusion without increasing myocardial oxygendemand. Once a minimum BP is achieved, oxygendelivery is improved by increasing cardiac output,hemoglobin concentration, or oxyhemoglobinsaturation. Hemoglobin concentration can beincreased by blood transfusion and oxyhemoglobinsaturation by the administration of supplementaloxygen and/or mechanical ventilation.Therapeutic intervention with fluids, vasopressoragents, inotropic agents, and, rarely, vasodilators(cardiogenic shock) is utilized to optimize BP andcardiac output.FluidsMost types of shock benefit from initial volumeresuscitation. Even patients with cardiogenic shockmay require fluids to increase ventricular fillingpressures in order to optimize myocardial performance(the Frank Starling principle). Isotonic crystalloidsolutions or colloid solutions are equivalentas long as an appropriate quantity is administeredto achieve hemodynamic goals. Although somestudies have suggested that specific groups ofpatients may benefit more from a specific typeof fluid, further study is needed to confirm thesehypotheses. In general, crystalloids require two tothree times greater volume than colloids to achievesimilar goals. Hypotonic crystalloids should notbe used for volume resuscitation. Colloid solutionshave the potential advantage of achievingadequate volume resuscitation more quickly andwith smaller volumes. Some studies have suggestedthat hetastarch solutions may be associatedwith more renal dysfunction and increased mortality,but the different properties of these hetastarchsolutions preclude extrapolation of the results tothe entire class of fluids. Therapy with titratedboluses of fluids (500 to 1,000 mL of crystalloidand 300 to 500 mL of colloid) is recommended inpatients who are in shock, with close monitoringof the response until the hemodynamic goals areachieved.Vasoactive AgentsVasoactive medications for the treatment ofshock include agents with vasopressor, intropic,and/or vasodilator effects. Some agents havecombined effects (vasopressor and inotropic),and effects may vary with the dose. An agentshould be selected based on the physiology of thetype of shock, the effect desired, and the specificassessment of the patient. There have been noclinical trials that have established the superiorityof a single agent or combination of agents intreating shock. Common vasoactive agents, dosingranges, and adrenergic effects are listed inTable 9. In general, agents with predominantlyα-adrenergic properties will have greater vasopressoreffects, and agents with greater β-adrenergicproperties will have more inotropic andchronotropic effects. Therapy with potent vasopressoragents will generally result in increasesin afterload, MAP, and preload, with decreases incardiac output and relative bradycardia. Agentswith potent inotropic effects increase myocardialcontractility, cardiac output, and heart rate, andmay decrease BP.Severe Sepsis and Septic ShockConsensus guidelines have been developed forthe management of severe sepsis and septic shock.The clinician should be familiar with theserecommendations and the evidence supportingthe interventions. A recent revision of theinternational guidelines resulted in a total of 73recommendations, as follows: initial resuscitation354 Hemodynamic Monitoring and Shock (Zimmerman)ACC-PULM-09-0302-022.indd 3547/10/09 11:10:53 PM


Table 9. Vasoactive Agents in Shock: Adrenergic EffectsDrug Usual IV Dose α β 1β 2DopaminergicDopamine 1–2 μg/kg/min 1 0–1 0–1 32–10 μg/kg/min 2 1–2 0–1 310–30 μg/kg/min 3 1–2 0–1 3Dobutamine 2–30 μg/kg/min 0–1 3 1 0Norepinephrine 0.5–80 μg/min 3 2 0 0Epinephrine Low: 0.005–0.02 0 3 2 0μg/kg/minHigh: 0.02 2 3 1–2 0μg/kg/minPhenylephrine 20–200 μg/min 3 0 0 0Vasopressin 0.01– 0.04 U/kg/min 0 0 0 0(3 recommendations); infection issues (15 recommendations);hemodynamic support (13 recommendations);adjunctive therapy (9 recommendations);and supportive therapy (33 recommendations).Recommendations of particular interest are listedbelow. Recommendations with a grade of 1 arestrong recommendations, and those with a ratingof 2 are suggested recommendations. The letterdesignations refer to the quality of the evidencebehind them. Additional information on theguidelines can be found at www.survivingsepsis.org or in the published article.Initial Resuscitation• Initiate resuscitation immediately in patientswith hypotension or lactate concentration of 4 mmol/L. Grade of recommendation, 1C• Resuscitation goals are central venous pressureof 8 to 12 mm Hg, MAP of 65 mm Hg, urineoutput of 0.5 mL/kg/h, Scvo 2of 70% or Svo 2of 65%. Grade of recommendation, 1C• If the goals for Scvo 2and Svo 2are not achieved,consider therapy with additional fluid, transfusionto a hematocrit of 30%, or dobutamineinfusion to a maximum of 20 μg/kg/min. Gradeof recommendation, 2CInfection Issues• Begin therapy with IV antibiotics as soon aspossible but always within 1 h of recognizingthe presence of severe sepsis (grade of recommendation,1D) and septic shock (grade ofrecommendation, 1B)• Reassess the antibiotic regimen daily to optimizeefficacy, prevent resistance, avoid toxicity, anddecrease costs. Grade of recommendation, 1C• Consider using combination therapy for 3to 5 days and deescalate therapy when susceptibilitiesare known. Grade of recommendation,2D• Implement source control measures as soonas possible after initial resuscitation. Grade ofrecommendation, 1CHemodynamic Support• Use crystalloids or colloids for fluid resuscitation.Grade of recommendation, 1B• Maintain MAP at 65 mm Hg. Grade of recommendation,1C• Norepinephrine and dopamine are the initialvasopressors of choice. Grade of recommendation,1C• Epinephrine, phenylephrine, or vasopressinshould not be the initial vasopressor. Grade ofrecommendation, 2C• Epinephrine is the first alternative therapeuticagent to be used when BP is poorly responsiveto norepinephrine or dopamine. Grade of recommendation,2BAdjunctive Therapy• Consider therapy with IV hydrocortisone whenhypotension responds poorly to adequate fluidresuscitation and vasopressors. Grade of recommendation,2C• An adrenocorticotropic hormone stimulationtest is not recommended to identify patients who<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 355ACC-PULM-09-0302-022.indd 3557/10/09 11:10:53 PM


should receive hydrocortisone. Grade of recommendation,2B• Consider therapy with activated protein C inpatients with sepsis-induced organ dysfunctionand a clinical assessment of the high risk of death(APACHE [acute physiology and chronic healthevaluation] II score of 25 or multiple organfailure) if there are no contraindications. Gradeof recommendation, 2B• Patients with severe sepsis and a low risk of death(APACHE II score of 20 or single organ failure)should not receive activated protein C. Grade ofrecommendation, 1ASupportive Care• After initial resuscitation, transfuse RBCs whenhemoglobin concentration decreases to 7 g/dLto target a hemoglobin concentration of 7 to 9 g/dL. Grade of recommendation, 1B• Use IV insulin to control hyperglycemia followingstabilization. Grade of recommendation, 1B• Control glucose concentration to a target value of 150 mg/dL using a protocol for insulin adjustment.Grade of recommendation, 2CAnnotated BibliographyAmerican Society of Anesthesiologists Task Force on<strong>Pulmonary</strong> Artery Catheterization. Practice guidelinesfor pulmonary artery catheter. Anesthesiology 2003;99:888–1014An updated guideline with recommendations and rationalefor the role of PAC in the perioperative setting.Botero M, Kirby D, Lobato EB, et al. Measurement ofcardiac output before and after cardiopulmonary bypass:comparison among aortic transit-time ultrasound,thermodilution, and noninvasive partial CO 2rebreathing.J Cardiothoracic Vasc Anesth 2004; 18:563–572A study comparing three methods of determining cardiacoutput in cardiopulmonary bypass patients.Brunkhorst F, Engel C, Bloos F, et al. Intensive insulintherapy and pentastarch resuscitation in severe sepsis.N Engl J Med 2008; 358:125–139A randomized, controlled trial demonstrating no benefit oftight glucose control in patients with severe sepsis and anassociation of renal dysfunction with the use of pentastarch.Cohen MG, Kelly RV, Kong DF, et al. <strong>Pulmonary</strong> arterycatheterization in acute coronary syndromes: insightsfrom the GUSTO IIb and GUSTO III trials. Am J Med2005; 118:482–488A retrospective analysis of randomized international trialsin acute coronary syndromes showing increased mortalityassociated with PAC use except in patients in cardiogenicshock.Connors AF Jr, Speroff T, Dawson NV, et al. The effectivenessof right heart catheterization in the initial care ofcritically ill patients. JAMA 1996; 276:889–897This article, which revived the controversy of benefit of PAC,reports a retrospective study with extensive matching thatfound a higher mortality and use of resources in patients witha PA catheter.Cruz K, Franklin C. The pulmonary artery catheter: usesand controversies. Crit Care Clinics 2001; 17:271–291A review of the principles and interpretation of hemodynamicdata as well as discussion of controversies on use of PAC.Dark PM, Singer M. The validity of trans-esophagealDoppler ultrasonography as a measure of cardiac outputin critically ill adults. Intensive Care Med 2004; 30:2060–2066A systematic review of clinical studies to evaluate the validityof transesophageal Doppler ultrasonography.Dellinger RP, Levy MM, Carlet JM, et al. Surviving sepsiscampaign: international guidelines for management ofsevere sepsis and septic shock: 2008. Crit Care Med 2008;36:296–327A revision of the 2004 guidelines for severe sepsis and septicshock based on new evidence and consensus.ESCAPE investigators and ESCAPE study coordinators.Evaluation study of congestive heart failure and pulmonaryartery catheterization effectiveness. JAMA 2005;294:1625–1633Randomized controlled trial of PAC in patients with heartfailure (ejection fraction 30%) found no benefit on mortalityand hospitalization, and PAC use increased adverseevents.Harvey S, Stevens K, Harrison D, et al. An evaluationof the clinical and cost-effectiveness of pulmonaryartery catheters in patient management in intensivecare: a systematic review and a randomised controlledtrial. Health Technol Assess 2006; 10:iii-iv, ix-xi, 1–133An extensive review of PAC clinical trials and detailed findingsof the PAC-Man trial that concludes that withdrawal of PACfrom the United Kingdom would save money.Harvey S, Young D, Brampton W, et al. <strong>Pulmonary</strong> arterycatheters for adult patients in intensive care (review).Cochrane Database of Systematic <strong>Review</strong>s (databaseonline). Issue 3, 2006356 Hemodynamic Monitoring and Shock (Zimmerman)ACC-PULM-09-0302-022.indd 3567/10/09 11:10:54 PM


A comprehensive review of the quality and outcomes of clinicaltrials of PAC use.Hollenberg S. Vasopressor support in septic shock. Chest2007; 132:1678–1687A review of the evidence and experience with specific vasopressorsin treating patients with septic shock.Marik PE, Baram M. Noninvasive hemodynamic monitoringin the intensive care unit. Crit Care Clin 2007;23:383–400A review of newer noninvasive monitoring techniques andglobal indicators of perfusion.Marx G, Reinhart K. Venous oximetry. Curr Opin CritCare 2006; 12:263–268A review of the applications and limitations of SvO 2and ScvO 2as indicators of tissue oxygenation.Michard F, Teboul J. Predicting fluids responsiveness inICU patients. Chest 2002; 121:2000-2008A review of studies predicting response to fluids thatsuggests that dynamic rather than static parameters are moreuseful.National Heart, Lung, and Blood Institute Acute RespiratoryDistress Syndrome (ARDS) Clinical Trials Network.<strong>Pulmonary</strong> artery versus central venous catheterto guide treatment of acute lung injury. N Engl J Med2006; 354:2213–2224A randomized clinical trial of 1,000 patients with ARDS showingthat PAC-guided therapy did not improve outcomes butwas associated with complications as compared to therapyguided by a central venous catheter.PACEP Collaborative. <strong>Pulmonary</strong> Artery Catheter EducationProject. Available at: www.pacep.org. AccessedApril 2, 2008A self-paced educational program on basic principles of hemodynamics,correct analysis of waveforms, and interpretationof data developed by several critical care organizations.Polanczk CA, Rohde LE, Goldman L, et al. Right heartcatheterization and cardiac complications in patientsundergoing noncardiac surgery: an observational study.JAMA 2001; 286:309–314This study found more postoperative major cardiac and noncardiacevents in patients undergoing right-heart catheterizationeven when matched for multiple variables.Rhodes A, Cusack RJ, Newman PJ. A randomized, controlledtrial of the pulmonary artery catheter in criticallyill patients. Intensive Care Med 2002; 28:256–264This study found no difference in mortality in patients managedwith or without PAC.Richard C, Warszawski J, Anguel N, et al. Early use of thePAC and outcomes in patients with shock and acute respiratorydistress syndrome. JAMA 2003; 290:2713–2720This randomized trial in France found no harm but also nomortality or morbidity benefit from PAC.Russell JA, Walley KR, Singer J, et al. Vasopressin versusnorepinephrine infusion in patients with septic shock.N Engl J Med 2008; 358:877–887Randomized, controlled trial demonstrating the equivalenceof therapy with norepinephrine and therapy with vasopressinplus norepinephrine in treating patients with septic shock.Sakr Y, Vincent J-L, Reinhart K, et al. Use of the pulmonaryartery catheter is not associated with worse outcomein the ICU. Chest 2005; 128:2722–2731An observational study in European ICUs showing that PACuse was not associated with increased mortality.Sandham JD, Hull RD, Brant RF, et al. A randomized,controlled trial of the use of pulmonary-artery catheters inhigh-risk surgical patients. N Engl J Med 2003; 348:5–14This Canadian randomized trial of PAC use with goal-directedtherapy found no difference in mortality but a higher rate ofpulmonary embolism in the PAC group.Shah MR, Hasselblad V, Stevenson LW, et al. Impact ofthe pulmonary artery catheter in critically ill patients.JAMA 2005; 294:1664–1670A metaanalysis of randomized trials of PAC use showing PACuse did not increase mortality or benefit patients.Sharkey SW. Beyond the wedge: clinical physiology andthe Swan-Ganz catheter. Am J Med 1987; 83:111–122An older article with a good review and illustrations of hemodynamicwaveforms in a variety of clinical conditions.Sprung CL, Annane D, Keh D, et al. Hydrocortisonetherapy for patients with septic shock. N Engl J Med 2008;358:111–124Randomized, controlled trial of hydrocortisone therapy in thetreatment of patients with septic shock showing no benefit,even in nonresponders to an adrenocorticotropic hormonestimulation test.Tuman KJ, Carroll GC, Ivankovich AD. Pitfalls in interpretationof pulmonary artery catheter data. J CardiothoracicAnesth 1989; 3:625–641A thorough review of conditions that affect and sometimesinvalidate interpretation of PAC data.Uchino S, Bellomo R, Morimatsu H, et al. <strong>Pulmonary</strong>artery catheter versus pulse contour analysis: a prospectiveepidemiological study. Crit Care 2006; 10:R174;available at: http://ccforum.com/content/10/6/R174.Accessed May 19, 2007An observational study showing that the choice of monitoringdid not influence major outcomes in critically ill patients.Wiedemann HP, Matthay MA, Matthay RA. Cardiovascular-pulmonarymonitoring in the intensive care unit:part 1. Chest 1984; 85:537–549<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 357ACC-PULM-09-0302-022.indd 3577/10/09 11:10:54 PM


Wiedemann HP, Matthay MA, Matthay RA. Cardiovascular-pulmonarymonitoring in the intensive care unit:part 2. Chest 1984; 85:656–658Still a classic series covering PAC technique, complications,and physiologic correlates of measurements.Weiniger CF, Ginosar Y, Sprung C, et al. Arterial and pulmonaryartery catheters. In: Parrillo JE, Dellinger RP, eds. Criticalcare medicine: principles of diagnosis and management inthe adult. 2nd ed. St. Louis, MO: Mosby, 2001; 36–63A good textbook chapter on PAC use, including insertion,complications, and interpretation of data.Yu DT, Platt R, Lanken PN, et al. Relationship of pulmonaryartery catheter use to mortality and resource utilizationin patients with severe sepsis. Crit Care Med 2003;31:2734–2741A case-control study finding that use of a PAC was not associatedwith a change in mortality rate or resource utilization.358 Hemodynamic Monitoring and Shock (Zimmerman)ACC-PULM-09-0302-022.indd 3587/10/09 11:10:54 PM


Community-Acquired Pneumonia:Advances in ManagementRonald R. Grossman, MD, FCCPObjectives:• Identify the common pathogens causing communityacquiredpneumonia• Identify the risk factors for mortality to improve siteof-caredecisions• Improve antibiotic selection in the empiric therapy ofcommunity-acquired pneumonia• Recognize the risk of avian influenza and communityacquiredmethicillin-resistant Staphylococcus aureusKey words: community-acquired methicillin-resistant Staphylococcus aureus; CURB-65; Patient Outcomes Research Team(PORT) score; Streptococcus pneumoniaeCommunity-acquired pneumonia (CAP) affects5.6 million adults annually in the United Statesand causes 1.7 million hospitalizations per year. 1Combined mortality rates for pneumonia andinfluenza indicate that this is the sixth-leadingcause of death in the United States, accountingfor 83,000 deaths annually. It is the cause of 46%of all deaths from infectious disease. 2 Althoughcausative pathogens are identified in 50% ofcases, 3 the specific organisms that require coverageinclude typical organisms such as Streptococcuspneumoniae, Haemophilus influenzae, and Moraxellacatarrhalis. Atypical organisms, including Mycoplasmapneumoniae, Chlamydia pneumoniae, andLegionella pneumophila may have a presentation thatis more subacute, with nonproductive cough anda chest radiograph that appears characteristicallyworse than the patient’s clinical appearance, butthis presentation is not uniform by any means anddoes not allow clinical identification from patientswith infection from usual organisms. 4When the latest Infectious Diseases Society ofAmerica/American Thoracic Society (ATS) CAPguideline is used as a framework, patients withCAP are classified into one of two groups, eachwith a list of likely pathogens. 5 Stratification isbased on an assessment of the need for a specificsite of therapy (outpatient, inpatient ward, or ICU),the presence of comorbidities, and the likelihoodfor drug-resistant S pneumoniae (DRSP), entericGram-negative organisms, and Pseudomonas aeruginosa.Not every patient should be considered atrisk for infection with DRSP. Specific risk factorshave been identified, and there are regional differencesin resistance rates. The role of enteric Gramnegativeorganisms in CAP is controversial, butthese organisms do not need to be consideredunless specific risk factors are present; one of theserisk factors includes residence in a nursing home.S pneumoniae is the most common pathogen,and it may even account for pneumonia in patientswho have no pathogen identified by routine diagnostictesting (Table 1). Although the incidence ofDRSP is increasing, available data show that therate of mortality in patients with CAP is adverselyaffected by drug-resistant pneumococci only whenminimum inhibitory concentration values topenicillin are 4 mg/L. 6 The impact of organismsat lower levels of resistance remains uncertain.Some studies 7 have been unable to detect treatmentTable 1. Organisms Associated with Community-AcquiredPneumonia*Patient TypeOutpatientInpatient (non-ICU)Inpatient (ICU)*Adapted from Mandell et al. 5EtiologyS pneumoniaeM pneumoniaeH influenzaeC pneumoniaeRespiratory virusesS pneumoniaeM pneumoniaeC pneumoniaeH influenzaeLegionella sppAspirationRespiratory virusesS pneumoniaeS aureusLegionella sppGram-negative bacilliH influenzae<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 359ACC-PULM-09-0302-023.indd 3597/10/09 8:36:43 PM


failures when patients with drug-resistant strainsof S pneumoniae are treated with antibiotics that donot have in vitro activity against the pathogen.Treatment failures have been described amongpatients with infection treated with macrolideswhen pneumococci exhibit macrolide resistanceand fluoroquinolones when pneumococci exhibitfluoroquinolone resistance. 8,9 The mechanism ofmacrolide resistance does not appear to influencethe risk of macrolide failures. 10All patients with CAP could potentially beinfected with C pneumoniae, M pneumoniae, andLegionella sp (the “atypical” pathogens), eitheralone or as part of a mixed infection; thus, allpatients should receive therapy to account for thispossibility. 11 A meta-analysis 12 has suggested thatclinical outcomes are satisfactory when outpatientsare treated with -lactams alone without atypicalcoverage. However, most of the trials selected forthis meta-analysis were for antibiotic registrationpurposes, and the patients were highly selected.These trials examined late clinical outcomes, anddifferences in rates of recovery could easily bemissed by this analysis. This meta-analysis appliedonly to outpatients and does not speak to theappropriate management of hospitalized patientswith CAP. There are considerable data to suggestthat atypical coverage (either with macrolides orfluoroquinolones) is associated with better clinicaloutcomes, including reduced lengths of hospitalstay and rates of mortality. 13,14When patients with CAP are admitted to theICU, the organisms responsible include S pneumoniae,Staphylococcus aureus, Legionella sp, andenteric Gram-negative organisms. 15 P aeruginosahas been recovered from some patients with severeCAP, but this organism should only be consideredwhen patients have well-identified risk factorspresent. In a prospective study, Arancibia andcoworkers 16 identified that probable aspiration,previous hospital admission, previous antimicrobialtreatment, and the presence of pulmonarycomorbidity were independent predictors of Gramnegativepneumonia. Similarly, pulmonary comorbidityand previous hospital admission wereindependent predictors of pneumonia caused byP aeruginosa.Patients should be investigated for the presenceof microorganisms when the usual empiric regimenwould not be sufficient, especially when thepatient has clinical or epidemiologic considerationssuggesting the presence of unusual organisms. 5 Allpatients with CAP should undergo chest radiographyto establish the diagnosis and the presenceof complications (pleural effusion, multilobar disease).5 All outpatients should have a careful assessmentof disease severity, but sputum culture andGram stain are not required. All admitted patientswith CAP should have an assessment of gasexchange (oximetry or arterial blood gas), routineblood chemistry and blood counts, and a collectionof two sets of blood cultures. The standard recommendationfor blood cultures has recently beenchallenged. 17The risk of bacteremia is increased amongpatients with underlying liver disease, demonstratinghypotension (systolic BP 90 mm Hg),tachycardia (pulse rate 125 beats/min), extremesof temperature (< 35C or 40C), and exhibitingcertain abnormal laboratory features (BUN 30 mg/dL, sodium < 130 mmol/L, WBC count< 5,000 L or 20,000 L). The risk of bacteremia issignificantly decreased among patients who havereceived antibiotics previously. When researchersused a decision support tool whereby no samplesfor blood cultures would be drawn if the risk ofbacteremia was low, one blood culture if the riskof bacteremia was moderate, and two blood culturesif the risk of bacteremia was high, 88% ofbacteremias would be detected and 38% fewerblood cultures would be drawn. 17If a drug-resistant pathogen or an organism notcovered by usual empiric therapy is suspected,sputum culture should be obtained, and a Gramstain should be used to guide interpretation ofculture results. Routine serologic testing is notrecommended for any population with CAP. Forpatients with severe CAP, Legionella and pneumococcalurinary antigen should be measured, andaggressive efforts at establishing an etiologic diagnosisshould be made, including the collection ofbronchoscopic samples of lower respiratory secretionsin selected patients, although the benefit ofsuch efforts has not been proven. 5The admission decision can be difficult, andprognostic scoring rules (the Patient OutcomesResearch Team [PORT] and British Thoracic Society[BTS] rules) are adjunctive tools to support, but notreplace, this process. 18–20 In general, hospitalizationis needed if patients have multiple risk factors for360 Community-Acquired Pneumonia (Grossman)ACC-PULM-09-0302-023.indd 3607/10/09 8:36:44 PM


a complicated course. Patients may also need to behospitalized for a variety of nonmedical reasons,and such social factors should also be incorporatedinto the admission decision process.Admission to the ICU is needed for patientswith severe CAP, defined as the presence of eitherone of two major criteria or the presence of two orthree minor criteria (modified ATS). The majorcriteria include need for mechanical ventilation orseptic shock. The minor criteria include systolic BP 90 mm Hg, multilobar disease, and Pao 2/fractionof inspired oxygen ratio < 250. 5 Patients who havetwo of four criteria from the modified BTS rule alsohave more severe illness and should be consideredfor admission to the ICU. These criteria includerespiratory rate 30 breaths/min, diastolic BP 60mm Hg, BUN 7.0 mmol/L ( 19.1 mg/dL), andconfusion (ie, CURB [confusion, elevated urea,increased respiratory rate, decreased BP]). Themodified ATS rule achieved a sensitivity of 69%and specificity of 97% in predicting admission tothe ICU and a sensitivity of 94% and specificity of93% for predicting mortality. 21 When two variableswere used to predict outcome, the BTS-CURB criteriadid not perform nearly as well as the modifiedATS rules. 21Patients should initially be treated empiricallybased on the likely pathogens for each of the sitesof care, although when culture results become available,organism-specific therapy may be possible forsome patients. All populations should be treatedfor the possibility of atypical pathogen infection(Table 2). For outpatients or non-ICU inpatientswith risk factors for these other organisms, therapyshould be with either a -lactam/macrolide combinationor an antipneumococcal fluoroquinolonealone (Table 2). Although both regimens appeartherapeutically equivalent, particularly amonginpatients, in the outpatient treatment of themore complicated patient, an antipneumococcalfluoroquinolone may be more convenient than a-lactam/macrolide combination.All admitted patients should receive their firstdose of antibiotic therapy as soon as possible afterarrival to the hospital. 22 Although 4 h has been therecommended standard, it is not clear that everypatient must be treated that quickly because thediagnosis of pneumonia often takes 4 h to establishand treatment should not be started until otherdiagnostic possibilities have been excluded. In theICU-admitted patient, current data do not supportthe use of an antipneumococcal fluoroquinolonealone, and therapy should be with a -lactam pluseither a macrolide or fluoroquinolone, using aregimen with two antipseudomonal agents inappropriate, at-risk patients (Table 3). The morerecent Infectious Diseases Society of Americaguidelines use a similar format to the ATS guidelinesand have similar antibiotic recommendations.16 The additional twist is the recognition thatthe use of antibiotics in the previous 3 months isa significant predictor of antibiotic resistance tothat class.Accordingly, the guidelines recommend use ofan alternative class of therapy if an agent hasbeen used within the past 3 months. Retrospectivestudies 23 have suggested that combination antibiotictherapy (usually a -lactam plus a macrolide)is superior to single-agent therapy (usually a-lactam alone) among severely ill patients withbacteremic pneumococcal pneumonia. In aprospec tive, international, observational study, 24Table 2. Outpatient Treatment*1. Previously healthy and no use of antimicrobials within the previous 3 moA macrolideDoxycyline2. Presence of comorbidities such as chronic heart, lung, liver, or renal disease; diabetes mellitus; alcoholism; malignancies;asplenia; immunosuppressing conditions or use of immunosuppressing drugs; or use of antimicrobials within the previous3 mo (in which case an alternative from a different class should be selected)A respiratory fluoroquinolone (moxifloxacin, gemifloxacin, or levofloxacin [750 mg])A -lactam plus a macrolide3. In regions with a high rate ( 25%) of infection with high-level (minimum inhibitory concentration, 16 g/mL) macrolideresistantS pneumoniae, consider use of alternative agents listed above in (2) for patients without comorbidities (moderaterecommendation; level III evidence)*Adapted from Mandell et al. 5<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 361ACC-PULM-09-0302-023.indd 3617/10/09 8:36:44 PM


Table 3. Inpatient Treatment*Inpatient, non-ICUA respiratory fluoroquinoloneA -lactam plus a macrolideInpatient, ICUA -lactam (cefotaxime, ceftriaxone, or ampicillin-sulbactam) plus either azithromycin (level II evidence) or a respiratoryfl uoroquinolone (for penicillin-allergic patients, a respiratory fluoroquinolone and aztreonam are recommended)Special concernsIf Pseudomonas is a consideration:An anti-pneumococcal, anti-pseudomonal -lactam (piperacillin-tazobactam, cefepime, imipenem, or meropenem) pluseither ciprofloxacin or levofloxacin (750 mg); orThe above -lactam plus an aminoglycoside and azithromycin; orThe above -lactam plus an aminoglycoside and an anti-pneumococcal fluoroquinolone (for penicillin-allergic patients,substitute aztreonam for above -lactam)*Adapted from Mandell et al. 5combination antibiotic therapy was superiorto monotherapy but only for severely ill patients.Whether a respiratory fluoroquinolone is equivalentto two agents cannot be ascertained from theavailable data.Most patients with CAP will have an adequateclinical response within 3 days; when the patienthas met appropriate criteria, a switch to oraltherapy should be made. 25 Criteria for switchinclude improvement in cough and dyspnea, afebrile(< 37.8C) on two occasions 8 h apart, decreasingWBC count, and functioning GI tract withadequate oral intake. Even if the patient is febrile,a switch in therapy can occur if other clinical featuresare favorable. If the patient has met criteriafor a switch, oral therapy can be started and thepatient discharged on the same day if other medicaland social factors permit. Switch therapy issimpler and quicker if a respiratory fluoroquinoloneis used rather than a -lactam/macrolidecombination. Outcomes are identical to patientsstaying in hospital longer to complete their courseof therapy. 26Length of stay has been shown to be related tothree quality-of-care measures: initial administrationof antibiotics in the emergency department,appropriate antibiotic selection, and shorteneddoor to needle time. 27–30 Initiation of antibiotictherapy in the emergency department led to shorterhospital stays for CAP by approximately 2 days inone report. 28The following have been listed as qualityindicators in the assessment of care of the CAPpatient:• Antibiotics administered in a timely way,within either 4 h or 8 h of hospital admission;• Oxygen assessment or therapy within 8 to 24 hof hospital arrival;• Blood specimen for culture drawn before theadministration of antibiotics in the hospitalizedpatient;• Administration of antibiotics with activityagainst all likely causative pathogens, preferablythe least-expensive efficacious regimen;• Counseling patients regarding smoking cessation;• Switching from IV to oral antibiotics if thepatient is clinically improving and hemodynamicallystable, with discharge within 24 h ofswitching to oral therapy;• Chest radiography within 24 h of hospitaladmission; and• Use of methods to increase vaccination ratesagainst influenza and pneumococcus; and Nodischarge home for patients who are unstableon the day of discharge. 30In a metaanalysis, 3 the overall mortality rateamong 33,148 patients reported in 127 studies was13.7%, ranging from 5.1% in a population thatincluded both hospitalized and ambulatorypatients, to 13.6% for hospitalized patients. In theelderly, the mortality rate was 17.6%, whereasamong nursing home patients, it was 30.8%, andamong those admitted to the ICU was 36.5%. WhenS pneumoniae was responsible, the mortality ratewas 12.3%, a rate similar to that found when nopathogen could be identified (12.8%). Certain362 Community-Acquired Pneumonia (Grossman)ACC-PULM-09-0302-023.indd 3627/10/09 8:36:44 PM


pathogens such as P aeruginosa, other Gram-negativeorganisms (such as Klebsiella pneumoniae), andS aureus had greater associated mortality rates. Anepisode of CAP in young adults without significantcomorbidity is not a predictor of poor mediumtermsurvival. 31 Increasing age, comorbid cerebrovascularand cardiovascular disease, an alteredmental status, and increasing blood glucose areindependent predictors of decreased medium-term(ie, up to 3 years) survival. 31Treatment failure rates with guideline-drivenempiric therapy have been reported as high as15%. 32 Risk factors for treatment failure that havebeen identified include liver disease, pneumoniarisk class, leukopenia, multilobar CAP, pleuraleffusion, and radiographic signs of cavitation.Independent factors associated with a lower riskof treatment failure include influenza vaccination,initial treatment with fluoroquinolones, and aconcurrent diagnosis of COPD. Failure of empirictherapy increases the rate of mortality in patientswith CAP 11-fold after adjustment for risk class.Modification of these risk factors is possible andshould encourage prospective randomized trials.Emerging ThreatsCommunity-Acquired Methicillin-ResistantS aureusCommunity-acquired methicillin-resistantS aureus (MRSA) has emerged as a cause of CAPover the past several years. 33 It is quite distinct fromhospital-acquired MRSA from an epidemiologic,genetic, and clinical perspective. CommunityacquiredMRSA tends to be less resistant to antimicrobialagents than hospital-acquired MRSAstrains, and it almost always contains a type IVStaphylococcus cassette chromosome mec type IVgene (SCC mec IV) in addition to the genes forPanton-Valentine leukocidin, an extracellular cytotoxinthat is a virulence factor for primary skininfections and pneumonia. Although communityacquiredMRSA is not common, it may causepneumonia that is typically characterized by ashort duration of illness and focal necrotizing infiltrates.The disease may be rapidly progressive, soconsideration of this entity is important in themanagement of severely ill patients with CAP.In general, these strains are usually susceptiblein vitro to vancomycin, linezolid, trimethoprimsulfamethoxazole,doxycycline, and fluoroquinolones.In addition, they are often susceptible toclindamycin, but the presence of in vitro inducibleresistance should be determined before the drug isadministered. The initial choice in most publishedseries has been IV vancomycin, although linezolidappears to be at least equivalent to vancomycin inthe treatment of nosocomial MRSA pneumonia. Theefficacy of treatment with trimethoprim-sulfamethoxazolein systemic infections caused byMRSA was comparable with vancomycin, and ithas been effective alone and in combination withrifampin in patients with soft-tissue infectionscaused by community-acquired MRSA.Avian InfluenzaAvian influenza is an infectious disease ofbirds caused by type A strains of the influenzavirus. Migratory waterfowl, most notably wildducks, are the natural reservoir of avian influenzaviruses, and these birds are also the most resistantto infection. Domestic poultry, including chickensand turkeys, are particularly susceptible to epidemicsof rapidly fatal influenza. H5N1 variantsdemonstrated a capacity to directly infect humansin 1997 and have done so again in Vietnam inJanuary 2004. The spread of infection in birdsincreases the opportunities for direct infection ofhumans. If more humans acquire infection overtime, the likelihood also increases that humanswith concurrent infection with human and avianinfluenza strains could serve as a “mixing vessel”for the emergence of a novel subtype with sufficienthuman genes to be easily transmitted fromperson to person. Such an event would mark thestart of an influenza pandemic. In publishedseries, symptoms of fever, sore throat, cough and,in several of the fatal cases, severe respiratorydistress secondary to viral pneumonia developed.34 Previously healthy adults and childrenand some with chronic medical conditions wereaffected. Antiviral drugs, some of which can beused for both treatment and prevention, areclinically effective against influenza A virusstrains in otherwise-healthy adults and childrenbut have some limitations.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 363ACC-PULM-09-0302-023.indd 3637/10/09 8:36:44 PM


Newer StudiesA metaanalysis was performed on randomizedcontrolled trials of antibacterials for CAP in outpatientsaged 18 years. 35 Clinical success and rateof mortality were compared between different oralantibiotic classes, and antibacterials with atypicalcoverage (macrolides and fluoroquinolones) werespecifically compared with other antibacterials. Intotal, 13 eligible studies involving a total of 4,314patients were included. No significant differencewas detected regarding clinical success or mortality,regardless of atypical coverage or betweenantibacterial classes with similar atypical coverage.It was not possible to demonstrate any advantageof specific antibacterials for mild CAP in relativelyhealthy outpatients.In contrast, a retrospective cohort study 36 wasconducted in patients with pneumonia and clinicalcriteria of severe sepsis. Severe sepsis was presentin 30.1% of subjects, of whom 43.9% received macrolides.The rate of mortality was 20.3% at 30 daysand 24.5% at 90 days. In the multivariable analysis,the use of macrolide was associated with adecreased rate of mortality at 30 days and at 90days in patients with severe sepsis and in patientswith macrolide-resistant pathogens.A prospective multinational observationalstudy 37 was conducted to develop a propensityscore for combination therapy vs monotherapybased on baseline patient and infection characteristics.Patients were matched by the propensityscore and the 30-day mortality and hospital staywere examined. Patients treated with monotherapy(-lactam alone) were older, had a greater chronicdiseases score, and had a different clinical presentationcompared with patients treated with combinationtherapy (-lactam, macrolide). The unadjustedrate of mortality was significantly greater whenmonotherapy was used (22% vs 7%). Amongpatients in the monotherapy group matched topatients in the combination group, the mortality inthese groups was identical when the propensityscore was used. The benefit of combination therapyvs monotherapy cannot be reliably assessed inobservational studies because the propensity toprescribe these regimens differs markedly.In a retrospective analysis, 2,209 Medicarepatients with bacteremic pneumonia, admitted tohospitals from either home or a nursing facility,were stratified according to the type of antibiotictreatment. 38 Multivariate modeling was performedto assess the relationship between the class of antibioticused and several outcome variables. Theinitial use of any antibiotic active against atypicalorganisms was independently associated with adecreased risk of 30-day mortality and hospitaladmission within 30 days of discharge. Furtheranalysis revealed that the benefits of atypical coveragewere associated with the use of macrolidesbut not the use of fluoroquinolones or tetracyclines.Initial antibiotic treatment including a macrolideagent is associated with improved outcomes inMedicare patients hospitalized with bacteremicpneumonia. However, this study did not addressthe possibility that a fluoroquinolone as monotherapyis equally as effective as combinationtherapy with a macrolide/-lactam.To compare clinical outcomes of patients withCAP treated with and without atypical coverage,a secondary analysis was performed with the useof two comprehensive international databases. 39Patients treated with atypical coverage had decreasedtime to clinical stability, decreased length ofstay, decreased rate of total mortality, and decreasedCAP-related mortality. These findings would supportempiric therapy for all hospitalized patientswith CAP with a regimen that covers atypicalpathogens.In a prospective observational study of 2,457patients, 57 patients (2.3%) died 48 h afteradmission. 40 The overall rate of mortality was 7.7%.Independent factors associated with early deathwere increased age, altered mental status at presentation,multilobar pneumonia, shock at admission,pneumococcal bacteremia, and discordantempiric antibiotic therapy. Currently, early mortalityis relatively low and is caused by pneumoniarelatedfactors. Because pneumococcal bacteriemiaand discordant antibiotic therapy, mainly attributableto a lack of coverage against P aeruginosa, aresignificant risk factors, appropriate early interventionmay improve clinical outcome.In a prospective multicenter trial, CRP levelswere measured on admission and on days 3 and 7. 41Etiology could be determined in 47.4%; in 38.8%,initial antibiotic therapy was appropriate. Adecrease of < 60% in CRP levels in 3 days and a364 Community-Acquired Pneumonia (Grossman)ACC-PULM-09-0302-023.indd 3647/10/09 8:36:45 PM


decrease of < 90% in CRP levels in 7 days were bothassociated with an increased risk of having receivedinappropriate empiric antibiotic treatment. ConsecutiveCRP measurements may be useful in thefirst week in follow-up of antibiotic treatment forsevere CAP, and delayed normalization of CRPlevels is associated with a greater risk of havingreceived inappropriate antibiotic treatment.A retrospective national cohort study in whichthe authors used the Department of VeteransAffairs administrative data involving mainlyelderly men (n 8,652) examined the role of statinsand angiotensin-converting enzyme (ACE) inhibitorson the rate of mortality among patientsadmitted to hospital with pneumonia. 42 A total of9.9% of subjects died within 30 days of presentation.In this cohort, 18.1% of subjects were receivingstatins and 33.9% were receiving ACE inhibitors.After adjusting for potential confounders, currentstatin use and ACE inhibitor use were significantlyassociated with decreased 30-day mortality.In a prospective observational study 43 ofpatients admitted to the hospital with CAP, the useof statins, ACE inhibitors, -blockers, and aspirinwas recorded. On multivariate logistic regression,statin use was associated with significantly lower30-day mortality and development of complicatedpneumonia. There was no effect on requirement ofmechanical ventilation or inotropic support. Onmultivariate logistic regression, statin use wasindependently protective against a CRP that failedto decrease by 50% at day 4.In a secondary analyses of the Community-Acquired Pneumonia Organization database ofhospitalized patients with CAP and pneumococcalbacteremia and patients with CAP and negativeblood culture findings, investigators modeled allcausemortality and CAP-related mortality by theuse of logistic regression analysis and time-toclinicalstability and length of hospital stay by theuse of Cox proportional hazards models. 44 Themultivariable regression analysis revealed a lackof association of pneumococcal bacteremic CAPand time to clinical stability, length of hospital stay,all-cause mortality, and CAP-related mortality. Thisfinding suggests that factors related to severity ofdisease are confounders of the association betweenpneumococcal bacteremia and poor outcomes andthe presence of pneumococcal bacteremia by itselfshould not be a contraindication for the de-escalationof therapy in clinically stable hospitalizedpatients with CAP.In a prospective, observation study involving193 patients, 39% had a pathogen identified. Ofthese pathogens, 15% were viruses, 20% were bacteria,4% were mixed, and the rest were unknown. 45Influenza, human metapneumovirus, and respiratorysyncytial virus accounted for most viral infections.Compared with patients with bacterialinfection, patients with viral infection were older,more likely to have cardiac disease, and were morefrail. There were few clinically meaningful differencesin presentation and no differences in outcomesaccording to the presence or absence of viralinfection. This study suggests that viral infectionsare common in adults with pneu monia, raisingissues of incidental nosocomial transmission.In a secondary analysis 46 of a multicenter observationalstudy conducted among 457 patients withICU-admitted CAP, the presence of radiographicprogression and bacteremia was determined. Logisticregression analysis indicated that the groupswith radiographic progression plus bacteremia orradiographic progression alone had a greater riskfor shock than patients without either finding,whereas bacteremic patients had no increased risk.In addition, patients with radiographic progressionplus bacteremia or radiographic progression alonehad an increased risk of ICU death compared withpatients without either finding, whereas bacteremicpatients had no increased risk. This finding suggeststhat in patients with severe CAP, radiologic progressionof pulmonary infiltrates in the first 48 h isa poor prognostic feature, whereas bacteremia doesnot affect outcomes.In an observational retrospective study 47 of 500consecutive CAP patients, those who experiencedclinical failure were identified. Clinical failureoccurred in 13%. The most common etiologies forclinical failure related to CAP were severe sepsis(33%), acute myocardial infarction (28%), and progressivepneumonia (19%). All cases of severe sepsisoccurred in the first 72 h of hospitalization. The mostcommon etiology for clinical failure unrelated toCAP was the development of hospital-acquiredpneumonia (45%). The development of severesepsis early during hospitalization is the primaryetiology for clinical failure related to CAP.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 365ACC-PULM-09-0302-023.indd 3657/10/09 8:36:45 PM


Clinical, laboratory, and functional data wereprospectively collected on 1,813 adults with CAPadmitted to hospital outside of influenza season. 48The cohort consisted of 352 vaccine recipients and352 matched control subjects. Influenza vaccinationwas associated with a 51% reduction of mortalityoutside influenza season. Adjustment for age, sex,and comorbidities did not alter these findings.More complete adjustment for confounding (eg,functional and socioeconomic status) markedlyattenuated these benefits. This result would suggestthat previous observational studies may haveoverestimated the mortality benefits of influenzavaccination.A prognostic index was derived in 1,117 adultCAP patients and was validated among 646 consecutivepatients discharged from hospitals at alater time. 49 In the derivation cohort, preillnessfunctional status, Charlson index, and the severityon admission were independently associated with90-day mortality, and a scoring system was developed.The 90-day mortality was 0.7% in the lowriskgroup, 3.5% in the intermediate-risk group,and 17.3% in the high-risk group. The results in thevalidation cohort were very similar. The prognosticindex accurately stratified patients hospitalizedfor CAP into low-, intermediate-, and high-riskgroups for 90-day mortality on discharge.In a randomized controlled trial 50 in 596patients with COPD, the role of pneumococcalvaccination in preventing pneumonia was determined.There were 58 first episodes of CAP causedby pneumococcus or of unknown etiology. Therewas no significant difference between the interventionand nonintervention arms among all patients.However, the efficacy of pneumococcal vaccinationin younger patients with severe airflow obstructionwas 91%. Pneumococcal vaccination was effectivein preventing CAP in patients with COPD < 65years of age and in those with severe airflowobstruction but was ineffective among the othergroups of patients with COPD.In a database analysis 51 of 62,918 adults hospitalizedwith CAP, 12% had a record of previouspneumococcal vaccination. Vaccine recipientswere less likely to die of any cause during hospitalizationthan were individuals with no record ofvaccination, even after adjustment for the presenceof comorbid illnesses, age, smoking, and influenzavaccination and under varying assumptions aboutmissing vaccination data. Vaccination also loweredthe risk of respiratory failure and other complicationsand reduced median length of stay by 2 dayscompared with no vaccination.In a systematic review 52 of 15 randomized controlledtrials comprising 2,796 total subjects, therole of short-course therapy was examined. Shortcourseregimens primarily studied the use ofazithromycin (n 1) also were analyzed. Of theextended-course regimens, three studies used thesame antibiotic, whereas nine involved an antibioticof the same class. Overall, there was no differencein the risk of clinical failure, mortality, orbacterial eradication between the short-course andextended-course regimens. This finding wouldsuggest that adults with mild-to-moderate CAPcan be safely and effectively treated with an antibioticregimen 7 days.ConclusionSignificant progress has been made in the managementof patients with CAP. With the widespreaddevelopment of clinical practice guidelinesand the identification of processes of care that leadto well-documented improvement in outcomes,the prognosis of this disease should improve. Physiciansneed to be cognizant of the threats posedby community-acquired MRSA and possible pandemicinfluenza.References1. Niederman MS, McCombs JS, Unger AN, et al. Thecost of treating community acquired pneumonia.Clin Ther 1998; 20:820–8372. Merchant S, Mullins CD, Shih YT. Factors associatedwith hospitalization costs for communityacquiredpneumonia. Clin Ther 2003; 25:593–6103. Fine MD, Smith MA, Carson CA, et al. Prognosisand outcomes of patients with communityacquired pneumonia: a meta-analysis. JAMA 1996;275:134–1414. Fang GD, Fine M, Orloff J, et al. New and emergingetiologies for community-acquired pneumoniawith implications for therapy: a prospective multicenterstudy of 359 cases. <strong>Medicine</strong> 1990; 69:307–3165. Mandell LA, Wunderink RG, Anzueto A, et al.Infectious Diseases Society of America/American366 Community-Acquired Pneumonia (Grossman)ACC-PULM-09-0302-023.indd 3667/10/09 8:36:45 PM


Thoracic Society consensus guidelines on the managementof community-acquired pneumonia inadults. Clin Infect Dis 2007; 44:S27–S726. Feikin D, Schuchat A, Kolczak M, et al. Mortalityfrom invasive pneumococcal pneumonia in theera of antibiotic resistance, 1995–1997. Am J PublicHealth 2000; 90:223–2297. Yu VL, Chiou CC, Feldman C, et al. An internationalprospective study of pneumococcal bacteremia:correlation with in vitro resistance, antibioticsadministered, and clinical outcome. Clin Infect Dis2003; 37:230–2378. Vanderkooi OG, Low DE, Green K, et al. Predictingantimicrobial resistance in invasive pneumococcalinfections. Clin Infect Dis 2005; 40:1288–12979. Davidson R, Cavalcanti R, Brunton JL, et al. Resistanceto levofloxacin and failure of treatment ofpneumococcal pneumonia. N Engl J Med 2002;346:747–75010. Daneman N, McGeerA, Green K, et al. Macrolideresistance in bacteremic pneumococcal disease:implications for patient management. Clin InfectDis 2006; 43:432–43811. Marston BJ, Plouffe JF, File TF Jr., et al. Incidence ofcommunity-acquired pneumonia requiring hospitalization:results of a population-based active surveillancestudy in Ohio. The Community-BasedPneumonia Incidence Study Group. Arch InternMed 1997; 157:1709–171812. Mills GD, Oehley MR, Arrol B. Effectiveness of -lactam antibiotics compared with antibiotics activeagainst atypical pathogens in non-severe communityacquired pneumonia: meta-analysis. BMJ2005; 330:456–46013. Martinez JA, Horcajada JP, Almeld M, et al. Additionof a macrolide to a [beta]-lactam-based empiricalantibiotic regimen is associated with lowerin-hospital mortality for patients with bacteremicpneumococcal pneumonia. Clin Infect Dis 2003;36:396–39814. Gleason PP, Meehan TP, Fine JM, et al. Associationsbetween initial antimicrobial therapy and medicaloutcomes for hospitalized elderly patients withpneumonia. Arch Intern Med 1999; 159:562–57215. Ruiz M, Ewig S, Marcos M, et al. Etiology ofcommunity-acquired pneumonia: impact of age,comorbidity, and severity. Am J Respir Crit CareMed 1999; 160:397–40516. Arancibia F, Bauer TT, Ewig S, et al. Community-acquiredpneumonia due to Gram-negativebacteria and Pseudomonas aeruginosa: incidence, riskand prognosis. Arch Intern Med 2002; 162:1849–185817. Metersky ML, Ma A, Bratzler DW, et al. Predictingbacteremia in patients with community-acquiredpneumonia. Am J Respir Crit Care Med 2004;169:342–34718. Fine MJ, Auble TE, Yealy DM, et al. A predictionrule to identify low-risk patients with communityacquiredpneumonia. N Engl J Med 1997; 336:243–25019. Lim WS, van der Eerden MM, Laing R, et al. Definingcommunity acquired pneumonia severity onpresentation to hospital: an international derivationand validation study. Thorax 2003; 58:377–38220. Ewig S, Ruiz M, Mensa J, et al. Severe communityacquiredpneumonia: assessment of severity criteria.Am J Respir Crit Care Med 1998; 158:1102–110821. Ewig S, de Roux A, Bauer T, et al. Validation of predictiverules and indices of severity for communityacquired pneumonia. Thorax 2004; 59:421–42722. Houck PM, Bratzler DW, Nsa W, et al. Timing of antibioticadministration and outcomes for Medicarepatients hospitalized with community-acquiredpneumonia. Arch Intern Med 2004; 164:637–64423. Waterer GW, Somes GW, Wonderink RG. Monotherapymay be suboptimal for severe bacteremicpneumococcal pneumonia. Arch Intern Med 2001;161:1837–184224. Baddour LM, Yu VL, Klugman KP, et al. Combinationantibiotic therapy lowers mortality amongseverely ill patients with pneumococcal pneumonia.Am J Respir Crit Care Med 2004; 170:440–44425. Rhew DC, Tu GS, Ofman J, et al. Early switch andearly discharge strategies in patients with community-acquiredpneumonia: a meta-analysis. ArchIntern Med 2001; 161:722–72726. Fine MJ, Stone RA, Lave JR, et al. Implementationof an evidence-based guideline to reduce durationof intravenous antibiotic therapy and lengthof stay for patients hospitalized with communityacquiredpneumonia: a randomized controlledtrial. Am J Med 2003; 115:343–35127. Meehan TP, Fine MJ, Krumholz HM, et al. Qualityof care, process, and outcomes in elderly patientswith pneumonia. JAMA 1997; 278:2080–208428. Battleman DS, Callahan M, Thaler HT. Rapidantibiotic delivery and appropriate antibiotic selectionreduce length of hospital stay of patients withcommunity-acquired pneumonia. Arch Intern Med2002; 162:682–688<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 367ACC-PULM-09-0302-023.indd 3677/10/09 8:36:45 PM


29. Marrie TJ, Lau CY, Wheeler SL, et al. A controlledtrial of a critical pathway for treatment of community-acquiredpneumonia. JAMA 2000; 283:749–75530. Rhew DC, Goetz MB, Shekelle PG. Evaluatingquality indicators for patients with communityacquiredpneumonia. Jt Comm J Qual Improv2001; 27:575–59031. Waterer GW, Kessler LA, Wunderink RG. Mediumterm survival after hospitalization with community-acquiredpneumonia. Am J Respir Crit CareMed 2004; 169:910–91432. Menendez R, Torres A, Zalacain R, et al. Risk factorsof treatment failure in community acquiredpneumonia: implications for disease outcome.Thorax 2004; 59:960–96533. Vandenesch F, Naimi T, Enright MC, et al. Community-acquiredmethicillin-resistant Staphylococcusaureus carrying Panton-Valentine leukocidingenes: worldwide emergence. Emerg Infect Dis2003; 9:978–98434. The Writing Committee of the World Health Organization(WHO). Consultation on human influenzaA/H5: avian influenza A (H5N1) infectionin humans. N Engl J Med 2005; 353:1374–138535. Maimon N, Nopmaneejumruslers C, Marras TK.Antibacterial class is not obviously important inoutpatient pneumonia: a meta-analysis. Eur RespirJ 2008; 31:1068–107636. Restrepo MI, Mortensen EM, Waterer GW, et al.Impact of macrolide therapy on mortality forpatients with severe sepsis due to pneumonia. EurRespir J 2009; 33:153–15937. Paul M, Nielsen AD, Gafter-Gvili A, et al. The needfor macrolides in hospitalised community-acquiredpneumonia: propensity analysis. Eur Respir J 2007;30:525–53138. Metersky ML, Ma A, Houck PM, Bratzler DW.Antibiotics for bacteremic pneumonia. Chest 2007;131:466–47339. Arnold FW, Summersgill JT, LaJoie AS, et al. Aworldwide perspective of atypical pathogens incommunity-acquired pneumonia. Am J Respir CritCare Med 2007; 175:1086–109340. Garcia-Vidal C, Fernández-Sabé N, Carratalà J, etal. Early mortality in patients with communityacquiredpneumonia: causes and risk factors. EurRespir J 2008; 32:733–73941. Bruns AHW, Oosterheert JJ, Hak E, et al. Usefulnessof consecutive C-reactive protein measurementsin follow-up of severe community-acquired pneumonia.Eur Respir J 2008; 32:726–73242. Mortensen EM, Pugh PJ, Copeland LA, et al.Impact of statins and angiotensin-convertingenzyme inhibitors on mortality of subjects hospitalisedwith pneumonia. Eur Respir J 2008; 31:611–61743. Chalmers JD, Singanayagam A, Murray MP, et al.Prior statin use is associated with improved outcomesin community-acquired pneumonia. Am JMed 2008; 121:1002–100744. Bordón J, Peyrani P, Brock GN, et al. The presenceof pneumococcal bacteremia does not influenceclinical outcomes in patients with communityacquiredpneumonia. Chest 2008; 133:618–62445. Johnstone J, Majumdar SR, Fox JD, et al. Viralinfection in adults hospitalized with communityacquiredpneumonia. Chest 2008; 134:1141–114846. Lisboa T, Blot S, Waterer GW, et al. Radiologic progressionof pulmonary infiltrates predicts a worseprognosis in severe community-acquired pneumoniathan bacteremia. Chest 2009; 135:165–17247. Aliberti S, Amir A, Peyrani P, et al. Incidence, etiology,timing, and risk factors for clinical failurein hospitalized patients with community-acquiredpneumonia. Chest 2008; 134:955–96248. Eurich DT, Marrie TJ, Johnstone J, et al. Mortalityreduction with influenza vaccine in patients withpneumonia outside “flu” season: pleiotropic benefitsor residual confounding? Am J Respir CritCare Med 2008; 178:527–53349. Capelastegui A, España PP, Quintana JM, et al.Development of a prognostic index for 90-daymortality among patients discharged after hospitalizationfor community-acquired pneumonia.Thorax 2009; 64:496–50150. Alfageme I, Vazquez R, Reyes N, et al. Clinical efficacyof anti-pneumococcal vaccination in patientswith COPD. Thorax 2006; 61:189–19551. Fisman DN, Abrutyn E, Spaude KA, et al. Priorpneumococcal vaccination is associated withreduced death, complications, and length of stayamong hospitalized adults with communityacquiredpneu monia. Clin Infect Dis 2006; 42:1093–110152. Li JZ, Winston LG, Moore DH, Bent S. Efficacy ofshort-course antibiotic regimens for communityacquiredpneumonia: a meta-analysis. Am J Med2007; 120:783–790368 Community-Acquired Pneumonia (Grossman)ACC-PULM-09-0302-023.indd 3687/10/09 8:36:46 PM


PneumoconiosisDavid W. Kamp, MD, FCCPObjectives:• <strong>Review</strong> the epidemiology, pathophysiology, and clini calmanifestations of silicosis and coal worker’s pneumoconiosis• Understand the pulmonary manifestations of asbestosinduceddiseases of the lung (asbestosis and roundedatelectasis) and of the pleura (effusions and plaques)• Discuss the malignancies associated with asbestos exposure(bronchogenic carcinoma and mesothelioma)• Briefly review a variety of other pneumoconioses (talcosis,berylliosis, and hard metal lung disease)Key words: asbestos; coal; lung cancer; mesothelioma;pulmonary fibrosis; silicaPneumoconiosis is a 19th century Greek term(pneumo, meaning “breath”; konis, meaning “dust”)that describes lung diseases associated with mineraldust exposure. The term has evolved to imply theputative dust (eg, silicosis/silica, asbestosis/asbestos,berylliosis/beryllium, stannosis/tin) or workerswho are at risk (coal workers’ pneumoconiosis[CWP], hard metal lung disease). The InternationalLabor Organization (ILO) defines pneumoconiosisas the accumulation of mineral dusts inciting tissuereactions that range from minimal stromal reactionsthat are reversible to interstitial fibrosis that resultsin permanent scarring. 1 The term is often furtherrestricted to non-neoplastic pathology excludingasthma, chronic bronchitis, and emphysema, allof which can occur with occupational dust exposure.The patient with pneumoconiosis typicallypresents with nonspecific respiratory symptoms(eg, cough and dyspnea) and an abnormal chestradiograph finding. A careful occupational historyto ascertain the cause of a particular patient’s complaints(ie, whether acute or chronic, malignant ornonmalignant) is absolutely essential. This historyis especially important, given the large number ofworkers exposed and the significant rate of morbidityand mortality associated with mineral dustexposure.SilicosisDefinitionSilicosis is a chronic lung disease that is causedby the inhalation of crystalline silica (usuallyquartz and, less commonly, cristobalite, tridymite,coesite, and stishovite) and is characterized byprogressive parenchymal nodules and pulmonaryfibrosis. Silica or silicon dioxide is the most abundantmineral in the crust of the earth and is usedin a wide range of industrial products. Amorphousor noncrystalline silica particulates (eg, diatomiteand vitreous silica) are relatively less fibrogenic,but when they are combined with metal complexesto form silicates, as occurs with asbestos, mica, ortalc, they induce unique forms of pulmonary toxicity,each of which will be reviewed separately.Silica-associated lung disease is among theearliest lung diseases described and has been themost intensively studied occupational lung disease.In 1556, Agricola described the pulmonaryeffects of inhaled dusts from mining and the productionof metals; van Dumerbroek in 1672 notedsimilar problems in stone cutters. 2EpidemiologySilicosis is the most prevalent chronic occupationaldisease worldwide. The risk of silicosis isdirectly proportional to the particle concentration,the duration of exposure, and the silica content ofdifferent rock types, which ranges from nearly100% (sandstone and flint) to 10% (shale). In2006, the National Institute for Occupational Safetyand Health (NIOSH) estimated that each year 1million US workers are at risk for the developmentof silicosis and that 200 workers die from silicosis,with hundreds more becoming disabled. 3 Thedust control measures and workplace permissibleexposure limits (PELs) that have been establishedby most industrial countries have significantly<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 369ACC-PULM-09-0302-024.indd 3697/10/09 8:49:30 PM


decreased the number of deaths caused by silicosis.However, the number of silicosis-associated deathsamong persons 15 to 44 years of age has notdecreased significantly in the United States. 4,5 Further,30 to 50% of the workplaces inspected between1979 and 1982 had at least one air sample containinga concentration of 100 g/m 3 , whichis greater than the Occupational Safety andHealth Administration (OSHA) PEL for an 8-hwork exposure.Industries/Occupations at RiskSome of the major industries in which workersare at high risk for silica exposure are listed inTable 1, along with some examples of hazardousoccupations. These examples include industriesthat involve work in mines, in quarries, with stonework, and in foundries; in the use of abrasives; andin ceramic production. Another high-risk groupincludes persons working in road maintenance thatinvolves “cut and repair,” where potentiallyincreased levels of silica exposure occur during thecutting, breaking up, and removal of concrete. Acareful occupational history for silica exposure isessential for any patient presenting with pulmonarynodules, masses, or interstitial lung disease.This history is particularly important given thelong latency between toxic exposure and appearanceof the disease, the large number of workersexposed in past years before stringent work safetymeasures were invoked in the 1970s, and the factthat silicosis, like most pneumoconioses, can progressin the absence of additional exposure.PathologyThere are three distinct forms of silicosis, asfollows: chronic (the most common), acute, andaccelerated. The chronic form is characterized bysilicotic nodules that typically are located in theperibronchial regions with interstitial extension, 1 cm in diameter, well-formed, spherical, hard,and light gray (Fig 1, A). This is in contrast to theheavily black-pigmented, stellate nodules that areassociated with coal exposure (Fig. 1, B). The silicoticnodule has three components, as follows:(1) a central area of dense, acellular, hyalinizedcollagen-containing silica particles that are visiblewith polarized light; (2) a mid-zone of concentricallylayered collagen; and (3) a peripheral thickcapsule consisting of dust-laden macrophages andlymphocytes mixed with collagen. Silicotic nodules,which can develop decades after exposure,Table 1. Some High-Risk Industries and Occupations Associated With Silica Exposure*Industries (Examples)Mining, tunneling, and excavatingUnderground: gold, copper, iron, tinSurface: coal, iron, foundation excavationQuarryingGranite, slate, sandstoneStoneworkGranite sheds, monument masonryFoundriesIron and noniron metalsAbrasivesProduction: metal polish, paint fillersSandblasting, oil rigs, tombstonesCeramicsPottery, stoneware, oven bricksOthers (glass making, boiler scaling, gemstone worker, dental technician)OccupationsMiner, driller, tunnelerDrill operatorDigger, driller, hammererCutter, dresser, polisher, grinderMolder, caster, knockout manCrusher, mixer, abrasive workOven-brick maker*Modified from Becklake and Cowie . 1370 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3707/10/09 8:49:30 PM


Figure 1. A, silicotic nodule. B, coal macule.typically are scattered throughout the upper lungzones, sparing the majority of the lung parenchymaand thereby causing minimal symptoms (Fig 2).Simple silicosis, which is the earliest stage ofchronic silicosis, can become complex silicosis asthe nodules expand, affecting the bronchioles andvasculature and eventually coalescing into largemasses (progressive pulmonary fibrosis [PMF])that can undergo necrosis (Fig 3). However, theFigure 3. Progressive massive pulmonary fibrosis due tosilica exposure.presence of central cavitation should raise concernabout tuberculosis (TB). The adjacent lung mayshow signs of retraction and emphysema. If silicais mixed with other dusts (eg, coal, mica, andkaolin), then the nodules become less distinct andmore stellate in appearance. Silicotic nodules mayalso form in the hilar and mediastinal lymphnodes, which can become calcified (so-called eggshellcalcification, as shown in Fig 4) and mayimpinge on the airways (broncholith). Silicoticnodules can form outside the lungs (eg, liver,spleen, bone marrow, and kidney). In the kidney,silicon nephropathy can have a variable presentationfrom mild renal insufficiency to rapidly progressiverenal failure associated with necrotizingvasculitis.Acute silicosis is characterized by diffuse fluidfilledalveolar spaces that consist of eosinophilic,proteinaceous, and surfactant-containing material.The interstitial space typically shows extensivealveolar epithelial cell damage with hypertrophicalveolar type II cells as well as small focal nodulesand fibrosis.Pathogenic MechanismsFigure 2. Simple silicosis.The risk for silicosis depends on the level ofparticle exposure (dose and duration) as well as thecontent and type of silica inhaled. 6 The preciserelationship between specific silica subtypes andthe development of pulmonary fibrosis is unclear,but quartz, cristobalite, and tridymite are morefibrogenic than amorphous silica. Silica particles<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 371ACC-PULM-09-0302-024.indd 3717/10/09 8:49:30 PM


Figure 4. A, chest radiograph from a patient with silicosis showing left hilar adenopathy with egg-shell calcification.B, chest CT scan demonstrating egg-shell calcification.that are < 5 m in size can make their way to thedistal alveoli, where they are ingested by alveolarmacrophages (AMs), which are the first line ofalveolar defense. Silica-activated AMs subsequentlyrelease reactive oxygen species (ROS), reactivenitrogen species, fibronectin, and various cytokinesand growth factors. In particular, silica and otherparticulates activate the expression of transforminggrowth factor-, resulting in excess extracellularmatrix deposition. Silica also can directly injurealveolar type I cells, which leads to the formationof hyperplastic type II cells. If extensive alveolardamage occurs, as seen with intense exposure tohigh levels of silica, surfactant production andprocessing are altered, resulting in acute silicosis.PMF, which can occur with both silica and coalexposure, may ensue over time but may also occurin association with a mycobacterial infection.Figure 5 depicts some of the key pathogenicevents leading to silicosis and highlights the overlappingmolecular pathways that are activated byFigure 5. A simplified hypothetical model underlying mineraldust-induced pulmonary toxicity.other mineral dusts such as coal and asbestos. Theprecise mechanisms underlying the distinct clinicalmanifestations of each mineral dust are unknownbut likely relate in part to the differences in thedeposition patterns of the various dusts into thedistal airways, the chemical composition of eachparticulate, and the genetic background of theworker in whom the host response is regulated. Arecent case−control study involving 183 silicosispatients and 111 silica-exposed miners in Chinaidentified several interesting genetic polymorphismsthat will need to be verified by a largercohort. 7Clinical FeaturesPatients with silica-induced lung disease canpresent with chronic, accelerated, or acute onset ofdisease, depending on the intensity and durationof the silica exposure. Chronic silicosis, the mostcommon form, typically occurs after low-dosesilica exposure for 20 years. The prevalence estimatesof silicosis in exposed workers varies from0.4 to 11% after 20 years of exposure to the OSHAPEL of 100 g/m 3 to 1.2 to 53% after a cumulativeexposure of 4 mg/m 3 . 8 A study of US foundryworkers 9 showed a 12% prevalence of chest radiographchanges after 30 years and that black patientswere twice as likely to have abnormalities as whitepatients. Most patients with chronic simple silicosisare relatively asymptomatic, and the diagnosisof disease is established based on the occupationalexposure history and characteristic chest radiographfindings. A lung biopsy generally is reserved372 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3727/10/09 8:49:31 PM


for patients with atypical occupational exposureor chest radiograph findings. Silicotic nodules aregenerally small (1 to 3 mm), symmetrical, upperlobe—predominant lesions that favor the posterioraspects of the lungs but can involve other lungzones. Calcification is evident in 10 to 20% of thepulmonary nodules.Complicated silicosis consists of conglomeratednodules that can eventually form mass-likePMF lesions. The PMF lesions, unlike those inbronchogenic carcinoma, typically are symmetricalin location and may have an “angel wing”appearance. Eggshell calcification (Fig 4) is infrequent;if present, it strongly suggests a silicosisdiagnosis but is not pathognomonic. Over time,the mass lesions tend to contract, especially in theupper lobes, resulting in a rim of emphysemasurrounding the mass. Although central necrosissimilar to TB may develop in the mass, active TBshould be suspected whenever a lesion changesin size or cavitates. As the disease progresses,patients may note exertional dyspnea, cough, andsputum production. Inspiratory crackles anddigital clubbing are unusual and suggest an alternativecause such as another interstitial lungdisease, bronchogenic carcinoma, or infection. Ina study 10 of 243 patients with silicosis/mixed-dustpneumoconiosis and 62 patients with idiopathicpulmonary fibrosis (IPF), the prevalence ofchronic interstitial fibrosis, as defined by chest CTscan, was 12% in the silicosis/mixed-dust exposuregroup, of which 75% of the lesions were IPFlike.Cor pulmonale and cyanosis can occur in theadvanced phase of silicosis.Patients with accelerated and acute forms ofsilicosis typically present within months to a fewyears after intense silica exposure rather thandecades. Dyspnea occurs after 5 to 15 years of highlevels of occupational exposure to material withmoderate levels of quartz-containing silica. Thesepatients often progress to respiratory failure anddeath. Patients with acute silicosis generally presentwithin several months to years after intense exposureto high levels of silica (eg, from sandblasting,surface drilling, tunneling, silica flour milling, andceramic production), often with little or no respiratoryprotection. The chest radiograph is characterizedby a diffuse alveolar pattern similar tothat found in patients with alveolar proteinosis. Inthese patients, dyspnea, cough, and weight lossappear, which rapidly progress to respiratory failureand death.<strong>Pulmonary</strong> function tests (PFTs) correlatepoorly with the presence of nodules as assessed bychest CT scan. Patients with early silicosis mayshow normal volumes, air flows, and diffusingcapacity of the lung for carbon monoxide (Dlco).As the disease progresses or the level of silicaexposure increases, a progressive restrictive patternwith a reduced Dlco typically is found inthese patients. 11,12 Occasionally, an obstructive ormixed pattern may be evident. In patients withsimple silicosis, the chest radiographic abnormalitiesare typically more impressive than the PFTchanges. However, when compared with patientswith simple silicosis, patients with conglomerate,accelerated, or acute silicosis invariably demonstratesignificant and progressive restrictivephysiology and reduced Dlco.ComplicationsTB can occur at any stage of silicosis, but generallythe rate increases as silicosis progresses. <strong>Pulmonary</strong>TB must always be considered in patientswith PMF, a group with a significantly increasedrisk of morbidity and mortality. There is a threefoldgreater risk of pulmonary TB in patients withchronic silicosis than in healthy individuals. 13 Silicaexposure alone in the absence of silicosis is also arisk factor for pulmonary TB, and the increasedrisk persists after exposure to silica has stopped. 14Current smokers have a dose-dependent increasedrisk of developing TB compared with other silicoticpatients after correcting for confounders. 15 Becausesilica impairs AM function, the risk for variousmycobacterial infections such as Mycobacteriumtuberculosis, Mycobacterium kansasii, and Mycobacteriumavium-intracellulare is increased.A clinical suspicion of pulmonary TB may bethwarted by relatively nonspecific complaints suchas fever, weight loss, hemoptysis, malaise, orincreased dyspnea, all of which can be found inpatients with PMF or a variety of other infections.Because of impaired host defense mechanisms anddecreased drug penetration into silicotic nodules,antituberculous therapy should be extended for atleast 2 months beyond the typical 6-month coursefor patients who demonstrate no cavities on thechest radiograph and who have a smear result that<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 373ACC-PULM-09-0302-024.indd 3737/10/09 8:49:32 PM


is negative for acid-fast bacilli. 13 For patients withsilicosis who present with active pulmonary TBwith cavities and have smears positive for acid-fastbacilli, the recommended treatment duration is for11 months rather than 9 months. TB prophylaxisshould be continued for 12 months with isoniazidalone or for 4 months if a four-drug regimen consistingof isoniazid, rifampin, ethambutol, andpyrazinamide is used. Close posttreatment followupof patients with silica-mycobacterial infectionsis necessary because of the high rate of relapse. 13The risk for lung cancer in workers who havebeen exposed to silica but do not have disease iscontroversial. In 1997, the International Agency forResearch on Cancer reviewed the available experimentaland human epidemiologic data andconcluded that there was sufficient evidence implicatingsilica in the form of quartz or cristobalite asa group I carcinogen (ie, carcinogenic in humans). 8,16At least 20 studies have suggested that men withsilicosis have a twofold to fourfold increased riskof dying of lung cancer. However, the risk fromconfounding variables (eg, smoking and exposuresto other dust, such as asbestos) in some of thesestudies 12,17 is unclear. It is also uncertain whetherthe lung cancer risk is restricted to workers withsilicosis because pulmonary fibrosis per se is a wellknownrisk factor for lung cancer. Some studies 18,19have noted a dose−response relationship betweenlung cancer mortality rate and cumulativesilica exposure among workers without silicosis.In contrast, a 2006 analysis 20 of 28 cohort, 15case−control, and 2 proportionate mortality ratiostudies concluded that the carcinogenic role ofsilica alone in the absence of silicosis remainsunclear.Crystalline silica exposure has also been causallyattributed to a variety of other diseases, suchas various collagen vascular diseases (eg, rheumatoidarthritis, scleroderma, vasculitis, and systemiclupus erythematosus); sarcoidosis; glomerulonephritis;and hepatosplenic silicosis. 12,16 However,these associations are relatively rare and have notbeen well established.ManagementThere is no established treatment that can preventor reduce the pulmonary toxicity associatedwith silica exposure. Because pulmonary fibrosisis irreversible in workers with chronic silicosis andis directly related to the level of silica exposure, anyadditional silica exposure must be avoided. NIOSHrecommends an exposure limit of 50 g/m 3 ,whereas OSHA recommends a PEL of 100 g/m 3despite the fact that these levels have been associatedwith the development of silicosis. 8,9,12 Thereare several NIOSH-approved air-purifying respiratorswith replaceable N95 filters that have beenrecommended for silica-exposed workers. For theworkers who have the greatest risk and who areinvolved with abrasive blasting, the only approvedrespirator is the type CE abrasive blasting respiratorthat completely covers the head and operatesin a positive-pressure mode. Table 2 lists the recommendedchest radiograph surveillance guidelinesfrom various groups for silica-exposed workers. 21Treatment may be indicated to prevent silica-associatedcomplications, as described previously, aswell as assistance with pulmonary disability thatmay ensue in the advanced phases of silicosis.Lung transplantation has been performed in asmall number of patients with advanced silicosis,but it has a limited role in the management of silicosis,given the late age of onset of severe pulmonarydisability.Coal Workers’ PneumoconiosisDefinitionCWP, also known as black lung, is a distinctpathologic entity that results from chronic exposureto coal dust, graphite, and other forms ofcarbon. Some of the mineral and elemental contaminantscommonly found within coal includepyrite; kaolinite; ankerite; quartz; titanium dioxide;calcite; and trace metals, such as cadmium,copper, nickel, iron, lead, and zinc. CWP is distinguishedfrom silicosis partly by the unique pathologicchanges and the fact that carbon tends to beless fibrogenic than quartz. However, mixed-dustexposures are well documented. The Mine Safetyand Health Administration respirable coal dustPEL is 2 mg/m 3 , whereas NIOSH has set a standardof 1 mg/m 3 . 16 Coal is ranked into four typesbased on the carbon content, geologic age, andpresence of contaminating minerals and oxygen.The four types of coal are lignite, subbituminous,374 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3747/10/09 8:49:33 PM


Table 2. Pneumoconiosis Chest Radiograph Surveillance Guidelines*Particulate Frequency †Silica• OSHA• WHO• ACOEMAsbestos• OSHA• WHOPreplacement; every 5 years if 20 years of exposure; every 2 years if 20 years of exposure;at employment terminationPreplacement; then after 2–3 years of silica exposure; then every 2–5 years of silica exposureand thereafterPreplacement; every 3 years if 10 years of exposure; every 2 years if 10 years of exposure;at employment terminationPreplacement; every 5 years if 10 years of exposure; every 5 years if 10 years of exposureup to age 35 then every 2 years up to age 45; then annually thereafter or at employmenttermination if 0.1 fiber/cm 3 exposurePreplacement; every 3–5 years if 10 years of exposure; every 1–2 years if 10 years ofexposure; annually if 20 years of exposure*Adapted from Schwerha JJ. Occupational <strong>Medicine</strong> Forum. J Occup Environ Med 2008; 50:101–104.†Frequency of recommended chest radiographs by either the Occupational Safety and Health Administration (OSHA), WorldHealth Organization (WHO) or American College of Occupational and Environmental <strong>Medicine</strong> (ACOEM); may be increasedat the discretion of the managing physician.bituminous, and anthracite. Bituminous coalaccounts for nearly 43% of the US coal reserves; itis contaminated with low levels of silica and, inthe Pittsburgh area, surface-active iron (0.119mg/100 mg), which may account for its greaterpathogenicity. 22EpidemiologyUnderground miners, compared with surfaceor strip miners, have the greatest risk for the developmentof CWP. In the 1950s and 1960s, there wasan increase in mortality rate among coal miners inpart because of CWP. 1,2,16 The extensive dust-controlmeasures invoked in the 1960s through the 1970sin the United States and in other industrializedcountries resulted in a substantial reduction in theprevalence and mortality rate of CWP. It is estimatedthat 100,000 workers currently are employedin the coal-mining industry in the UnitedStates and that the estimated prevalence of CWP,defined as an ILO chest radiograph abnormalityof 1/0 or greater, varies from 3 to 14%, dependingon the mining region, with the highest prevalenceoccurring in workers with 30 years of miningwork. 2,16 Fortunately, complicated CWP or PMFdevelops in 0.5% of workers with CWP. 2 The riskfor PMF is directly proportional to the radiographicprofusion score and the levels of quartz in the coalbeing mined.PathologyThe initial lung tissue reaction to coal dust isthe formation of a coal macule and, when combinedwith surrounding focal emphysema, defines simpleCWP. Coal macules are typically up to 4 mm in sizeand tend to form at the junctions of respiratorybronchioles in the upper lungs (Fig 1, B). Themacule, consisting of coal dust–laden macrophagesand fibroblasts, enlarge, loosen the connectionsof the alveolar walls to the bronchioles, and therebycause focal areas of emphysema. Unlike the concentriccollagen arrangement in silicotic nodules,the nodules associated with CWP have haphazardlyarranged collagen bundles filled with dark-staininganthracotic pigment and lack the birefringentparticles found in silicosis tissue, which can be seenwith polarized light. However, coal-induced maculesare similar to silicosis in that they can enlargeand coalesce, resulting in masses that encroach onthe alveolar ducts and alveoli. In patients with eitherCWP or silicosis, when these masses are 1 cmradiographically, the PMF label can be used.PathogenesisCWP and silicosis are both closely related yetdistinct. The severity of CWP is closely associatedwith the levels of coal dust rather than quartz.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 375ACC-PULM-09-0302-024.indd 3757/10/09 8:49:33 PM


The prevalence of CWP is greater in anthracite coalminers than in bituminous coal miners in partbecause of the greater levels of crystalline silica andin part because of the greater levels of carbon-centeredfree radicals. Coal mine dust fractured bygrinding contains more free radicals than unfractureddust and is a better predictor of the toxicpotential of the dust. 22 As outlined in Figure 5,many of the patho-physiologic events implicatedin mediating CWP are similar to those involvedwith silica. Although coal and carbon are not toxicto AMs, the combination of coal and quartz is. Coaldust is generally less fibrogenic than silica andtends to blunt the effects of quartz. 23 In a study 24 of253 coal miners, genetic polymorphisms for genesencoding for tumor necrosis factor- and lymphotoxin-were associated with an increased risk forCWP. A role for interleukin (IL)-18, a lymphokinethat regulates neutrophil activation and the productionof proinflammatory cytokines and ROS, indecreasing coal-induced fibrotic lung disease wassuggested in a longitudinal study of 200 miners. 25These findings suggest that interactions betweenenvironmental dusts and certain genes are criticallyinvolved in determining disease expression inpatients with CWP, which is similar to the situationwith other pneumoconioses.Clinical FeaturesIn 1831, the first case of CWP was described,in a British coal miner. Although coal was believedto be relatively inert and the pulmonary toxicitylargely caused by silica, subsequent studies establishedCWP as a distinct entity. 16 The clinicalmanifestations of CWP that are similar to silicosisinclude simple CWP, complicated CWP, and PMF.However, unlike silica, chronic bronchitis is a moreprominent feature, and an accelerated or acuteonsetform of the disease does not occur. SimpleCWP is the most common and typically occursafter low-dose coal exposure for 20 years. Mostpatients with simple CWP are relatively asymptomaticbut can develop cough and sputum production.The diagnosis is based on the occupationalexposure history and characteristic chest radiographfindings. Nodules associated with coalexposure are generally small (ie, 1 to 3 mm), haphazardlyarranged lesions that favor the posterioraspects of the upper lobes but can involve otherlung zones. A lung biopsy generally is reserved forpatients with an atypical occupational exposure orchest radiograph findings.The onset of dyspnea on exertion typically isseen in patients with complicated CWP or PMF. Arim of emphysema surrounding the large nodulesis a prominent feature in patients with CWP and,at this stage of the disease, PFTs generally reveal amixed obstructive and restrictive defect. There isa well-described association between CWP andrheumatoid arthritis (Caplan syndrome), so anyclinical evidence of rheumatoid arthritis shouldalso be carefully evaluated. Similar to silicoticmasses, central necrosis may develop, raising thespecter of a complicating mycobacterial infection.The presence of crackles or clubbing is unusual andsuggests an alternative cause for the lesion, suchas another interstitial lung disease, bronchogeniccarcinoma, or infection. Cor pulmonale and cyanosiscan occur in the advanced phase of CWP.As reviewed in detail elsewhere, 26 the radiographicfindings of CWP are better detected byhigh-resolution CT (HRCT) of the chest as comparedwith chest radiograph. The characteristicHRCT findings include: (1) focal emphysema, (2)punctate opacities, (3) subpleural nodules, (4) PMF,(5) cavitating masses, (6) diffuse interstitial pulmonaryfibrosis, and (7) bronchiectasis.As in silicosis, PFT abnormalities correlatepoorly with the presence of nodules as assessed bychest imaging. Patients with early CWP may shownormal volumes, flows, and Dlco. As the diseaseprogresses or the levels of coal dust exposureincrease, a progressive restrictive pattern with areduced Dlco or a mixed obstructive/restrictivepattern is often evident resulting from the focalemphysematous changes around the coal macule.11,12 Expiratory airflow limitation occurs independentlyof the stage of CWP. Coal dust levelshave an inverse relationship with FEV 1. Patientswith complicated CWP or PMF caused by coalexposure invariably demonstrate a severe andrapidly progressive combined obstructive andrestrictive defect with a reduced Dlco.ManagementAs with silicosis, there is no established treatmentto prevent or reduce the fibrogenic effects ofcoal exposure and the complicating mycobacterial376 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3767/10/09 8:49:33 PM


infections that occur in both diseases. An unexplaineddichotomy is that the risk of lung cancerin coal miners is lower, but in patients with silicosis,the risk is greater than that in the generalpopulation when adjusted for age and smokingstatus. 12 Premature death is limited to patients withcomplicated CWP or PMF. A 23-year follow-upstudy 27 of 8,899 coal miners who were initiallyexamined between 1969 and 1971 at 31 US coalmines showed an increase in mortality from nonmalignantrespiratory diseases (eg, pneumoconiosisand COPD) but not from lung cancer.Management is also centered on the diagnosis andprevention of the coal-associated complicationsthat were described previously as well as assistancewith pulmonary disability that may ensue in theadvanced phases of CWP.Asbestos-Related DiseasesDefinitionAsbestos is a generic term for a group of naturallyoccurring hydrated silicate fibers, the tensilestrength and resilient structural and chemicalproperties of which are ideal for a variety of constructionand insulation purposes. The word asbestos,which is derived from the Greek word for“inextinguishable” or “unquenchable,” was firstused in the late 1300s. 28 However, industrial productionof asbestos did not occur until the 1850s.All forms of asbestos cause nonmalignant inflammatorypulmonary diseases (ie, pleural effusions,pleural plaques, rounded atelectasis, and asbestosis)and malignant pulmonary diseases (ie, mesotheliomaand broncho-genic carcinoma). 29–31EpidemiologyThe first cases of asbestos-associated fibrosiswere described in the early 1900s, and the termasbestosis was coined by Cooke 32 in 1927. Asbestosassociatedbronchogenic carcinoma in patientswith asbestosis was well-recognized by the mid-1950s; the association with mesothelioma wasestablished by the 1960s. Despite a dramaticdecrease in its use since the 1970s, asbestos-inducedpulmonary diseases continue to be a significanthealth concern for multiple reasons. First, 30million tons of asbestos have been mined, processed,and used in the United States since the early1900s. 29 Second, an estimated 27 million workersin the United States were exposed to aerosolizedasbestos fibers between 1940 and 1979. 33 Third,there is a long interval (called the latency period)between fiber exposure and the development ofasbestosis or mesothelioma (15 to 40 years), requiringlong-term, careful follow up of occupationallyexposed workers. Fourth, exposures from consumerproducts and from fibers released duringstructural renovation are a source of possible morbidityand mortality for both occupationallyexposed workers and the general population. 30,34Although the prevalence of asbestosis in theUnited States has not been established, in 2000there were an estimated 20,000 hospital dischargeslisting “asbestosis” and 2,000 deaths in whichasbestosis was the primary or contributing cause. 34Surveys 28 in Europe and the United States haveshown a doubling of the prevalence of asbestosassociatedpleural disease, including mesothelioma,from the early 1970s until 2000. It is estimatedthat the total number of asbestos-associated deathsin the United States may exceed 200,000 by the year2030. 35 Given these facts, it is not surprising thatasbestos-related diseases have inundated our legalsystem, resulting in recent very large class-actionlawsuits (68,000 individual claims were made inthe year 2000 alone) and the bankruptcy of manyold-line industrial companies. 28Asbestos is a well-recognized human carcinogen,but whether a threshold level exists that doesnot increase the risk of malignancy is unknown. 28,30,36OSHA established a PEL for fibers 5 mm longwith a 3:1 aspect ratio assessed by phase-contrastlight microscopy of 0.1 fibers per cubic centimeterover an 8-h period for all fiber types. Epidemiologicstudies have convincingly shown that all types offibers increase the malignant risk associated withoccupational exposure greater than the OSHA PEL.There is considerable controversy regarding themalignant risks associated with nonoccupationalexposure, especially to chrysotile, which accountsfor 95% of global asbestos consumption. 28,30,36Few well-designed epidemiologic studies havedirectly assessed the risk of malignancy from lowdoseasbestos exposure. A large-scale, retrospectivepopulation study 37 of 405 hospital-based patientsand control subjects concluded that 5 years ofexposure to the current OSHA PEL would produce<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 377ACC-PULM-09-0302-024.indd 3777/10/09 8:49:33 PM


a fourfold excess of pleural mesotheliomas. Asreviewed in detail elsewhere, 36 several recent studieshave documented more precisely an increasedrisk from environmental asbestos exposure thatappears to be approximately 10 times lower thanthat observed with occupational asbestos exposure.Industries/Occupations at RiskAsbestos exposure arises from the followingthree principal sources: (1) mining and milling ofthe fibers; (2) industrial applications of asbestos(eg, textiles, cement, friction materials, insulation,shipbuilding, repairing brake linings, lagging, orpipe cutting); and (3) nonoccupational exposure toairborne asbestos. The development of asbestosisis directly associated with both the magnitude andduration of the asbestos exposure. 6,29,33 Although asingle well-documented case 38 of asbestosis causedby brief inhalational exposure has been described,the overwhelming majority of patients have hadsignificant occupational asbestos exposure over aprolonged period. Airborne asbestos levels in publicbuildings are generally several orders of magnitudebelow the current OSHA standard, butgreater levels occur during renovation and demolition.Although asbestos-induced lung cancer andmesothelioma were well-recognized in the UnitedStates by the early 1960s, industrial productioncontinued until 1973 when it reached a peak of803,000 metric tons. 28 By 1998, it had decreased to16,000 metric tons.PathophysiologyThe toxic effects of asbestos depend on thecumulative dose and the time since the initialexposure. Essentially, all adverse effects of asbestosarise from inhalation. There are two classes ofasbestos fibers: serpentine and amphibole fibers.Serpentine fibers (eg, chrysotile) are curly-stranded,curved structures, whereas amphiboles (eg, crocidolite,amosite, and tremolite) are straight, rodlikefibers.The physical properties of the different asbestosfibers have been implicated in the pathogenesisof asbestos-induced diseases for many years.Amphibole fibers may be more toxic than theserpentine chrysotile, but this area is one ofconsiderable controversy because tremolite, whichis an amphibole, is a frequent contaminant that hasbeen implicated as a contributor to the toxicity ofchrysotile. Compared with chrysotile, amphibolefibers accumulate more readily in the distal lungparenchyma, are not cleared as effectively, and aremore durable (their estimated half-life in the lungsis on the order of months vs decades, respectively).6,30 Chrysotile-induced asbestosis typicallyrequires a threefold greater lung fiber concentrationcompared with amphiboles, but chrysotileinjures lung parenchymal and pleural cells and caninduce asbestosis, lung cancer, and mesotheliomain humans. 6,31Although fiber length has been shown to beimportant in the fibrogenic and malignant capacityof asbestos in animal and in vitro models, the resultsof human studies have been less impressive. 6,30The discrepancies between animal and human dataare in part attributed to the confounding effectsof cigarette smoke, which reduces lung fiberclearance. 6 Asbestos-induced lung cancer has beenattributed to the “amphibole hypothesis” that fiberstructural characteristics (ie, length, diameter, andaspect ratio) are the basis for lung malignancies. 6,30This concept has been questioned because fiberphysical properties alone appear to be insufficientto account for asbestos pulmonary toxicity, but theymay partly contribute to lung injury.Extensive investigations conducted during thelast 20 years have convincingly shown that ROSand reactive nitrogen species are important secondmessengers of toxicity caused by asbestos. 6,30,36,39The mechanisms underlying free radical generationby asbestos are in part caused by reactions occurringat surface sites on the mineral dusts, by mitochondrialdysfunction, and by the activation ofAMs or neutrophils attempting to take up the fibers.Once inside macrophages, asbestos fibers maybecoated with iron that is derived from hemosiderinand can be redox-activated. These coated fibers,called ferruginous bodies, can be seen with varioustypes of fibers, including asbestos, and as such arealso known as asbestos bodies. Asbestos fibers causepulmonary toxicity in a manner similar to silica andcoal via many of the same pathogenic pathways asshown in Figure 5. By mecha nisms that remainuncertain, asbestos-induced mitochondrial ROSsignaling in lung epithelial cells stabilizes p53and promotes p53-dependent transcription of378 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3787/10/09 8:49:34 PM


various proteins involved in tumor suppression,cell cycle arrest, apoptosis, and cell survival. 36Benign Asbestos Pleural EffusionsThe latent period between asbestos exposureand the development of a benign asbestos pleuraleffusion varies from 1 year (shortest latency ofall the asbestos-related diseases) to 40 years.The clinical presentation ranges from beingasymptomatic with total resolution or a bluntedcostophrenic angle to pleuritic chest pain associatedwith fever, dyspnea, and bloody pleuralfluid. The pleural fluid is usually unilateral, leftsided,and exudative yet rarely contains ferruginousbodies. The erythrocyte sedimentation rateoften is elevated. A diagnosis can be made onlyafter the appropriate exposure history, the exclusionof all other causes (especially malignancy),and close follow-up for 2 to 3 years. Although thishas not been firmly established, benign asbestospleural effusion has no clear prognostic implicationsfor the development of pleural plaques ormalignancy. 28Pleural PlaquesPleural plaques occur in 20 to 60% of occupationallyexposed workers 40,41 and in 2 to 6% ofnonoccupationally exposed individuals. 42 The mostcommon lesions are circumscribed or localizedpleural plaques, which are discrete areas of fibrosison the parietal pleura. Plaques serve as a markerof asbestos exposure and little else. They consist ofdense collagenous material that develops in themid-lower ribs and on the diaphragm (Fig 6). Electronmicroscopic studies may reveal noncoatedfibers and, rarely, ferruginous bodies. Over time,typically 30 years, these lesions calcify. The frequencywith which plaques occur varies widelybased on the sex (male female), age (up to 50%of shipyard workers older than the age of 70 years),and the number of years of occupational and nonoccupationalexposure. 28 The clinical consequencesof circumscribed plaques are minimal. In mostpatients, plaques are an incidental finding on aroutine chest radiograph. PFT results are invariablynormal unless they are confounded by anotherprocess. Conventional chest CT scans combinedwith selective high-resolution cuts are often usefulFigure 6. Asbestos-induced pleural plaques demonstratedon the diaphragm, parietal pleura, and interlobar fissure.to better define the extent of the plaques and toclarify whether confounding processes (eg, asbestosis,lung mass, fat tissue, or serratus anterior andexternal oblique muscles) contribute to the chestradiographic abnormalities. 28,43There is no firm evidence that pleural plaquesincrease the risk of mesothelioma or lung cancer.In a review 44 of 13 studies of the relationshipbetween pleural plaques and the risk for lung cancer,10 of the studies were deemed appropriate toaddress this issue; in none of the 10 was there adirect relationship found. In 1997, an InternationalExpert Meeting 45 concluded that the presence ofparietal pleural plaques alone were insufficient forattributing lung cancer to asbestos. Only 6.5% of234 patients with lung cancer and some asbestosexposure had pleural plaques without histologicasbestosis. 46 However, a review 47 of 38 asbestoscohort studies demonstrated a direct relationshipbetween the cumulative asbestosis rate and the rateof excess lung cancers. Thus, patients with asbestos-inducedpleural plaques without asbestosishave a minimally elevated risk for the developmentof bronchogenic carcinoma. A reasonableperiodic monitoring program for such patientswould include assessing current asbestos exposure<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 379ACC-PULM-09-0302-024.indd 3797/10/09 8:49:34 PM


isks, advising the patient of the importance ofsmoking cessation, and obtaining an intermittentchest radiograph (Table 1).Diffuse pleural plaques are a more widespreadform of pleural fibrosis involving the visceral andparietal pleural surfaces, resulting in blunting ofthe costophrenic angle. These lesions, which canbe unilateral or bilateral, are relatively infrequentand, unlike circumscribed plaques, diffuse pleuralplaques may result in restrictive pulmonary physiologyand functional impairment.Rounded AtelectasisRounded atelectasis is a distinct form of pleural/parenchymalthickening that is characterizedby its major radiographic finding, the “comet tail”sign. As pleural fibrosis progresses, it can entrapthe underlying healthy lung and bronchovasculartissue. These patients present with a pleural-basedmass that appears similar to bronchogenic carcinomabut can be distinguished from it based onthe characteristic asbestos exposure history, therelative stability of the lesion 2 to 3 years, otherradiographic signs of asbestos exposure, and thecomet tail sign pointing toward the hilum on thechest CT scan. The uncertain nature of the lesionmay occasionally necessitate a biopsy by means ofpercutaneous needle aspiration or, rarely, videoassistedthoracoscopic surgery. The mechanismsby which rounded atelectasis occurs are unclear,but the pathways implicated include the localfusion of the visceral and parietal pleura by thefibrotic process, the regional shrinkage of connectivetissue along the visceral pleura, and/or theabsorption of pleural effusions. 28Malignant MesotheliomaMalignant mesothelioma is a rare tumor thatis caused by all forms of asbestos. It occurs in thefollowing two sites: the parietal pleura and theperitoneum. It is much more frequent in the parietalpleura, possibly because inhalation is thetypical route of pathogenicity. The incidence ofmalignant mesothelioma has increased for decades,largely as a result of the extensive use of asbestosfrom World War II until the 1970s; the incidence isexpected to peak sometime between 2010 and 2020because of the extraordinarily long interval fromthe time of exposure to the onset of disease. 48,49Malignant mesothelioma occurred in nearly 2% ofasbestos textile workers exposed to approximatelyone fiber per cubic centimeter 50 years andaccounts for approximately 8% of the deathsamong asbestos workers. 50 In industrial countries,the yearly incidence of mesothelioma is 2 cases permillion person-years among women and 10 to 30cases per million person-years among men but asmuch as 270 to 366 cases per million person-yearsin men who have been exposed to crocidoliteasbestos. 49,51 Although occupational asbestos exposurecan be documented in 50 to 85% of patientswith malignant mesotheliomas, nonoccupationalexposure, radiation, chemical carcinogens, viruses,and chronic pleural disease also have beenimplicated. 30,36 Unlike bronchogenic carcinoma,there is no synergistic interaction between asbestosand cigarette smoke affecting the incidence ofmesothelioma.The typical clinical presentation of a patientwith malignant pleural mesothelioma is an elderly(age range, 50 to 70 years) man (male/female ratio,5:1) with an insidious onset (median time fromsymptom onset to diagnosis, 3 months to 2 years)of nonspecific complaints such as fever, weight loss,night sweats, and fatigue. 49 Dyspnea (50 to 70%),nonpleuritic chest pain (approximately 60%), andcough (approximately 30%) also may be present.Physical examination and chest radiographicfindings consistent with a pleural effusion are seenin 80 to 95% of patients. 28 A mediastinal shift towardthe side of the pleural effusion is seen occasionallyand indicates a potentially trapped lung. Chest CTscanning is more sensitive than a chest radiographfor detecting pleural plaques, differentiating apleural effusion from a tumor, assessing the extentof diaphragmatic as well as hilar and mediastinallymph node involvement, and identifying areas ofasbestosis. MRI of the chest is a promising approachfor distinguishing between benign pleural plaquesand malignant mesothelioma. 52The chest radiographic findings of malignantpleural mesothelioma, although often very suggestive,are not specific enough to establish a definitivediagnosis. Differentiating malignant mesotheliomafrom metastatic adenocarcinoma, benign pleuralmesothelioma, or reactive mesothelial cells ischallenging. No single immunohistochemicalmarker or ultrastructural feature is pathognomonic.380 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3807/10/09 8:49:35 PM


A diagnosis of mesothelioma begins with a strongclinical suspicion based on the appropriate exposurehistory. Pleural fluid cytology is diagnostic inonly 25 to 33% of patients, closed needle pleuralbiopsy is diagnostic in 21 to 77% of patients, andvideo-assisted thoracoscopic surgery is occasionallydiagnostic. 28 A panel of immunohistochemicalstains of pleural tissue is often helpful.Adenocarcinomas, with which mesotheliomais often confused, show periodic acid-Schiff-positivevacuoles and stain positively with carcinoembryonicantigen and Leu-M1 (CD15 antigen). Thefindings of these stains typically are negative inmesotheliomas. However, mesothelioma cells generallystain positively for vimentin. Mesotheliomasmay also elevate pleural fluid hyaluronic acidlevels, but this test is not always readily available.Newer markers of mesothelioma include calretinin,a calcium-binding protein that strongly stains 88%of malignant mesotheliomas, as opposed to adenocarcinoma,where it binds weakly in approximately50%. 48 Also, simian virus-40 DNA is detected inapproximately 66% of malignant mesotheliomasbut not in surrounding normal mesothelial cells. 48Simian virus-40 is not necessary for inducing mesotheliomasbut may contribute to its development,especially in the nearly 20% of patients with mesotheliomawho have not knowingly been exposedto asbestos.<strong>Review</strong>s published during the past severalyears examining whether various biomarkers (eg,mesothelin, megakaryocyte potentiating factor,osteopontin, soluble mesothelin-related protein,and others) are useful for discriminating betweenasbestos-exposed individuals with malignantmesothelioma as compared with other benign (eg,asbestosis, pleural plaques, etc) and malignantconditions (eg, lung cancer) have been limited bya significant number of false-positive results andnegative studies 53-55 Complementary DNA microarraytechnology has been used to identify nearly300 genes of the 6,500 genes surveyed that couldserve as useful markers of human mesothelioma. 56The translational significance of these recentlydescribed biomarkers awaits the outcome ofongoing large prospective studies investigatingthe diagnostic accuracy and the relationshipbetween biomarker levels and mortality. A combinationof these pathologic techniques and thetraditional history, clinical presentation, and chestradiographic findings should establish a definitivediagnosis.Unlike lung cancer, there is no widely acceptedstaging system for patients with diffuse malignantpleural mesothelioma. There are five different stagingsystems, but none of them has been adequatelyvalidated in comparative trials involving a sufficientnumber of patients. 28,57 Some of the favorableprognostic factors identified include the following:(1) absence of a 5% total body weight loss at thetime of diagnosis; (2) tumor confined to the ipsilateralparietal pleura; (3) epithelioid cell type;(4) good performance status; (5) young age; and(6) platelet count 400,000 cells/ L.No standard therapeutic regimen has beenclearly shown to substantially alter the mediansurvival time, which remains a dismal 6 to 18months, with a 5-year survival rate of 5%. 28,48Single-modality therapy, including various chemotherapeuticagents, radiation therapy, intrapleuralchemotherapy, intrapleural immunotherapy (eg,interferon- or IL-2), or surgery (eg, pleurectomyor extrapleural pneumonectomy), for highlyselected patients rarely prolongs survival beyondseveral months. 48,49,57 A large phase III study 58,59showed that cisplatin, which is the single mostefficacious chemotherapeutic agent (response rate,approximately 17 to 23%; median survival time,9 months), is more effective when combined withpemetrexed (response rate, 41%; median survivaltime, 12 months). Multimodality therapy withvarious combinations of chest surgery, chemotherapy,and radiation has been adapted. In 183patients who were treated with taxol, carboplatin,extrapleural pneumonectomy, and radiation therapy,the overall median survival time was17 months, and the 5-year survival rate was 14%. 60However, 2-year and 5-year survival rates of 68%and 46%, respectively, were observed for highlyselected patients with stage I disease and epithelialtypehistopathology. Although select patients maysurvive longer when managed with multimodalitytherapy, no randomized, multicenter, phase IIIstudies have shown that this approach is superiorto any other therapy.AsbestosisAsbestosis is a slowly progressive, diffuse pulmonaryfibrotic process caused by the inhalation<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 381ACC-PULM-09-0302-024.indd 3817/10/09 8:49:35 PM


of asbestos fibers. There is a latency period of atleast 20 to 40 years from the time of initial exposureto the development of respiratory symptoms. Theearliest symptom of asbestosis is insidiously progressiveexertional dyspnea that often progressesinexorably regardless of further asbestos exposure.Cough with sputum production is generally attributedto exposure to cigarette smoke rather thanasbestos. Patients with asbestosis may have bibasilarfine end-inspiratory crackles (32 to 64%), clubbing(32 to 42%) and, in the advanced stages, signsof cor pulmonale. 29,33The chest radiograph usually reveals smallparenchymal opacities in the lower lung zones witha nodular and/or reticular pattern that is similar tothat seen in patients with IPF. Unlike IPF, pleuralinvolvement is a hallmark of asbestos exposure.Pleural plaques are usually present but may beabsent in 10 to 20% of patients. 43 In the early stagesof asbestosis, the interstitial and pleural involvementbegin in the mid to lower lung zones, resultingin a hazy, ground-glass appearance that blurs theheart border (called the shaggy heart sign) andthe diaphragm on the chest radiograph. 40 Neitherthe mediastinal nor hilar lymph nodes are enlarged.Honeycombing and upper lobe involvementdevelop in the advanced stages of asbestosis.Because the chest radiographic changes associatedwith asbestosis may be subtle, a HRCT scanis frequently obtained because of its greater sensitivity.The characteristic HRCT scan findings ofasbestosis, although nonspecific, include the following:(1) subpleural linear densities parallel to thepleura; (2) parenchymal bands 2 to 5 cm in length,often contiguous with the pleura; (3) thickenedinterlobular septal lines; and (4) honeycombing. 43In a study 61 of 169 asbestos-exposed workers witha normal chest radiograph findings, HRCT scanfindings consistent with asbestosis were noted in57 patients (33.7%). These workers also had reductionsin both vital capacity and Dlco and anincreased dyspnea score in comparison with workerswho had normal HRCT scan findings.<strong>Pulmonary</strong> function abnormalities in patientswith asbestosis demonstrate restricted lung volumesand diminished Dlco, but otherwise, theirspirometry results are normal. 29,33,62 Abnormal spirometryresults (FEV 1, 75% predicted) generallyreflect concomitant exposure to cigarette smoke.Although controversial, some groups havesuggested 62 that asbestos alone may cause airwaysobstruction in part attributable to the large airwaysinflammation resulting from fiber deposition alongthe respiratory bronchioles and alveolar ductbifurcations.A diagnosis of asbestosis is based on severalcharacteristic features and does not require histopathologicevaluation for compensation purposes.29,33 An American Thoracic Society (ATS)expert panel concluded originally in 1986 33 andagain in 2004 29 that a clinical diagnosis of asbestosisis established, in the absence of lung tissue, bythe following: (1) a reliable exposure history, (2) anappropriate latency period, (3) characteristic chestradiograph abnormalities of 1/1 or greater basedon the ILO scale, (4) reduced lung volumes and/orDlco, and (5) end-inspiratory crackles. The 2004ATS statement 29 encompassed chest radiograph aswell as HRCT scan abnormalities in the clinicaldefinition and established the role of BAL fluid (ie,showing more than one asbestos body per milliliter,which is indicative of a high probability ofsubstantial occupational asbestos exposure),whereas asbestos fiber analysis was deemphasized.For patients with an atypical presentation, the mostdefinitive evidence supporting the diagnosis ofasbestosis is lung tissue demonstrating asbestosbodies in the setting of diffuse interstitial pulmonaryfibrosis. Asbestosis can have a patchy distributionthat favors the lower lung zones, similar tothat of IPF. The early phase of asbestosis is characterizedby peribronchiolar fibrosis with normaldistal alveoli. In the advanced stages of asbestosis,the fibrotic process, similar to that in IPF, extendsinto the distal airspaces and is associated with tractionbronchiectasis and honeycombing. Conglomeratelesions are unusual unless the worker hasbeen exposed to a mixed dust containing silica.Asbestosis is generally a slowly progressiveprocess that culminates in respiratory failure. Inseven studies 63 involving 2,200 asbestos workerswho were followed up from 1 to 17 years after theirinitial diagnosis of asbestosis, progression occurredin 5 to 31% of the patients, as assessed by changesin the chest radiograph or PFT results. Risk factorspredicting the progression of asbestosis include thefollowing: cumulative asbestos exposure (totalamosite/crocidolite fiber burden, 100,000 fibersper gram of dry lung); duration of exposure; andcrocidolite exposure. 63 A study 64 of the mortality of382 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3827/10/09 8:49:35 PM


655 male workers with asbestosis found that of the233 workers who died, 20% died from asbestosis,39% died from lung cancer, 9% died from mesothelioma,and 32% died from other causes. Giventhe long latency between exposure and disease aswell as the direct relationship between asbestosconsumption in 33 countries and mortality fromasbestos-related diseases (eg, asbestosis, mesothelioma,and lung cancer), the ILO has called for atotal worldwide asbestos ban. 65Lung CancerBy the mid-1950s, epidemiologic data firmlysupported a relationship between asbestosis andlung cancer. In the mid-1980s, the EnvironmentalProtection Agency and the World Health Organization’sInternational Agency for Research declaredasbestos a proven human carcinogen. From a 2001review 66 of 23 studies of the associations amongasbestos, cigarette smoke, and lung cancer, it wasconcluded that asbestos causes lung cancer innonsmokers despite the small numbers of suchworkers available for study. Moreover, a multiplicativeor synergistic interaction rather than anadditive model better described the associationbetween asbestos and cigarette smoke. This distinctionimplies that the combined attributable lungcancer risk exceeds the sum of the risk of eachagent. The mechanisms underlying this synergisticeffect are not fully known but are caused in partby impaired lung fiber clearance and enhancedDNA damage. 6,39 Early studies suggested thattumor location and histology differ in asbestos- andcigarette smoke-exposed individuals, but this hasbeen refuted by others as reviewed elsewhere.Asbestos-induced lung cancers can occur in anylobe of the lung, and the distribution of the fourmajor histopathologic lung cancer types is similarto the distribution pattern among patients withcigarette smoke-induced lung cancer. 67,68Considerable controversy surrounds thehypothesis that excess lung cancer risk in thosepersons with an occupational asbestos exposure islimited to workers with asbestosis. There is generalagreement that histologically and radiographicallydefined asbestosis, in addition to other forms ofpulmonary fibrosis, significantly increase the riskof lung cancer. Weiss 47 reviewed 34 cohort studiesand reported a direct correlation between the rateof asbestosis and lung cancer, suggesting thatasbestosis is a better predictor of excess lung cancerrisk than measures of asbestos exposure. Others 69have challenged this conclusion, arguing thatasbestos exposure, and not asbestosis, causes lungcancer.Because it is not necessary to have emphysemato implicate cigarette smoke as a cause of lungcancer, it might not be necessary to have asbestosisfor implicating asbestos as the attributable causeof lung cancer. This controversy is not likely to beresolved soon because of the nonuniform definitionof asbestosis used in the various studies (eg, clinicalvs histopathologic) and the uncertain biologicalscenario whereby the molecular mechanismsunderlying interstitial fibrosis are required todevelop a malignancy. Oksa et al 70 showed thatlung cancer developed in 11 of 24 patients withprogressive asbestosis (46%), whereas lung cancerdeveloped in only 5 of 54 patients with stableasbestosis (9%). They postulated that the progressionof asbestosis, in addition to cigarette smokeand asbestos exposure, is an independent predictorof lung cancer risk in patients with asbestosis.Asbestos can act at all the critical steps in theformation of a malignant clone of cells (eg, initiation,promotion, and progression) and, as such, isnot dependent on the presence of fibrosis. Asbestosexposedworkers can have mutations in the k-rasgene at codon 12 in lung cancers without radiographicdetermination of asbestosis, suggestingthat these two events are not necessarily linked. 30Thus, it is unclear whether asbestosis is simply amarker of high-dose asbestos exposure or a necessaryrequirement for attributing an individual’slung cancer to asbestos. Until more definitive studieshave clarified this controversy, lung cancerattribution should be based on the merits of eachpatient’s carcinogen exposure history combinedwith the appropriate clinical history and laboratoryfindings.Talc PneumoconiosisDefinition/OccupationsTalc is a heterogeneous term that includeshydrated magnesium silicates, and it is used commerciallyto describe mixed materials that maycontain minimal amounts of talc. Relatively pure<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 383ACC-PULM-09-0302-024.indd 3837/10/09 8:49:36 PM


talc, which is widely used in the cosmetic industry,is mined primarily in Vermont, the FrenchPyrenées, and Italy. Less pure talc (approximately60%) that is contaminated with asbestos, carbonates,and silica is mined from other areas (eg, NewYork, Texas, and California) and is used extensivelyfor a variety of industrial purposes. Talc is foundnear the surface of the earth and is mined in openpits. Rock containing talc is crushed into a finepowder, bagged, and shipped for numerous commercialuses in the manufacturing of ceramics,rubber products, chemicals, cosmetics, and pharmaceuticalsand as a filler in paint, paper, soaps,and roofing material.EpidemiologyTalc pneumoconiosis, also known as talcosis,typically develops in heavily exposed occupationalworkers. 2 The first case of talc pneumoconiosis wasdescribed in 1896. With modern dust control measuresin mines and in industrial talc plants, theincidence of talcosis is a relatively rare one today.Because talc is frequently contaminated with othermineral dusts, other lung diseases occur in talcexposedworkers. Although widely used by consumers,talc rarely causes disease except in adultsexposed to massive amounts of baby powder. Also,talcosis can develop in drug abusers who inject orinhale crushed tablets containing talc.PathophysiologyThe toxic manifestations of talc depend on thepurity of the material to which the workers areexposed. Nodular lesions in the respiratory bronchiolessimilar to those seen in silicosis are characteristicbut, unlike silicosis, talc is rarely seen in thenodule, nor are birefringent silica crystals visible.Multinucleated giant cells can occur, but focal anddiffuse fibrosis occurs in the later stages. A mixednodular and diffuse fibrotic process may be evidentpathologically as well as radiologically. In IV drugabusers, talc granulomas may be seen in the interstitiumas well as in the pulmonary arteries, resultingin pulmonary hypertension and cor pulmonale.Drug abusers may also present with very severeemphysematous changes associated with areas ofgranulomatous inflammation and fibrosis surroundingthe talc. 71Clinical Features/ManagementThe nodular form of talcosis is similar, bothclinically and radiographically, to silicosis in thatpatients are typically asymptomatic despite substantialchest radiographic abnormalities. A nonproductivecough and dyspnea may develop manyyears after the initial exposure. The onset of diffusefibrosis results in cough and dyspnea, bibasilarinspiratory crackles on examination, and chestradiographic abnormalities that are similar to thoseseen in silicosis and other forms of pulmonaryfibrosis. PFT abnormalities are minimal in thenodular form of talcosis but can demonstrate arestrictive defect with a reduced Dlco once fibrosisoccurs. There is no proven treatment for talcpneumoconiosis; therefore, the prevention of furtherexposure is the key to management. Therehave been anecdotal reports of a benefit from steroidtherapy in symptomatic patients with talcosiswhose disease is progressing.Beryllium Lung DiseaseDefinition/OccupationsBeryllium is a rare light metal that is extractedfrom beryl ore that is mined in the United States,Brazil, and China, and is refined in the UnitedStates, Japan, and Russia. Beryllium has a very highmelting point and a very low density, elasticity, andcoefficient of thermal expansion. It also is permeableto radiation. These properties make it ideallysuited for use in heat shields, rotor blades, radiographtubes, and parts for microwave equipmentand nuclear reactors. 2,72 Shortly after beryllium wasfirst used commercially in phosphors for fluorescentlights in the United States in the 1930s, its toxicproperties were recognized. Since then, an estimated800,000 workers have had past or currentberyllium exposure, resulting in chronic berylliosisin nearly 1 to 21% of the workers. 2,72 Beryllium cancause the following two types of pathologic conditionsacute pneumonitis caused by short-term,high-level exposure and chronic disease caused bylong-term, low-level exposure causing a granulomatousinflammation similar to sarcoidosis.Worker exposure to beryllium occurs primarily inthe aerospace, nuclear, computer, and electronicsindustries.384 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3847/10/09 8:49:36 PM


PathophysiologyAcute beryllium pneumonitis is caused bydirect acute lung injury resulting from the inhalationof high levels of beryllium particles. Berylliumalso incites an antigen-specific immuneresponse that can cause a noncaseating granulomatousinflammatory reaction that is histopathologicallysimilar to that seen in patients withsarcoidosis. Persistent low-level beryllium exposureresults in chronic berylliosis, which is manifestedby chronic interstitial fibrosis, often withbullous changes, as well as systemic involvementin the skin, liver, spleen, lymph nodes, myocardium,kidney, bones, and salivary glands. T lymphocytesfrom the lung and blood of patients withberylliosis proliferate when exposed in vitro toberyllium salts, which act as an antigen or hapten.This reaction serves as an important early markerfor diagnosing berylliosis and also has been implicatedin the pathogenesis of lung fibrosis. 2,73 Upto 16% of beryllium workers are sensitized toberyllium as assessed in the lymphocyte proliferationtest. 2Clinical Features/ManagementAcute beryllium disease is now relatively rarebecause of improved industrial hygiene methods,but when it does occur, beryllium acts as a directirritant. The diagnosis should be suspected whenthere is high-level beryllium exposure in associationwith acute pneumonitis, conjunctivitis, periorbitaledema, nasopharyngitis, and tracheobronchitis.These patients typically present with dyspnea,cough, sputum production, chest pain, tachycardia,crackles, and hypoxemia. Chest radiographs typicallyreveal diffuse alveolar infiltrates, whereas PFTresults show a restrictive process with a reducedDlco. Patient management includes eliminatingfurther beryllium exposure, providing supplementaloxygen if needed, supportive measures and,in more advanced forms, a therapeutic trial ofsteroids. Most patients fully recover, althoughchronic berylliosis may develop in up to 25% ofpatients. 2Chronic berylliosis develops after a latentperiod ranging from months to as long as 10 yearsafter the onset of low-level beryllium exposure. 2,72The diagnosis should be suspected in patientswith an appropriate exposure history who presentwith dyspnea (the most common complaint),cough, chest pain, weight loss, fatigue, andarthralgias and who have an abnormal chestradiograph finding, with a reticular-nodularinfiltrate seen predominately in the upper lunglobes (although all lobes can be involved). Thechest radiographic pattern is similar to that foundwith sarcoidosis, including mediastinal and bilateralhilar adenopathy. A beryllium lymphocytetransformation test should be performed todocument sensitization. Although an estimated50% of beryllium-sensitized workers have evidenceof chronic beryllium disease at the time ofan initial presentation, a 2005 longitudinal study 74showed that chronic berylliosis will develop in31% of disease-free sensitized workers (especiallymachinists) within 3.8 years (range, 1.0 to 9.5years), whereas the disease never develops in theremaining 69% 4.8 years of follow-up. However,there is insufficient evidence to recommendperforming routine beryllium lymphocyte transformationtests for screening asymptomaticworkers. 75Large calcified masses and pleural reactionscan be seen, whereas nearly one-third of thepatients have bilateral hilar adenopathy. PFTresults show reduced volumes and Dlco. Althoughthe clinical course of chronic berylliosis is variable,most patients have a slow, inexorable decline thatcan result in cor pulmonale in nearly one-third ofpatients. Treatment options are sparse but of primeimportance is the elimination of further berylliumexposure; often, a course of steroids is administered.Although there are no controlled studiesdemonstrating their efficacy, most reports haveshown that removal from further beryllium exposureand the use of steroids result in a favorableresponse both clinically and radiographically;however, similar to the situation with sarcoidosis,many cases recur when steroid therapy is discontinued.Data from a large beryllium registry 2involving 689 patients demonstrated excess mortalityfrom lung cancer and nonmalignant berylliumdisease. Notably, a recent study documented16 cases of nonoccupational, environmental chronicberyllium disease in residents surrounding a berylliummanufacturing facility. 76<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 385ACC-PULM-09-0302-024.indd 3857/10/09 8:49:36 PM


Hard Metal Lung DiseaseDefinition/OccupationsAs shown in Table 3, various hard metals caninduce a wide range of lung diseases. Althoughthe majority of the hard-metal lung diseases arisefrom occupational exposure, environmental exposureshave been reported (eg, mercury-inducedchemical pneumonitis). Metal-induced pulmonarytoxicity can be caused by either antigen-specificimmunologic reactions or nonspecific immune/inflammatory responses caused in part by oxidativestress similar to that depicted in Figure 5. 2,77CobaltThe main components of commercially availablehard metal include tungsten and titaniumcarbides (70 to 95%); cobalt (5 to 25%); and smallamounts of various other metals (eg, nickel, chromium,tantalum, and vanadium). Because tungstenis inert, cobalt is the primary culprit that causespulmonary toxicity associated with hard-metalexposure. Cobalt is used as a hardener and binderof metal alloys of, for example, tungsten, aluminum,chrome, molybdenum, and beryllium. Cobaltexposure may occur in occupations such asmachinist, metal tool maker, polisher, grinder, sawsharpener, and dental driller.Occupational exposure to cobalt is associatedwith the following three pulmonary manifestations:(1) airways obstruction that resemblesoccupational asthma; (2) acute interstitial pneumonitiswith features of hypersensitivity pneumonitis;and (3) chronic diffuse interstitial fibrosis that canprogress to end-stage fibrosis. Patients with cobaltinducedchronic interstitial fibrosis present withthe insidious onset of exertional dyspnea, cough,weight loss, and restrictive pulmonary physiology.Pathologic findings include desquamative interstitialpneumonitis, giant cell interstitial pneumonitis,and varying degrees of bronchiolitis obliterans,interstitial fibrosis, and sarcoid-like granulomas.A characteristic diagnostic feature is the presenceof odd-appearing giant multinucleated cells. Animmune mechanism involving suppressor T lymphocytesis implicated in mediating cobalt pulmonarytoxicity. 76 Also, genetic factors involving theallelic substantiation of a glutamic acid in position69 of the human leukocyte antigen-DP chain hasbeen identified in workers with cobalt-inducedchronic interstitial lung disease. 78 Furthermore, thecombination of cobalt and other metallic carbidescan induce an oxidative stress on the lungs. Limitingexposure below the PEL of 50 mg/m 3 is mandatory.Once cobalt-related lung disease has occurred, theworker should be moved to other sites at whichno cobalt exposure can occur. Progression canoccur without further cobalt exposure; occasionally,cor pulmonale and death may occur many yearsafter the initial exposure. In a 13-year follow-up 79of lung function changes in 122 workers exposedto cobalt-containing compounds, pulmonary fibrosisdid not develop in one person, whereas asignificant decrease in FEV 1occurred only in workerswho were cigarette smokers. There were noTable 3. Some Representative Hard Metal-Induced Lung Diseases*Agent Histopathology Occupational ExposureAluminum Interstitial fibrosis; sarcoid-like Aluminum miner/refineriesgranulomaBeryllium Sarcoid-like granulomas; ILD Beryllium miner/aerospace, nuclear,and electronics workCadmium Acute pneumonitis; ILD; emphysema Lead or zinc smelter; batteries(Brazier disease)Cobalt Giant-cell ILD; sarcoid-like granuloma; Metal hardener for tungsten carbide (toolobstructive airwaysgrinders, metal polisher, dental drills)Iron Dust macule, ILD Iron miner/welderMercury Tracheobronchitis; pneumonitis; Electrical industry; scientific equipment;pulmonary edemacatalyst work*ILD = interstitial lung disease.386 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3867/10/09 8:49:36 PM


associations with polymorphisms in glutamic acidin position. 69SiderosisOccupations involving iron exposure (eg, mining,smelting, welding, and steel making) can leadto siderosis. It is often a relatively benign condition,usually noted as an incidental finding or as anabnormality on a chest radiograph with denselinear opacities in an asymptomatic iron worker.Occasionally, iron can induce respiratory symptoms,restrictive pulmonary physiology, and pulmonaryfibrosis in heavily exposed workers. 80Pathologic findings range from the characteristiciron dust macule to interstitial fibrosis. There is nospecific treatment other than the prevention offurther iron exposure.Metal Fume FeverMetal fume fever is a set of flu-like symptomsthat occur within 4 to 12 h after the start of thework shift in metal workers after exposure to highlevels of metal fumes, especially zinc oxide. Fumesof metal oxides of arsenic, boron, cadmium, chromium,copper, magnesium, manganese, nickel,and titanium can cause similar symptoms. 77 It wasinitially defined in brass foundry workers in themid-1800s and in steel welders in the early 1900s. 76These patients present with fever at temperaturesfrom 102 to 104F associated with a metallic tastein the mouth, throat irritation, cough, dyspnea,malaise, fatigue, myalgias, arthralgias, sweats, andshaking chills. The syndrome typically is associatedwith mild airflow obstruction and minimalchest radiographic findings and resolves in 24to 48 h. 77 Metal fume fever is caused by doserelatedtoxicity rather than immunologic sensitizationbecause there is no latency period. Studieshave shown that workers with metal fume feverhave increased levels of various cytokines (eg,tumor necrosis factor-, IL-6, and IL-8) in theirBAL fluid, presumably released from AMs, thatlikely mediate this syndrome. 77 An unexplainedfeature of this syndrome is the development oftolerance so that symptoms diminish on successivedays of exposure but recur after a brief absenceaway from work.Acknowledgment: This work was supportedin part by a Merit <strong>Review</strong> grant from the Departmentof Veterans Affairs. The author is gratefulfor the comments of David Cugell.Annotated References1. Cowie RL, Murray J, Becklake MR. Pneumoconioses.In: Mason RJ, Broaddus C, Murray JF,et al, eds. Textbook of respiratory medicine. 4th ed.Philadelphia, PA: WB Saunders, 2005; 1748−1782An excellent overview of the topic. Includes references.2. King TE, Kamp DW, Panos RJ. Pneumoconiosis,chronic interstitial fibrosis and bronchiolitis. In:Witorsch P, Spagolo SV, eds. Air pollution andlung disease in adults. Boca Raton, FL: CRC Press,1994; 207−284A concise summary of the relevant diseases. Includes273 references.3. National Institute for Occupational Safety andHealth. Silicosis. Publication No. 2006-110 Availableat: http://www.cdc.gov/niosh/docs/2006-110/.Accessed April 9, 20084. National Institute for Occupational Safety andHealth, Division of Respiratory Disease Studies.Silicosis deaths among young adults: UnitedStates, from 1968 to 1994. JAMA 1998; 280:13−145. National Institute for Occupational Safety andHealth. Hazard review: health effects of occupationalexposure to respirable crystalline silica.Washington, DC: Department of Health andHuman Services, NIOSH, 2002; 1−127; PublicationNo. 2002-1296. Mossman BT, Churg A. Mechanisms in the pathogenesisof asbestosis and silicosis. Am J Respir CritCare Med 1998; 157:1666−16807. Wu F, Xia Z, Qu Y, et al. Genetic polymorphism ofIL-1A, IL-1, IL-1RN, NFKB1, FAS, and FASL andrisk of silicosis in a Chinese occupational population.Am J Indust Med 2008; 51:843−518. Finkelstein MM. Silica, silicosis, and lung cancer: arisk assessment. Am J Ind Med 2000; 38:8−189. Roseman KD, Reill MJ, Rice C, et al. Silicosisamong foundry workers: implications for the needto revise the OSHA standard. Am J Epidemiol 1996;144:890−90010. Arakawa H, Johkoh T, Honma K, et al. Chronicinterstitial pneumonia in silicosis and mix-dustpneumoconiosis. Chest 2007; 131:1870−1876<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 387ACC-PULM-09-0302-024.indd 3877/10/09 8:49:36 PM


11. Wang X-R, Christiani DC. Respiratory symptomsand functional status in workers exposed to silica,asbestos, and coal mine dusts. J Occup EnvironMed 2000; 42:1076−108412. Greenberg MI, Waksman J, Curtis J. Silicosis: areview. Dis Mon 2007; 53:394−416A relatively current review of silicosis. Includes 79references.13. American Thoracic Society, Centers for DiseaseControl and Prevention, Infectious Diseases Societyof America. Treatment of tuberculosis. Am JRespir Crit Care Med 2003; 167:603−66214. Hnizdo E, Murray J. Risk of pulmonary tuberculosisrelative to silicosis and exposure to silica dustin South American gold miners. Occup EnvironMed 1998; 55:496−50215. Leung CC, Yew WW, Law WS, et al. Smoking andtuberculosis among silicotic patients. Eur Respir J2007; 29:745−75016. Castranova V, Vallyathan V. Silicosis and coalworkers’ pneumoconiosis. Environ Health Perspect2000; 108(suppl):675−684An excellent review of silicosis and CWP emphasizingrecent clinical developments and pathogenic mechanisms.Includes 130 references.17. Weill H, McDonald JC. Exposure to crystalline silicaand risk of lung cancer: the epidemiologic studies.Thorax 1996; 51:97−10218. Checkoway H, Hughes JM, Weill H, et al. Crystallinesilica exposure, radiologic silicosis, and lungcancer mortality in diatomaceous earth industryworkers. Thorax 1999; 54:56−5919. Cassidy A, Mannetje A, van Tongeren M, et al.Occupational exposure to crystalline silica andrisk of lung cancer. Epidemiology 2007; 18:36−4320. Pelucchi C, Pira E, Piolatto G, et al. Occupationalsilica exposure and lung cancer risk: a review ofepidemiologic studies from 1996 to 2005. AnnOncol 2006; 17:1039−105021. Schwerha JJ. Occupational <strong>Medicine</strong> Forum. JOccup Environ Med 2008; 50:101−104.An excellent reference documenting the recommendedradiographic follow-up in patients with pneumoconiosis.22. Kamp DW, Vallyathan V. Influence of mineral dustsurface characteristics and generation of reactivespecies. In: Vallyathan V, Castranova V, Shi X, eds.Oxygen/nitrogen radicals: lung biology in healthand disease series (vol 187). New York: MarcelDekker, 2004; 139−16023. Wallace WE, Harrison JC, Grayson RL, et al. Aluminosilicatesurface contamination of respirablequartz particles from coal mine dusts and from clayworks dusts. Ann Occup Hyg 1994; 38:439−4524. Nadif R, Jedlicka A, Mintz M, et al. Effect of TNFand LTA polymorphisms on biological markersof response to oxidative stimuli in coal miners: amodel of gene-environment interaction; tumournecrosis factor and lymphotoxin . J Med Genet2003; 40:96−10325. Nadif R, Mintz M, Marzec J, et al. IL18 and IL18R1polymorphisms, lung CT and fibrosis: a longitudinalstudy in coal miners. Eur Respir J 2006;28:1100−110526. Blum T, Kollmeier J, Ott S, et al. CT for diagnosisand grading of dust-induced occupational lungdisease. Curr Opin Pulm Med 2008; 14:135−14027. Attfield MD, Kuempel ED. Mortality amongUnited States underground coal miners: a 23-yearfollow up. Am J Ind Med 2008; 51:231−24528. Cugell DW, Kamp DW. Asbestos and the pleura: areview. Chest 2004; 125:1103−1117A summary of the various effects of asbestos on thepleura. Includes 120 references.29. Guidotti TL, Miller A, Christiani D, et al. Diagnosisand initial management of nonmalignant diseasesrelated to asbestos. Am J Respir Crit Care Med2004; 170:691−715The ATS 2004 consensus statement about the criteriafor diagnosing and managing asbestos-induced nonmalignantdiseases containing excellent representativehistopathologic and radiographic examples as well as160 references.30. Kamp DW, Mossman BT. Asbestos-associatedcancers: clinical spectrum and pathogenic mechanisms.Clin Occup Environ Med 2002; 2:753−77831. Hein MJ, Staynor LT, Lehman E, et al. Follow upstudy of chrysotile textile workers: cohort mortalityand exposure-response. Occup Environ Med2007; 64:616−62532. Cooke WE. <strong>Pulmonary</strong> asbestosis. BMJ 1927;2:1024−102533. Murphy RL, Becklake MR, Brooks SM, et al. Thediagnosis of nonmalignant diseases related toasbestos. Am Rev Respir Dis 1986; 134:363−368The ATS consensus statement about the criteria fordiagnosing asbestos-induced nonmalignant diseases.34. O’Reilly KMA, McLaughlin AM, Beckett WS, et al.Asbestos-related lung disease. Am Fam Physician2007; 75:683−688388 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3887/10/09 8:49:37 PM


35. Nicholson WJ, Perkel G, Selikoff IJ. Occupationalexposure to asbestos: population at risk and projectedmortality; from 1980 to 2030. Am J Ind Med1982; 3:259−31136. Kamp DW. Asbestos-induced lung diseases: anupdate. Transl Res 2009; 153:143−152.A recent review emphasizing epidemiology and pathogenesis.Includes 106 references.37. Iwatsubo Y, Pairon JC, Boutin C, et al. Pleuralmesothelioma, dose-response relation at low levelsof asbestos exposure in a French populationbasedcase-control study. Am J Epidemiol 1998;148:133−14238. Barbers RG, Abraham JL. Asbestos occurring aftera brief inhalational exposure: usefulness of BAL indiagnosis. Br J Ind Med 1989; 46:106−11039. Shukla A, Gulumian M, Hei T, et al. Multipleroles of oxidants in the pathogenesis of asbestosinduceddiseases. Free Radic Biol Med 2003; 34:1117−112940. Merchant JA. Human epidemiology: a review offiber type and characteristics in the developmentof malignant and nonmalignant disease. EnvironHealth Perspect 1990; 88:287−29341. Miller A, Lilis R, Godbold J, et al. Relationship ofpulmonary function to radiographic interstitialfibrosis in 2,611 long-term asbestos insulators. AmRev Respir Dis 1992; 145:263−27042. Rogan WJ, Ragan NB, Dinse GE. Radiograph evidenceof increased asbestos exposure in the UnitedStates population from NHANESI and NHANE-SII, from 1973 to 1978. Cancer Causes Control 2000;11:441−44943. Kim JS, Lynch DA. Imaging of nonmalignantoccupational lung disease. J Thorac Imaging 2002;17:238−26044. Weiss W. Asbestos-related pleural plaques andlung cancer. Chest 1993; 103:1854−185945. Henderson DW, Rantanen J, Barnhart S, et al.Asbestos, asbestosis, and cancer: the Helsinki criteriafor diagnosis and attribution; a consensusreport of an international expert group. Scand JWork Environ Health 1997; 28:311−31646. Roggli VL, Sanders LL. Asbestos content of lungtissue and carcinoma of the lung: a clinicopathologiccorrelation and mineral fiber analysis of 234cases. Ann Occup Hyg 2000; 44:109−11747. Weiss W. Asbestosis: a marker for the increasedrisk of lung cancer among workers exposed toasbestos. Chest 1999; 115:536−54948. Robinson BWS, Lake RA. Advances in malignantmesothelioma. N Engl J Med 2005; 353:1591−1603Excellent update on the clinical-pathologic features ofmesothelioma as well as more recent treatment strategies.Includes 98 references.49. Boutin C, Schlesser M, Frenay C, et al. Malignantpleural mesothelioma. Eur Respir J 1998;12:972−98150. Peto J, Doll R, Herman C, et al. Relationship ofmortality to measures of environmental asbestospollution in an asbestos textile factory. Ann OccupHyg 1985; 29:305−35551. McDonald JC, Sebastien P, McDonald AD, et al.Epidemiologic observations on mesothelioma andthe implications for nonoccupational exposure. In:Bignon J, Peto J, Saracchi R, eds. Nonoccupationalexposure to mineral fibers. Lyon, France: InternationalAgency for Research on Cancer, 1989;420−427; IARC Science publication No. 9052. Boraschi P, Neri S, Braccini G, et al. Magnetic resonanceappearance of asbestos-related benign andmalignant pleural disease. Scand J Work EnvironHealth 1999; 25:18−2353. Creaney J, Yeoman D, Demelker Y, et al. Comparisonof osteopontin, megakaryocyte potentiatingfactor, and mesothelin proteins as markers in theserum of patients with malignant mesothelioma.J Thorac Oncol 2008; 3:851−85754. Amati M, Tomasetti M, Mariotti L, et al. Assessmentof biomarkers in asbestos-exposed workersas indicators of cancer risk. Mutat Res/Gene Toxicoland Environ Mutag 2008; 655:52−5855. van Meerbeeck JP, Hillerdal G. Screening for mesothelioma:more harm than good. Am J Respir CritCare Med 2008; 178:781−782.56. Rihn BH, Mohr S, McDowell SA, et al. Differentialgene expression in mesothelioma. FEBS Lett 2000;480:95−10057. Sterman DH, Kaiser LR, Albelda SM. Advances inthe treatment of malignant mesothelioma. Chest1999; 116:504−52058. Vogelzang NJ, Rusthoven JJ, Symanowski J, et al.Phase III study of pemetrexed in combination withcisplatin vs cisplatin alone in patients with malignantpleural mesothelioma. J Clin Oncol 2003;21:2636−264459. Janne PA. Comprehensive review: chemotherapyfor malignant pleural mesothelioma. Clin LungCancer 2003; 5:98−106<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 389ACC-PULM-09-0302-024.indd 3897/10/09 8:49:37 PM


60. Sugarbaker DJ, Flores RM, Jaklitsch MT, et al.Resection margins, extrapleural nodal status, andcell type determine postoperative long-term survivalin trimodality therapy of malignant pleuralmesothelioma. J Thorac Cardiovasc Surg 1999;117:54−6561. Staples CA, Gamsu G, Ray CS, et al. High resolutionCT and lung function in asbestos-exposedworkers with normal chest radiographs. Am RevRespir Dis 1989; 139:1502−150862. Miller A, Lilis R, Godbold J, et al. Relationship ofpulmonary function to radiographic interstitialfibrosis in 2,611 long-term asbestos insulators. AmRev Respir Dis 1992; 145:263−27063. Becklake MR. Asbestos and other fiber-relateddiseases of the lungs and pleura. Chest 1991;100:248−25464. Berry G. Mortality of workers certified by pneumoconiosismedical panel as having asbestosis.Br J Ind Med 1981; 38:130−13765. Lin R-T, Takahashi K, Karjalainen A, et al. Ecologicalassociations between asbestos-related diseasesand historical asbestos consumption: an internationalanalysis. Lancet 2007; 369:844−84966. Lee PN. Relation between exposure to asbestosand smoking jointly and the risk of lung cancer.Occup Environ Med 2001; 58:145−15367. Roggli VL, Sanders LL. Asbestos content of lungtissue and carcinoma of the lung: a clinicopathologiccorrelation and mineral fiber analysis of 234cases. Ann Occup Hyg 2000; 44:109−11768. Weiss W. Asbestosis and lobar site of lung cancer.Occup Environ Med 2000; 57:358−36069. Henderson DW, Roggli VL, Shilkin KB, et al. Isasbestosis an obligate precursor for asbestosrelatedlung cancer? In: Peters GA, Peters BJ, eds.Asbestos health effects, treatment and control (vol11). Charlottesville, VA: The Michie Co., 1995;103−16970. Oksa P, Klockars M, Karjalainen A, et al. Progressionof asbestosis predicts lung cancer. Chest 1998;113:1517−152171. Schmidt RA, Glenny RW, Godwin JD, et al. Panlobularemphysema in young IV ritalin abusers.Am Rev Respir Dis 1991; 143:649−65672. Kreiss K, Day GA, Schuler CR. Beryllium: a modernindustrial hazard. Annu Rev Public Health2007; 28:259−277A recent update of beryllium-induced pulmonarytoxicity. Includes 75 references.73. Newman LS. Significance of the blood berylliumlymphocyte proliferation test. Environ Health Perspect1996; 104(suppl):953−95674. Newman LS, Mroz MM, Balkissoon R, et al. Berylliumsensitization progresses to chronic berylliumdisease. Am J Respir Crit Care Med 2005;171:54−6075. Borak J, Woolf SH, Fields CA. Use of the berylliumlymphocyte proliferation testing forscreening of asymptomatic individuals: an evidence-basedassessment. J Occup Environ Med2006; 48:937−94776. Maier LA, Martyny JW, Liang J, et al. Recentchronic beryllium disease in residents surroundinga beryllium facility. Am J Respir Crit Care Med2008; 177:1012−101777. Kelleher P, Pacheco K, Newman LS. Inorganicdust pneumoconiosis: the metal-related parenchymaldisorders. Environ Health Perspect 2000; 108(suppl):685−696An excellent comprehensive review of the area. Includes196 references.78. Potolicchio I, Mosconi G, Forni A, et al. Susceptibilityto hard metal lung disease is strongly associatedwith the presence of glutamate 69 in HLA-DP chain. Eur Respir J 1997; 27:2741−274379. Verougstraete V, Mallants A, Buchet J-P, et al.Lung function changes in workers exposed tocobalt compounds. Am J Respir Crit Care Med2004; 170:162−16680. Akbar-Khanzadeh F. Short term respiratory functionchanges in relation to work shift weldingfume exposure. Int Arch Occup Environ Health1993; 64:393−397390 Pneumoconiosis (Kamp)ACC-PULM-09-0302-024.indd 3907/10/09 8:49:37 PM


Hospital-Acquired and Ventilator-Associated PneumoniaRonald R. Grossman, MD, FCCPObjectives:• Identify the key respiratory pathogens causing hospitalacquiredpneumonia/ventilator-associated pneumonia• Develop a systematic approach to the diagnosis ofhospital-acquired pneumonia/ventilator-associatedpneumonia• Improve the initial therapeutic choices for the managementof hospital-acquired pneumonia/ventilator-associatedpneumoniaKey words: clinical pulmonary infection score; hospitalacquiredpneumonia; Pseudomonas aeruginosa; Staphylococcusaureus; ventilator-associated pneumoniaDefinitionPneumonia that develops 48 h after admissionto the hospital and was not incubating at thetime of admission is defined as hospital-acquiredpneumonia (HAP ). 1 Ventilator-associated pneumonia(VAP) refers to pneumonia that occurs 48 to 72 hafter endotracheal intubation among patientstreated in an ICU setting. 1,2 Health-care-associatedpneumonia refers to patients admitted to the hospitalwith pneumonia that may have been receivedvia IV therapy at home, wound care, or specializednursing care through a health-care agency, family,or friends or via self-administered IV medical therapyin the 30 days before pneumonia. 3,4 They mayhave attended a hospital or hemodialysis clinic orreceived IV chemotherapy in the 30 days beforepneumonia. They also may have been admittedto an acute care hospital for 2 or more days in the90 days before pneumonia or resided in a nursinghome or a long-term care facility.EpidemiologyHAP is the second most common nosocomialinfection in the United States and is associatedwith high rates of mortality and morbidity. 1 Hospitallength of stay increases by an average of 7 to9 days per affected patient and is associated withcosts $40,000 per patient. 3,5 HAP occurs at a rateof between 5 and 10 cases per 1,000 hospital admissions,with the incidence increasing by as muchas 6- to 20-fold in patients receiving mechanicalventilation. 3,6,7HAP accounts for up to 25% of all ICU infectionsand for 50% of the antibiotics prescribed. 8VAP occurs in 9 to 27% of all intubated patients. 3,5The risk of VAP is greatest early in the course ofhospital stay and then abates somewhat. The riskhas been estimated to be 3% per day during the first5 days of ventilation, 2% per day during days 5 to10 of ventilation, and 1% per day after this. 9 Nosocomialpneumonia is less common when patientsare managed with noninvasive ventilation. 10–12Previous guidelines 13 identified that early-onsetpneumonia frequently was caused by communitypathogens. Early-onset HAP and VAP, defined asoccurring within the first 4 days of hospitalization,are more likely to be caused by antibioticsensitivebacteria. Late-onset HAP and VAP ( 5days) are more likely to be caused by multidrugresistant(MDR) pathogens and are associatedwith increased patient mortality and morbidity.However, patients with early-onset HAP who havereceived antibiotics or who have had hospitalizationwithin the past 90 days have similar pathogensto those patients with late-onset HAP or VAP andshould be treated similarly (Table 1). 14The crude mortality rate for HAP may be ashigh as 30 to 70%, but the mortality related to HAPor “attributable mortality” has been estimated tobe between 33% and 50% in several case-matchingstudies of VAP. 15 Increased mortality rates are associatedwith bacteremia, especially with Pseudomonasaeruginosa or Acinetobacter species; medical ratherthan surgical illness; and treatment with ineffectiveantibiotic therapy. 15,16 Strangely, the authors of somestudies 17–19 using similar methodology have not identifiedany attributable mortality caused by VAP.EtiologyThe most frequent isolated pathogens fromrespiratory secretions are aerobic Gram-negative<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 391ACC-PULM-09-0302-025.indd 3917/10/09 8:50:15 PM


Table 1. Risk Factors for MDR Pathogens Causing HAP,Health-Care-Associated Pneumonia, and VAP*Antimicrobial therapy in preceding 90 dCurrent hospitalization of 5 dHigh frequency of antibiotic resistance in the communityor in the specific hospital unitPresence of risk factors for health-care-associatedpneumoniaHospitalization for 2 d in the preceding 90 dResidence in a nursing home or extended-care facilityHome infusion therapy (including antibiotics)Long-term dialysis within 30 dHome wound careFamily member with MDR pathogenImmunosuppressive disease and/or therapy*Adapted from American Thoracic Society/InfectiousDiseases Society of America. 34bacilli, such as P aeruginosa, Escherichia coli, Klebsiellapneumoniae, and Acinetobacter sp. 3–6,20–25 Infectionscaused by Gram-positive cocci, such as Staphylococcusaureus, particularly methicillin-resistantS aureus (MRSA), are increasing, particularly amongpatients with diabetes mellitus, head trauma,and those hospitalized in ICUs. 26,27 Rates of polymicrobialinfection vary widely but appear to beincreasing and are especially high in patients withARDS. 28–30The etiology of HAP and the frequency ofspecific MDR pathogens vary by hospital, patientpopulation, exposure to antibiotics, type of ICUpatient, and changes over time. 31,32 Local surveillancedata are crucial in monitoring the emergenceof MDR pathogens and should drive the formulationof local guidelines. Anaerobic lung infectionmay follow aspiration in nonintubated patients butis very uncommon in patients with VAP. 21,33Elderly residents of long-term care facilitieshave been found to have a spectrum of pathogensthat more closely resemble late-onset HAP andVAP. 23,24 This observation has led to the developmentof the term health care-associated pneumonia.This designation essentially identifies patientsfrom health-care facilities who are at increased riskfor infection with resistant pathogens. 34 Rates ofLegionella pneumophila vary considerably betweenhospitals, and disease occurs more commonly withserogroup 1 when the water supply is colonizedor there is ongoing construction. 35,36 Nosocomialvirus and fungal infections are uncommon causesof HAP and VAP in immunocompetent patients.Outbreaks of influenza have occurred sporadically,and the risk of infection can be substantiallyreduced with widespread effective infectioncontrol, vaccination, and use of anti-influenzaagents. 37–39PathogenesisThe severity of the patient’s underlying disease,previous surgery, exposure to antibiotics,other medications, and exposure to invasiverespiratory devices and equipment may leadto airway colonization and contribute to thepathogenesis of HAP and VAP. 40,41 Aspiration oforopharyngeal pathogens or leakage of secretionscontaining bacteria around the endotracheal tubecuff are the primary routes of bacterial entry intothe lower respiratory tract. 42–44 Inhalation or directinoculation of pathogens into the lower airway,hematogenous spread from infected IV catheters,and bacterial translocation from the GI tract lumenare uncommon pathogenic mechanisms. 44 Infectedbiofilm in the endotracheal tube, with subsequentembolization to distal airways, is another sourceof entry of infected material into the lung. 45 Thestomach and sinuses may be potential reservoirs ofnosocomial pathogens that contribute to bacterialcolonization of the oropharynx, but the importanceof these routes is debated. 46,47DiagnosisThe standard diagnostic criteria in VAP includeat least two of the following three findings: fever,leukocytosis, and purulent tracheal secretions,usually with abnormal findings on chest radiographicstudies. When these conditions occur,the likelihood of VAP is high. 48 The presence of aradiographic infiltrate in a patient with fever, leukocytosis,or purulent tracheobronchial secretionshas high diagnostic sensitivity but low specificity.When all four criteria are present, specificityimproves but sensitivity decreases to 50%, whichis clinically unacceptable. 49 The only study 48 examininginterobserver diagnostic reliability found nomajor differences between individual physicians orthose grouped by level or training.Pugin and coworkers 50 developed the clinicalpulmonary infection score (CPIS), which combines392 Hospital-Acquired and Ventilator-Associated Pneumonia (Grossman)ACC-PULM-09-0302-025.indd 3927/10/09 8:50:15 PM


clinical, radiographic, physiologic (Pao 2/fractionof inspired oxygen), and microbiologic data into asingle numerical result to improve the specificityof clinical diagnosis. When the CPIS exceeded 6, agood correlation with positive quantitative culturefindings of bronchoscopic and nonbronchoscopicBAL fluid specimens was observed. These findingshave not been reproduced by other investigators,but the accuracy of the CPIS score is improved ifa Gram stain of a deep respiratory tract culture isadded to the evaluation. 51,52These findings suggest that the presence ofabnormal clinical manifestations, combined withabnormal radiographic findings, can be used forinitial screening for VAP. However, the lack ofspecificity with this method suggests that additionalprocedures are needed, such as cultures oflower respiratory tract secretions.All patients with suspected VAP should haveblood cultures collected. The physician orderingthe culture should recognize that a positive resultcan indicate the presence of either pneumonia orextrapulmonary infection. 53 A diagnostic thoracentesisto rule out a complicating empyema orparapneumonic effusion should be performed ifthe patient has a large pleural effusion or if thepatient with a pleural effusion appears toxic. Samplesof lower respiratory tract secretions should beobtained from all patients with suspected HAP andshould be collected before changes of antibiotic.Samples can include an endotracheal aspirate,BAL fluid sample, or protected specimen brushsample. 54–56 A sterile culture of respiratory secretionsin the absence of a new antibiotic in the past72 h virtually rules out the presence of bacterialpneumonia, but viral or Legionella sp infection arestill possible. If these patients have clinical signsof infection, an extrapulmonary site of infectionshould be investigated. 57A reliable tracheal aspirate Gram stain can beused to direct initial empiric antimicrobial therapyand may increase the diagnostic value of the CPIS. 58The presence of a new or progressive radiographicinfiltrate and at least two of three clinical features(fever 38°C, leukocytosis or leukopenia, andpurulent secretions) represent the most accurateclinical criteria for starting empiric antibiotictherapy. 51 If a clinical strategy is used, reevaluationof the decision to use antibiotics based on theresults of semiquantitative lower respiratory tractcultures and serial clinical evaluations, by day 3 orsooner, is appropriate. 59–61Singh and coworkers 60 demonstrated that amodified CPIS of 6 for 3 days is an objectivecriterion to select low-risk patients for early discontinuationof empiric treatment of HAP. Teststhat identify pathogens on the basis of qualitativecultures will lead to treatment for more organismsthan diagnostic techniques based on quantitativecultures. 57,62–64 Nonbronchoscopic sampling canreliably obtain lower respiratory tract secretionsfor quantitative cultures if bronchoscopic samplingis not immediately available. 65 The use of a bronchoscopicbacteriologic strategy has been shown toreduce 14-day mortality but not the 28-day mortality,compared with a clinical strategy, in one study 57of suspected VAP.Three other smaller studies demonstrated noadvantage of an invasive diagnostic approach.A randomized controlled trial 66 comparing theutility of bronchoscopic lavage with quantitativeculture of BAL fluid or endotracheal aspirate withnonquantitative culture of the aspirate identifiedno significant difference in the 28-day mortalityrate (the primary outcome), the rate of targetedtherapy, days alive without antibiotics, hospital orICU length of stay, or maximum organ dysfunctionscores. Delays in the initiation of appropriateantibiotic therapy can increase the rate of mortalityof VA and, thus, therapy should not be delayed forthe purpose of performing diagnostic studies inpatients who are clinically unstable. 37,67,68TherapyInitial empiric therapy should be selectedbased on the absence or presence of risk factorsfor MDR pathogens. These risk factors includeprolonged duration of hospitalization ( 5 days),admission from a health-care-related facility, andrecent prolonged antibiotic therapy (Table 1). 14Choice of specific agents should be dictated bylocal microbiology, cost, availability, and formularyrestrictions (Tables 2 and 3). 69,70 Patients withhealth-care-related pneumonia should be treatedfor potentially drug-resistant organisms regardlessof when during the hospital stay the pneumoniabegins. Inappropriate therapy (failure of the etiologicpathogen to be sensitive to the administeredantibiotic) is a major risk factor for excess mortality<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 393ACC-PULM-09-0302-025.indd 3937/10/09 8:50:15 PM


Table 2. Initial Empiric Antibiotic Therapy for HAP or VAP inPatients With No Known Risk Factors for MDR Pathogens, EarlyOnset, and Any Disease SeverityPotential PathogensStreptococcus pneumoniaeHaemophilus influenzaeMethicillin-sensitive SaureusAntibiotic-sensitive entericGram-negative bacilliE coliK pneumoniaeEnterobacter speciesProteus speciesSerratia marcescensRecommended AntibioticCeftriaxone; or levofloxacin,moxifloxacin, or ciprofloxacinor ampicillin/sulbactamor ertapenemTable 3. Initial Empiric Therapy for HAP, VAP, and Health-Care-Associated Pneumonia in Patients With Late-Onset Diseaseor Risk Factors for MDR Pathogens and All Disease SeverityPotential PathogensPathogens listed in Table 2and MDR pathogensP aeruginosaK pneumoniae (extendedspectrumβ-lactamase positive)Acinetobacter speciesMRSAL pneumophilaCombination AntibioticTherapyAntipseudomonal cephalosporin(cefepime, ceftazidime);orAntipseudomonal carbepenem(imipenem or meropenem);orβ-lactam/β-lactamase inhibitor(piperacillintazobactam)plusAntipseudomonal fluoroquinolone(ciprofloxacin orlevofloxacin; orAminoglycoside (amikacin,gentamicin, or tobramycin)Plus linezolid or vancomycinand length of stay for patients with HAP, andantibiotic-resistant organisms are the pathogensmost commonly associated with inappropriatetherapy. 67 In selecting empiric therapy for patientswho have recently received an antibiotic, an effortshould be made to use an agent from a differentantibiotic class because recent therapy increasesthe probability of inappropriate therapy andcan predispose to resistance to that same class ofantibiotics. 70 Initial antibiotic therapy should beadministered promptly because delays in administrationmay add to excess mortality resultingfrom VAP. 68,71,72Empiric therapy of patients with severe HAPor VAP requires the use of antibiotics at optimaldoses to ensure maximum efficacy. 73 Initial therapyshould be administered IV to all patients, with aswitch to oral/enteral therapy in selected patientswith a good clinical response and a functioningintestinal tract. Highly bioavailable agents, suchas the quinolones and linezolid, may be easilyswitched to oral therapy in such patients. Combinationtherapy should be used if patients are likelyto have infection with MDR pathogens. 68 No datahave documented the superiority of this approachcompared with monotherapy, except to enhancethe likelihood of initially appropriate empirictherapy. If patients receive combination therapywith an aminoglycoside-containing regimen, theaminoglycoside can be stopped after 5 to 7 days inresponding patients. 74 Monotherapy with selectedagents can be used for patients with severe HAPand VAP in the absence of resistant pathogens. 73Patients in this risk group should initially receivecombination therapy until the results of lowerrespiratory tract cultures are known and confirmthat a single agent can be used. If patients receivean initially appropriate antibiotic regimen, therapycan be shortened from the traditional 14 to 21 daysto periods as little as 7 days, provided that theetiologic pathogen is not P aeruginosa and that thepatient has a good clinical response. 75,76If P aeruginosa pneumonia is documented, combinationtherapy is recommended. The principaljustification is the high frequency of developmentof resistance on monotherapy. 74 Although combinationtherapy will not necessarily prevent thedevelopment of resistance, combination therapyis more likely to avoid inappropriate and ineffectivetreatment of patients. 67 If Acinetobacter sp aredocumented to be present, the most active agentsare the carbapenems, sulbactam, colistin, andpolymyxin. There are no data documenting animproved outcome if these organisms are treatedwith a combination regimen. 77 If extended-spectrum-lactamase—positive Enterobacteriaceae sp areisolated, the use of third-generation cephalosporinsshould be avoided and carbapenems considered. 78Linezolid is an alternative to vancomycin for thetreatment of MRSA VAP. A subset analysis of twoprospective randomized trials 79,80 has indicated thatthe use of linezolid may be superior to vancomycin.This agent may also be preferred if patients have394 Hospital-Acquired and Ventilator-Associated Pneumonia (Grossman)ACC-PULM-09-0302-025.indd 3947/10/09 8:50:16 PM


enal insufficiency or are receiving other nephrotoxicagents, but more data are needed. Antibioticrestriction can limit epidemics of infection withspecific resistant pathogens. The heterogeneity ofantibiotic prescriptions, including formal antibioticcycling, may be able to reduce the overall frequencyof antibiotic resistance. However, the long-termimpact of this practice is unknown. 81,82A serial assessment of clinical parametersshould be used to define the response to initialempiric therapy. 59,61 Modifications of empirictherapy should be made on the basis of this informationin conjunction with microbiologic data.Clinical improvement usually takes 48 to 72 h, andthus, therapy should not be changed during thistime unless there is rapid clinical decline. Nonresponseto therapy is usually evident by day 3 andcan be ascertained by an assessment of clinicalparameters. 59,61 Serum procalcitonin may act as aprognostic marker in VAP, with increased levelsdespite the use of therapy predicting a poor outcome.83 The responding patient’s antibiotics shouldbe reduced and his or her therapy narrowed tothe most focused regimen possible on the basis ofculture data. 69 The nonresponding patient shouldbe evaluated for noninfectious mimics of pneumonia,unsuspected or drug-resistant organisms,extrapulmonary sites of infection, and complicationsof pneumonia and its therapy. Diagnostictesting should be directed to whichever of thesecauses is likely. 84New DevelopmentsTwice-weekly quantitative surveillance culturesof endotracheal aspirates may assist in the earlyprescription of appropriate antibiotic treatmentsfor patients who acquire VAP. 85 This strategy mayimprove clinical outcomes, potentially may reduceantibiotic resistance in patients in critical careunits and related complications, and may reducehospitalization costs. Antibiotic selection basedon the available results of pre-VAP endotrachealaspirate cultures was adequate in 38 of 40 patients(95%) in a recent small study. 86 In contrast, had theguidelines of the American Thoracic Society/InfectiousDiseases Society of America guidelines 34 andTrouillet et al 14 been used, empiric antibiotic treatmentwould have been adequate in 68% and 83%of patients, respectively. The main reason for theinadequate coverage using the published guidelineswas the failure of the guidelines to informthe empiric treatment selection in the coverageof highly resistant pathogens. Before this strategycan be widely adopted, prospective randomizedtrials are needed to confirm this observation. Thisapproach to the diagnosis and management ofnosocomial pneumonia has been summarized inthe recently published American Thoracic Societyguidelines for the management of adults with HAP,VAP, and health-care--associated pneumonia. 34In a prospective, observational, cohort study,variables independently associated with ICU rateof mortality were derived from 441 patients withVAP. 85 Results were converted into a four- variablescore based on the PIRO (predisposition, insult,response, organ dysfunction) concept for ICU mortalityrisk stratification in VAP patients. The VAPPIRO score identified four independent variablesassociated with mortality (comorbidities [COPD,immunocompromise, heart failure, cirrhosis, orchronic renal failure], bacteremia, systolic BP 90mm Hg, and ARDS), allowing assessment of severity.On the basis of observed mortality for each VAPPIRO score, patients were stratified into three levelsof risk: (1) mild, 0 to 1 points; (2) high, 2 points;and (3) very high, 3 to 4 points. The VAP PIROscore was associated with greater risk of death inthe high-risk group and the very-high-risk group.The use of this scoring may improve prediction ofoutcomes, allowing the more appropriate use ofmedical resources.Sixty episodes of VAP treated with appropriatetherapy (H influenzae, 15 episodes; methicillinsensitiveS aureus, 15 episodes; P aeruginosa,15 episodes; and MRSA 15 episodes), and 30episodes with initial inappropriate therapy, allcaused by P aeruginosa, were compared in a prospectiveobservational study. 87 A significant delayin the resolution of hypoxemia was observed inVAP episodes caused by MRSA and P aeruginosawith inappropriate antibiotic therapy when comparedwith the remaining pathogens. MRSA VAPwas associated with longer respiratory supportindependent of the appropriateness of initial antibiotictherapy. More aggressive management maybe required for these difficult to treat pathogens.A prospective randomized trial compared conventionaland continuous aspiration of subglotticsecretions (CASS) procedures in ventilated patients<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 395ACC-PULM-09-0302-025.indd 3957/10/09 8:50:16 PM


after major heart surgery. 88 CASS was a safe procedurethat reduced the use of antimicrobial agentsin the overall population and the incidence of VAPin at-risk patients (ventilated for 48 h). The costof the CASS tube was $12.60 (€9) vs $2.10 (€1.5)for the conventional tube. This is more evidence tosupport the use of these more expensive endotrachealtubes, at least in patients undergoing majorheart surgery.In a prospective observational study involving105 consecutive patients receiving mechanicalventilation and undergoing BAL, BAL fluid solubletriggering receptor expressed on myeloid cells-1concentration was measured to determine thediagnostic utility of this measurement. 89 Receiveroperating curve analysis and multivariate logisticregression analysis demonstrated that measurementof soluble triggering receptor expressed onmyeloid cells-1 was inferior to clinical parametersfor the diagnosis of VAP. Previous studies hadsuggested this measurement may assist in thediagnosis of VAP, and therefore, these findings aredisappointing.In a prospective, randomized, controlled,unblinded, multicenter study, patients withnosocomial tracheobronchitis were randomlyassigned to receive or not receive IV antibioticsfor 8 days. 90 Antimicrobial treatment wasassociated with a greater number of days free ofmechanical ventilation and lower rates of VAPand ICU rate of mortality. However, antibiotictreatment had no significant impact on totalduration of mechanical ventilation. This studyand previous ones demonstrate the importanceof nosocomial tracheobronchitis as a precursor toVAP and suggest therapy at this early stage mayimprove outcomes.References1. Tablan OC, Anderson LJ, Besser R, et al. Guidelinesfor preventing health-care-associated pneumonia,2003: recommendations of the CDC andthe Healthcare Infection Control Practices AdvisoryCommittee. MMWR Recomm Rep 2004;53(RR-3):1−362. Craven DE, Kunches LM, Kilinsky V, et al. Riskfactors for pneumonia and fatality in patientsreceiving continuous mechanical ventilation. AmRev Respir Dis 1986; 133:792−7963. Chastre J, Fagon JY. Ventilator-associated pneumonia.Am J Respir Crit Care Med 2002; 165:867−9034. Fagon JY, Chastre J, Hance AJ, et al. Nosocomialpneumonia in ventilated patients: a cohort studyevaluating attributable mortality and hospital stay.Am J Med 1993; 94:281−2885. Rello J, Ollendorf DA, Oster G, et al. Epidemiologyand outcomes of ventilator-associated pneumoniain a large US database. Chest 2002; 122:2115−21216. Celis R, Torres A, Gatell JM, et al. Nosocomialpneumonia: a multivariate analysis of risk andprognosis. Chest 1988; 93:318−3247. Torres A, Aznar R, Gatell JM, et al. Incidence, risk,and prognosis factors of nosocomial pneumonia inmechanically ventilated patients. Am Rev RespirDis 1990; 142:523−5288. Richards MJ, Edwards JR, Culver DH, et al.Nosocomial infections in medical ICUs in theUnited States: National Nosocomial InfectionsSurveillance System. Crit Care Med 1999; 27:887−8929. Cook DJ, Walter SD, Cook RJ, et al. Incidence ofand risk factors for ventilator-associated pneumoniain critically ill patients. Ann Intern Med 1998;129:433−44010. Brochard L, Mancebo J, Wysocki M, et al. Noninvasiveventilation for acute exacerbations of chronicobstructive pulmonary disease. N Engl J Med 1995;333:817−82211. Antonelli M, Conti G, Rocco M, et al. A comparisonof noninvasive positive-pressure ventilation andconventional mechanical ventilation in patientswith acute respiratory failure. N Engl J Med 1998;339:429−43512. Hilbert G, Gruson D, Vargas F, et al. Noninvasiveventilation in immunosuppressed patients withpulmonary infiltrates, fever, and acute respiratoryfailure. N Engl J Med 2001; 344:817−82213. American Thoracic Society. Hospital-acquiredpneumonia in adults: diagnosis, assessment ofseverity, initial antimicrobial therapy, and preventivestrategies. Am J Respir Crit Care Med 1996;153:1711−172514. Trouillet JL, Chastre J, Vuagnat A, et al. Ventilatorassociated pneumonia caused by potentially drugresistantbacteria. Am J Respir Crit Care Med 1998;157:531−53915. Heyland DK, Cook DJ, Griffith L, et al. The attributablemorbidity and mortality of ventilator-associated396 Hospital-Acquired and Ventilator-Associated Pneumonia (Grossman)ACC-PULM-09-0302-025.indd 3967/10/09 8:50:16 PM


pneumonia in the critically ill patient. Am J RespirCrit Care Med 1999; 159:1249−125616. Rello J, Ausina V, Ricart M, et al. Impact of previousantimicrobial therapy on the etiology and outcomeof ventilator-associated pneumonia. Chest1993; 104:1230−123517. Chastre J, Trouillet JL, Vuagnat A, et al. Nosocomialpneumonia in patients with acute respiratorydistress syndrome. Am J Respir Crit Care Med1998; 157:1165−117218. Bregeon F, Ciais V, Carret V, et al. Is ventilatorassociatedpneumonia an independent risk factorfor death? Anesthesiology 2001; 94:554−56019. Papazian L, Bregeon F, Thirion X, et al. Effect ofventilator-associated pneumonia on mortalityand morbidity. Am J Respir Crit Care Med 1996;154:91−9720. Craven DE, Steger KA. Epidemiology of nosocomialpneumonia: new perspectives on an old disease.Chest 1995; 108:1S−16S21. Dore P, Robert R, Grollier G, et al. Incidence ofanaerobes in ventilator-associated pneumoniawith use of a protected specimen brush. Am JRespir Crit Care Med 1996; 153:1292−129822. Yu VL, Kroboth FJ, Shonnard J, et al. Legionnaires’disease: new clinical perspective from a prospectivepneumonia study. Am J Med 1982; 73:357−36123. El Solh AA, Sikka P, Ramadan F, et al. Etiology ofsevere pneumonia in the very elderly. Am J RespirCrit Care Med 2001; 163:645−65124. El Solh AA, Aquilina AT, Dhillon RS, et al. Impactof invasive strategy on management of antimicrobialtreatment failure in institutionalized olderpeople with severe pneumonia. Am J Respir CritCare Med 2002; 166:1038−104325. Lim WS, Macfarlane JT. A prospective comparisonof nursing home acquired pneumonia with communityacquired pneumonia. Eur Respir J 2001;18:362−36826. Fridkin SK. Increasing prevalence of antimicrobialresistance in intensive care units. Crit Care Med2001; 29:N64−N6827. Rello J, Torres A, Ricart M, et al. Ventilatorassociatedpneumonia by Staphylococcus aureus:comparison of methicillin-resistant and methicillin-sensitiveepisodes. Am J Respir Crit Care Med1994; 150:1545−154928. Torres A, Puig de la Bellacasa J, Xaubet A, et al.Diagnostic value of quantitative cultures of bronchoalveolarlavage and telescoping pluggedcatheters in mechanically ventilated patients withbacterial pneumonia. Am Rev Respir Dis 1989;140:306−31029. Luna CM, Vujacich P, Niederman MS, et al.Impact of B AL data on the therapy and outcomeof ventilator-associated pneumonia. Chest 1997;111:676−68530. Markowicz P, Wolff M, Djedaini K, et al. ARDSStudy Group: multicenter prospective study ofventilator-associated pneumonia during acuterespiratory distress syndrome: incidence, prognosis,and risk factors. Am J Respir Crit Care Med2000; 161:1942−194831. Johanson WG Jr., Pierce AK, Sanford JP, et al. Nosocomialrespiratory infections with Gram-negativebacilli: the significance of colonization of the respiratorytract. Ann Intern Med 1972; 77:701−70632. Rello J, Sa-Borges M, Correa H, et al. Variations inetiology of ventilator-associated pneumonia acrossfour treatment sites: implications for antimicrobialprescribing practices. Am J Respir Crit Care Med1999; 160:608−61333. Marik PE, Careau P. The role of anaerobes inpatients with ventilator-associated pneumonia andaspiration pneumonia: a prospective study. Chest1999; 115:178−18334. American Thoracic Society, Infectious DiseasesSociety of America. Guidelines for the managementof adults with hospital-acquired, ventilatorassociatedand healthcare-associated pneumonia.Am J Respir Crit Care Med 2005; 171:388−41635. Stout J, Yu VL, Vickers RM, et al. Potable watersupply as the hospital reservoir for Pittsburghpneumonia agent. Lancet 1982; 1:471−47236. Yu VL, Kroboth FJ, Shonnard J, et al. Legionnaires’disease: new clinical perspective from a prospectivepneumonia study. Am J Med 1982; 73:357−36137. El-Ebiary M, Torres A, Fabregas N, et al. Significanceof the isolation of Candida species from respiratorysamples in critically ill, non-neutropenic patients.Am J Respir Crit Care Med 1997; 156:583−59038. Krasinski K, Holzman RS, Hanna B, et al. Nosocomialfungal infection during hospital renovation.Infect Control 1985; 6:278−28239. Pachucki CT, Pappas SA, Fuller GF, et al. Influenzaamong hospital personnel and patients: implicationsfor recognition, prevention, and control. ArchIntern Med 1989; 149:77−8040. Kollef MH. The prevention of ventilator-associatedpneumonia. N Engl J Med 1999; 340:627−634<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 397ACC-PULM-09-0302-025.indd 3977/10/09 8:50:16 PM


41. Craven DE, Steger KA. Nosocomial pneumonia inmechanically ventilated adult patients: epidemiologyand prevention in 1996. Semin Respir Infect1996; 11:32−5342. Valles J, Artigas A, Rello J, et al. Continuous aspirationof subglottic secretions in preventing ventilator-associatedpneumonia. Ann Intern Med 1995;122:179−18643. Cook D, De Jonghe B, Brochard L, et al. Influenceof airway management on ventilator-associatedpneumonia: evidence from randomized trials.JAMA 1998; 279:781−78744. Bergmans DC, Bonten MJ, van Tiel FH, et al.Cross-colonisation with Pseudomonas aeruginosaof patients in an intensive care unit. Thorax 1998;53:1053−105845. Inglis TJ, Millar MR, Jones JG, et al. Tracheal tubebiofilm as a source of bacterial colonization of thelung. J Clin Microbiol 1989; 27:2014−201846. Bonten MJ. Controversies on diagnosis and preventionof ventilator-associated pneumonia. DiagnMicrobiol Infect Dis 1999; 34:199−20447. Rouby JJ, Laurent P, Gosnach M, et al. Risk factorsand clinical relevance of nosocomial maxillarysinusitis in the critically ill. Am J Respir Crit CareMed 1994; 150:776−78348. Fagon J, Chastre J, Hance A. Evaluation of clinicaljudgment in the identification and treatmentof nosocomial pneumonia in ventilated patients.Chest 1993; 103:547−55349. Sutherland K, Steinberg K, Maunder R, et al.<strong>Pulmonary</strong> infection during the acute respiratorydistress syndrome. Am J Respir Crit Care Med1995; 152:550−55650. Pugin J, Auckenthaler R, Mili N, et al. Diagnosis ofventilator-associated pneumonia by bacteriologicanalysis of bronchoscopic and nonbronchoscopic“blind” bronchoalveolar lavage fluid. Am RevRespir Dis 1991; 143:1121−112951. Fabregas N, Ewig S, Torres A, et al. Clinicaldiagnosis of ventilator associated pneumoniarevisited: comparative validation using immediatepost-mortem lung biopsies. Thorax 1999;54:867−87352. Fartoukh M, Maitre B, Honore S, et al. Diagnosingpneumonia during mechanical ventilation: theclinical pulmonary infection score revisited. Am JRespir Crit Care Med 2003; 168:173−17953. Luna CM, Videla A, Mattera J, et al. Blood cultureshave limited value in predicting severity of illnessand as a diagnostic tool in ventilator-associatedpneumonia. Chest 1999; 116:1075−108454. Marquette CH, Herengt F, Mathieu D, et al. Diagnosisof pneumonia in mechanically ventilatedpatients: repeatability of the protected specimenbrush. Am Rev Respir Dis 1993; 147:211−21455. Michaud S, Suzuki S, Harbarth S. Effect ofdesign-related bias in studies of diagnostic testsfor ventilator-associated pneumonia. Am J RespirCrit Care Med 2002; 166:1320−132556. Souweine B, Veber B, Bedos JP, et al. Diagnosticaccuracy of protected specimen brush and bronchoalveolarlavage in nosocomial pneumonia:impact of previous antimicrobial treatments. CritCare Med 1998; 26:236−24457. Fagon JY, Chastre J, Wolff M, et al. Invasive andnoninvasive strategies for management of suspectedventilator-associated pneumonia: a randomizedtrial. Ann Intern Med 2000; 132:621−63058. Cook D, Mandell L. Endotracheal aspiration inthe diagnosis of ventilator-associated pneumonia.Chest 2000;117:195S−197S59. Dennesen PJ, Van der Ven AJ, Kessels AG, et al.Resolution of infectious parameters after antimicrobialtherapy in patients with ventilatorassociatedpneumonia. Am J Respir Crit CareMed 2001; 163:1371−137560. Singh N, Rogers P, Atwood CW, et al. Short-courseempiric antibiotic therapy for patients with pulmonaryinfiltrates in the intensive care unit: aproposed solution for indiscriminate antibioticprescription. Am J Respir Crit Care Med 2000;162:505−51161. Luna CM, Blanzaco D, Niederman MS, et al.Resolution of ventilator-associated pneumonia:prospective evaluation of the clinical pulmonaryinfection score as an early clinical predictor of outcome.Crit Care Med 2003; 31:676−68262. Sanchez-Nieto JM, Torres A, Garcia-Cordoba F,et al. Impact of invasive and noninvasive quantitativeculture sampling on outcome of ventilatorassociatedpneumonia: a pilot study. Am J RespirCrit Care Med 1998; 157:371−37663. Ruiz M, Torres A, Ewig S, et al. Noninvasive versusinvasive microbial investigation in ventilatorassociatedpneumonia: evaluation of outcome.Am J Respir Crit Care Med 2000; 162:119−12564. Sole Violan J, Fernandez JA, Benitez AB, et al.Impact of quantitative invasive diagnostic techniquesin the management and outcome of398 Hospital-Acquired and Ventilator-Associated Pneumonia (Grossman)ACC-PULM-09-0302-025.indd 3987/10/09 8:50:17 PM


mechanically ventilated patients with suspectedpneumonia. Crit Care Med 2000; 28:2737−274165. Campbell GD. Blinded invasive diagnostic proceduresin ventilator-associated pneumonia. Chest2000;117:207S−211S66. The Canadian Critical Care Trials Group. A randomizedtrial of diagnostic techniques for ventilatorassociatedpneumonia. N Engl J Med 2006; 355:2619−263067. Kollef MH, Sherman G, Ward S, et al. Inadequateantimicrobial treatment of infections: a risk factorfor hospital mortality among critically ill patients.Chest 1999; 115:462−47468. Luna CM, Aruj P, Niederman MS, et al. Appropriatenessand delay to initiate therapy in ventilator-associatedpneumonia. Eur Respir J 2006;27:158−16469. Ibrahim EH, Ward S, Sherman G, et al. Experiencewith a clinical guideline for the treatment of ventilator-associatedpneumonia. Crit Care Med 2001;29:1109−111570. Namias N, Samiian L, Nino D, et al. Incidence andsusceptibility of pathogenic bacteria vary betweenintensive care units within a single hospital: implicationsfor empiric antibiotic strategies. J Trauma2000; 49:638−64571. Trouillet JL, Vuagnat A, Combes A, et al. Pseudomonasaeruginosa ventilator-associated pneumonia:comparison of episodes due to piperacillinresistantversus piperacillin-susceptible organisms.Clin Infect Dis 2002; 34:1047−105472. Garnacho-Montero J, Garcia-Garmendia JL, Barrero-AlmodovarA, et al. Impact of the outcomeof adequ ate empirical antibiotherapy in patientsadmitted to the ICU for sepsis. Crit Care Med 2003;31:2742−275173. Leone M, Bourgoin A, Cambon S, et al. Empiricalantimicrobial therapy of septic shock patients:adequacy and impact on the outcome. Crit CareMed 2003; 31:462−46774. Fink MP, Snydman DR, Niederman MS, et al.Treatment of severe pneumonia in hospitalizedpatients: results of a multicenter, randomized,double-blind trial comparing intravenous ciprofloxacinwith imipenem-cilastatin. AntimicrobAgents Chemother 1994; 38:547−55775. Gruson D, Hilbert G, Vargas F, et al. Rotationand restricted use of antibiotics in a medicalintensive care unit: impact on the incidenceof ventilator- associated pneumonia caused byantibiotic- resistant Gram-negative bacteria. Am JRespir Crit Care Med 2000; 162:837−84376. Chastre J, Wolff M, Fagon JY, et al. Comparison of8 vs 15 days of antibiotic therapy for ventilatorassociatedpneumonia in adults: a randomizedtrial. JAMA 2003; 290:2588−259877. Garnacho-Montero J, Ortiz-Leyba C, JimenezJimenez FJ, et al. Treatment of multidrugresistantAcinetobacter baumannii ventilatorassociatedpneumonia (VAP) with intravenouscolistin: a comparison with imipenem-susceptibleVAP. Clin Infect Dis 2003; 36:1111−111878. Paterson DL, Ko WC, Von Gottberg A, et al. Outcomeof cephalosporin treatment for seriousinfections due to apparently susceptible organismsproducing extended-spectrum -lactamases:implications for the clinical microbiology laboratory.J Clin Microbiol 2001; 39:2206−221279. Wunderink RG, Rello J, Cammarata SK, et al.Linezolid vs vancomycin: analysis of two doubleblindstudies of patients with methicillin-resistantStaphylococcus aureus nosocomial pneumonia.Chest 2003; 124:1789−179780. Rubinstein E, Cammarata S, Oliphant T, et al.Linezolid (PNU-100766) versus vancomycin inthe treatment of hospitalized patients with nosocomialpneumonia: a randomized, double blind,multicenter study. Clin Infect Dis 2001; 32:402−41281. Kollef MH, Vlasnik J, Sharpless L, et al. Scheduledchange of antibiotic classes: a strategy to decreasethe incidence of ventilator-associated pneumonia.Am J Respir Crit Care Med 1997; 156:1040−104882. Gruson D, Hilbert G, Vargas F, et al. Strategy ofantibiotic rotation: long-term effect on incidenceand susceptibilities of Gram-negative bacilliresponsible for ventilator associated pneumonia.Crit Care Med 2003; 31:1908−191483. Luyt C-E, Guerin V, Combes A, et al. Procalcitoninkinetics as a prognostic marker of ventilatorassociatedpneumonia. Am J Respir Crit Care Med2005; 171:48−5384. Meduri GU, Mauldin GL, Wunderink RG,et al. Causes of fever and pulmonary densitiesin patients with clinical manifestations ofventilator-associated pneumonia. Chest 1994;106:221−23585. Lisboa T, Diaz E, Sa-Borges M, et al. The ventilatorassociatedpneumonia PIRO score. Chest 2008;134:1208−1216<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 399ACC-PULM-09-0302-025.indd 3997/10/09 8:50:17 PM


86. Michel F, Franceschini B, Berger P, et al. Earlyanti biotic treatment for BAL-confirmed ventilator-associatedpneumonia: a role for routine endotrachealaspirate cultures. Chest 2005; 127:589−59787. Vidaur L, Planas K, Sierra R, et al. Ventilatorassociatedpneumonia. Chest 2008; 133:625−63288. Bouza E, Pérez MJ, Muñoz P, et al. Continuousaspiration of subglottic secretions in the preventionof ventilator-associated pneumonia in thepostoperative period of major heart surgery. Chest2008; 134:938−94689. Anand NJ, Zuick S, Klesney-Tait J, et al. Diagnosticimplications of soluble triggering receptorexpressed on myeloid cells-1 in bal fluid of patientswith pulmonary infiltrates in the ICU. Chest 2009;135:641−64790. Nseir S, Favory R Jozefowicz E, et al. Antimicrobialtreatment for ventilator-associated tracheobronchitis:a randomized, controlled, multicenterstudy. Critical Care 2008; 12:R62400 Hospital-Acquired and Ventilator-Associated Pneumonia (Grossman)ACC-PULM-09-0302-025.indd 4007/10/09 8:50:17 PM


Hypoxemic Respiratory FailureCurtis N. Sessler, MD, FCCPObjectives:• Examine the causes of hypoxemic respiratory failure• <strong>Review</strong> the pathophysiology, diagnostic criteria, clinicalfeatures, and epidemiology for acute lung injury (ALI) andthe ARDS• Examine management strategies for ALI/ARDS, includingmechanical ventilation, medical management, and rescuetherapy for refractory hypoxemiaKey words: acute lung injury; ARDS; hypoxemia; mechanicalventilationHypoxemic Respiratory Failure<strong>Pulmonary</strong> Gas ExchangeThe disruption of gas exchange, as a result ofone or more of numerous pathophysiologic mechanisms,can cause a significant elevation in Paco 2,ie, hypercapnia, and/or a reduction in Pao 2, ie,hypoxemia. The normal range of Pao 2varies withage as a result of the loss of effective alveoli duringadvancing age and can be approximated from thefollowing equation: Pao 2 100.1 0.32 (age inyears). This equation assumes the individual isbreathing ambient air and is near sea level. Pao 2will decline progressively with increasing altitudegreater than sea level as a result of decreasingbarometric pressure and, thus, lower Po 2of inhaledgas.Hypoxemia and Hypoxia: Mechanismsand MeasurementHypoxemia usually arises as a result of disruptedalveolar ventilation and/or perfusion thatcauses relative impaired or absent ventilation ofperfused alveoli. The range of ventilation−perfusionmismatching of the 300 million or so alveoli runsthe spectrum from near absence of ventilation(shunt-like) to absence of perfusion (dead space).In fewer cases, hypoxemia is the result of trueshunting of venous blood that admixes with oxygenatedblood, producing hypoxemia. These conditionscause an increased gradient betweencalculated Pao 2and measured Pao 2, as calculatedin the equation:P(a a)O 2 (barometric pressure 47 mm Hg) 0.21 Paco 2/RQ Pao 2,where 47 mm Hg represents the partial pressureof water vapor, RQ is the respiratory quotient(generally regarded to be 0.8 to 1.0), and P(a-a)O 2is alveolar-arterial oxygen pressure difference.While one is breathing ambient air at sea level,Pao 2 150 (Paco 2/0.8), and P(a-a)O 2should be 10 to 15 mm Hg. In contrast, generalized alveolarhypoventilation without ventilation/perfusionmismatch can result in hypercapnia plus hypoxemiawith a normal P(a-a)O 2. Supplemental oxygenoften is administered for hypoxemia, and in thissetting, the relative degree of hypoxemia is moreaccurately estimated by use of the ratio of Pao 2tothe fraction of oxygen of inhaled gas (Fio 2), or Pao 2:Fio 2, than with P(a-a)O 2. In the presence of diffuseparenchymal lung disease, the ratios 300 mm Hgand 200 mm Hg are general thresholds for moderateand severe hypoxemia, respectively.Although hypoxemia represents low oxygentension in blood, tissue hypoxia can occur throughvarious mechanisms that impair oxygen deliveryor utilization from a cellular perspective. Reductionsin oxygen delivery include hypoxemichypoxia from reduced oxygen tension in arterialblood, anemic hypoxia (in which reductions inoxygen content of blood is caused by reducedoxygen-carrying capacity via loss of hemoglobinor hemoglobin dysfunction), and circulatoryhypoxia (in which depressed cardiac output orvascular factors impair oxygen delivery to tissue).Finally, metabolic forms of hypoxia occur atthe cellular level when oxygen utilization isdisrupted.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 401ACC-PULM-09-0302-026.indd 4017/11/09 2:09:15 PM


Clinical Manifestations and CausativeConditions of HypoxemiaA variety of symptoms and clinical signs mayaccompany hypoxemia, including dyspnea, tachypnea,tachycardia, hypertension, cardiac arrhythmias(including bradycardia progressive to asystolein extreme cases), tremor, anxiety, delirium, andagitation. Often additional clinical manifestationsare present that reflect the underlying causativecondition(s). The pace of onset of clinical manifestationscan be more gradual, as with slow progressionof long-standing illness, or may be abruptfrom rapidly progressive cardiopulmonary disease.Common clinical conditions which are associatedwith hypoxemic respiratory failure are listedin Table 1. Among these associated clinical conditions,acute lung injury (ALI)/ARDS represents acontinuum of a syndrome of parenchymal lungdisease characterized by high-protein alveolaredema and diffuse inflammation that usually hasa specific underlying cause(s).Management Principles for HypoxemicRespiratory FailureCare of the patient who has hypoxemic respiratoryfailure includes initial evaluation and stabilization,assessment and management of theunderlying conditions, support of oxygenation andventilation, application of interventions identifiedas being associated with improved outcomes,consideration of use of rescue therapies for refractoryhypoxemia, and supportive care. GeneralTable 1. Common Clinical Conditions Associated WithHypoxemic Respiratory FailureAcute lung injury (ALI)/ARDS<strong>Pulmonary</strong> edemaDiffuse alveolar hemorrhage<strong>Pulmonary</strong> embolismInterstitial lung diseaseInfectious pneumoniaPneumonitisNeoplasm<strong>Pulmonary</strong> contusionAtelectasisEmphysemaAsthmaChronic bronchitisBronchiolitisconsiderations for initial evaluation and stabilizationand support of oxygenation and ventilationin hypoxemic respiratory failure are discussedhere, whereas issues more directly related toALI / ARDS, such as evidence-based interventionsand rescue therapy for refractory hypoxemia, areaddressed below.Initial Evaluation and Stabilization: Basic lifesupportmeasures, including management of theairway and breathing, should be rapidly initiated.Life-threatening respiratory distress will oftenrequire urgent endotracheal intubation andmechanical ventilation. Assessment of oxygenationand ventilation by arterial blood gas analysis isuseful, and pulse oximetry may provide continuousdisplay of oxygenation, including response totherapy. Initial evaluation should also includetimely detection and management of urgent conditions,such as airway obstruction or tension pneumothorax.Supplemental Oxygen and Artificial Ventilation:Supplemental oxygen can be administered by avariety of interfaces, including face mask, face tent,and nasal cannula, with greater effective Fio 2beingdelivered with a tight-fitting nonrebreather mask.Mechanical ventilation, with delivery of positivepressure breaths for alveolar ventilation, as wellas positive end-expiratory pressure (PEEP) foralveolar recruitment and distention, can be providedby the use of a ventilator and delivered viaan endotracheal or tracheostomy tube or with atightly fitting mask applied to the face or nose. Theaddition of PEEP is particularly effective forimproving oxygenation in the setting of diffuselung disease and poorly compliant lungs as seenin ALI/ARDS and similar conditions. Detailsregarding positive pressure ventilation, includingnoninvasive positive pressure ventilation are discussedmore extensively in other chapters.ALI and ARDSDefinitionsARDS was first described in 1967 as the acuteonset of hypoxemic respiratory failure accompaniedby diffuse pulmonary infiltrates in the absenceof a cardiac failure and after an inciting event.Although broadly debated, specific diagnosticcriteria based on readily available information402 Hypoxemic Respiratory Failure (Sessler)ACC-PULM-09-0302-026.indd 4027/11/09 2:09:16 PM


form the basis for the American European consensusconference definitions, the essential elementsof which are displayed in Table 2. It is noteworthythat exclusion of left atrial hypertension as thecause for pulmonary edema is inexact. For example,measurement of pulmonary capillary wedgepressure (PCWP), or more properly pulmonaryartery occlusion pressure, is performed infrequentlyin current practice, and elevations in PCWP mayreflect volume expansion in the setting of ALI asthe result of sepsis-induced capillary leak, ratherthan heart failure, per se. In the ARDS Network(ARDSNet) randomized controlled trial (RCT) thatcompared fluid management guided by measurementsfrom a central venous catheter vs a pulmonaryartery catheter (PAC), 29% of patients withALI had PCWP 18 mm Hg, nearly all of whomhad normal or elevated cardiac output. Althoughincreased alveolar capillary leak for protein as wellas pulmonary inflammation, factors that mightmore accurately reflect the pathophysiology of ALIthan other criteria, have been demonstrated inresearch settings, such measurements are not availabilityfor widespread clinical use.Clinical FeaturesClinical features of ALI/ARDS at the onset ofillness include manifestations of hypoxemia asdescribed previously as well as signs and symptomsrelated to the underlying cause(s). When datafrom large RCTs are used, it is shown that mostpatients have sepsis, pneumonia, gastric aspiration,and/or trauma as the cause. It is noteworthythat the manifestations that are directly related toALI/ARDS can change over time. In the earlyphase of ARDS, progressive hypoxemia andincreased work of breathing caused by the developmentof pulmonary edema typically results inhypoxemia that is partially responsive to supplementaloxygen but may also lead to frank respiratorydistress and ventilatory failure. The pace ofonset varies, but in general, is faster with aspirationlung injury (90% of cases of ARDS develop within48 h of insult) than sepsis (75% within 48 h) ortrauma (65% within 48 h). Some patients recoverquickly with effective transport of sodium andwater from alveoli and re-epithelialization withtype II pneumocytes, whereas others lapse into acondition characterized by ongoing inflammationand disorganized fibrosis. This “fibroproliferative”phase of ARDS often is heralded by ongoing fever,leukocytosis, and acute alveolar inflammation,with a predominance of neutrophils found in BALfluid.The chest radiograph typically demonstratesbilateral infiltrates that are patchy or confluent andoften are described as alveolar or ground glass. Incontrast to cardiogenic pulmonary edema, featuressuch as cardiomegaly, pleural effusions, and wideningof the vascular pedicle are not prominent.Although these infiltrates usually appear diffuseon chest radiograph, cross-sectional imaging withthe use of CT reveals heterogeneity of infiltrates.Often there is consolidation involving dependentlung units with areas of near-normal appearing lungand atelectatic lung units in mid and upperlung zones, as seen in Figure 1. Interestingly,quickly after moving the patient from supine toTable 2. American European Consensus Criteria for ALI andARDSAll features must be present:1. Acute onset ( 7 days)2. Hypoxemiaa. ALI: Pao 2:Fio 2 300 mm Hgb. ARDS: Pao 2:Fio 2 200 mm Hg3. Diffuse bilateral pulmonary infiltrates on frontal chestradiograph consistent with pulmonary edema; infiltratescan be patchy and/or asymmetric4. absence of left atrial hypertension based upon clinicalassessment or PCWP 18 mm Hg if measured.Figure 1. CT section of chest in a patient with ARDS thatdemonstrates dependent consolidation and areas of normallung and atelectatic lung in the mid and anterior zones.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 403ACC-PULM-09-0302-026.indd 4037/11/09 2:09:16 PM


prone position, CT may demonstrate migration ofconsolidation and infiltrates to the ventral lungunits which are now in a dependent position.Causes of ALI/ARDSThe causes of ALI/ARDS often are categorizedbased on the mechanism of injury: direct (or epithelial)or indirect (or vascular). The more commoncauses are listed in Table 3 using this categorization.When one considers comprehensive epidemiologystudies, sepsis, pneumonia, trauma, andgastric aspiration account for the majority ofcases.Pathophysiology of ALI / ARDSThe acute phase of ALI/ARDS is characterizedby the accumulation of protein-rich edema fluidwithin the lung interstitium spaces and the alveolias a result of increased permeability of the alveolarcapillary endothelium and the alveolar epithelium,and a highly inflammatory lung injury. Net fluidaccumulation is influenced by the influx of fluidand proteins as well as clearance which is normallyprimarily from lymphatic drainage of peribronchovascularedema fluid. Epithelial disruption is aprimary cause of alveolar flooding, and injury totype I pneumocytes contributes to impaired transportof sodium, chloride, and water from alveoli.Further, surfactant is inactivated and diluted. Asepithelium is sloughed, hyaline membranes areformed.The injury to epithelium and endothelium isinterwoven with an inflammatory injury. Neutrophilsmigrate from the vasculature into interstitiumand alveoli, where they release inflammatorymediators, proteases, and oxidants. Neutrophilsare recruited and activated in part because ofrelease by alveolar macrophages of cytokines suchas tumor necrosis factor-, interleukin (IL)-1, IL-6,and IL-8. Other cellular sources of mediators alsoare important, because ALI/ARDS occurs in neutropenicpatients. Additionally, fibroblasts areactivated to produce extracellular matrix. Fibrin isabundant, as is collagen, particularly in a state ofprolonged inflammation and fibrosis. Manypatients quickly progress to a recovery phase characterizedby phagocytosis of apoptotic neutrophils,clearance of sodium by the sodium pump oftype II epithelial cells, and water efflux via aquaporins.The presence of alveolar and interstitialedema, fibrosis, and inflammation contributes tostiff poorly compliant lungs. Many alveoli are notfunctional gas exchange units because they arefilled with fluid and/or inflammatory exudates, orare prone to collapse, and thus contribute to shuntlikegas exchange resulting in refractory hypoxemia.The reader is referred to excellent reviews byWare and Matthay for further information.Table 3. Common Causes of ALI/ARDSDirect causes<strong>Pulmonary</strong> infectionAspiration injury (gastric contents, near drowning, blood,toxins)Inhalation injury (smoke, toxins)Trauma (contusion)Embolism (amniotic fluid, fat, air)Re-expansion injuryReperfusion injuryIndirect causesSevere sepsisShockNonpulmonary traumaTransfusion-related lung injuryCardiopulmonary bypassAnaphylaxisMedications (opioids, salicylates, amiodarone, tocolytics,chemotherapy)Acute pancreatitisEpidemiology of ALI/ARDSEpidemiologic estimates of the incidence ofALI and ARDS vary widely from 1.5 to 75 per100,000 annually but are highly dependent on thepopulation studied, the definitions used, andother factors. Some of the best data regarding UScases comes from studies conducted in Seattle, inwhich Rubenfeld et al estimated crude incidenceof 79.8 and 58.7 per 100,000 for ALI and ARDS,respectively. Extrapolated to the US population,190,600 and 141,500 cases are estimated to occurannually.The authors of these same studies reported inhospitalmortality rates of 38.5% and 41.1% for ALIand ARDS, respectively, by using data collected in1999 and 2000. Of note, mortality rates tend to belower for ALI/ARDS from trauma compared with404 Hypoxemic Respiratory Failure (Sessler)ACC-PULM-09-0302-026.indd 4047/11/09 2:09:17 PM


cases arising after sepsis or gastric aspiration. Ananalysis of published reports suggests that mortalityrates for ALI/ARDS have declined approximately1% per year from 1994 to 2006. Mortalityrates in recent large RCTs of patients with ALI,performed primarily in AMCs, reveal 30- to 60-daymortality rates of 25 to 30%.<strong>Pulmonary</strong> function tends to return to nearnormal during the ensuing 12 months for survivors,with residual reduced diffusing capacity andexercise-induced hypoxemia seen in some patients.In contrast, quality of life is often impaired afterARDS and the causative event(s), with a high proportionof patients having neuropsychologicalproblems and neuromuscular weakness as prominentcomponents.An Overview of Management of ALI/ARDSManagement of ALI/ARDS incorporates treatmentof the underlying cause; support for gasexchange; supportive and preventative measuressuch as adequate nutrition and prophylaxisagainst deep venous thrombosis, stress gastriculceration, and ventilator-associated pneumonia;application of evidence-based strategies for ventilationand medical care ALI/ARDS; and rescuetherapies for refractory hypoxemia in selectedcases. There is a long history of proposed therapiesfor ARDS, many subjected to large-scaleRCTs. Yet very few interventions have been demonstratedto improve key outcomes, such assurvival, or reduced duration of mechanical ventilationand length of stay (LOS) in ICU. There isgeneral agreement that use of lung protectiveventilation, including limiting tidal volumes (Vts)to 6 mL/kg of predicted body weight (PBW) isa key component of contemporary management.Some interventions improve physiologic parameters,like oxygenation, that have relevance forclinical management. Interestingly, there is notconsistent concordance between physiologicimprovement and enhanced outcome. Despiteconsiderable investigation, some interventionssuch as systemic corticosteroids, prone positioning,and high levels of PEEP remain controversialdespite extensive study. Management will bediscussed as ventilatory management and medicalmanagement.Ventilatory Management of ALI/ARDSVentilatory management goals for ALI/ARDSinclude enhancing the likelihood of survival, avoidanceof ventilator- and airway-related complications,reducing the duration of mechanicalventilation and critical illness, supporting gasexchange, and relieving distress and excessivework of breathing. Key principles of ventilatorymanagement are guided by observations fromexperimental results of clinical and animal-modelinvestigations, and imaging studies that help tocorrelate structure and function. First, multiplelines of evidence demonstrate that effective alveolarvolume is reduced substantially in ARDS, mimicking“baby” lungs in size, and that overdistentionof functional alveoli impairs gas exchange andinduces a state of inflammatory lung injury, socalled ventilator-associated lung injury (VALI).Alveolar overdistention is typically produced bycreation of excessive transpulmonary distendingpressure during positive pressure breath delivery.Second, many alveoli are prone to collapseeither throughout the respiratory cycle or as lungvolume and airway pressure decrease during exhalation.Recruiting these alveoli and maintainingthem in an open state improves gas exchange andreduces VALI that is related to injury caused byrepetitive alveolar collapse and recruitment. Third,high concentrations of oxygen are typically necessaryto achieve sufficient cellular oxygenation, butif administered for prolonged periods are associatedwith pulmonary fibrosis in animal models.Thus, current ventilatory strategies for ARDS focuson using small Vts, sufficient PEEP to splint openalveoli, and supplemental oxygen. Also see thechapter on “Mechanical Ventilatory Support” foradditional information about VALI and mechanicalventilation in ARDS.Although earlier RCTs that compared low Vtspositive pressure ventilation to conventional ventilationin patients with ALI/ARDS were not conclusive,the pivotal trial performed by the ARDSNetinvestigators in 861 patients with ALI/ARDS demonstrateda 10% absolute reduction in mortality.Patients randomized to receive Vts of 6 mL/kgPBW also had more ventilator-free and organ failure-freedays compared with patients who received12 mL/kg PBW Vts. Interestingly, there was nosignificant difference in barotrauma rates.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 405ACC-PULM-09-0302-026.indd 4057/11/09 2:09:17 PM


ARDSNet recommendations for mechanicalventilation in ARDS include the following: (1) usethe volume-assist control mode, (2) set inspiratoryflow greater than patient demand (usually 80L/min), (3) start with Vt 8 mL/kg PBW andreduce Vt by 1 mL/kg until 6 mL/kg PBW isreached, (4) aim for plateau inspiratory airwaypressure (Pplat) 30 cm H 2O, (5) increase respiratoryrate, to a maximum of 35 breaths/min, toachieve pH 7.3 to 7.45, (6) consider adding IVbicarbonate if necessary to achieve pH goals, (7)adjust Fio 2and PEEP to achieve Pao 2 55 to 80mm Hg or pulse oxygen saturation 88 to 93%according to Table 4. Note that PBW (in kilograms)can be calculated as 50 2.3 (height in inches –60) for men and 45.5 2.3 (height in inches – 60)for women.Although the results are compelling for adoptionof a low Vt approach to mechanical ventilation,examination of actual clinical practice at ARDSNethospitals and other academic medical centers(AMCs) years after publication of ARDSNet resultsindicates that clinicians often do not reduce lungvolumes adequately. In fact, in one three-centerstudy only 16% of patients with ALI/ARDS hadVt 8 mL/kg PBW. Barriers to adoption includeclinician reluctance to relinquish control of theventilator; lack of recognition of ARDS; and concernsfor contraindications, discomfort, andimpaired gas exchange. However, a recent reportdemonstrated that higher rates of compliance canbe achieved by the use of a written protocol. Theuse of nomograms to easily identify hypoxemia thatreaches the threshold for ALI or ARDS (Fig 2) andthat quickly converts height in inches or cm to predictedbody weight and to the target Vt of 6 mL/kg(Fig 3) may help raise awareness and streamline theprocess of providing lung-protective ventilation.It is noteworthy that the original ARDSNet6 mL/kg Vt approach (described previously) actuallyled to slightly worse oxygenation comparedwith 12 mL/kg Vt. This finding contrasts withother smaller studies, such as those by Amatoet al and Ranieri and coworkers, in which a moreaggressive PEEP strategy was associated with betteroxygenation. As a result of these observationsand the considerable evidence from animal modelsof lung injury that greater PEEP may protectagainst injury from repetitive alveolar collapse andrecruitment (the “atelectrauma” component ofVALI), three large ( 500 subjects) multicenterRCTs designed to compare a high PEEP strategy toTable 4. Adjustments to PEEP and FIO 2During Mechanical Ventilation for ALI/ARDS According to the ARDS Network*Fio 20.3 0.4 0.4 0.5 0.5 0.6 0.7 0.7 0.8 0.9 1.0 1.0PEEP, cm H 2O 5 8 8 10 10 10 14 14 14 18 24*PEEP is increased in increments of 2 cm H 2O, Fio 2by 0.1. From the ARDS Network N Engl J Med 2000; 342:1301−1308.PaO2(mmHg)3503002502001501005000.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1FiO2ALIwithoutARDSARDSFigure 2. Graph of Pao 2vs Fio 2to easily identify if oxygenation criteria is present for ALI or ARDS. Find the point where thehorizontal line for measured Pao 2intersects the vertical line for Fio 2and determine if this point is within the area that definesARDS or ALI without ARDS.406 Hypoxemic Respiratory Failure (Sessler)ACC-PULM-09-0302-026.indd 4067/11/09 2:09:17 PM


Height ininchesMaleHeight incmHeight ininchesFemaleHeight incmPredictedbody weight(in kg)6 ml/kgtidalvolume(in ml)58 147 60 152 45.5 27260 152 62 157 50 30062 157 64 163 54.7 32864 163 66 168 59 35566 168 68 173 64 38368 173 70 178 68.5 41070 178 72 183 73 43872 183 74 188 78 46674 188 76 193 82 493Figure 3. Table to identify Vts that corresponds to 6 mL/kgof PBW based on patient sex and height in inches or centimeters.Find the cell that most closely matches the heightin inches or centimeters for a male (left) or female (right)and identify the corresponding Vt listed in the same row.The PBW is also listed for the corresponding heights. Calculationsare based on the following equations: man: PBW(in kg) 50 2.3 (height in inches −60); woman: PBW(in kg) 45.5 2.3 (height in inches −60).traditional ARDSNet PEEP have been completed.All studies were similar in that PEEP averagedabout 8 to 10 cm H 2O in the low PEEP arm and 10 to16 cm H 2O in the high PEEP arm over the courseof the first week of ARDS. Pao 2:Fio 2ratios wereuniformly greater and rescue therapy for hypoxemiaused less frequently in patients randomized tothe more aggressive PEEP strategy in these studies.None of the studies demonstrated a differencein rates of mortality or barotrauma events; however,more organ failure-free and ventilatorfreedays were seen with greater PEEP in the studyby Mercat and colleagues. In this study, thevolume-assist control mode with Vt 6 mL/kgPBW was used and the PEEP was increased accordingto the Table 5 as long as the pressure plateau(Pplat) remained 30 cm H 2O. A word of cautionregarding the more aggressive PEEP approach ofMercat et al; there was a trend for lower mortalityfor the three quartiles with more severe hypoxemia,but a trend for higher mortality in the quartilewith Pao 2:Fio 2was 180 mm Hg. Accordingly,we would not support this approach if Pao 2:Fio 2is 180 mm Hg. In a recent study, esophagealpressure was used as a surrogate for pleural pressureto guide setting PEEP while avoiding excessivetranspulmonary pressure. Better oxygenationand lung compliance was achieved, withoutincreasing barotrauma in this pilot study.Although the pivotal ARDSNet study prescribedusing the volume-assist control mode,application of a pressure-targeted mode, as doneby Amato et al, in his lung-protective strategy, isof interest to clinicians. A pressure-targeted modemay permit more fine-tuning of Pplat, PEEP, andinspiratory time to improve oxygenation while alsoadhering to Vt 6 mL/kg PBW and acceptablePplat. The “open ventilation” mode tested byMeade and colleagues incorporated pressuretargetedventilation with I:E up to 1:1, greater Pplatto a maximum of 40 cm H 2O, higher PEEP (discussedpreviously), and recruitment maneuversafter disconnection from the ventilator. With thisapproach, they demonstrated superior oxygenation,fewer deaths caused by refractory hypoxemia, andno difference in mortality, barotrauma, or LOS inthe ICU compared with traditional ARDSNet ventilation.A significant limitation of using pressuretargetedventilator modes in ALI/ARDS is the lossof “control” over the Vt and, thus, violation of theimportant goal of using low Vt ventilation. For thesame inspiratory pressure and inspiratory timesettings, a more vigorous inspiratory effort orTable 5. Adjustments to PEEP and FIO 2During Mechanical Ventilation for ALI/ARDS According to the “Increased Recruitment” Armof the EXPRESS Trial*Fio 20.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0PEEP, cm H 2O 5−10 10−18 18−20 20 20 20−22 22 22−24*PEEP is adjusted upwards according to the table, unless Pplat 30 cm H 2O while the patient is receiving volume-assist controlventilation with Vt = 6 mL/kg PBW. PEEP is increased in increments of 2 cm H 2O, Fio 2by 0.1. Note that a trend for greatermortality was noted for patients with Pao 2: Fio 2 180 mm Hg; thus, this strategy is not recommended for ALI without ARDS(ie, Pao 2: Fio 2 200 mm Hg. From Mercat et al. JAMA 2008; 299:646−655.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 407ACC-PULM-09-0302-026.indd 4077/11/09 2:09:18 PM


improvement in lung compliance will result inlarger, and potentially excessive, Vts.A modification of pressure-targeted controlledventilation that permits inspiratory efforts to occurregardless of the phase of ventilation (ie, the socalledbilevel ventilation), may permit the use of lesssedation. Additionally, there is evidence that unrestrictedspontaneous breathing, particularly if thebreaths are not pressure supported, promotes betteraeration of dependent juxtadiaphragmatic lungzones. Investigators have demonstrated better gasexchange with measured decreases in shunt fractionand dead space, as well as increased cardiac indexwith the addition of unsupported spontaneousbreaths throughout ventilation using a pressuretargetedmode. Some clinicians employ a form ofbilevel ventilation with a very long inspiratoryplateau and a very brief ( 0.8 s) exhalation (theso-called airway pressure release ventilation [APRV]).Although oxygenation may be improved withAPRV, there are few data and no adequately poweredRCTs that address important outcomes withthis technique. Further, the very short exhalationtime may be inadequate to fully empty the lungs,particularly if bronchospasm or secretions impedeexhalation, potentially leading to life-threateningauto-PEEP. If a longer inspiratory time is used, weprefer using an expiratory time of 1 s to reducethe likelihood of this dangerous situation fromoccurring. Recruitment maneuvers are used bysome clinicians to open alveoli that are thensplinted open with PEEP. However, large-scalestudies suggested the benefit is generally shortlived.Application of a recruitment maneuver, suchas applying an inspiratory pressure at 40 cm H 2Ofor 40 s, may be reasonable after ventilator disconnectionwith the accompanying loss of alveolardistention.High frequency oscillatory ventilation (HFOV)is perhaps the ultimate form of lung protectiveventilation for ARDS since tidal volumes of 150mL are delivered at rates of 180 to 300 breaths/min.Although HFOV has been demonstrated in RCTsto be equivalent to conventional ventilation, suchstudies were performed with the use of larger(10 t 12 mL/kg PBW) Vts and an RCT comparingHFOV to conventional ventilation with Vt 6 mL/kg PBW is needed. Use of HFOV adds significantcomplexity to patient management anduse should be limited to centers with experienceand immediately available on-site expertise totroubleshoot problems.Nonventilatory Management of ALI/ARDSMany pharmacologic interventions, as well asother nonventilatory treatment strategies, havebeen used in uncontrolled settings or tested inRCTs, most with negative results. Some experimentaltherapies, such as surfactant replacement, continueto undergo active investigation and are notreadily available; thus, they are not discussed.Rather, we will focus on readily available interventionsthat have either compelling or controversialevidence from RCTs, including conservative fluidmanagement with or without albumin administration,specialized nutritional formulations, andcorticosteroids. Additional interventions, such asinhaled nitric oxide (INO) and prone positioning,will be considered within the context of rescuetherapy for refractory hypoxemia.Conservative Fluid Management: Typical managementof a patient with ALI results in a net positivefluid balance of about 1 L/d. There is evidencefrom a large multicenter RCT from the ARDSNetinvestigators that a protocol-directed conservativeapproach to fluid management for patientswith ALI leads to shorter duration of mechanicalventilation (more ventilator-free and ICU-freedays) without negatively impacting renal function,when compared with a “liberal fluid management”conventional approach. There was nodifference in mortality between groups but moreelectrolyte disorders (hypernatremia, hypokalemia,alkalosis) were noted.The research protocol for conservative fluidmanagement is complex, however. Each of 20separate management cells is defined by the centralvenous pressure (CVP) using a central venouscatheter (CVC) or PCWP using a PAC range andone or more of the following clinical characteristics:(1) the presence vs absence of shock (i.e., meanarterial pressure 60 mm Hg or any need for vasopressor),(2) urinary output ( 0.5 mL/kg/h vs 0.5 mL/kg/h), or (3) ineffective circulation (cardiacoutput 2.5 L/min/m 2 or cold, mottled skinwith delayed capillary refill time) or effective circulation.A simplified version (by collapsing severalCVP ranges into one CVP range, and eliminationof the “ineffective circulation” parameter because408 Hypoxemic Respiratory Failure (Sessler)ACC-PULM-09-0302-026.indd 4087/11/09 2:09:18 PM


Table 6. Conservative Fluid Management Strategy for ALI*CVP range,mm HgShock PresentOliguricShock AbsentNonoliguric 9 Vasopressors Diuretic † Diuretic †4–8 Fluid bolus Fluid bolus Diuretic † 4 Fluid bolus Fluid bolus KVO fluids*Shock indicates mean arterial pressure 60 mm Hg or useof any vasopressor. CVP central venous pressure. Oliguriaindicates urine output 0.5 mL/kg/h. KVO “keep veinopen” IV fluid rate.†Diuretics are not given if a vasopressor was used within 12 h.this was rarely present) is presented in Table 6. Ofnote, 75% of interventions performed during thestudy were when the patient was “stable,” ie, normotensiveand nonoliguric; this is the opportunetime to eliminate excessive fluid by diuresis. Workby other investigators suggest that the additionof albumin infusion to patients with ALI and low( 5 mg/dL) serum protein as they are receivingfurosemide diuresis leads to improved oxygenation,fluid balance, and hemodynamics.Nutritional Supplementation: ALI/ARDS is aninflammatory disorder in which oxidants alsoplay a role in lung injury. Altering the fatty acidsthat are incorporated into tissue lipids can modifythe nature of lipid inflammatory mediators thatare subsequently released. The arachidonic acidmetabolites, in particular, are proinflammatoryand may contribute to worsening lung injury. Severalsmaller RCTs have demonstrated that enteraladministration of a nutrition product that containsgamma-linoleic acid, and eicosapentaenoic acid,and antioxidants to patients with ARDS is associatedwith better oxygenation, along with moreventilator-free days, more organ dysfunction-freedays, and more ICU-free days. Additionally, theearlier multicenter RCT by Gadek and colleaguesalso demonstrated reduced alveolar inflammationwith the specialized nutritional formulation.Although addition study in a larger multicenterRCT is needed, the simplicity, low cost, and safetyof this intervention is attractive and such a nutritionalproduct is on the market.Corticosteroids: The use of glucocorticoids inthe treatment of ALI/ARDS as a condition characterizedby intense inflammatory and disorderedfibrotic repair spans decades and has neverbeen free of controversy. In the 1980s, four RCTsdemonstrated the lack of benefit of high-dose,short-duration corticosteroids in preventing thedevelopment of ARDS in high risk patients. Duringthe past two decades, four studies addressingthe administration of corticosteroids early in thecourse of ALI/ARDS suggest that corticosteroidsmay be beneficial in reducing mortality but did notreach statistical significance (odds ratio 0.48, 95%confidence interval, 0.40 to 1.09) when pooled. Alow-dose (1 mg/kg/d methylprednisolone to start,then taper) long-course (28-day) regimen yieldedpositive results (higher Pao 2:Fio 2, shorter durationof mechanical ventilation, shorter ICU LOS)vs placebo in an RCT; however, the results may beclouded by methodologic concerns of the study.Some clinicians have given corticosteroids inmodest doses (such as methylprednisolone [MP]40 mg q6h) to treat unresolving or “fibroproliferative”ARDS that has been present 7 days. Thecondition is typically recognized by fever andleukocytosis without evidence of infection andpersistent pulmonary infiltrates and hypoxemiarequiring ongoing mechanical ventilation. A multicenterRCT performed by the ARDSNet investigatorsyielded mixed results, interpreted by theauthors as showing no benefit. These results, however,deserve further review, and the study designbears closer scrutiny. Proponents of corticosteroidadministration have long emphasized a longercourse of therapy slowly tapered over severalweeks, since more rapid tapering has been linkedto relapse. The ARDSNet trial prescribed taperingthe MP over only 2 days under selected circumstancesincluding septic shock or successful extubationfor 2 days. Critics argue that this rapid tapermay have led to relapse and reintubation or toworsening shock since the administration of corticosteroidsshortens the duration of vasopressor usein septic shock. The results show significantly moreventilator-free days through day 28 with MP vsplacebo (11.2 vs 6.8 days), but higher rates of reintubation.Patients randomized to MP were firstextubated 10 days earlier on average (14 days vs24 days)\compared with the use of placebo.Patients randomized to MP also had significantlylower rates of shock and pneumonia, but did havesignificantly more adverse effects attributed tomyoneuropathy.Interestingly, structured evaluation for myoneuropathydisclosed no difference in the rates<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 409ACC-PULM-09-0302-026.indd 4097/11/09 2:09:18 PM


etween groups. Randomization to MP was alsoassociated with better oxygenation, lower Pplat,and better lung compliance than placebo. Mortalitywas similar between groups at 28, 60, and 180days. Subset analysis demonstrated that amongpatients enrolled 14 days after onset of ARDS,those randomized to placebo had a significantlylower 60-day mortality rate (8%) than MP (35%),accounting for the difference in survival for the latepersistent ARDS group. Although this is a subsetanalysis on a small number of patients (n 48), theauthors conclude (and we concur), that corticosteroidsshould not be initiated to treat unresolvingARDS after day 13, until more data are available.When RCT data are limited to patients with unresolvingARDS who had onset 1 to 13 days earlierand who received treatment for at least 1 week, ameta-analysis by Meduri et al shows better survivalwith MP (p 0.01), although the populationis limited to 245 subjects and includes studies withmethodologic concerns.Further research in well-designed studies andre-analysis of existing data are needed. Of note,measurement of procollagen peptide III in BAL fluidmay prove helpful as a screening test for treatment,since patients who received MP had a significantlower mortality rate compared with placebo whenthis peptide was found in high amounts beforetreatment in the ARDSNet study. If MP is administeredfor unresolving ARDS, the following caveatsshould be considered: (1) starting MP 13 daysafter onset of ARDS should be avoided, (2) the MPdosage should be modest (0.5 mg/kg every 6 h tostart) and tapered slowly, (3) active infection surveillanceshould be performed during MP therapy, and(4) the use of neuromuscular blocking agents(NMBA) concomitantly should be avoided.Rescue Therapy for Refractory HypoxemiaThe incidence of refractory and/or lifethreateninghypoxemia in ALI/ARDS is notknown, although “refractory hypoxemia” wasfound in about 10% and rescue therapies (INO,prone positioning, almitrine) were used in 10 to35% of patients in several recent multicenter RCTsfor mechanical ventilation in ALI/ARDS. Althoughmost patients with fatal ALI/ARDS die frommultiple organ failure, some deaths are the resultof refractory hypoxemia. Typically, these patientshave oxygen desaturation with minimal movement,suctioning, or other minor stimulation thatis very slow to recover. A number of the interventionsdiscussed earlier have been associated withsignificantly improved oxygenation acutely, suchas increased PEEP and APRV. Additionally severalnonventilatory interventions are commonly usedas rescue therapy for refractory and life-threateninghypoxemia, including neuromuscular blockade(NMB), INO, and prone positioning.NMB: The administration of a NMB agent(NMBA) in ALI/ARDS can result in improvedoxygenation, primarily when there is persistentpatient-ventilator dyssynchrony. The frequencywith which NMBA are used in ARDS is highlyvariable, ranging from 10 to 50% of patientsin large multicenter RCTs. The dosage of NMBAcan be titrated to achieve a range of effects, fromsynchronous breathing with the ventilator, toapnea, or even cessation of all movement. WhenNMBAs are used, the lowest effective dose is recommended,and the duration of therapy shouldbe as brief as possible. Further, concomitantadministration of corticosteroids is to be avoidedbecause of the increased likelihood of persistentweakness caused by acute quadriplegic myopathyunder these circumstances.INO: NO is a gas with vasodilator propertiesin minute quantities (ie, 5 to 20 parts per million[ppm]). Mechanistically, INO is an attractive intervention.NO is delivered by inhalation to alveoli,which are ventilated, and then diffuses into capillaries,feeding these alveoli and thus increasingtheir perfusion while not entering nonfunctionalalveoli. Increasing perfusion to all alveoli, such aswith an IV vasodilator, would indiscriminatelyincrease perfusion, thus potentially increasing theproportion of blood flow to shunt-like alveoli andworsening oxygenation. Additionally, NO is inactivatedquickly; thus, it does not produce vasodilatationof systemic vessels that might result insystemic hypotension. In clinical studies, INOincreases Pao 2and decreases pulmonary arterypressure. In dose ranging studies, these effects tendto peak at concentrations of 20 to 40 ppm. Theauthors of RCTs demonstrate improved oxygenationin most patients, sometimes dramatically, butthe effects tend to decline 2 to 4 days. Noneof the multicenter RCTs have shown that the410 Hypoxemic Respiratory Failure (Sessler)ACC-PULM-09-0302-026.indd 4107/11/09 2:09:18 PM


administration of INO to patients with ALI/ARDSimproves chances of survival or reduces ventilatortime. Thus, the role of INO is largely as rescuetherapy for refractory and life-threatening hypoxemia.Potential adverse effects of INO includemethemoglobinemia and increased levels of nitrogendioxide, neither of which is common unlesshigh doses of INO (80 ppm) are used. Because theacquisition cost is quite high, use of INO shouldbe carefully considered, and unless a clinicallyimportant improvement in oxygenation demonstrated,it should be discontinued in a timely fashion.Other inhaled vasodilators are beinginvestigated as alternatives.Prone Positioning: Placing hypoxemic patientswho have ALI/ARDS in the prone positionimproves oxygenation in most. Principle mechanismsby which this occurs include (1) creatinga more uniform distribution of ventilation, withmore ventilation of dependent lung units, therebymore effectively matching ventilation and perfusion;(2) lifting the weight of the heart off of theleft lower lobe, as the heart is now in a dependentposition; and (3) promoting more effectivedrainage of secretions from the airways. Movingthe patient from supine to prone and maintainingthe patient in the prone position for an extendedtime period creates new concerns and potentialcomplications. For example, deeper sedation isneeded, and NMB is required in more patients,the incidence of new or worsening pressure soresincreases, the likelihood of airway obstructionincreases, and the likelihood of accidental removalof critical tubes and lines during position changeis present. RCTs demonstrate a clear increase inoxygenation that is partially lost when the patientis returned to the supine position. Several largemulticenter RCTs failed to demonstrate improvementin survival or reduction in LOS, althoughMancebo et al noted a trend for improved survivaland in post-hoc analysis Gattinoni and coworkersobserved improved mortality amongthe sickest quartile of patients, both by severityof illness and by severity of hypoxemia. Pronepositioning can be useful for severely hypoxemicpatients with ALI/ARDS, but a healthy respectfor the challenging process and use of measuresto safeguard against complications is necessary.Additional RCTs that may further inform clinicalpractice are underway.Summary and ConclusionsHypoxemic respiratory failure has manycauses, but ALI/ARDS is chief among these.Supplemental oxygen, often combined with positivepressure ventilation, usually is necessary asthe underlying cause(s) is treated. Management ofALI/ARDS focuses on ventilatory strategies thatsupport oxygenation and ventilation while limitingthe potential injurious effects of mechanical ventilationon the vulnerable lung. Specifically, low Vtventilation (6 mL/kg PBW) and increased levelsof PEEP are the cornerstones. Nonventilatory measuresinclude avoidance of excessive lung water,and consideration for using specialized nutritionalproducts and corticosteroids in selected circumstances.Ancillary interventions such as administrationof NMBAs, INO, and prone positioningmay be necessary to combat life-threateninghypoxemia in selected cases. Future research willhelp define the role for novel interventions andnew treatment strategies.Annotated BibliographyAdhikari NK, Burns KE, Friedrich JO, et al. Effect ofnitric oxide on oxygenation and mortality in acutelung injury: systematic review and meta-analysis. BMJ2007; 334:779Meta-analysis shows little benefit from INO for routine usein ARDS, although oxygenation usually improves.Amato MB, Barbas CS, Medeiros DM, et al. Beneficialeffects of the “open lung approach” with low distendingpressures in acute respiratory distress syndrome. Aprospective randomized study on mechanical ventilation:am J Respir Crit Care Med 1995; 152:1835−1846First RCT that promoted an “open lung” or lung protectiveapproach. Criticized for high mortality rates. Meade studyreplicates study design and demonstrates equivalency toARDSNet study parameters now 13 years later.Angus DC, Clermont G, Linde-Zwirble WT, et al.Healthcare costs and long-term outcomes after acuterespiratory distress syndrome: a phase III trial ofinhaled nitric oxide. Crit Care Med 2006; 34:2883−2890Poor quality of life after ARDS described.Arroliga AC, Thompson BT, Ancukiewicz M, et al.Use of sedatives, opioids, and neuromuscular blockingagents in patients with acute lung injury and acuterespiratory distress syndrome. Crit Care Med 2008;36:1083−1088<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 411ACC-PULM-09-0302-026.indd 4117/11/09 2:09:19 PM


Use of sedatives, opioids, and NMBAs was similar in lowand high PEEP, and sedatives and opioids were associatedwith delayed weaning.Ashbaugh DG, Bigelow DB, Petty TL, et al. Acute respiratorydistress in adults. Lancet 1967; 2:319–323Original description of ARDS.Bernard GR, Artigas A, Brigham KL, et al. The American-European Consensus Conference on ARDS. Definitions,mechanisms, relevant outcomes, and clinicaltrial coordination: am J Respir Crit Care Med 1994;149:818−824Consensus definitions for ALI/ARDS.Brower RG, Lanken PN, MacIntyre N, et al. Higher versuslower positive end-expiratory pressures in patientswith the acute respiratory distress syndrome. N Engl JMed 2004; 351:327−336ARDSNet RCT found better oxygenation but no other significantdifferences with a more aggressive PEEP strategycompared to the original ARDSNet protocol.Calfee CS, Matthay MA. Nonventilatory treatments foracute lung injury and ARDS. Chest 2007; 131:913−920<strong>Review</strong> of medical management. Substantial focus on fluidmanagement.Classic ARDSNet RCT showing lower mortality with lowerVTs.Dellinger RP, Zimmerman JL, Taylor RW, et al. Effectsof inhaled nitric oxide in patients with acute respiratorydistress syndrome: results of a randomized phaseII trial. Inhaled Nitric Oxide in ARDS Study Group.Crit Care Med 1998; 26:15−23Multicenter RCT showing improved oxygenation but noeffect on mortality or ventilation duration with INO.Derdak S, Mehta S, Stewart TE, et al. High-frequencyoscillatory ventilation for acute respiratory distresssyndrome in adults: a randomized, controlled trial.Am J Respir Crit Care Med 2002; 166:801−808HFOV had outcomes similar to conventional ventilation,but with larger VTS than currently used. Complexity muchhigher than conventional ventilation.de Wit M, Sessler CN. Acute respiratory failure. In:Rakel RE, Bope ET, ed. Conn’s current therapy 2005.Philadelphia, PA: Elsevier Saunders, 2005; 247−253<strong>Review</strong> of respiratory failure.Gadek JE, DeMichele SJ, Karlstad MD, et al. Effect ofenteral feeding with eicosapentaenoic acid, gammalinolenicacid, and antioxidants in patients with acuterespiratory distress syndrome. Enteral Nutrition inARDS Study Group. Crit Care Med 1999; 27:1409−1420Modest sized multicenter RCT showing less inflammation,better oxygenation, and shorter ventilator time for patientswho received feeding modified for reducing lipid mediatorsand supplemented with antioxidants.Gainnier M, Roch A, Forel JM, et al. Effect of neuromuscularblocking agents on gas exchange in patientspresenting with acute respiratory distress syndrome.Crit Care Med 2004; 32:113−119NMBA improved gas exchange substantially.Gattinoni L, Chiumello D, Cressoni M, et al. <strong>Pulmonary</strong>computed tomography and adult respiratory distresssyndrome. Swiss Med Wkly 2005; 135:169−174One of a number of papers by Gattinoni examining structure-functionrelationships for ARDS using CT.Gattinoni L, Tognoni G, Pesenti A, et al. Effect of pronepositioning on the survival of patients with acute respiratoryfailure. N Engl J Med 2001; 345:568−573Early RCT showing better oxygenation and subset analysessuggesting possible mortality benefit for the most severelyill patients.Girard TD, Bernard GR. Mechanical ventilation inARDS: a state-of-the-art review. Chest 2007;131:921−929<strong>Review</strong>, focusing on low VT ventilation.Guerin C, Gaillard S, Lemasson S, et al. Effects of systematicprone positioning in hypoxemic acute respiratoryfailure: a randomized controlled trial. JAMA 2004;292:2379−2387Multicenter RCT demonstrates improved oxygenation butno other benefit to 8 h/d prone positioning.Mancebo J, Fernandez R, Blanch L, et al. A multicentertrial of prolonged prone ventilation in severe acuterespiratory distress syndrome. Am J Respir Crit CareMed 2006; 173:1233−1239Another multicenter RCT showing proning improves oxygenationand has a trend for lower mortality when appliedfor 17 h/day x 10 d. They recommend early and prolongedproning for severe ARDS.Marik PE, Pastores SM, Annane D, et al. Recommendationsfor the diagnosis and management of corticosteroidinsufficiency in critically ill adult patients: consensusstatements from an international task force bythe American College of Crit Care Med Crit Care Med2008; 36:1937−1949SCCM consensus statement concludes that moderate doseglucocorticoids should be considered for early severe ARDSand before day 14 with unresolving ARDS and that steroidsshould be tapered and not stopped abruptly – as 2B recommendations.Note that no ARDSNet investigators were taskforce members.Martin GS, Moss M, Wheeler AP, et al. A randomized,controlled trial of furosemide with or without albumin412 Hypoxemic Respiratory Failure (Sessler)ACC-PULM-09-0302-026.indd 4127/11/09 2:09:19 PM


in hypoproteinemic patients with acute lung injury.Crit Care Med 2005; 33:1681−1687Addition of albumin to furosemide led to better oxygenationand hemodynamic stability.Meade MO, Cook DJ, Guyatt GH, et al. Ventilation strategyusing low tidal volumes, recruitment maneuvers,and high positive end-expiratory pressure for acutelung injury and acute respiratory distress syndrome: arandomized controlled trial. JAMA 2008; 299:637−645Large multicenter RCT of mostly Canadian centers, publishedin the same JAMA issue with the Mercat paper. Both comparehigher PEEP strategies to the traditional ARDSNet approach.All patients received 6 mL/kg VTs. Additionally, higher Pplat,pressure-targeted ventilation, and recruitment maneuvers(Amato-like parameters). Similar outcomes but better oxygenationwith this approach, compared to ARDSNet.Meduri GU, Golden E, Freire AX, et al. Methylprednisoloneinfusion in early severe ARDS: results of arandomized controlled trial. Chest 2007; 131:954−963Recent study of methylprednisolone for early ARDS—positiveresults, but significant methodologic concerns with thestudy.Meduri GU, Marik PE, Chrousos GP, et al. Steroidtreatment in ARDS: a critical appraisal of the ARDSnetwork trial and the recent literature. Intensive CareMed 2008; 34:61−69Essentially a rebuttal and meta-analysis to the SteinbergARDSNet RCT for corticosteroids and unresolvingARDS. Raises important questions about interpretation ofthe ARDSNet data and looks at reasons for re-examining.Mercat A, Richard JC, Vielle B, et al. Positive end-expiratorypressure setting in adults with acute lung injuryand acute respiratory distress syndrome: a randomizedcontrolled trial. JAMA 2008; 299:646−655Multicenter French RCT showing higher PEEP strategyhad better oxygenation and more ventilator-free days, butno more barotrauma compared to the ARDSNet PEEP strategy.Mortality was unchanged.Milberg JA, Davis DR, Steinberg KP, et al. Improvedsurvival of patients with acute respiratory distress syndrome(ARDS): 1983-1993. JAMA 1995; 273:306−309Older paper that showed improving ARDS outcomes at onecenter over a decade.Peter JV, John P, Graham PL, et al. Corticosteroids in theprevention and treatment of acute respiratory distresssyndrome (ARDS) in adults: meta-analysis. BMJ 2008;336:1006−1009Meta-analysis concludes that a possibility of reduced mortalityand increased ventilator days with steroids startedafter the onset of ARDS is suggested.Pontes-Arruda A, Aragao AM, Albuquerque JD. Effectsof enteral feeding with eicosapentaenoic acid, gammalinolenicacid, and antioxidants in mechanically ventilatedpatients with severe sepsis and septic shock. CritCare Med 2006; 34:2325−2333Compared to isocaloric feeding, this enteral feed was associatedwith better oxygenation and other outcomes in patients withsepsis and hypoxemic respiratory failure (ie, ALI/ARDS).Ranieri VM, Suter PM, Tortorella C, et al. Effect ofmechanical ventilation on inflammatory mediators inpatients with acute respiratory distress syndrome: arandomized controlled trial. JAMA 1999; 282:54−61Small but fascinating study that demonstrates beneficialeffects of lung protective ventilation – low VTs + highPEEP on course of BAL and blood cytokine levels and physiologicand outcomes effects as compared to conventionalventilation.Rubenfeld GD, Caldwell E, Peabody E, et al. Barriers toproviding lung-protective ventilation to patients withacute lung injury. Crit Care Med 2004; 32:1289−1293Large epidemiology study of ALI in Seattle area.Rubenfeld GD, Cooper C, Carter G, et al. Incidenceand outcomes of acute lung injury. N Engl J Med 2005;353:1685−1693Identifies reasons why physicians do not do low VT ventilationfor ALI.Sessler CN. Mechanical ventilation of patients withacute lung injury. Crit Care Clin 1998; 14:707−729<strong>Review</strong> of mechanical ventilation for ALI/ARDS, includingmechanisms of VALI in the context of ARDS pathophysiology.Sessler CN. Sedation, analgesia, and neuromuscularblockade for high-frequency oscillatory ventilation.Crit Care Med 2005; 33:S209−S216<strong>Review</strong> of sedation and NMB during ARDS and HFOV.Steinberg KP, Hudson LD, Goodman RB, et al. Efficacyand safety of corticosteroids for persistent acuterespiratory distress syndrome. N Engl J Med 2006;354:1671−1684Controversial ARDSNet placebo-controlled RCT showingmethylprednisolone possible benefit (physiology and markedlymore ventilator-free days) when started 7 to 13 daysafter ARDS onset, but higher mortality when started > 13days into ARDS.Talmor D, Sarge T, Malhotra A, et al. Mechanical ventilationguided by esophageal pressure in acute lunginjury. N Engl J Med 2008; 359:2095−2104In a pilot RCT, guidance of setting PEEP by measuredesophageal pressure as a surrogate for pleural pressure toestimate transalveolar pressure led to better oxygenation and<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 413ACC-PULM-09-0302-026.indd 4137/11/09 2:09:19 PM


lung compliance without increasing barotrauma. Althoughthere are technical concerns regarding accuracy of esophagealpressure, the concept of individualized setting of PEEP(rather than by a table) is attractive.Taylor RW, Zimmerman JL, Dellinger RP, et al. Lowdoseinhaled nitric oxide in patients with acute lunginjury: a randomized controlled trial. JAMA 2004;291:1603−1609Multicenter RCT demonstrating no outcomes benefits toroutine use of INO in ALI.Umoh NJ, Fan E, Mendez-Tellez PA, et al. Patient andintensive care unit organizational factors associatedwith low tidal volume ventilation in acute lung injury.Crit Care Med 2008; 36:1463−1468Use of a written protocol was highly associated with improvedcompliance with low VT ventilation.Ware LB, Matthay MA. Clinical practice. Acute pulmonaryedema: N Engl J Med 2005; 353:2788−2796Mechanisms of edema formation and clearance, plus otherinformation regarding pulmonary edema.Ventilation with lower tidal volumes as compared withtraditional tidal volumes for acute lung injury and theacute respiratory distress syndrome. The Acute RespiratoryDistress Syndrome Network. N Engl J Med 2000;342:1301−1308Ware LB, Matthay MA. The acute respiratory distresssyndrome. N Engl J Med 2000; 342:1334−1349Excellent review, including detailed analysis of mechanisms.Wheeler AP, Bernard GR, Thompson BT, et al. <strong>Pulmonary</strong>-arteryversus central venous catheter to guidetreatment of acute lung injury. N Engl J Med 2006;354:2213−2224Management no better with PAC than CVC, but more minorcomplications with PAC. About 30% of patients with ALIhad PCWP > 18 mm Hg, but normal cardiac output.Wheeler AP, Bernard GR. Acute lung injury and theacute respiratory distress syndrome: a clinical review.Lancet 2007; 369:1553−1564<strong>Review</strong> of current concepts.Wiedemann HP, Wheeler AP, Bernard GR, et al. Comparisonof two fluid-management strategies in acutelung injury. N Engl J Med 2006; 354:2564−2575ARDSNet trial with complex conservative fluid managementprotocol resulted in shorter duration of ventilation by2 days without worsening renal function.Young MP, Manning HL, Wilson DL, et al. Ventilationof patients with acute lung injury and acute respiratorydistress syndrome: has new evidence changed clinicalpractice? Crit Care Med 2004; 32:1260−1265No, only a small percentage of patients had VTs < 8 mL/kgPBW and [mL/kg] virtually none < 6 mL/kg.Zambon M, Vincent JL. Mortality rates for patientswith acute lung injury/ARDS have decreased overtime. Chest 2008; 133:1120−1127By pooling many different studies, the authors conclude thatmortality has fallen about 1% per year from 1994 to 2006.414 Hypoxemic Respiratory Failure (Sessler)ACC-PULM-09-0302-026.indd 4147/11/09 2:09:19 PM


Symptoms of Respiratory DiseaseRichard S. Irwin, MD, FCCPObjectives:• Address the symptoms of respiratory disease commonlyencountered by the pulmonologist• <strong>Review</strong> the physiology, pathophysiology, complications,differential diagnosis, pathogenesis, diagnosis, and treatmentof cough• <strong>Review</strong> the physiology, differential diagnosis, pathogenesis,and diagnosis of wheeze• <strong>Review</strong> the pathogenesis, differential diagnosis, diagnosis,and treatment of hemoptysis• <strong>Review</strong> the physiology, differential diagnosis and pathophysiology,diagnosis, and treatment of dyspneaKey words: cough; dyspnea; hemoptysis; wheezeRespiratory symptoms are among the most commonreasons for which patients seek medical care.The National Ambulatory Medical Care Survey,published in 2006 by the US Department of Healthand Human Services, 1 showed that respiratorysystem-related symptoms were the most commonsymptoms for which patients made office visitsin 1998, accounting for approximately 12% of thetotal. Other common system-related symptomsare shown in Table 1. Of all symptoms reported,cough of undifferentiated duration was the singlemost common complaint for which patients soughtmedical care of primary care physicians. Of theestimated 1.1 billion visits to office-based physiciansin the United States from 2001 to 2002,3.1%, or 34.1 million visits, were for cough. Becausethe common cold is the most common affliction ofmen and women, and it is almost always accompaniedby an acute (ie, 3 weeks in duration) andusually self-limited cough, it is likely that the greatestmajority of the coughs seen by primary carephysicians are acute in duration. With respect tothe pulmonologist, referrals of patients with persistentlytroublesome chronic cough of unknownetiology have been shown to account for 10 to 38%of the outpatient practice. 2,3CoughHealthy people rarely cough during wakinghours or especially during sleep, and when theydo, it is essentially devoid of any clinical importance.The lesser frequency of cough during sleepcompared with wakefulness in normal subjects aswell as in patients with chronic bronchitis andemphysema is likely the result of greater thresholdsto coughing stimuli during sleep. 4 However, whencough is present and persistent, it can assume greatclinical importance. Cough can be an importantdefense mechanism that helps clear excessivesecretions and foreign material from the airway, itcan be an important factor in the spread of infection,and it is one of the most common symptomsfor which patients seek medical attention andspend health-care dollars. Voluntary coughing canbe used as a means of providing cardiopulmonaryresuscitation.Table 1. Spectrum/Frequency of Symptoms by System for WhichPatients Sought Medical Care in the Office Setting in the UnitedStates in 2001 to 2002*Reason for Visit% of SymptomsSymptoms related to respiratory system 11.8Symptoms related to nervous system andeyes and ears. 8.6Symptoms related to circulatory system 8.0Symptoms related to musculoskeletal andconnective tissue 7.9Symptoms of a general nature 7.5Symptoms related to endocrine, nutritional/metabolic/immunitydisorders 5.3Symptoms related to mental disorders 4.8Symptoms related to skin and subcutaneoustissues 4.8Symptoms related to genitourinary system 4.7Symptoms related to digestive system 3.7*Adapted from Schappert and Burt. 1<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 415ACC-PULM-09-0302-027.indd 4157/10/09 10:22:23 PM


Physiology/Pathophysiology of CoughThere are usually three phases involved in theactive cough mechanism. Cough is usually precededby a deep inspiration (inspiratory phase).This initial inspiration is important in producingan effective cough by permitting both expiratorypressure and flow to be maximal during the ensuingexpiration. This high lung volume allowsmaximal expiratory flow rates. During the secondcompressive phase, intrathoracic pressure isincreased sufficiently to produce flow rates necessaryfor effective cough during the expiratoryphase. It is during this latter phase that the defensemechanism function of cough is carried out, thatis, the removal of undesired material from thelower respiratory tract.Although dynamic changes are taking place inthe glottis (eg, vocal cords separate and vibrate, andthe width of the glottis narrows at the aryepiglotticfolds that shake secretions loose from the larynx),the same thoracic and abdominal muscles that wereactive during the compressive phase of coughcontract further. The continued shortening of thesemuscles after the opening of the vocal cords servesto maintain the rapid flow of air by ensuring ahigh-pressure gradient between intrathoracic airwaysand the mouth. Numerous studies havenoted that maximal intrathoracic pressures duringa cough occur after the opening of the glottis, attestingto the contribution of the expiratory musclesin maintaining high pressures and therefore highflow. Along with the larynx and expiratory musculature,the tracheobronchial tree also undergoesdynamic changes that ensure an effective cough.The single most important parameter in theproduction of an effective cough is the linear velocityof the moving column of air. The determinantsof linear velocity are described by the followingformula: velocity = flow/cross-sectional area.Therefore, an effective cough depends primarilyon a small cross-sectional area and high flow rates. 5Most important in maintaining the small crosssectionalarea is the dynamic compression that thetracheobronchial tree undergoes during the forcedexpiration. The most important factor necessaryfor achieving high flow rates is the initial deepbreath of a cough.All clinically important disorders that mitigatethe effectiveness of a cough interfere with theinspiratory or expiratory phases of cough, withmost conditions affecting both, and respiratorymuscles of sufficient strength are key to bothphases. Disorders or interventions that might predominantlyaffect the compressive phase in anegative way are probably not clinically important.For instance, although vocal cord closure is animportant component of the compressive phase, itis not essential for the production of effective coughpressures. The muscles of expiration appear to bethe most important determinant in producingelevated intrathoracic pressures, and they arecapable of doing so even when an endotrachealtube is in place. 6Cough can be predicted to be ineffective at thebedside if patients are unable to or can barelycough on command. However, there are few datathat allow one to predict when cough is approachinga threshold for ineffectiveness that will prohibitivelypredict the risk of substantial gasexchange abnormalities, atelectasis, and/or suppurativedisease of the lower respiratory tract.Observations in patients with myasthenia gravishave suggested that when maximal expiratorypressures are 40 cm H 2O (31 mm Hg), patientsseemed to have difficulty in raising secretionswithout endotracheal suctioning. 7Complications of Cough/CoughCardiopulmonary ResuscitationDuring the expiratory phase of vigorouscoughing, intrathoracic pressures up to 300 mmHg, expiratory velocities up to 28,000 cm/s or 500miles/h (ie, 85% of the speed of sound), and from1 to 25 J of energy can be generated. 8 Althoughpressures, velocities, and energy of these magnitudesallow coughing to be an effective means ofproviding cardiopulmonary resuscitation in theconscious patient, they also cause a variety of complications,as summarized in Table 2.Cough cardiopulmonary resuscitation hasassumed an established place in clinical practiceand has been found to be beneficial in the ECGmonitored,conscious patient with the following:(1) asystole, (2) profound bradycardia with hypotension,and (3) ventricular tachycardia. 8 Althoughvoluntary coughing is unlikely to convert ventricularfibrillation to a more viable rhythm, it canmaintain BP and oxygenation while the defibrillator416 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4167/10/09 10:22:23 PM


Table 2. Complications of CoughCategoryCardiovascularConstitutional symptomsGIGenitourinaryMusculoskeletalNeurologicOphthalmologicPsychosocialQuality of lifeRespiratorySkinExamplesArterial hypotension; loss of consciousness; rupture of subconjunctival/nasal/analveins; massive intraocular suprachoroidal hemorrhage during pars plana vitrectomy;dislodgement/malfunctioning IV catheters; bradytachyarrhythmiasExcessive sweating; anorexia; exhaustionMalfunction of gastrostomy button; splenic rupture; inguinal herniation; cough-inducedgastroesophageal reflux events; gastric hemorrhage after percutaneous endoscopicgastrostomy; hepatic cyst rupture; herniations (eg, inguinal, through abdominal wall,small bowel through laparoscopic trocar site); Mallory-Weiss tearUrinary incontinence; inversion of bladder through urethraRanges from asymptomatic elevations of serum creatine phosphokinase to rupture of rectusabdominis muscles; rib fractures; diaphragmatic rupture; sternal wound dehiscenceCough syncope/headache/dizziness; cerebrospinal fluid rhinorrhea; acute cervical radiculopathy;cervical epidural hematoma associated with oral anticoagulation; malfunctioningventriculoatrial shunts; seizures; stroke caused by vertebral artery dissection; cerebral airembolismSpontaneous compressive orbital emphysema of rhinogenic origin; others are listed undercardiovascular categoryFear of serious disease; lifestyle changes; self-consciousnessDecreasedExacerbation of asthma; herniations of the lung (eg, intercostals and supraclavicular); hydrothoraxin peritoneal dialysis; laryngeal trauma (eg, edema and hoarseness); pulmonaryinterstitial emphysema, with potential risk of pneumatosis intestinalis; pneumomediastinum;pneumoperitoneum; pneumoretroperitoneum; pneumothorax; subcutaneousemphysema; tracheobronchial trauma (eg, bronchitis and bronchial rupture)Petechiae and purpura; disruption of surgical woundsis being readied. It has been shown that patientswith ventricular fibrillation, asystole, or heart blockcan maintain consciousness in catheterizationlaboratories or coronary care units with forceful,abrupt coughing at 1- to 3-s intervals for 39 to 92 s.Coughing produces hemodynamic changes thatcompare favorably with chest compressions. Duringthe expiratory phase of vigorous coughing,systolic pressures approach 140 mm Hg, comparedwith 75 mm Hg during chest compressions. Coughcardiopulmonary resuscitation cannot be continuedindefinitely because the act of coughingexpends more than a modest amount of energy.Differential Diagnosis and Pathogenesisof CoughCough can be caused by a multiplicity of disordersin a variety of locations. Because virtuallyany condition that stimulates cough receptors orafferent nervous pathways is capable of producingcough, only the most common disorders will bediscussed in detail. They will be categorizedaccording to the duration of cough into acute, subacute,and chronic types. Acute cough is one thatlasts 3 weeks and is most commonly transientand of minor consequence (eg, the common cold);however, it can occasionally be potentially lifethreatening (eg, pulmonary embolism, congestiveheart failure, or pneumonia). Subacute cough lastsfrom 3 to 8 weeks. It is most commonly caused bypostinfectious cough that encompasses whoopingcough and exacerbations of underlying conditionssuch as asthma and COPD. Chronic cough lasts 8weeks and is persistently troublesome. Becausethere are patients with respiratory infections (eg,pertussis) more severe than the common cold whocomplain of cough for 3 weeks and have it spontaneouslydisappear by 8 weeks, it is reasonable towithhold a diagnostic workup for cough that followsan obvious respiratory infection for 8 weeks.However, when cough lasts 3 weeks and doesnot follow an obvious respiratory infection, theworkup for cough should not wait 8 weeks. Finally,because acute cough may become chronic, thecategories are not mutually exclusive.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 417ACC-PULM-09-0302-027.indd 4177/10/09 10:22:23 PM


With respect to the pathogenicity of cough,there are limited data that do not appear at thistime to support one common mechanism. Forexample, although excessive mucus productionmay lead to cough by mechanically stimulating theafferent limb of the cough reflex and extraluminalmasses by compressing/distorting submucosalairway receptors, an increased sensitivity of theafferent limb of the cough reflex appears to be animportant pathogenetic mechanism common topatients who have a nonproductive cough causedby a variety of diseases (eg, asthma, upper respiratoryinfection, gastroesophageal reflux disease[GERD], and angiotensin-converting enzymeinhibitors [ACEIs]).Acute CoughThe limited published studies on the spectrumand frequency of acute cough suggest acute upperrespiratory tract infections primarily caused by thecommon cold 9,10 are the most common causes ofthe acute, transient cough (Table 3). Less commoncauses include potential life-threatening conditionsthat must always be considered. These conditionsinclude pneumonia, aspiration syndromes, congestiveheart failure, and pulmonary embolism.Subacute CoughThe most common causes of subacute cough inimmunocompetent adults include postinfectiouscough 11 that includes Bordetella pertussis infection,bacterial sinusitis, and exacerbations of underlyingconditions such as asthma, COPD, and bronchiectasis.When patients complain only of chroniccough after a respiratory tract infection and havenormal chest radiographic findings, some authors 12have referred to these coughs as postinfectious incausation. An important part of the definition isTable 3. Spectrum of the Most Common Causes of Acute Coughin Immunocompetent AdultsCommon coldAcute bronchitisB pertussis infectionExacerbations of underlying conditions (eg, COPD, asthma,bronchiectasis)Allergic rhinitisthat the cough eventually resolves, seemingly onit own; however, on occasion, the resolution of thecough may be hastened by a brief course of an oralcorticosteroids, inhaled corticosteroids, or ipratropium.The diagnosis is one of exclusion.In adults, during outbreaks of B pertussis infection,the frequency of postinfectious cough increasesto 25 to 50% in selected series. Although up to 28%of the cases reported to the Centers for DiseaseControl and Prevention on an annual basis occurin adults, it is clear that this is an underestimationbecause the disease is underappreciated. Whoopingcough should be considered in all subjects whopresent with a cough-vomit syndrome even if thetypical whoop (ie, the stridorous noise heard duringinspiration following a prolonged fit of coughing)may be absent. B pertussis-specific serum acute IgAantibody by enzyme-linked immunosorbent assayis a sensitive test (albeit not perfectly sensitive) forthe diagnosis and can distinguish between aresponse to natural infection and that from previousimmunization. Treatment with a macrolide (ortrimethoprim/sulfamethoxazole when macrolidecannot be administered) for the sick individual andprophylaxis for exposed persons have been foundto be effective in decreasing the severity and transmissionof the disease to others if therapy is begunearly, within the first 8 days of the infection. It isreasonable to consider adding corticosteroids.Chronic CoughThe most common causes of chronic cough inimmunocompetent adults include the following:upper airway cough syndrome (UACS), previouslyreferred to as postnasal drip syndrome (PNDS), whichis caused by variety of rhinosinus conditions; (2)asthma; (3) GERD; (4) chronic bronchitis fromcigarette smoking or other inhaled environmentalirritants; (5) nonasthmatic eosinophilic bronchitis;and (6) bronchiectasis. These conditions, singlyor in combination, have accounted for up to 94%of the causes of chronic cough in prospectivestudies. 2,3,13–15 Other conditions have constitutedno 6% of the causes in prospective studies. 2,3,13-15These have included bronchogenic carcinoma,chronic interstitial pneumonia, sarcoidosis, leftventricular failure, ACEI-induced cough, and aspirationfrom a condition associated with pharyngealdysfunction.418 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4187/10/09 10:22:24 PM


UACSBecause it is unclear whether the mechanismsof cough are postnasal drip itself or the direct irritationor inflammation of the cough receptorslocated in the upper airway, the 2006 <strong>ACCP</strong> CoughGuideline Committee recommended that the termUACS be used in preference to PNDS when referringto cough associated with upper airway conditions.The diagnosis of UACS is determined byprescribing medications that eliminate the dischargeor inflammation and the complaint, such ascough. Although the pathogenesis of chronic coughattributed to UACS is not known for certain, availabledata 16 suggest that it arises from stimuli irritatingthe afferent limb of the cough reflex located inthe hypopharynx and/or larynx. It is not knownwhy only some individuals in the population whohave a postnasal drip sensation and/or clear theirthroats have chronic cough, whereas most otherswith these same complaints do not. Any conditionwith the potential for irritating the upper respiratorytract may cause the UACS. Although GERDcan irritate the upper respiratory tract and mimicUACS, cough and throat irritation in this settingwill only resolve with treatment for GERD, notwith therapy directed at causes of UACS such asthose shown in Table 4.The treatment options are somewhat dependenton the subcategory of disease. For instance,in nonhistamine-mediated rhinitides such as thecommon cold and perennial nonallergic rhinitis,an older-generation antihistamine/decongestantcombination medication has been shown to beefficacious in double-blind, randomized, doubleblind,prospective studies, and in four prospective,descriptive studies. 2,3,19,13,15In contradistinction, newer-generation, relativelynonsedating antihistamines, such as terfenadine inTable 4. Spectrum of Conditions Causing Chronic Cough Dueto UACSRhinitisAllergic rhinitisDrug-induced rhinitisEnvironmental irritant rhinitisPerennial nonallergic rhinitisPostinfectious rhinitisVasomotor rhinitisSinusitistwo studies 17,18 and loratadine plus pseudoephedrinein one study, 19 were found to be ineffective in treatingthe nonhistamine-mediated acute cough associatedwith the common cold in randomized, placebocontrolledtrials. The older antihistamines are probablyeffective because of their anticholinergicproperties. Other treatment options include intranasalcorticosteroids and ipratropium; however, dataon their use are limited.On the basis of numerous controlled clinicaltrials in allergic rhinitis, there is good reason tobelieve that avoidance of allergens, intranasal corticosteroidsand cromolyn, and all antihistamineswill be efficacious for the treatment of the coughcaused by histamine-mediated UACS. Allergenimmunotherapy may be of long-term value but isnot for immediate help.Acute bacterial sinusitis in adults is most commonlycaused by Streptococcus pneumoniae andHaemophilus influenzae. Other organisms includeanaerobes, Streptococcal species, Moraxella catarrhalis(especially in children), and Staphylococcusaureus. Therapy includes antibiotics, intranasalcorticosteroids to decrease inflammation, anddecongestants such as oxymetazoline hydrochloride.It should be noted, however, that no prospective,randomized, double-blind studies haveproven either nasal or oral decongestants to beefficacious in patients with documented acute orchronic sinusitis.The treatment of chronic sinusitis is even lessclear cut. However, although the role of bacterialinfection and the importance of antibiotic therapyare controversial, four descriptive studies 2,3,13,15 haveshown the following treatment regimen to be effectivein eliminating the cough caused by UACSsecondary to sinusitis: a minimum of 3 weeks of anantibiotic effective against H influenzae, mouthanaerobes, and S pneumoniae; a minimum of3 weeks of oral treatment with an older-generationantihistamine/decongestant twice daily and 5 daysof intranasal decongestants twice daily. Whencough disappears with this therapy, intranasalcorticosteroids have been administered for3 months.AsthmaCough is a symptom that occurs in all asthmatics;sometimes, however, persisting cough can be<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 419ACC-PULM-09-0302-027.indd 4197/10/09 10:22:24 PM


the most troublesome symptom. In prospective,descriptive studies of patients with chronic coughattributable to asthma, cough has been the onlysymptom from 6.5 to 57% of the time. This is calledcough-variant asthma. A diagnosis of asthma shouldbe considered in the differential diagnosis of allpatients with chronic cough because it is a commoncause. Usually, patients with isolated cough do nothave variable airflow obstruction at the time ofpresentation. The diagnosis of cough-variantasthma is suggested by the presence of airwayhyperresponsiveness in a patient with chroniccough and only confirmed when cough goes awaywith asthma medications. The treatment of coughvariantasthma should be the same as that ofasthma presenting with other symptoms. The mostbenefit is likely to be obtained with corticosteroids,either orally initially followed by inhaled corticosteroidsif the symptoms are very severe, or withinhaled corticosteroids together with inhaled 2-agonists to relieve acute symptoms. The inhaledmedications should be delivered from a dry powderdevice or a pressurized metered-dose inhalertogether with a spacer. Inhalation from a pressurizedmetered-dose inhaler can exacerbate coughacutely in some patients with asthma. The maximumsymptomatic benefit of the inhaled corticosteroidsis often not seen for 6 to 8 weeks. 20A therapeutic trial with systemic corticosteroidshas sometimes been the only method usedby some investigators 21 to attempt to establish adiagnosis of cough-variant asthma. This methodfor diagnosis is not recommended because othertypes of chronic cough (eg, chronic nonasthmaticeosinophilic bronchitis 22 ) that are not associatedwith the physiologic abnormalities of asthma (eg,airway hyperresponsiveness or variable airflowobstruction) can respond to the same antiinflammatorymedication.Nonasthmatic Eosinophilic BronchitisThere are patients who present with chronic coughas an isolated symptom, normal spirometry results,normal airway responsiveness to inhaled methacholine,and eosinophilic airway inflammation (ie, 3% of nonsquamous epithelial cells in inducedsputumsamples are eosinophils). This conditionis called nonasthmatic eosinophilic bronchitis. 22 Thedifferences in functional association appear to berelated to the localization of mast cells within theairway wall, with airway smooth-muscle infiltrationoccurring in patients with asthma and onlyepithelial infiltration in patients with nonasthmaticeosinophilic bronchitis. A few cases havebeen associated with occupational exposures toacrylates. Although this condition is considered tobe distinct from asthma, it appears to be sensitiveto inhaled or oral corticosteroids. The dose andduration of treatment differ between patients. Thecondition can be transient, episodic, or persistentunless treated, and occasionally, patients mayrequire long-term prednisone treatment.GERDGERD, along with asthma and UACS, is oneof the three most common causes of chronic coughin adults. Although GERD can cause cough byaspiration, it most likely causes chronic cough inpatients with normal chest radiographic findingsby a vagally mediated distal esophagealtracheobronchialreflex mechanism. 23 When GERDis the cause of chronic cough, there may be no GIsymptoms up to 75% of the time. A cough refluxself-perpetuating cycle is likely to exist wherebycough from any cause may precipitate furtherreflux. Therefore, in the management of patientswith chronic cough, consideration should be givento breaking this cycle by searching for and treatingother simultaneously contributing conditions.Although far from perfect, the most sensitiveand specific test for GERD is 24-h esophageal pHmonitoring. In interpreting this test, it is essentialnot only to evaluate the duration and frequency ofthe reflux episodes, but also to determine the temporalrelationship between reflux and cough episodes.Patients with normal standard refluxparameters may still have reflux diagnosed as acause of cough if the temporal relationship exists.Although 24-h esophageal pH monitoring is themost sensitive and specific test in linking GERDand cough in a cause-effect relationship, it has itslimitations. For example, there is no general agreementon how to best interpret the test, and it cannotdetect nonacid reflux events that have been shownto be important in the pathogenesis of the disease.Therefore, in patients with chronic coughcaused by GERD, the 2006 <strong>ACCP</strong> Cough GuidelineCommittee recommends that the term acid reflux420 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4207/10/09 10:22:24 PM


disease, unless it can be definitively shown to apply,should be replaced by the more general term refluxdisease, so as not to mislead the clinicians into thinkingthat all patients with cough caused by GERDshould improve with acid-suppression therapy. 23The Committee also recommends empiric therapyfor GERD when patients fit the following clinicalprofile that has been prospectively validated to be92% accurate in predicting that cough will improveor disappear with antireflux therapy:• Cough 8 weeks in duration• Not exposed to environmental irritants or apresent smoker• Not receiving an ACEI• Chest radiographic findings are normal orshow nothing more than stable, inconsequentialscarring• Symptomatic asthma has been ruled out:– Cough has not improved with asthmatherapy or– Methacholine inhalation challenge is negative• UACS caused by rhinosinus diseases has beenruled out:– First-generation histamine type 1 antagonisthas been used and cough failed to improve,and– “Silent” sinusitis has been ruled out; and• nonasthmatic eosinophilic bronchitis has beenruled out:– Properly performed sputum study resultsare negative or– Cough has not improved with inhaled/systemic corticosteroidsWhen using empiric therapy as a diagnostictest, it is important to be aware that if empirictherapy fails, it cannot be assumed that GERD hasbeen ruled out as a cause of cough because theempiric therapy may not have been intensiveenough or medical therapy may have failed. Theobjective of therapy is to decrease the frequencyand duration of reflux events and decrease theirritative nature of gastric secretions. Conservative,dietary, and lifestyle measures should be tried inall patients. These measures, in addition to acidsuppressingagents with or without prokineticagents, have resulted in the resolution of cough in70 to 100% of adults. 23 In patients who failed torespond to this therapy, antireflux surgery may besuccessful. 23Because it is not known what constitutes theminimum effective medical therapy for treatingchronic cough caused by GERD, the followingcaveats should be considered: (1) drug therapyshould not be used to the exclusion of dietary andlifestyle changes; (2) acid-suppression therapyalone may fail; (3) patients may not start toimprove until 2 to 3 months of intensive medicaltherapy has been administered, including diet,lifestyle changes, and acid-suppressing and prokineticagents; and (4) antireflux surgery may besuccessful when intensive medical therapy hasfailed. Failure of medical therapy can bedetermined by performing 24-h esophageal pH/impedance monitoring while the patient is continuingthe most intensive medical therapy. Thisrelatively new test has the ability to detect allreflux events by changes in impedance and candetermine the percentage that is acidic and nonacidicin nature.Chronic BronchitisCough is one of the main features of chronicbronchitis. It is caused by inhalation of irritantsand associated with airway inflammation, mucushypersecretion, and impaired mucociliary clearance.It is important not to overdiagnose this conditionas the cause of a cough-phlegm syndromein patients who are not smokers and who are notexposed to environmental irritants because UACS,asthma, and GERD are more common causes ofthis syndrome. 15 Although most smokers have achronic cough, they are not the group of patientswho most commonly seek medical attention complainingof cough. Treatment should mainly targeta reduction of sputum production and airwayinflammation by removing environmental irritants,particularly through smoking cessation. Cough hasbeen shown to disappear or markedly decrease in94 to 100% of patients with smoking cessation; 54%of the time, cough resolution occurred within4 weeks. 24 Ipratropium can decrease sputum production,but mucolytics are of no apparent benefit. 25Although the effectiveness of systemic corticosteroidsand antibiotics on cough has not been specificallystudied, they are likely to be helpful indecreasing cough during the most severe exacerbationsof COPD. 25<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 421ACC-PULM-09-0302-027.indd 4217/10/09 10:22:24 PM


BronchiectasisAs a cause of chronic cough, bronchiectasis hasbeen diagnosed in prospective studies 3,15 with afrequency of approximately 4%. Its diagnosis isestablished by compatible clinical history, chestradiographs, high-resolution CT scans of the thorax,and cough disappearance with specific therapy.Cough associated with flares of the diseasecan be treated successfully with a combination ofchest physiotherapy, drugs to stimulate mucociliaryclearance, and systemic antibiotics. Aerosolizedantibiotics have been shown to be effective in cysticfibrosis (CF) patients with bronchiectasis, buttheir use in non-CF patients has not been provento be of benefit.Bronchogenic CarcinomaBronchogenic carcinoma is not a common causeof chronic cough (0 to 2%). It is very unlikely inpatients who have never smoked. Chest radiography,sputum cytology, and flexible bronchoscopyare the most important initial tests to consider inevaluating for this entity. The chest radiograph isthe most important initial diagnostic test for predictingwhether a bronchogenic carcinoma is not apotential cause of chronic cough. The chest radiographhas a positive predictive value of 36 to 38%and a negative predictive value of 100%. 26 Whenthe radiograph suggests that bronchogenic carcinomaor an inflammatory pulmonary parenchymalprocess are present in the context of chronic cough,bronchoscopy has a positive predictive value of 50to 89% and negative predictive value of 100%. 26 Innonsmokers with chest radiographic findings thatare normal or show nothing more than a stable,inconsequential scar, bronchogenic carcinoma is sounlikely to be the cause of chronic cough that othercauses should initially be preferentially pursued. 26ACEI-Induced CoughCough caused by ACEIs is typically nonproductiveand is associated with an irritating, tickling,or scratchy sensation in the throat. Coughappears to be a class effect of these drugs and notdose related. Cough has been shown to be causedby ACEIs from 0 to 3% of the time. 27 Cough mayappear within a few hours to weeks or months aftertaking the first dose of the ACEI. Because no laboratorytest will predict who will have an ACEIinducedcough, the diagnosis should be consideredin any patient who has a cough while receiving anACEI. Although cough caused by ACEIs typicallycomes on after variable periods of time after initiationof therapy, patients can present with coughinitially as the result of another condition beforeinitiation of the ACEI therapy and have coughsubsequently be attributed solely to an ACEI whenthe original cough goes away. Therefore, no matterwhen cough starts in association with an ACEI,consider the diagnosis of ACEI-induced cough.Cough caused by ACEIs will disappear or substantiallyimprove within 4 weeks of discontinuing thedrug. 27 Although sulindac, indomethacin, nifedipine,picotamide, and inhaled sodium cromoglycatewill provide symptomatic relief in some patients,definitive treatment is discontinuation of thedrug. 27The incidence of cough associated with therapywith angiotensin-receptor blockers has been shownto be similar to that of control drugs such as hydrochlorothiazide.Therefore, these agents in mostinstances can be used in place of ACEIs when theACEIs have to be stopped because of cough.Habit Cough, Tic Cough, and PsychogenicCough in Adult and Pediatric PopulationsIn adults, these are very uncommon causes ofchronic cough. The methodologies used and therigor of the diagnostic and therapeutic interventionsreported in the literature are inconsistent. Theputative clinical characteristics of habit cough andpsychogenic cough, for the most part, have notbeen prospectively or systematically studied.Therefore, on the basis of expert opinion, the diagnosesof habit cough or psychogenic cough can bemade only after an extensive evaluation is performedthat includes ruling out tic disorders anduncommon causes of chronic cough and aftercough improves with behavior modification orpsychiatric therapy. 28 Short of a therapeutic trial, itis hard to distinguish habit cough from coughcaused by UACS. Although the pediatric literature28 suggests that the patients with psychogeniccough will have a barking or honking character totheir coughs, the presence or absence of these characteristicsis not diagnostically helpful in adults.422 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4227/10/09 10:22:25 PM


Chronic Interstitial <strong>Pulmonary</strong> DiseaseIn series of patients, this is an uncommoncause of chronic cough. In patients with chroniccough, before diagnosing interstitial lung diseaseas the sole cause, common etiologies such as UACS,asthma, and GERD should be considered. Thecause of cough may be caused by a diagnosis otherthan interstitial lung disease at least 50% of thetime. Therefore, the diagnosis of interstitial lungdisease as the cause of cough should be considereda diagnosis of exclusion. 29Guidelines for Evaluating Acute CoughA clinical approach is recommended for evaluatingacute cough. 30 The approach initially consistsof obtaining history, physical examination, andconsidering the need for laboratory investigations.Figure 1 is the suggested algorithm for managingadult patients with an acute cough. The mostimportant initial clinical decision is to decidewhether the patient does or does not have a potentiallife-threatening condition.Guidelines for Evaluating Subacute CoughA clinical approach is recommended for evaluatingsubacute cough. 30 The approach initiallyconsists of obtaining history, physical examination,and considering the need for laboratory investigations.Figure 2 is the suggested algorithm for managingadult patients with a subacute cough. Themost important initial clinical decision is to decidewhether the subacute cough did or did not followan obvious respiratory infection.Guidelines for Evaluating Chronic CoughA great deal is now known about how to successfullydiagnose and treat chronic cough. Forinstance, the relative frequency of the disorders,alone and in combination, that can cause cough, aswell as the sensitivity and specificity of variousfindings on medical history, physical examination,and the pertinent diagnostic tests (Table 5), areknown from a number of clinical studies. 31 On thebasis of this composite body of information, the2006 <strong>ACCP</strong> Cough Guidelines Committee 30 recommendsan initial empiric integrative approach tothe management of chronic cough. Figure 3 is thesuggested algorithm for managing adult patientswith a chronic cough. After obtaining a history andperforming a physical examination and chest radiography,the most important initial clinical decisionsare to decide whether the etiology of thechronic cough is associated with an abnormal chestradiographic finding, smoking, or an ACEI. If it isnot associated with any of these, the most commoncauses (UACS, asthma, nonasthmatic eosinophilicbronchitis, and GERD) should then be systematicallyconsidered.Acute CoughLife-threatening DxHistory,examination± investigationsNon-life-threatening DxPneumonia, severeexacerbationof asthma or COPD,PE, heart failure,other serious diseaseInfectiousExacerbation ofpreexisting conditionEnvironmental/occupationalURTI LRTI Asthma Bronchiectasis UACS COPDEvaluateand treatFigure 1. Acute cough algorithm for the management of patients 15 years of age with cough lasting 3 weeks. Dx diagnosis;LRTI lower respiratory tract infection; PE pulmonary embolism; UACS upper airway cough syndrome; URTI upperrespiratory tract infection.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 423ACC-PULM-09-0302-027.indd 4237/10/09 10:22:25 PM


Subacute CoughPostinfectiousHistory,and physicalexamNon-postinfectiousPneumonia & otherserious diseasesNew onset or exacerbation ofpre-existing conditionWorkupsame aschronic coughEvaluateand treatUACS Asthma GERD BronchitisPertussisNAEBAECBBronchitisFigure 2. Subacute cough algorithm for the management of patients 15 years of age with cough lasting 3 to 8 weeks.AECB acute exacerbation of chronic bronchitis; NAEB nonasthmatic eosinophilic bronchitis; UACS upper airway coughsyndrome.Table 5. Testing Characteristics of Diagnostic Protocol*Tests Sensitivity, % Specificity, % PPV, % NPV, %Chest radiograph 100 54–76 36–38 100Sinus radiograph 97–100 75–79 57–81 95–100MIC 100 67–71 60–82 100BaE 48–92 42–76 30–63 63–9324-h esophageal pH 100 66–100 89–100 100Bronchoscopy 100 50–92 50–89 100*Data are from Irwin et al 3 and Smyrnios et al. 15 BaE modified barium esophagography; MIC = methacholine inhalationchallenge; NPV negative predictive value; PPV positive predictive value.If the patient’s cough persists after the empiricintegrative evaluation, the following systematic,diagnostic approach that has been validated inimmunocompetent patients with chronic coughis recommended: (1) <strong>Review</strong> the patient’s historyand perform a physical examination, concentratingon the anatomy of the cough reflex (Fig 4) andspecifically the most common causes of chroniccough. The character of the cough (eg, paroxysmal,loose, and self propagating; productive ordry) and timing (eg, nocturnal, with meals) havenot been shown in a prospective study 14 to be ofdiagnostic help. (2) Order a chest radiograph innearly all patients. It is extremely useful forinitially ranking differential diagnostic possibilitiesand directing laboratory testing. (3) Donot order additional laboratory tests in presentsmokers or patients receiving an ACEI until theresponse to cessation of smoking or discontinuationof the drug for 4 weeks can be assessed. (4)Depending on the results of the initial evaluation,smoking cessation, or discontinuation of the ACEI,the following may be obtained: sinus radiographsand allergy evaluation; spirometry prebronchodilatorand postbronchodilator or methacholineinhalation challenge; modified barium esophagographyand/or 24-h esophageal pH/impedancemonitoring; sputum for microbiology and/or424 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4247/10/09 10:22:25 PM


Chronic CoughInvestigateand treatA cause ofcough issuggestedHistory,examinationchest x-raySmokingACE-1DiscontinueInadequateresponse tooptimal RXUpper airway cough syndrome (UACS)Empiric treatmentAsthmaIdeally evaluate (spirometry, bronchodilatorreversibility, bronchial provocation challenge)or empiric treatmentNonasthmatic eosinophilic bronchitis (NAEB)Ideally evaluate for sputum eosinophiliaor empiric treatmentGastroesophageal reflux disease (GERD)Empiric treatmentNo responseInadequateresponseto optimalRXFor initial treatments see box belowFurther investigations to consider:• 24h esophageal pH monitoring• Endoscopic or video fluoroscopicSwallow evaluation• Barium esophagram• Sinus imaging• HRCT• Bronchoscopy• Echo cardiogram• Environmental assessment• Consider other rare causesImportant general considerationsOptimize therapy for each diagnosisCheck complianceDue to the possibility of multiple causes,maintain all partially effective treatmentInitial treatmentsUACS-A/DAsthma-ICS, BD, LTRANAEB-ICSGERD-PPI, diet/lifestyleFigure 3. Chronic cough algorithm for the management of patients 15 years of age with cough lasting 8 weeks.A/D antihistamine/decongestant; BD bronchodilator; HRCT high-resolution chest CT; ICS inhaled corticosteroid;LTRA leukotriene receptor antagonist; PPI proton pump inhibitor.cytology; flexible bronchoscopy; chest CT scan;and noninvasive cardiac studies. (5) Determinethe cause(s) of cough by observing which specifictherapy eliminates cough as a complaint. If theevaluation suggests more than one possible cause,therapies should be initiated in the same sequencein which the abnormalities were discovered.Because cough can be simultaneously caused bymore than one condition, do not stop therapy thatappears to be partially successful; rather, sequentiallyadd to it.Summary of Expected ResultsThe following is a summary of expected resultswhen the aforementioned systematic, diagnosticapproach is used 31 :1. The cause can be determined from 88 to 100%of the time, leading to “specific therapy” withsuccess rates between 84% and 98%.2. Chronic cough is often simultaneously causedby more than one condition. A single causehas been found from 38 to 82% of the time,<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 425ACC-PULM-09-0302-027.indd 4257/10/09 10:22:25 PM


Figure 4. Schematic representation of the anatomy of thecough reflex. This representation comes from an amalgamationof clinical observations in man, and histologic and physiologicalresults in experimental animal studies.multiple causes from 18 to 62%. Multiplycaused cough has been the result of three diseasesup to 42% of the time.3. Although most smokers have a cough, theyhave not been the group of patients who mostcommonly seek medical attention complainingof cough.4. In adults of all ages and the elderly, UACS,asthma, and/or GERD are the three mostcommon causes of chronic cough.5. In prospective studies, 31 chronic cough inadults is most commonly (up to 94%) causedby six disorders: UACS, asthma, GERD,chronic bronchitis, noneosinophilic bronchitis,and bronchiectasis.6. Cough has been shown in prospective studies31 to be caused by four conditions (UACS,asthma, nonasthmatic eosinophilic bronchitis,and/or GERD) > 99% of the time in nonsmokingadults who are not receiving an ACEI andwho have a normal or nearly normal and stablechest radiographic findings.7. Cough can be the sole clinical manifestationof asthma and GERD up to 57% and 75% ofthe time, respectively, and nonspecific bronchoprovocationchallenge testing and 24-hesophageal pH monitoring have been shownto be extremely useful in making a diagnosisin these patients.8. Unless the chest radiographic findings areabnormal, and this is an uncommon occurrence(no > 7%), flexible bronchoscopy willhave a very low diagnostic utility (approximately4%).9. The principal strength of the diagnostic protocolis in ruling out suspected possibilities.The principal limitation is that a positive testresult cannot necessarily be relied on to establishthe diagnosis because it has not alwaysbeen shown to predict a favorable response tospecific therapy (Table 5).10. A carefully obtained history with detailedquestioning of the character, timing, and complicationsof cough are not likely to be helpfulin predicting the cause of chronic cough.TreatmentWe have learned a great deal about treatmentof chronic cough. It can be divided into two maincategories: (1) therapy that controls, prevents, oreliminates cough (antitussive therapy); and (2)therapy that makes cough more effective (protussivetherapy). Antitussive therapy can be either specificor nonspecific. Specific therapy is directed at theetiology or operant pathophysiologic mechanismresponsible for cough. Nonspecific therapy isdirected at the symptom; it is indicated when specifictherapy cannot be administered. There is alimited role for nonspecific therapy because of thehigh probability of one being able to determine thecause of chronic cough and prescribe specific therapythat can be successful 84 to 98% of the time,Recommendations for specific therapy can befound in the section on specific diseases. Whennonspecific antitussive therapy for chronic coughis indicated, dextromethorphan, codeine, andipratropium bromide aerosol in patients withchronic bronchitis appear to be the agents of choice;they have been well studied and have been shownto be effective in double-blind, placebo-controlledstudies. 32 To date, the clinical utility of protussivetherapy remains to be determined in future studiesthat assess short-term and long-term effects ofthese agents on the patient’s condition. The use ofaerosolized hypertonic saline solution in patientswith CF appears promising.426 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4267/10/09 10:22:26 PM


WheezeWheeze is a continuous musical sound that lasts 80 to 100 ms. 33 The pathophysiologic mechanism(s)that generate wheezing are still not known.Although movement of airway secretions may playa role, flutter of airway walls narrowed almost tothe point of closure is the most likely mechanism. 34Wheeze can be high or low pitched, consist ofsingle or multiple notes, occur during inspirationor expiration, and originate from airways of anysize, from the large extrathoracic upper airwaydown to the intrathoracic small airways. Stridorrefers to inspiratory wheezing loudest over thecentral airways.Although the pitch of a wheeze is not determinedby the site of its origin, 34 its timbre maytheoretically provide a clue to its location. A polyphonicwheeze, consisting of multiple musicalnotes, is typically produced by dynamic compressionof the large, more central airways. 34 Althoughmonophonic wheezes, consisting of single musicalnotes, typically reflect disease in small airwaysand suggest asthma, especially when they aremultiple, they have also been produced by diseasesinvolving the extrathoracic large airways. 34Although recognizing that the polyphonic ormonophonic nature of expiratory wheezes maybe helpful in suggesting the location of thewheeze, there are no studies that have evaluatedtheir positive and negative predictive values indoing so.Diagnostic Value of History and PhysicalExaminationExpiratory wheezing appreciated either byhistory or physical examination is lacking in sensitivityand specificity in diagnosing asthma. 35–38Symptomatic asthma can present without wheeze,and wheezing associated with other conditions canmimic asthma. In a prospective study 35 of patientsreferred to a pulmonary clinic because of wheeze,a history of wheeze was predictive of asthma 35%of the time, and the physical finding of expiratorymonophonic wheeze was predictive 43% of thetime.Inspiratory wheezing on physical examinationis not specific for extrathoracic upper airway conditions,and it is not a sensitive sign of extrathoracicupper airway obstruction. Although inspiratorywheezing frequently accompanies expiratorywheezing during acute asthma, wheezing duringasthma also may be heard only during inspiration.36 Patients will have dyspnea on exertion whenthe site of upper airway obstruction is 8 mm indiameter, and stridor when the diameter is 5 mm. 39,40Because a wheeze can be caused by obstructionof any airway and because there is no characteristicof the wheeze of asthma that reliably distinguishesit from other conditions, it follows thatwheezes should never be considered a priori causedby asthma. This finding has been confirmed by astudy 35 that prospectively evaluated the acousticspectrum and frequency of the causes of wheeze.This study also revealed that UACS, not asthma,was the most common cause of wheeze of unknownetiology in patients referred to a pulmonary outpatientclinic (Fig 5). Data from other studies 41,42suggest that the expiratory wheeze associated withthe UACS originated from the extrathoracic airwayand most likely from the vocal cords.An Approach to the Diagnosis of WheezeIn evaluating patients with wheeze, it is importantto be aware that “all that wheezes is notasthma; all that wheezes is obstruction.” Anapproach to evaluating wheeze is to localize thesite of the obstruction to either large or small intrathoracicairways or the extrathoracic airway usinghistory, physical examination, lung function studies,and knowledge of the spectrum of differentialdiagnostic possibilities, especially those that havebeen shown to be the most common (Fig 5). Oncediagnostic possibilities have been narrowed byhistory and physical examination, pulmonaryfunction testing can be quite helpful in confirminga diagnosis.<strong>Pulmonary</strong> Function Testing in the Evaluationof Wheeze<strong>Pulmonary</strong> function testing can be helpful inlocalizing the site of obstruction causing thewheeze. Because the airway can be divided intothree anatomic areas with different physiologiccharacteristics, obstructions in these three areas<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 427ACC-PULM-09-0302-027.indd 4277/10/09 10:22:26 PM


Figure 5. Spectrum and frequency of causes of wheeze in patients referred to a pulmonary clinic. PNDS is now referred to asUACS.can be physiologically differentiated. The threeareas include the following: (1) the extrathoracicupper airway that includes the nose, mouth, pharynx,larynx, and extrathoracic trachea; (2) theintrathoracic upper airways, including the intrathoracictrachea and bronchi down to the level ofthe 2-mm airways; and (3) the small airways thatare 2 mm in diameter (ie, airways that arebeyond the eighth and ninth generations from thetrachea). 43 From a physiologic standpoint, thedistinguishing characteristics of these three anatomicareas are as follows: (1) the extrathoracicand intrathoracic large upper airways undergodifferent and opposite transmural pressurechanges during the respiratory cycle (Fig 6); (2)airflow is predominantly turbulent through largeairways and mostly laminar through small airways;44 and (3) obstruction in the large airwaysimpedes all the air in a uniform manner whenthere is one point of narrowing, whereas obstructionin small airways is almost always in multiple,scattered sites impeding airflow in a nonuniformmanner.Because spirometry and flow-volume loopsduring helium and air breathing are influenced bythese phenomena, they can be routinely utilized tolocalize airway obstruction. Spirometry repeatedafter bronchodilator or systemic corticosteroidadministration may demonstrate the presence of asubstantial component of reversible airways diseaseconsistent with asthma. In patients with normalor nearly normal baseline spirometry results,nonspecific bronchoprovocation challenge testingmay show clinically significant bronchial hyperresponsivenessconsistent with asthma.Upper airway obstructing lesions are bestidentified by flow-volume loops (Fig 7). Whenthere is only one site of obstruction, as usuallyoccurs in the trachea, airflow is likely to remainconstant during the middle portion of a maximumrespiratory effort. If the obstruction is variable(ie, it allows the airway to respond to thenormal transmural pressure change), it will bepossible to distinguish an extrathoracic locationfrom an intrathoracic location. A variable extrathoracicobstruction will typically only be seen428 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4287/10/09 10:22:26 PM


Figure 6. The effect of transmural pressure changes during inspiration and expiration on the severity of obstruction at differentairway sites. Top: extrathoracic upper airway obstruction. During expiration, intrapleural pressure (Ppl) and subsequentlyintratracheal pressure (Ptr) are greater than atmospheric pressure (Patm). Therefore, the site of the obstruction widens. Duringinspiration, because Ppl and Ptr are less than Patm, the site narrows. Bottom: intrathoracic upper airway obstruction. Duringexpiration, because Ppl is greater than Ptr, the site narrows. During inspiration, because Ptr is greater than Ppl, the site widens.during maximal inspiratory effort (Fig 7).Because the extrathoracic airway will dilate duringexpiration, the maximal expiratory flowvolume curve and spirometry will typically benormal. A variable intrathoracic obstruction willonly be observed during a maximal expiratoryeffort (Fig 7), and spirometry will reveal valuesconsistent with expiratory airflow obstruction.Spirometry alone cannot distinguish large fromsmall intrathoracic airway diseases, becauseFEV 1, peak expiratory flow rate, and FEV 1/FVCpercentage will be reduced in both situations.Because inspiratory flow rates are normallygreater than expiratory flow rates when comparedat lung volumes in the lower two thirds ofthe vital capacity (VC), the ratio of the two maydistinguish variable extrathoracic from intrathoracicupper airway lesions. 45 When a variableextrathoracic lesion is present, inspiratory areless than expiratory flow rates (eg, forced inspiratoryflow at 50% of VC /forced expiratory flowat 50% of VC is 1).<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 429ACC-PULM-09-0302-027.indd 4297/10/09 10:22:27 PM


Figure 7. Schematic flow-volume loop configurationsin a spectrum of airway lesions. A, normal;B, variable extrathoracic upper airway obstruction;C, variable intrathoracic upper airway lesion; D,fixed upper airway obstruction; and E, small airwaysobstruction.If the upper airway obstruction is fixed (ie, itdoes not allow the airway to respond to normaltransmural pressure changes), it cannot be localizedphysiologically to an extrathoracic or intrathoraciclocation. Nevertheless, the flow-volume loop willhave a characteristic shape (Fig 7) that will suggestthat the obstruction is in large rather than smallairways (Fig 7). Airflow will be uniformly impededduring inspiration and expiration because there willbe no difference between the amount of narrowingin inspiration and expiration. Because expiratoryas well as inspiratory flow rates are decreased in afixed upper airway obstruction, spirometry willreflect the expiratory flow limitation.By comparing maximal expiratory flowvolumecurves while patients first breathe roomair and then a 20% oxygen to 80% helium mixture,it may also be possible to detect coexisting largeand small airway intrathoracic obstructions. 44Expiratory flow will substantially increase with thehelium mixture when an obstruction occurs inlarge airways because the turbulent flow thatoccurs in this region increases when patientsrespire the less dense helium mixture. 46Differential Diagnosis of WheezeA large number of conditions located in a varietyof anatomic airway locations (Table 6) canproduce airway obstruction and present with expiratoryor inspiratory wheezing. Asthma should beconsidered likely when patients present with episodicwheezing and other symptoms that respondfavorably to conventional asthma medications (eg,inhaled bronchodilators). Although asthmaticswith wheeze typically will present with reversibleexpiratory airflow obstruction demonstrated byspirometry, a restrictive ventilatory pattern alsomay be seen occasionally. A reversible restrictivelung disease has been described that is clinicallyindistinguishable from asthma. 47,48 The diagnosisof wheezing conditions other than asthma 49 shouldbe considered when the initial evaluation suggeststheir presence or when wheeze does not respondto conventional asthma medications. Table 7 provideshelpful hints 50,54 to aid in the differentialdiagnosis of the causes of wheezing.TreatmentThe treatment of wheezing depends on itscause (Tables 6, 7). Because the differential diagnosisof wheezing includes diseases that are potentiallylife threatening, it is important that thephysician arrives at a specific diagnosis and institutesspecific therapy. The failure of a therapy tocorrect wheezing should prompt the physicianeither to adjust the therapy or to search for anadditional cause. The number of patients who havemore than one simultaneous cause for wheezingis unknown.SummaryWheezes are not synonymous with asthma;they may occur with a variety of diseases. Theprotocol for the evaluation of wheeze places greatemphasis on findings during history and physicalexamination. Unless the patients appear to bein imminent danger of respiratory failure, the430 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4307/10/09 10:22:28 PM


Table 6. Differential Diagnosis of WheezeUpper airway obstructionExtrathoracic causesUACSVocal cord dysfunction syndromeHypertrophied tonsilsEpiglottitisLaryngeal edemaLaryngostenosisPostextubation granulomaRetropharyngeal abscessNeoplasmsAnaphylaxisObesityKlebsiella rhinoscleromaMobile supraglottic soft tissueRelapsing polychondritisLaryngoceleAbnormal arytenoid movementVocal cord hematomaBilateral vocal cord paralysisCricoarytenoid arthritisWegener granulomatosisIntrathoracic causesTracheal stenosis due to intubationForeign body aspirationBenign tracheal/bronchial tumorsMalignanciesIntrathoracic goiterTracheobronchomegalyAcquired tracheomalaciaHerpetic tracheobronchitisRight-sided aortic archLower airway obstructionAsthmaCOPD<strong>Pulmonary</strong> edemaAspiration<strong>Pulmonary</strong> embolismBronchiolitisCFCarcinoid syndromeBronchiectasisLymphangitic carcinomatosisParasitic infectionsphysician should, in stepwise fashion, addressthe following: (1) whatever is found in the historyand physical examination, (2) the common causesof wheezing, and (3) less common causes evaluatedin a physiologically oriented manner whencommon causes do not explain the patient’ssymptoms.HemoptysisHemoptysis is the spitting of blood derivedfrom the lungs or bronchial tubes. Hemoptysis maybe scant, producing the appearance of streaks ofbright red blood in the sputum; or profuse, withexpectoration of a large volume of blood. Massivehemoptysis is defined as the expectoration of 600mL of blood within 24 to 48 h and may occur in 3to 10% of all patients with hemoptysis. 55 Gross orfrank hemoptysis produces a quantity smaller thanmassive hemoptysis and greater than blood streaking.Dark red clots may also be expectorated whenblood has been present in the lungs for days.Pseudohemoptysis is the expectoration ofblood from a source other than the lower respiratorytract. It may cause diagnostic confusion whenpatients cannot clearly describe the source of theirbleeding. Pseudohemoptysis 56 may occur whenblood from the oral cavity, nares, pharynx, ortongue drains to the back of the throat and initiatesthe cough reflex; when blood is aspirated into thelower respiratory tract in patients who havehematemesis; and when the oropharynx is colonizedwith the red, pigment-producing, aerobic,Gram-negative rod, Serratia marcescens. This colonizationmay occur in hospitalized patients whohave received broad-spectrum antimicrobial agentsand mechanical ventilation. Other rare causes ofpseudohemoptysis are self-inflicted injuries orother bizarre tactics in the malingering patientseeking hospitalization, and rifampin overdose(red man syndrome).EtiologyHemoptysis can be caused by a wide varietyof disorders (Table 8). 56 The etiologies are consideredhere in three general categories: nonmassive,massive, and idiopathic.Nonmassive Hemoptysis: Although bronchitis,bronchiectasis, lung carcinoma, and tuberculosishave always been among the most common causesof hemoptysis, their incidence varies depending onthe study population. The more recent studies 55,57have shown that bronchiectasis and tuberculosishave become less common and bronchitis morecommon. In 148 urban, adult, mostly indigentpatients studied in Kansas City between 1977 and1985, the causes of hemoptysis were bronchitis in<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 431ACC-PULM-09-0302-027.indd 4317/10/09 10:22:29 PM


Table 7. Diagnosis of Selected Wheezing Illnesses Other Than AsthmaIllnessesUpper airway diseasesUACSEpiglottitisVocal cord dysfunctionsyndromeRetropharyngeal abscessLaryngotracheal injurycaused by trachealcannulationNeoplasmsAnaphylaxisSmall airway diseasesCOPD<strong>Pulmonary</strong> edemaAspiration<strong>Pulmonary</strong> embolismBronchiolitisCFCarcinoid syndromeBronchiectasisLymphangitis carcinomatosisParasitic infectionsDistinguishing FeaturesHistory of postnasal drip, throat clearing, nasal discharge; physical examination showsoropharyngeal secretions or cobblestone appearance to the mucosaHistory of sore throat out of proportion to pharyngitis; evidence of supraglottitis onendoscopy or lateral neck radiographsLack of symptomatic response to bronchodilators; presence of stridor plus expiratorywheeze in absence of increased alveolar-arterial pressure gradient; extrathoracicvariable obstruction on flow-volume loops; paradoxical inspiratory and/or earlyexpiratory adduction of vocal cords on laryngoscopy during wheezing; thissyndrome can masquerade as asthma, be provoked by exercise, and often coexistswith asthmaHistory of stiff neck, sore throat, fever, trauma to posterior pharynx; swelling noted bylateral neck or CT radiographsHistory of cannulation of trachea by endotracheal or tracheostomy tube; evidence ofintrathoracic or extrathoracic, variable obstruction on flow-volume loops, neck/chestradiographs, laryngoscopy or bronchoscopy, or all of these. Flow-volume loops withhelium and air may reveal large airway obstruction superimposed on underlyingsmall airway disease (eg, COPD).Carcinoid tumor is suspected when there is hemoptysis, unilateral wheeze, or evidence oflobar collapse on chest radiography or combinations of these; bronchogenic carcinomain the same setting in a cigarette smoker; diagnosis is confirmed by bronchoscopyAbrupt onset of wheezing with urticaria, angioedema, nausea, diarrhea, and hypotension,especially after insect bites or administration of drug or IV contrast, or familyhistoryHistory of dyspnea on exertion and productive cough in cigarette smoker; because productivecough is nonspecific, it should only be ascribed to COPD when other coughphlegm syndromes have been excluded 47 ; forced expiratory time 4 s, and there isdecreased breath sound intensity; unforced wheezing during auscultation andirreversible, expiratory airflow obstruction on spirometry 47,48History and physical examinations consistent with passive congestion of the lungs,ARDS, impaired lung lymphatics; abnormal chest radiograph, echocardiogram, radionuclideventriculography, cardiac catheterization, or combinations of theseHistory of risk for pharyngeal dysfunction or GERD; abnormal modified barium swallowand/or 24-h esophageal pH monitoringHistory of risk for thromboembolic disease; positive confirmatory test resultsHistory of respiratory infection, connective tissue disease, transplantation, ulcerativecolitis, development of chronic airflow obstruction over months to few years ratherthan over many years in a nonsmoker; mixed obstructive and restrictive pattern onpulmonary function tests, and hyperinflation and fine nodular infiltrates on chestradiographyCombination of productive cough, digital clubbing, bronchiectasis with Pseudomonascolonization/infection obstructive azoospermia; family history; pancreatic insufficiencyand two sweat chloride determinations 60 mEq/L; some diagnoses are notmade until adulthood; in one instance as late as age 69 yr 51 ; when sweat test result isoccasionally normal, definitive diagnosis may require nasal transepithelial voltagemeasurements and genotyping 52History of episodes of flushing and watery diarrhea; elevated 5-hydroxyindoleaceticacidlevel in a 24-h urine specimenHistory of episodes of productive cough, fever, or recurrent pneumonias; suggestivechest radiographs or typical chest CT findings; allergic pulmonary aspergillosis shouldbe considered when bronchiectasis is centralHistory of dyspnea or previous malignancy; reticulonodular infiltrates with or withoutpleural effusions; suggestive high-resolution chest CT scan; bronchoscopy withbiopsiesConsider in a nonasthmatic who has traveled to an endemic area and complains of fatigue,weight loss, fever; peripheral blood eosinophilia; infiltrates on chest radiograph;stools for ova and parasites for nonfilarial causes; blood serologic studies for filarialcauses432 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4327/10/09 10:22:29 PM


Table 8. Causes of Hemoptysis According to Category of Disease*Tracheobronchial disorders LeukemiaAcute tracheobronchitis ThrombocytopeniaAmyloidosisCardiovascular disordersAspiration of gastric contents Aortic aneurysmBronchial adenoma Bronchial artery ruptureBronchial endometriosis Congenital heart diseaseBronchial telangiectasia Congestive heart failureBronchiectasis Coronary artery bypass graftBronchogenic carcinoma Fat embolizationBroncholithiasis Hughes-Stovin syndromeChronic bronchitis Mitral stenosisCF Postmyocardial infarction syndromeEndobronchial hamartoma <strong>Pulmonary</strong> arteriovenous fistulaEndobronchial metastases <strong>Pulmonary</strong> artery aneurysmEndobronchial tuberculosis <strong>Pulmonary</strong> embolismForeign body aspiration <strong>Pulmonary</strong> venous varixMucoid impaction of bronchus SchistosomiasisTracheobronchial trauma Subclavian artery aneurysmTracheoesophageal fistula Superior vena cava syndromeTracheoartery fistula Tumor embolizationLocalized parenchymal diseasesDiffuse parenchymal diseasesAcute and chronic nontuberculous pneumonia Capillaritis (with or without systemic vasculitis)Actinomycosis Disseminated angiosarcomaAmebiasis Farmer lungAscariasis Goodpasture syndromeBronchopulmonary sequestration Idiopathic pulmonary hemosiderosisCoccidioidomycosis IgA nephropathyCongenital and acquired cyst Inhaled isocyanatesCryptococcosis Legionnaires diseaseExogenous lipoid pneumonia Mixed connective tissue diseaseHistoplasmosis Mixed cryoglobulinemiaHydatid mole Polyarteritis nodosaLung abscess SclerodermaLung contusion Systemic lupus erythematosusMetastatic cancer Systemic vasculitisMucormycosis TMA toxicityNocardiosis Viral pneumonitisParagonimiasis Wegener granulomatosis<strong>Pulmonary</strong> endometriosisMiscellaneous disorders<strong>Pulmonary</strong> tuberculosis IdiopathicSporotrichosis IatrogenicHematologic disorders BronchoscopyAnticoagulant therapy Cardiac catheterizationDisseminated intravascular coagulation Needle lung biopsy* This listing is not meant to be all-inclusive.37%, bronchogenic carcinoma in 19%, tuberculosisin 7%, pneumonia in 5%, and bronchiectasis in1%. 57 In 59 patients with hemoptysis 200 mL/dstudied at Duke University between the years 1972and 1981, the causes included bronchitis or bronchiectasisin 28%, cancer in 36%, CF in 7%, anticoagulationin 7%, and tuberculosis in 5%. 55 In aretrospective study 58 of hemoptysis in patients withHIV infection, the cause was identified for 78% ofepisodes. Of these, it was thought to be the result ofpneumonia in 70%, Kaposi sarcoma in 12%, bronchiectasisin 5%, pulmonary embolism in 5%, bronchitisin 3%, endocarditis in 3%, and lung cancer in3%. Of the 28 presumed pneumonias, disease was<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 433ACC-PULM-09-0302-027.indd 4337/10/09 10:22:29 PM


caused by bacterial etiology in 16 cases, Mycobacteriumtuberculosis in 3 cases, Pneumocystis jiroveciin 3 cases, Mycobacterium avium-intracellulare in3 cases, and fungal causes in 3 cases. Hemoptysisin this series when described was blood streakingin 77%, frank in 19%, and massive in 4% (asthe result of Aspergillus sp and Pseudomonas spinfections). Although bleeding from tracheoarteryfistula complicating tracheostomy, rupture of pulmonaryartery from a balloon flotation catheter,and diffuse intrapulmonary hemorrhage may besubmassive, they are discussed in the section onmassive hemoptysis.Massive Hemoptysis: The more frequent causesof massive hemoptysis are listed in Table 9. Virtuallyall causes of hemoptysis may result in massivehemoptysis, but it is most frequently causedby infection and cancer. Several series 56 reportthat 60% of cases are caused by tuberculosis,bronchiectasis, and lung abscess. Infection is alsothe cause of bleeding from aspergilloma and CF.In one series 57 of 22 patients, 27% of cases werecaused by bronchitis, 18% were caused by tuberculosis,9% were caused by aspergilloma, 4% wereTable 9. Common Causes of Massive HemoptysisCardiovascularArteriobronchial fistulaCongestive heart failure<strong>Pulmonary</strong> arteriovenous fistulaDiffuse intrapulmonary hemorrhageDiffuse parenchymal diseaseIatrogenicMalposition of chest tube<strong>Pulmonary</strong> artery ruptureTracheoartery fistulaInfectionsAspergillomaBronchiectasisBronchitisCFLung abscessSporotrichosisTuberculosisMalignanciesBronchogenic carcinomaLeukemiaMetastatic cancerTraumacaused by bronchiectasis, and 4% were caused bysporotrichosis. In a series 55 from Duke University,17 of 42 patients (40%) with massive hemoptysishad cancer and only 8 patients (19%) had bronchitis,bronchiectasis, or tuberculosis as a cause.Idiopathic hemoptysis is less frequent in patientswith massive hemoptysis and usually constitutes 5% of cases. 57Rupture of a pulmonary artery complicatesballoon flotation catheterizations in 0.2% ofcases. 59 It is fortunate that this is rare because itcarries a mortality approximating 40%. 59 Tracheoarteryfistula is also an unusual but devastatingcondition, complicating approximately 0.7% oftracheostomies. 60 These two complications of proceduresand diffuse intrapulmonary hemorrhage,usually attributable to an immunologically mediateddisease, are most likely to be considered in thedifferential diagnosis of massive rather than submassivehemoptysis in ICU patients.Idiopathic Hemoptysis: When one uses the systematicdiagnostic approach outlined in Tables 10and 11, the cause of hemoptysis can be found inmost instances. In 2 to 32% of patients (average,12%), 61,62 the cause cannot be determined. Thiscondition, called idiopathic or essential hemoptysis,is seen most commonly in men between the agesof 30 and 50 years. Prolonged follow-up studiesalmost always fail to reveal the source of bleeding,although 10% continue to have occasional episodesof hemoptysis. 63 Consider Dieulafoy diseaseof bronchus (superficial bronchial artery)in the context of idiopathic hemoptysis; it canbe involved in a subset of patients yet to bedetermined. 64Table 10. Routine Evaluations for HemoptysisHistoryPhysical examinationCBCUrinalysisCoagulation studiesECGChest radiography Bronchoscopy**Although bronchoscopy is not indicated in some conditions(eg, pulmonary embolism, aortopulmonary fistula, congestiveheart failure), it should routinely be considered.434 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4347/10/09 10:22:29 PM


Table 11. Examples of Special Evaluations for HemoptysisAccording to Category of Disease*Tracheobronchial disordersExpectorated sputa for tuberculosis, parasites, fungi, andCytologyBronchoscopy (if not done)BronchographyHigh-resolution chest CT scanLocalized parenchymal diseasesExpectorated sputa for tuberculosis, parasites, fungi, andCytologyChest CT scanLung biopsy with special stainsDiffuse parenchymal diseasesExpectorated sputa for cytologyBlood for BUN, creatinine, anti-nuclear antibody,Rheumatoid factor, complement, cryoglobulins,Antineutrophil cytoplasmic antibodyAntiglomerular basement membrane antibodyLung and/or kidney biopsy with special stainsCardiovascular disordersEchocardiogramArterial blood gas on 21% and 100% oxygenVentilation/perfusion scans<strong>Pulmonary</strong> arteriogramAortogram, contrast CT scansHematologic disordersCoagulation studiesBone marrow*This list is not meant to be all-inclusive.PathogenesisThe bronchial arteries are the chief source ofblood for the airways (from mainstem bronchi toterminal bronchioles), for the supporting frameworkof the lung that includes the pleura, intrapulmonarylymphoid tissue, and the large branchesof the pulmonary vessels, and for the nerves inthe hilar regions. The pulmonary arteries supplythe pulmonary parenchymal tissue, including therespiratory bronchioles. Communications betweenthese two blood supplies, bronchopulmonary arterialand venous anastomoses, occur in the proximityof the junction of the terminal and respiratorybronchioles. These anastomoses allow the twoblood supplies to complement each other. 65 Forinstance, if flow through one system is increasedor decreased, a reciprocal change occurs in theamount of blood supplied by the other system.Arteriographic studies in patients with activehemoptysis have shown that the systemic circulationis primarily responsible for the bleeding inapproximately 92% of cases.The pathogenesis of hemoptysis depends onthe type and location of the disease. 56 In general,if the lesion is endobronchial, the bleeding is fromthe bronchial circulation; if the lesion is parenchymal,the bleeding is from the pulmonary circulation.Moreover, in chronic diseases, repetitiveepisodes are most likely the result of increasedvascularity in the involved area.In bronchogenic carcinoma, hemoptysis resultsfrom necrosis of the tumor with its increased bloodsupply from bronchial arteries or from local invasionof a large blood vessel. In bronchial adenomas,bleeding is usually from rupture of the prominentsurface vessels. In bronchiectasis, hemoptysis isusually the result of irritation by infection of thegranulation tissue that has replaced the normalbronchial wall. In acute bronchitis, bleeding resultsfrom irritation of the unusually friable vascularmucosa.The mechanism of hemoptysis in mitral stenosisis controversial, but the most likely explanationis rupture of the dilated varices of the bronchialveins in the submucosa of large bronchi caused bypulmonary venous hypertension. <strong>Pulmonary</strong>venous hypertension may also be responsible forthe bleeding in congestive heart failure because itis associated with widening of the capillary anastomosesbetween the bronchial and pulmonaryarteries.Hemoptysis in pulmonary embolism may becaused by infarction, with necrosis of parenchymaltissue or hemorrhagic consolidation as the resultof an excessive influx of bronchial arterial blood atsystemic pressures through the bronchopulmonaryanastomoses into the pulmonary circulation distalto the obstructing clot.In tuberculosis, bleeding can occur for a varietyof reasons. In the acute parenchymal exudativelesion, scant hemoptysis may result from necrosisof a small branch of a pulmonary artery or vein. Inthe chronic, parenchymal fibroulcerative lesion,massive hemoptysis may result from rupture of apulmonary artery aneurysm bulging into thelumen of a cavity. The aneurysm occurs from tuberculousinvolvement of the adventitia and media ofthe vessel. When a healed and calcified tuberculous<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 435ACC-PULM-09-0302-027.indd 4357/10/09 10:22:30 PM


lymph node erodes the wall of a bronchus becauseof pressure necrosis, the patient may cough upblood as well as the calcified node (broncholith).In endobronchial tuberculosis, hemoptysis mayresult from acute tuberculous ulceration of thebronchial mucosa. In healed and fibrotic parenchymalareas of tuberculosis, bleeding may arise fromirritation of granulation tissue in the walls of bronchiectaticairways in the same areas.In traumatic rupture of the pulmonary arteryby a balloon flotation catheter, risk factors includepulmonary hypertension, distal location of thecatheter tip, excessive catheter manipulation inan attempt to obtain a pulmonary artery occludedpressure measurement, a large catheter loop inthe right ventricle, and advanced age. 59 In tracheoarteryfistula complicating tracheostomy, bleedingis caused by trauma from the tracheostomy cannulaand/or balloon. 60 Bleeding is usually causedby rupture of the innominate artery. The fistulacan form at three tracheal locations: the stoma,the intratracheal cannula tip, and the balloon.Trauma at the stoma is caused by pressure necrosis,usually because the tracheostomy was createdtoo low (below the fourth tracheal ring). It mayalso occur at the cannula tip because of excessiveangulation of the cannula and at the balloon sitecaused by pressure necrosis because of excessiveinflation pressures. Diffuse intrapulmonary hemorrhageattributable to immunologic diseases iscaused by an inflammatory lesion, usually ofcapillaries. 66DiagnosisGeneral Considerations: The success rate indetermining the cause of hemoptysis is excellentbut variable. If one accepts the diagnosis of idiopathic(essential) hemoptysis as a distinct entity,the cause of hemoptysis can be determined inclose to 100% of cases. If one does not accept thisdiagnosis, the cause can be determined in 68 to98% of cases (average, approximately 88%).The diagnostic workup of hemoptysis involvesroutine (Table 10) as well as special evaluations(Table 11). Routine evaluations are initially performedin every patient, whereas special studiesare ordered only when the clinical setting suggeststhey are indicated. In general, each category ofdisease (Table 8) has its special studies.Routine Evaluation: As in any diagnostic problem,a detailed history and physical examinationmust be performed. These should be performedin systematic fashion, not only to rule in the commoncauses of hemoptysis, but also to rule in thecategory of the cause (Table 8).Although the amount of bleeding is generallynot indicative of the seriousness of the underlyingdisease process, a history of the frequency, timing,and duration of hemoptysis may be helpful. Forexample, repeated episodes of hemoptysis occurringover months to years suggest bronchial adenomaand bronchiectasis, whereas small amountsof hemoptysis occurring every day for weeks aremore likely to be caused by bronchogenic carcinoma.Bleeding attributable to bronchogenic carcinomausually is of short duration because it isgenerally a late finding in these patients. In additionto having abnormal chest radiographic findings,these patients are usually 40 years of ageand are almost invariably cigarette smokers.Hemoptysis that coincides with the menses(catamenial) suggests the rare diagnostic possibilityof pulmonary endometriosis, whereas bleedingassociated with sexual intercourse or other formsof exertion suggests passive congestion of thelungs. Although hemoptysis may be a symptom atany age, it is distinctly uncommon in the young.When hemoptysis is present before the thirddecade of life, it suggests an acute tracheobronchitis,a congenital cardiac or lung defect, an unusualtumor, CF, a blood dyscrasia, or infectious pneumonia.No matter what the age, if a patient withpneumonia who is undergoing appropriate therapyhas hemoptysis that persists for longer thanthe usual 24 h, an endobronchial lesion or coagulopathyshould be suspected. 67A travel history can often be helpful in bringingcertain endemic diseases to mind. This is true ofcoccidioidomycosis and histoplasmosis in theUnited States, of paragonimiasis and ascariasis inthe Far East, and of schistosomiasis in SouthAmerica, Africa, and the Far East. Chronic sputumproduction before hemoptysis suggest the diagnosesof chronic bronchitis, bronchiectasis, and CF.The presence of orthopnea and paroxysmal nocturnaldyspnea suggest the diagnoses of passivecongestion of the lungs from mitral stenosis andleft ventricular failure. A history of anticoagulanttherapy suggests an intrapulmonary bleed from436 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4367/10/09 10:22:30 PM


too large a dose or recurrent pulmonary embolismfrom too small a dose. The possibility of pulmonaryembolism should always be considered when apatient who presents with hemoptysis has been atincreased risk for deep venous thrombosis. Thepossibility of traumatic rupture of a pulmonaryartery caused by to balloon flotation catheterizationshould always be considered when thesecatheters are used.Although tracheoartery fistula must be consideredin the differential diagnosis of hemoptysis inevery patient with a tracheostomy, 60 it is fortunatelyan infrequent cause in this setting. When itoccurs, the onset is almost always at least 48 h afterthe procedure. While the peak incidence is betweenthe first and second week, and 72% of fistulas willbleed during the first 21 days after tracheostomy,hemorrhage from this complication can also occuras late as 18 months after the procedure. There isa sentinel bleed in 34 to 50% of cases. Before 48 h,bleeding from the stoma is usually the result ofcapillary bleeding from inadequate hemostasis.Whenever hemoptysis occurs in a patient who hasan endotracheal tube or tracheostomy in place,trauma from suctioning, especially when coagulationis abnormal, should be considered a likelycause.Although patients with diffuse intrapulmonaryhemorrhage typically have hemoptysis, they mayoccasionally not expectorate at all but just complainof dyspnea 68 or the abrupt onset of fever, dyspnea,cough, and malaise. Therefore, lack of hemoptysisdoes not rule out a substantial intrapulmonaryhemorrhage. 68 The diagnosis of trimellitic anhydride(TMA)-induced pulmonary hemorrhageshould be suspected in workers exposed to highdoseTMA fumes. Exposure occurs when heatedmetal surfaces are sprayed with corrosion-resistantepoxy resin coatings. The syndrome requires alatent period of exposure and appears to be antibodymediated. 69 Respiratory failure with pulmonaryinfiltrates and hemoptysis have also beenreported in a patient with a documented exposureand antibodies to isocyanates. 70In a patient with the triad of known upper airwaydisease, lower airway disease, and renal disease,systemic Wegener granulomatosis should besuspected. Although the diagnosis of systemiclupus erythematosus is readily considered in apatient known to have this disease, pulmonaryhemorrhage can be the presenting manifestation.Goodpasture syndrome (antibasement membraneantibody-mediated disease) typically occurs inyoung men, and it has been reported to be associatedwith influenza infection, inhalation of hydrocarbons,and penicillamine ingestion. Therefore, itshould be considered in these historical contexts. 56Diffuse alveolar hemorrhage should be suspectedin patients who have undergone recentbone marrow transplantation when they presentwith cough, dyspnea, hypoxemia, and diffusepulmonary infiltrates. This finding typically occurswith marrow recovery. Diffuse alveolar hemorrhagehas been reported to occur in approximately20% of patients during autologous bone marrowtransplantation, and it has been associated with an80% mortality. 71 The pathogenesis is thought to berelated to damage caused by chemotherapy orradiation therapy compounded by return ofinflammatory cells with marrow recovery.Physical examination may be helpful in severalways. Inspection of the skin and mucous membranesmay show telangiectasias, suggestinghereditary hemorrhagic telangiectasia, or ecchymosesand petechiae, suggesting a hematologicabnormality. Pulsations transmitted to a tracheostomycannula should heighten suspicion of a tracheoarteryfistula or risk of one. Inspection of thethorax may show evidence of recent or old chesttrauma, and unilateral wheeze or crackles mayherald localized disease such as bronchial adenomaor carcinoma. Although pulmonary embolismis not definitively diagnosed on physicalexamination, tachypnea, phlebitis, and pleuralfriction rub suggest this disorder. If crackles areheard diffusely on chest examination, passivecongestion as well as other diseases causing diffuseintrapulmonary hemorrhage should be considered.Careful cardiovascular examination may rulein mitral stenosis, pulmonary artery stenosis, orpulmonary hypertension.The routine laboratory studies listed in Table10 are useful for the following reasons. The CBCcount may suggest presence of an infection, hematologicdisorder, or chronic blood loss. Idiopathichemosiderosis or other causes of diffuse intrapulmonaryhemorrhage may present only with diffusepulmonary infiltrates and iron deficiency anemiafrom chronic bleeding into the lungs. Urinalysismay reveal hematuria and suggest the presence of<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 437ACC-PULM-09-0302-027.indd 4377/10/09 10:22:30 PM


a systemic disease associated with diffuse parenchymaldisease. While there is simultaneous evidenceof clinical involvement of the lungs andkidneys in 33% of cases of Goodpasture syndrome,there can be clinical lung involvement withoutrenal disease in 33% and clinical renal involvementwithout lung disease in 33%. 56Coagulation studies may uncover a hematologicdisorder that is primarily responsible for thehemoptysis or that contributes to excessive bleedingfrom another disease. The ECG may help suggestthe presence of a cardiovascular disorder.Although as many as 30% of patients with hemoptysishave normal chest radiographic findings, 72routine posteroanterior and lateral radiographsmay be diagnostically valuable.When pulmonary tumor or infection is not readilyapparent, other radiographic signs that maysuggest the cause and source of bleeding includeradio-opaque foreign bodies that may give rise tohemoptysis even years after entry into the lungs;the disappearance of a calcified mediastinal lymphnode after it has eroded the bronchial wall and isexpectorated as a broncholith; aortic or pulmonaryaneurysms that by dissection may erode into thebronchial tree; single or multiple pulmonary cavitiesthat suggest pulmonary tuberculosis, fungal disease,parasitic disease, acute or chronic lung abscess,primary or metastatic neoplasm, septic pulmonaryemboli, or Wegener granulomatosis; an intracavitarymass lesion that is indicative of a fungus ball(aspergilloma) or blood clot; and localized honeycombingthat is consistent with bronchiectasis. Thepresence of a new infiltrate localized to the areasubtending a balloon flotation catheter suggests arupture of the pulmonary artery. The appearanceof a new air-fluid level in a preexisting cavity or cystsuggests the location of the source of bleeding, asdoes a nonsegmental alveolar pattern that clearswithin a few days. A solitary pulmonary nodulewith vessels going toward it suggests an arteriovenousmalformation. In patients with hemoptysiscaused by pulmonary embolism, a parenchymaldensity abutting a pleural surface with evidence ofpleural reaction and/or effusion is usually present.The cardiac silhouette, vascular or parenchymalpatterns, and presence of Kerley B lines may beuseful in documenting cardiovascular disease.When the chest radiograph shows diffuse pulmonaryinfiltrates, hemorrhage from bleedingdisorders (eg, thrombocytopenia in the compromisedhost), lung contusion from blunt chesttrauma, and passive congestion of the lungs shouldbe considered, in addition to the diseases listedunder diffuse alveolar hemorrhage in Table 8. Inthe earliest stages of diffuse intrapulmonary hemorrhage,chest radiographs may appear normal,but usually hemorrhage appears in a diffusealveolar pattern. This progresses to a mixedalveolar-interstitial pattern and, then, whenbleeding ceases entirely, an interstitial pattern, ashemosiderin deposition accumulates.Even when findings of the history, physicalexamination, and chest radiograph are normal orthere is an obvious cause of hemoptysis on thechest radiograph, bronchoscopy is invaluable fornot only accurate diagnosis but for precise localizationof the pulmonary hemorrhage. It is not uncommonfor bronchoscopy to establish sites of bleedingdifferent from those suggested by the chest radiograph.73,74 Bronchoscopy may not be needed inpatients with stable chronic bronchitis with oneepisode of blood streaking, particularly if it isassociated with an exacerbation of acute tracheobronchitisin which the site of bleeding was recentlydocumented by bronchoscopic examination; withacute lower respiratory tract infections; and withobvious cardiovascular causes of hemoptysis, suchas congestive heart failure and pulmonary embolism.In localizing the bleeding site, the best resultsare obtained when bronchoscopy is performedduring or within 24 h of active bleeding. The bleedingsite can be localized in up to 93% of patients 74,75with a flexible bronchoscope, and in up to 86% withthe rigid instrument. 76 When the procedure is donewithin 48 h, localization of bleeding can drop to51%. 77 When bronchoscopy is done after bleedinghas ceased, accurate localization is likely to bereduced even further. Although the flexible bronchoscopeis usually the instrument of choice indiagnosing lower respiratory tract problems, rigidbronchoscopy is preferred in cases of massiveuncontrolled hemorrhage because patency of theairway is maintained more effectively during thisprocedure. There are data that show that obtaininga high-resolution chest CT before bronchoscopymay enhance the yield of bronchoscopy. With theexception of tracheoartery fistula, the tracheobronchialdisorders that can be diagnosed by a bronchoscopicexamination are listed in Table 8.438 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4387/10/09 10:22:30 PM


Bedside bronchoscopy should not be performedto rule in the diagnosis of tracheoarteryfistula. 60 In tracheostomized patients with hemoptysis,bronchoscopy should be performed to ruleout other causes, such as bleeding from suctionulcers, tracheitis, or lower respiratory tract disorders.If no other cause for hemoptysis can be foundand bleeding has stopped or anterior and downwardpressure on the cannula on the stomal site oroverinflation of the tracheostomy balloon slowsdown or stops the bleeding, a surgical consultationshould be sought immediately and the patientbrought to the operating room for examination ina more controlled environment. In this setting,balloon deflation and removal of the tracheostomytube are not advised unless airway protection withan endotracheal tube can be performed nearlysimultaneously, followed by a definitive vascularrepair should the diagnosis be correct. As long astracheoartery fistula remains a diagnostic possibility,the tracheostomy balloon should not be deflatedand the tracheostomy tube should not be removedwithout protecting the airway below the tracheostomytube.When there is no active bleeding, bronchoscopywith BAL fluid can be helpful in suggestingdiffuse intrapulmonary hemorrhage. Return ofbright red or blood-tinged lavage fluid from multiplelobes from both lungs suggests an active,diffuse intrapulmonary hemorrhage; hemosiderinladenmacrophages (ie, siderophages) on cytologicanalysis from these same specimens suggest bleedingthat has been ongoing. Because normal subjectsmay have siderophages in their alveoli, the diagnosisof diffuse alveolar hemorrhage requires asubstantial number of siderophages to be recoveredby BAL fluid ( 20% of total alveolar macrophages).78 Because carbon monoxide diffusingcapacity is increased as the result of intra-alveolarRBCs binding carbon monoxide for 24 to 48 h afterbleeding stops, this test may be helpful in diagnosingintra-alveolar hemorrhage in the stable patientwithout hemoptysis.Although the diagnosis of unquestionablebronchiectasis can be made only with the aid of thebronchogram or high-resolution chest CT scan, apresumptive diagnosis can be made without bronchography.Bronchiectasis is visible on routine chestradiographs in up to 90% of cases, 79 and bronchoscopycan localize the bleeding to the correspondingabnormal areas. Angiography may determine thesite of bleeding in up to 90 to 93% of cases. 75 Whenperformed routinely, diagnostic angiography establishesa diagnosis not identified by bronchoscopyin only 4% of patients. 75 99m Tc-labeled colloid andRBC studies have been shown to be accurate andmay be positive in 6 to 10 min. 79 Although the useof angiography may not be initially helpful in confirmingrupture of the pulmonary artery caused byballoon flotation catheterzation if the rent hassealed, it can be extremely helpful in detecting apseudoaneurysm that has formed in the healingprocess. 59 Identification of an unstable lesion isimportant because it should be obliterated to preventfuture rupture and death. 59 Angiographyhas not been useful in diagnosing tracheoarteryfistula. 60Special Evaluations: Depending on the resultsof the initial evaluation and the possible categoriesof cause of hemoptysis (Table 8), additionaldiagnostic evaluations should be systematicallyperformed (Table 11). The diagnosis of Goodpasturesyndrome is made by demonstrating lineardeposition of IgG along the basement membraneof the lung or kidney and the presence of hightiters of circulating anti-basement membrane antibodyin the blood. Antibodies from patients withtraditional Goodpasture syndrome react with the3(IV) chain of type IV collagen. A patient withpulmonary hemorrhage from adenocarcinomawas initially misdiagnosed and initially incorrectlytreated because he had increased levels of serumanti-glomerular basement membrane antibodies. 80Once the cancer was subsequently diagnosed, itwas determined that the patient did not haveGoodpasture syndrome because antibodies didnot show anti-3(IV) reactivity. They only reactedto the 1(IV) chain. The choice of lung or kidneybiopsy depends on the clinical setting. However,anti-IgG immunofluorescent staining of the kidneymay be positive even in the absence of clinicalevidence of renal disease. 81 Although Goodpasturesyndrome is typically associated with IgG, thereare also reports of a pulmonary-renal hemorrhagicsyndrome associated with IgA. 82 The importanceof this observation is that the immunoserologictesting must be designed to include both Igs.Definitive diagnosis of the pulmonary vasculitidesdepends on histologic examination, includingspecial stains and cultures that rule out<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 439ACC-PULM-09-0302-027.indd 4397/10/09 10:22:31 PM


tuberculosis and fungal diseases. <strong>Pulmonary</strong> capillaritiswith hemorrhage has been reported in anever-increasing number of conditions. The diagnosiscan sometimes be made on transbronchialbiopsy, thus avoiding the need for open-lungbiopsy, but care must be taken to exclude infectiousetiologies by using special stains. While high levelsof IgG, IgA, and IgM antibody to trimellitic-coupledprotein and trimellitic-conjugated erythrocyteshave been found in patients with TMA-inducedpulmonary disease, the diagnosis can be madeclinically by obtaining a history of the exposureand ruling out other diseases (Table 8).It is important to be aware that diseases maybe considered (and therefore evaluated) in morethan one category. For instance, the case of a patientwith hemoptysis caused by overzealous anticoagulationmay be evaluated in three categories: (1)a hematologic disorder that may cause (2) localizedand (3) diffuse parenchymal disease. A patient withchronic bleeding from the tracheobronchial disorderof diffuse bronchial telangiectasis could presentwith diffuse as well as localized parenchymal disease(aspiration hemosiderosis). A patient withlong-standing passive congestion of the lungs or acardiovascular disorder might present with diffusepulmonary hemosiderosis, whereas a patient withacute pulmonary edema usually presents with diffusepulmonary infiltrates.Differential DiagnosisIn evaluating patients with hemoptysis, it isnecessary to rule out the causes of pseudohemoptysis.Unless the cause of pseudohemoptysis isdefinitively determined, the spitting up of bloodmust be assumed to be true hemoptysis. An upperairway lesion must not be assumed to be the causeof the bleeding unless it is seen bleeding activelyat the time of examination.TreatmentThe treatment of hemoptysis can be dividedinto supportive and definitive categories. In prescribingdefinitive therapy, it is important to considerthe cause, the amount of bleeding, and thepatient’s underlying lung function.Supportive Care: Supportive care usuallyincludes bed rest and mild sedation. Drugs withantitussive effects (all narcotics) should not beused. An effective cough may be necessary to clearblood from the airways and to avoid asphyxiation.Drugs with antiplatelet effects also should not beused. Depending on the results of oximetry and/orarterial blood gas analysis, supplemental oxygenshould be administered. If bleeding continuesand gas exchange becomes further compromised,endotracheal intubation and mechanical ventilationmay become necessary. To facilitate flexiblebronchoscopy with a sufficiently large suctionport, an endotracheal tube with an internal diameterof at least 8 mm should be used, if possible.Other respiratory adjunctive therapy, such as chestphysiotherapy and postural drainage, should beavoided. Fluid and blood resuscitation shouldbe administered when indicated. The amount ofhemoptysis should be continuously quantitateduntil it stops. The amount helps determine thepatient’s subsequent care.Definitive Care—Nonmassive Hemoptysis: Indefinitive care for patients with scant or frank (submassive)hemoptysis, treatment is directed at thespecific cause. For instance, suppurative bronchiectasisis treated with antibiotics plus a mucociliaryescalator drug (eg, theophylline, -adrenergicagonists). An acute exacerbation of chronic bronchitisassociated with cigarette smoking is treatedwith a mucociliary escalator and cessation ofcigarette smoking. Although the role of antibioticsin the context of hemoptysis caused by COPDhas not been specifically studied, antibiotics havebeen shown to be helpful in patients with themost severe exacerbations of COPD. CF is treatedwith broad-spectrum antibiotics plus a mucociliaryescalator drug. Bronchial adenoma and bronchogeniccarcinoma should be resected wheneverpossible. Congestive heart failure is treated withcombinations of drugs for preload and afterloadreductions and inotropic agents when appropriate,mitral stenosis with diuretics, and pulmonaryembolism with heparin. There are no data tosupport that patients with hemoptysis caused bypulmonary embolism bleed more with heparin.Therefore, do not initially withhold or undertreatthese patients with nonmassive hemoptysis. Theeffects of overzealous anticoagulation are treatedwith cessation of blood thinning and perhapsfresh frozen plasma and vitamin K. Tuberculosisis treated with antituberculous drugs; appropriate440 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4407/10/09 10:22:31 PM


antibiotic therapy is administered for acute infectiouspneumonias.Definitive Care—Massive Hemoptysis: In patientswith massive hemoptysis, treatment is directed notonly at the specific cause but also at abrupt cessationof bleeding. Death from massive hemoptysisis predominantly caused by asphyxiation, and thelikelihood of death appears directly related to therate of bleeding. Urgent management in all casesof massive hemoptysis must emphasize protectingthe uninvolved lung from aspiration of blood andtamponading of the bleeding site. When tracheoarteryfistula may be present, the following stepsshould be considered: 60 if bleeding is immediateand profuse, there may be time only to overinflatethe balloon, tamponading the potential bleedingsite at the balloon, and apply downward andforward pressure on the top of the tracheostomytube, tamponading the potential bleeding site atthe stoma. If the arterial rupture is at the cannulatip, these efforts will not be helpful. If bleedingstops or slows either by these efforts or spontaneously,an endotracheal tube should be placed distalto the tip of the tracheostomy tube and an immediatesurgical consultation requested. Ideally, anexperienced surgeon should be present whenthe tracheostomy tube is removed; should crispbleeding start again, the surgeon can attempt tofinger tamponade/compress the bleeding artery(eg, usually the innominate) by bluntly dissectingdown the anterior tracheal wall and behind thesternum to the vessel. The vessel, once reached,can be compressed against the back of the sternum.When the situation has been stabilized, clotscan be gently suctioned from the distal tracheaand the patient taken to the operating room fordefinitive repair.When bleeding originates from below the primarycarina, the bleeding lung should be keptdependent to minimize aspiration of expectoratedblood. Numerous techniques have been advocatedto help minimize aspiration and have proved helpful.A bronchoscopically positioned endobronchialballoon may provide effective tamponade. Hemoptysiscaused by bleeding from all but the rightupper lobe has been managed with balloon occlusion.Placement of a Carlen tube that intubates eachmainstem bronchus separately is helpful, but thetube can be difficult to place; once in position, itssmall diameter may prevent subsequent diagnosticflexible bronchoscopy. In cases of persistent massivehemoptysis, diagnostic considerations mayneed to be delayed because placement of a Carlentube may be necessary to ensure patient survival.Urgent treatment to stop massive hemoptysismay involve laser bronchoscopy, iced saline solutionlavage, angiographic embolization, supportivetreatment only, or surgical resection. Use of a laserto stop hemoptysis has been successful in 90%of patients with cancer, 83 but recurrence of bleedingwithin a few weeks is typical. No large studies ofpatients with massive hemoptysis have beenreported. Because the laser is useful only in patientswith proximal airway lesions and is difficult to useduring massive hemoptysis, laser therapy willprobably not evolve into a common therapeutictool for these patients. Bronchoscopically directediced saline solution lavage of the bronchi leadingto the site of hemorrhage has been reported to besuccessful in stopping hemorrhage in an uncontrolledseries. 84Angiography can identify the bleeding site in 90% of cases 75,85 and, when combined with anembolization procedure, has been successful instopping bleeding in massive hemoptysis in 90%of cases. 85 Several angiographic sessions may berequired, and both systemic and pulmonary vesselsmay need to be studied. Fourteen percent ofpatients bleed again within 1 to 4 days, and multipleprocedures are frequently necessary. 85,86 Onceactive bleeding ceases, 20% of patients bleed againduring the next 6 months 87 and 22% by 3 to 5years. 85 Angiographic embolization may beachieved with the use of gel foam or polyurethaneparticles and is aided by temporary balloon occlusionof the involved vessel. 85,88,89 Sclerosing agentshave led to subsequent massive lung necrosis andshould be avoided. 85Although early studies included several casescomplicated by embolization of spinal arteries withsubsequent paresis or paralysis, this complicationappears rare when the procedure is performed byexperienced angiographers. 85 Other complications,such as pleurisy or hematoma formation, are infrequentand usually are minor. 85 In patients withhemoptysis caused by trauma, urgent thoracotomyhas been advocated, with the recommendation thatit be performed with the patient in the supine positionto minimize aspiration and that the bronchovasculartrunk of the involved lung be clamped<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 441ACC-PULM-09-0302-027.indd 4417/10/09 10:22:31 PM


while the patient is stabilized to minimize thechance of air embolism while receiving positivepressure ventilation.Survival from iatrogenic rupture of the pulmonaryartery has been reported. 59 Several urgentmaneuvers may prove helpful, and balloon tamponadeand selective intubation should always beattempted. Balloon tamponade of the ruptured vesselwith the Swan-Ganz balloon has been helpful.With the balloon deflated, the catheter should bewithdrawn 5 cm, the balloon inflated with 2 mL ofair, and the balloon allowed to float back into thehemorrhaging vessel to occlude it. Ideally, patientsshould immediately be intubated in the mainstembronchus opposite the involved lung to minimizeaspiration. In most patients, death from pulmonaryartery rupture occurs before the bleeding lung canbe identified. Because the catheter usually floats tothe right pulmonary artery, when it is not knownwhich pulmonary artery has been ruptured, selectiveintubation of the left mainstem bronchus orplacement of a Carlen tube should be attempted.Selective intubation of the left mainstem bronchusmight be facilitated by using a bronchoscope orsuction catheter designed specifically to enter theleft lung. All patients who stop bleeding requireangiographic evaluation to help localize the arterialtear and check for the formation of a pseudoaneurysm. 59At the time of angiography, embolization of theaffected vessel should be performed if a pseudoaneurysmor a tear is found. Hemoptysis from apseudoaneurysm usually occurs in the first dayafter formation but may occur weeks later.Aside from repairing lesions such as a tracheoarteryfistula, the role of emergency surgery forhemoptysis remains controversial. Several studies 90have advocated emergency surgery for all patientswith massive hemoptysis when feasible, citingolder statistics that patients who have had surgeryhave a 19% mortality compared with 54% in thenonoperative group. A more recent study 55 challengesthis, citing the following findings in 84patients with hemoptysis of 200 mL/d: (1) therewere no deaths in patients with 1 L/d of hemoptysis,even they did not undergo surgery; (2) therate of mortality varied more with whether thepatient was operable than with whether the patientunderwent surgery (mortality in operable patientswith hemoptysis of 1 L/d was 30% vs 65% ininoperable patients); and (3) the rate of mortalitywas greatly affected by diagnosis, in that no patientdied from hemoptysis attributable to bronchitis,tuberculosis, bronchiectasis, or anticoagulation,whereas 80% of patients with cancer and hemoptysisof 1 L/d died. Others have advocatedconservative nonsurgical treatment when hemoptysisis caused by infectious causes. In patients withCF, even with normal lung function, resectionshould be avoided because repeated episodes inother areas are likely to occur.With respect to surgery, it is clear that no treatmentpreference can be recommended for allpatients on the basis of reported studies. The trialsof therapy span different decades of practice, havewidely differing causes of hemoptysis in theirpopulations, and employ several different definitionsfor massive hemoptysis. A review of the literaturesuggests the following strategy: (1) patientswho are not candidates for surgery because of theirpulmonary function, general medical condition, ordiffuse nature of their lesions should be treatedwith selective embolization; (2) resectional surgeryshould be performed in operable patients whensurgery is the definitive treatment for the underlyingdisease; and (3) all potentially operable patientswho continue to bleed at rates of 1 L/d despitesupportive, conservative care should undergoeither surgical resection or embolization. The correcttherapy in a given patient depends on thecause of the bleeding, lung function, availability ofresources, and local expertise.In patients with diffuse intrapulmonary hemorrhage,selective arterial embolization and surgeryare not options. For immunologicallymediated diseases, corticosteroids, cytotoxicagents, and other interventions are available. Withrespect to corticosteroid and cytotoxic drug therapies,the following generally apply 56 : (1) corticosteroidsare used alone if the syndrome isself-limited; (2) cytotoxic therapy should be addedwhen the disease is known to respond to cytotoxictherapy (eg, Wegener granulomatosis or Goodpasturesyndrome); (3) cytotoxic agents can be usedas steroid-sparing agents in patients whose diseasehas responded to corticosteroids but who are havingside effects caused by corticosteroids; and (4)cytotoxic agents may be considered in immunologiclung disease that has not responded asexpected to corticosteroids and for which no intercurrentdisease process can be identified.442 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4427/10/09 10:22:31 PM


When corticosteroid therapy is administeredalone for critically ill patients with immunologiclung diseases, the dose is 1 mg/kg/d of methylprednisoloneIV or the equivalent dose of anothercorticosteroid. Larger doses, on the order of 7 to 15mg/kg/d for 1 to 3 days, have been recommendedto control progressive pulmonary hemorrhage andhypoxemia of Goodpasture syndrome, systemiclupus erythematosus, and some of the vasculitides. 56In general, corticosteroids should be administeredinitially in around-the-clock divided doses untilsubstantial improvement has occurred. They canthen be administered once per day and tapered asthe patient’s condition dictates.In general, patients with Goodpasture syndromeare treated in the following manner: (1) if patientspresent with severe pulmonary hemorrhage, methylprednisoloneis initially administered at the higherdose of 7 to 15 mg/kg/d for 1 to 3 days; (2) plasmapheresismay be useful for the immediate and longtermreduction of anti-glomerular basementmembrane antibody levels; and (3) short-term andlong-term cyclophosphamide and plasmapheresistherapy are guided by the results of serially obtainedserum anti-glomerular basement membrane antibodylevels. If pharmacologic doses of corticosteroidscontrol severe pulmonary hemorrhage and the plasmapheresisand cyclophosphamide therapy makescirculating anti-glomerular basement membraneantibody disappear, bilateral nephrectomy is unnecessary.On the basis of favorable case reports in lifethreateningsituations, consider recombinantactivated factor VII as a temporizing measure inunstable patients without coagulopathic bleeding. 91DyspneaDyspnea is a distressing sensation of difficult,labored, or unpleasant breathing. The word distressingis very important to this definition becauselabored or difficult breathing may be encounteredby healthy individuals while exercising, whichdoes not qualify as dyspnea because it may not beperceived as distressing. The sensation is oftenpoorly or vaguely described by patients.PhysiologyThe physiology of dyspnea remains uncleardespite 40 years of active investigation. What isclear, however, is that a simplistic single neuralpathway theory has been superceded by a complex,multiple neural pathway paradigm. 92 Becausethe respiratory system is extremely complex andredundant in its control mechanisms, it is quitelikely that there are multiple stimuli, receptors,nerves, and neural pathways that mediate thesensation of dyspnea. This multiple neural pathwaymodel of dyspnea suggests that dyspnea mayarise caused by abnormalities in the afferent pathways,the efferent pathways, or the central controlcenters of the respiratory system (Fig 8). Becauseafferent pathways feed back to the CNS from virtuallyall levels of the efferent pathway, afferentdyspneic information from virtually all thoracicand upper abdominal organs (including the pharynx,larynx, airways, lung parenchyma, esophagus,heart, and stomach) may potentially impactthe sensation.On the basis of the vagus nerve, muscle afferent,and chemoreceptor experiments, it appearsthat hypercapnia, hypoxemia, and muscle afferentdata that relate to load, effort, and impedance arethe major dyspnogenic stimuli. An important rolefor vagal afferents has not been ruled out. Withrespect to the central integration of the afferent andefferent systems, it appears that the intensity ofdyspnea can be modulated by learning, experience,and emotional/behavioral states.Differential Diagnosis of DyspneaAs with cough, a multiplicity of causes locatedin a variety of anatomic locations where the dyspneapathways travel have been reported to cause dyspnea.Although the individual causes are toonumerous to list, 92 they can be grouped into categoriesof diseases (Table 12). The spectra and frequenciesof causes will vary depending on whetherdyspnea is acute or chronic, on the age group studied,and on whether the patient presents as aninpatient, in the emergency department, or in anoutpatient clinic (Fig 9) or office. Nevertheless, if thedata from all studies reported in these varied settingsare combined, one can conclude that cardiopulmonarydiseases will cause 75 to 92% of cases inemergency department and hospitalized patientsand 46 to 85% of cases in outpatients. 93,94 Moreover,in all settings, dyspnea appears to be attributable toone of five major causes 94% of the time: (1) cardiac,<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 443ACC-PULM-09-0302-027.indd 4437/10/09 10:22:32 PM


Figure 8. Schematic representation of the afferent and efferent motor nerves serving the respiratory system. The cortex integratessensory afferent (left side) and efferent (right side) data with experience and produces sensations of dyspnea, effort, fatigue,and weakness. Reprinted with permission from Curley. 92Table 12. Differential Diagnosis of Dyspnea According toCategory of DiseaseCardiacDeconditioning/obesityDermatologicEndocrineGIHematologicInfectiousLarynx/upper airwayNeuromuscularNutritionalOncologicPharmacologicPregnancy<strong>Pulmonary</strong>PsychiatricRenalRheumatologicVascular(2) pulmonary, (3) psychogenic/hyperventilation,(4) GERD, and (5) deconditioning disorders.Approximately two thirds of cases, chronic dyspneawill be caused by four diseases: COPD, asthma,interstitial lung disease, and cardiomyopathy. 95Pathophysiology of DyspneaThe mechanism by which diseases producedyspnea varies with the disease. Some affectFigure 9. Representative spectrum and frequency of causesof chronic dyspnea evaluated in an outpatient pulmonaryclinic. The 64 respiratory disorders included asthma in 25cases, COPD in 12 cases, interstitial lung disease in 12 cases,UACS in 6 cases, bronchiectasis in 2 cases, tracheal stenosisin 1 case, kyphoscoliosis in 1 case, fibrothorax in 1 case, costochondritisin 1 case, lung cancer in 1 case, bronchiolitis obliteransin 1 case, and unilateral hyperlucent lung syndromein 1 case. Upper respiratory tract disorders accounted for8% of the causes of dyspnea. Decond’g deconditioning;LRT lower respiratory tract; URT upper respiratory tract.Adapted from Pratter et al. 97444 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4447/10/09 10:22:32 PM


afferent pathways, others affect efferent pathways,and others affect the central integrative sensorypathways. In most diseases, the precise mechanismby which dyspnea occurs is unknown or unstudied.Virtually all studies performed to elucidate themechanism of dyspnea have been performed onpatients with COPD or asthma.The authors of studies 92 on the genesis of dyspneain asthma reveal that mechanical, blood gas,and psychological factors probably all play a role.The intensity of dyspnea does not appear to correlatewell with the degree of obstruction. It hasmost recently been shown 96 that asthmatic patientswith a history of near-fatal attacks had a reducedresponse to hypoxia and a reduced sensation ofdyspnea.Compared with asthma, Paco 2and Pao 2appearto play a greater role in COPD, whereas themechanical effects are similar. 92 Similar to COPDand asthma, dyspnea in interstitial lung disease isprobably caused by abnormal gas exchange,impaired muscle function, or andactivation ofintrapulmonary vagal afferents. 92 Compression ofthe tracheobronchial tree and laryngeal dysfunctionprobably cause dyspnea by creating airway obstructionsimilar to COPD and perhaps by triggeringvagal afferent pathways. 92 Compression of lungtissue by pneumothorax, tumor, or pleural effusionlikely leads to dyspnea by stimulating vagal afferentsand altering gas exchange. 92 Distortion of thethoracic cage by kyphoscoliosis and similar chestwall deformities may alter gas exchange andmuscle length-tension relationships. 92Cardiac diseases probably cause dyspnea byactivating afferents, muscle efferents, and/or bycausing gas exchange abnormalities. 92 The mechanismby which GERD causes dyspnea (it hasbeen reported to do so 2 to 5% of the time 96,97,98 ) isnot known. Because patients may have normallungs, exercise capacity, muscle strength, and gasexchange, dyspnea in patients who do not havelaryngospasm and who do not aspirate may beattributable to the stimulation of vagal afferents.Hyperventilation syndrome, anxiety, panic, andother psychological diseases can cause dyspnea.Because these patients typically have normalcardiopulmonary function, normal efferent pathways,and no obvious afferent pathway defects,it is tempting to explain dyspnea as a perceptualcentral integrative problem. However, becausedyspnea can be provoked during hyperventilation,disordered afferent pathways related toincreased minute ventilation and metabolic disturbancesassociated with reducetion in Paco 2mayplay a role. 92Deconditioning presumably increases the sensationof dyspnea by requiring an abnormally highminute ventilation to perform a task. This increasedeffort may be perceived as inappropriate to thelevel of difficulty of the task performed. Therefore,the cause is likely to be multifactorial involvingafferent, efferent, and integrative pathways. 92 Forspeculative mechanisms of dyspnea associatedwith anemia, high levels of carbohydrate ingestion,and lung cancer, the reader is encouraged to readthe chapter by Curley. 92Guidelines for Evaluating Acute DyspneaA clinical approach is recommended forevaluating acute dyspnea. It consists of obtaininga history and performing a physical examinationand laboratory tests, considering potential lifethreateningconditions first (eg, acute asthma,pulmonary embolism, pulmonary edema states,pneumonias).Guidelines for Evaluating Chronic Dyspnea ofUnknown EtiologyThe following systematic, diagnostic approachhas been validated in patients with chronic dyspneawho were evaluated in a university hospital, outpatientpulmonary clinic. 97 (1) <strong>Review</strong> the patient’shistory and perform a physical examination, concentratingon the anatomy of the potential originsof the dyspnea sensation (Fig 8) and, specifically,the most common causes of dyspnea: COPD,asthma, interstitial lung disease, cardiomyopathy,GERD, other respiratory diseases, and hyperventilationsyndrome. Consider doing a provocative testfor hyperventilation syndrome. 92 (2) Order a chestradiograph in nearly all patients. (3) Depending onthe results of the initial evaluation and chest radiographs, The following tests may be ordered:• <strong>Pulmonary</strong> function testing to include spirometrybefore bronchodilator and after bronchodilatoror methacholine, lung volumes, diffusingcapacity, pulse oximetry at rest and duringexercise, maximal inspiratory and expiratory<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 445ACC-PULM-09-0302-027.indd 4457/10/09 10:22:33 PM


mouth pressures, and flow-volume loops• Noninvasive cardiac studies to include ECG,echocardiography, and stress testing• Modified barium esophagography and/or 24-hesophageal pH monitoring• Chest CT scan• Comprehensive exercise tolerance test• Other more invasive tests such as cardiac catheterizationand lung biopsy(4) Whenever possible, the final determination ofthe cause of dyspnea is made by observing whichspecific therapy eliminates dyspnea as a complaint.Because dyspnea may be simultaneously the resultof more than one condition, 99 do not discontinuetherapy that appears to be partially successful;rather, sequentially add to it.Summary of Expected Results Using theAforementioned Systematic Diagnostic ProtocolIt should be possible to determine the cause ofchronic dyspnea in nearly all cases that can lead tospecific treatment that can be successful in 76% ofcases. 97 COPD, asthma, interstitial lung disease,and cardiomyopathy are the most common causesof chronic dyspnea, accounting for 66% of cases;however, 13 other discrete disorders, includingUACS, make up the remaining 33% of cases. Avariety of respiratory disorders are the most commoncauses of chronic dyspnea, accounting forapproximately 75% of the total.When asthma is the cause of chronic dyspnea,dyspnea can be the sole presenting manifestationof this disease 20% of the time. The diagnosis ofchronic dyspnea is more accurately determined byusing objective test results combined with clinicalimpression, compared with clinical impressionalone. Diagnoses of disease based only on history,physical examination, and chest radiographs areincorrect approximately 33% of the time.The history and physical examination are mosthelpful in ruling out common diagnoses but areless useful in predicting the diagnosis. A previousdiagnosis of asthma or a history of wheezing eachhave positive predictive values for asthma of 50%. For COPD, a negative smoking history ruledout the diagnosis, but only 20% of smokers hadCOPD. A history of a previous diagnosis of COPDhas a positive predictive value of only 45%. Forinterstitial lung disease, the presence of crackles onphysical examination has a positive predictivevalue of 79%. However, the absence of crackles hada negative predictive value of 98%. For cardiovasculardisease, crackles had a positive predictivevalue of only 21% but a negative predictive valueof 92%.The predictive values of positional or nocturnaldyspnea have not been studied. Nevertheless, theyare likely to be nonspecific. Although trepopnea (ie,dyspnea that increases when one side is dependent)suggests asymmetric disease, it is more likely to beassociated with common diseases such as COPDthan a paralyzed hemidiaphragm. 92 Platypnea (ie,an increase in dyspnea in the upright posture) mayresult from pericarditis, ventilation-perfusion mismatch,ileus, or right-to-left shunt. 92 Orthopnea (ie,an increase in dyspnea on recumbency) hasoccurred in COPD, passive congestion of the lungs,and respiratory muscle weakness. 92 Nocturnaldyspnea has been associated with COPD, asthma,passive congestion of the lungs, and GERD. 92Several tests have been helpful in diagnosingthe cause of dyspnea. Most are best at ruling outdiagnoses (Table 13). Tests reported to have highnegative predictive value include methacholineinhalation challenge that is 100% for rulingout asthma; spirometry that is 100% for ruling outCOPD; diffusing capacity that is 95% for ruling outinterstitial lung disease; and chest radiograph thatis 91% for ruling out all diagnoses. 97 Few tests havea high positive predictive value, including methacholineinhalation challenge, which is 95% forruling in asthma; diffusing capacity, which is 79%for ruling in interstitial lung disease; and chestradiography, which is 75% for ruling in all diagnoses.In addition to the known value of comprehensiveexercise tolerance test to distinguish respiratorycauses from cardiovascular causes of dyspnea, itis particularly helpful in diagnosing psychogenicdyspnea or deconditioning.TreatmentTreatment for dyspnea can be either diseasespecific or disease nonspecific. Specific therapytreats the underlying disease. Nonspecific therapytreats the symptom. Whenever possible, the underlyingcause of dyspnea should be identified andspecific treatment prescribed. In a prospective446 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4467/10/09 10:22:33 PM


Table 13. Predictive Values of Diagnostic Tests in the Evaluation of Chronic Dyspnea*Test Diagnosis No. PPV, % NPV, %Spirometry COPD 84 32 100Spirometry Asthma 84 18 72Spirometry All diagnoses 84 80 56Methacholine challenge Asthma 62 95 100Diffusing capacity Interstitial lung disease 32 79 95Echocardiogram Cardiac 11 44 0Radionuclide ventriculogram Cardiac 9 66 0Barium esophagram GERD 15 33 83Comprehensive endotracheal tube All diagnoses 15 93 0Chest radiograph All diagnoses 84 75 91*These data are adapted from Pratter et al. 97 See Table 5 for definition of abbreviations.study, 97 the results of specific treatment yielded thefollowing results: (1) chronic dyspnea improved in76% of patients; (2) all patients with asthma, UACS,hyperventilation syndrome, deconditioning, andGERD improved; (3) 33% of patients with COPD,58% with interstitial lung disease, and 78% withcardiomyopathy improved; (4) patients with interstitiallung disease and cardiomyopathy all continuedto have some dyspnea despite treatment; and(5) when one of the less common diseases (eg,kyphoscoliosis, laryngeal edema) was the cause ofdyspnea, therapy was helpful in only 12% ofpatients.Specific treatment of asthma, COPD, and interstitiallung disease will be discussed by otherauthors in this volume; see the section on cough forrecommendations for treatment of GERD. Withrespect to the hyperventilation syndrome, theauthors of controlled studies 100 suggest that breathingretraining (eg, slow abdominal breathing) is veryeffective. Both obesity and deconditioning shouldrespond to diet and exercise respectively. 101Nonspecific therapy for dyspnea has includedconditioning regimens, nutrition, stress reductionand coping skills, oxygen, mechanical ventilation,respiratory muscle training, vagotomy, musclevibration, acupuncture, and pharmacotherapywith narcotics, anxiolytics, and phenothiazines. 92With respect to efficacy for dyspnea, the best casehas been made for nutritional support. A randomized,controlled clinical trial 102 in patients withCOPD showed that dyspnea decreased when proteinmalnutrition was improved. Although the useof anxiolytics may work in patients with panicdisorders, the only two controlled clinical trials 103,104of benzodiazepines showed that dyspnea was notimproved in COPD patients. Although the resultsof studies with systemic narcotics are conflicting, 92randomized controlled trials do not support use ofnebulized morphine in COPD or interstitial lungdisease 105,106 or acupuncture in nonmalignantbreathlessness. 107 Finally, although a small doubleblindrandomized controlled clinical crossoverpilot study 108 suggests that inhaled furosemide mayalleviate the sensation of dyspnea in patients withCOPD, the results must be confirmed in a larger,multicenter study before this can be routinely recommended.Annotated ReferencesCough1. Schappert SM, Burt CW. Ambulatory care visitsto physicians’ offices, hospital outpatient departments,and emergency department: United States,2001–2002. National Center for Health statistics.Vital Health Stat 2006; 13:1–66This survey summarized data from an estimated 1.1billion ambulatory visits during a 12-month period.Cough was the single most common complaint forwhich patients sought medical care.2. Irwin RS, Corrao WM, Pratter MR. Chronic persistentcough in the adult: the spectrum and frequencyof causes and successful outcome of specific therapy.Am Rev Respir Dis 1981; 123:413–417This prospective, descriptive study was the first to useand validate the anatomic, diagnostic protocol. It stud-<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 447ACC-PULM-09-0302-027.indd 4477/10/09 10:22:33 PM


ied patients with chronic cough of unknown etiology,defined as cough that was at least 3 weeks in durationand persistently troublesome.3. Irwin RS, Curley FJ, French CL. Chronic cough: thespectrum and frequency of causes, key componentsof the diagnostic evaluation and outcomes of specifictherapy. Am Rev Respir Dis 1990; 141:640–647This prospective, descriptive study revalidated the anatomicdiagnostic protocol that had been modified toinclude 24-h esophageal pH monitoring to assess for“silent” GERD as a cause of chronic cough.4. Power JT, Stewart IC, Connaughton JJ, et al. Nocturnalcough in patients with chronic bronchitisand emphysema. Am Rev Respir Dis 1984;130:999–1001This study showed that cough decreases during sleep,even in patients with a pathologic reason to cough. It isan important study to quote when others say that psychogeniccough should be considered because the patientdoes not cough at night.5. Macklem PT. Physiology of cough. Ann Otol 1974;83:761–768Excellent review of the physiology of cough.6. Gal TJ. Effects of endotracheal intubation on normalcough performance. Anesthesiology 1980;52:324–329This study clearly shows that an endotracheal tube passingthrough the vocal cords does not prevent cough frombeing an effective clearance mechanism.7. Gracey DR, Divertie MB, Howard FM Jr. Mechanicalventilation for respiratory failure in myastheniagravis: two-year experience with 22 patients.Mayo Clin Proc 1983; 58:597–602One of very few studies that provides some objectivedata on how to assess when cough effectiveness is diminished.It suggests that maximal expiratory pressure maybe useful in this regard.8. Irwin RS. Complications of cough: <strong>ACCP</strong> evidence-basedclinical practice guidelines. Chest2006; 129(Suppl):54S–58SThis publication is the most comprehensive review ofcomplications of cough.9. French CT, Fletcher KE, Irwin RS. A comparisonof gender differences in health-related quality oflife in acute and chronic coughers. Chest 2005;127:1991–1998This prospective study suggests that 90% of the timeacute cough is caused by an upper respiratory infection.10. Curley FJ, Irwin RS, Pratter MR, et al. Coughand the common cold. Am Rev Respir Dis 1988;138:305–311This prospective, randomized, double-blind, placebocontrolledstudy demonstrated the prevalence of coughin the untreated common cold and also showed that afirst-generation antihistamine-decongestant medicationwas effective in decreasing the severity and prevalenceof cough caused by the common cold. It also providedevidence to support an upper airway mechanism in thepathogenesis of cough caused by PNDS.11. Kwon N-H, Oh M-J, Min T-H, et al. Causes andclinical features of subacute cough. Chest 2006;129:1142–1147This prospective study reports the spectrum and frequencyof causes of subacute cough. Postinfectious coughis the most common condition in this category of cough.12. Braman SS. Postinfectious cough: <strong>ACCP</strong> evidencebased clinical practice guidelines. Chest 2006;129(Suppl):138S–146SThis publication is the most up-to-date and comprehensivediscussion of postinfectious cough and includes adiscussion regarding whooping cough.13. Pratter MR, Bartter T, Akers S, et al. An algorithmicapproach to chronic cough. Ann Intern Med 1993;119:977–983This prospective, descriptive study validates the anatomicdiagnostic protocol. The protocol was modifiedby routinely and empirically treating for PNDS beforeordering any tests. It introduces the concept of “silent”PNDS.14. Mello CJ, Irwin RS, Curley FJ. Predictive values ofthe character, timing, and complications of chroniccough in diagnosing its cause. Arch Intern Med1996; 156:997–1003This prospective, descriptive study again revalidates theanatomic diagnostic protocol and showed that the character,timing, and complications of cough are not helpfulin diagnosing the cause of chronic cough. It also showedthat the spectrum and frequency of productive chroniccough were similar to those of dry cough.15. Smyrnios NA, Irwin RS, Curley FJ. Chronic coughwith a history of excessive sputum production: thespectrum and frequency of causes, key componentsof the diagnostic evaluation, and outcome ofspecific therapy. Chest 1995; 108:991–997This prospective, descriptive study revalidated theanatomic diagnostic protocol in patients with productivecough (cough-phlegm syndrome) and showed thatthe spectrum and frequency of productive cough weresimilar to those of dry cough and that the most commoncauses of a cough-phlegm syndrome were PNDS,448 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4487/10/09 10:22:34 PM


asthma, and GERD and not chronic bronchitis.16. Pratter MR. Chronic upper airway cough syndromesecondary to rhinosinus diseases (previouslyreferred to as postnasal drip syndrome):<strong>ACCP</strong> evidence-based clinical practice guidelines.Chest 2006; 129(Suppl):63S–71SThis provides a comprehensive discussion of the upperairway cough syndrome and justification for changingthe name from PNDS.17. Gaffey MJ, Kaiser DL, Hayden FG. Ineffectivenessof terfenadine in natural colds: evidence againsthistamine as a mediator of common cold symptoms.Pediatr Infect Dis J 1988; 7:223–228This randomized, controlled, clinical trial showed that anewer histamine type 1 antagonist was not effective innonhistamine-mediated PNDS.18. Berkowitz RB, Tinkleman DG. Evaluation of oralterfenadine for the treatment of the common cold.Ann Allergy 1991; 67:593–597This randomized, controlled, clinical trial showed that anewer histamine type 1 antagonist was not effective ina nonhistamine-mediated PNDS.19. Berkowitz RB, Connell JT, Dietz AJ, et al. The effectivenessof the nonsedating antihistamine loratadineplus pseudoephedrine in the symptomaticmanagement of the common cold. Ann Allergy1989; 63:336–339This randomized, controlled, clinical trial showed that anewer histamine type 1 antagonist was not effective ina nonhistamine-mediated PNDS.20. O’Byrne PM, Cuddy L, Taylor DW, et al. The clinicalefficacy and cost benefit of inhaled corticosteroidsas therapy in patients with mild asthmain primary care practice. Can Respir J 1996; 3:169–175This randomized, controlled, clinical trial showed thatinhaled corticosteroids were effective in treating coughcaused by asthma and how long it took before maximalbenefit was achieved.21. Dicpinigaitis PV. Chronic cough due to asthma:<strong>ACCP</strong> evidence-based clinical practice guidelines.Chest 2006; 129(Suppl):75S–79SThis is an evidence-based clinical practice guideline formanaging cough caused by asthma that was endorsed bythe American College of Chest Physicians, the AmericanThoracic Society, and the Canadian Thoracic Society.22. Brightling CE. Chronic cough due to nonasthmaticeosinophilic bronchitis. Chest 2006;129(Suppl):116S–121SThis publication discusses the latest information regardingnonasthmatic eosinophilic bronchitis.23. Irwin RS. Chronic cough due to gastroesophagealreflux disease: <strong>ACCP</strong> evidence-based clinical practiceguidelines. Chest 2006; 129(Suppl):80S–94SA comprehensive state of the art discussion of coughcaused by GERD.24. Wynder EL, Kaufman PK, Lerrer RL. A short-termfollow-up study on ex-cigarette smokers with specialemphasis on persistent cough and weight gain.Am Rev Respir Dis 1967; 96:645–655This prospective study showed how quickly coughcaused by chronic bronchitis disappears when patientsstop smoking cigarettes.25. Braman SS. Chronic cough due to chronic bronchitis:<strong>ACCP</strong> evidence-based clinical practice guidelines.Chest 2006; 129(Suppl):104S–115SComprehensive discussion on managing cough causedby chronic bronchitis.26. Kvale PA. Chronic cough due to lung tumors:<strong>ACCP</strong> evidence-based clinical practice guidelines.Chest 2006; 129(Suppl):147S–153SComprehensive discussion on managing cough causedby lung tumors.27. Dicpinigaitis PV. Angiotensin-convertingenzyme inhibitor-induced cough: <strong>ACCP</strong> evidence-basedclinical practice guidelines. Chest2006;129(Suppl):169S–173SComprehensive discussion on cough caused by ACEIs.28. Irwin RS, Glomb WB, Chang AB. Habit cough, ticcough, and psychogenic cough in adult and pediatricpopulations: <strong>ACCP</strong> evidence-based clinical practiceguidelines. Chest 2006; 129(Suppl):174S–179SComprehensive review of psychogenic, tic, and habitcoughs in adults and children.29. Brown KK. Chronic cough due to chronic interstitialpulmonary diseases: <strong>ACCP</strong> evidencebasedclinical practice guidelines. Chest 2006;129(Suppl):180S–185SA comprehensive review of cough caused by interstitiallung diseases such as idiopathic interstitialfibrosis, sarcoidosis,and hypersensitivity pneumonitis.30. Pratter MR, Brightling CE, Boulet L-P, et al. Anempiric integrative approach to the managementof cough: <strong>ACCP</strong> evidence-based clinical practiceguidelines. Chest 2006; 129(Suppl):222S–231SAlgorithms that provide a “road map” that the cliniciancan follow are useful and are presented for acute,subacute, and chronic cough.31. Diagnosis and management of cough: <strong>ACCP</strong> evidence-basedclinical practice guidelines. Chest<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 449ACC-PULM-09-0302-027.indd 4497/10/09 10:22:34 PM


2006; 129(Suppl):1S–292SThis document is the most comprehensive and up-todateclinical practice guidelines on managing cough.32. Bolser DC. Cough suppressant and pharmacologicprotussive therapy: <strong>ACCP</strong> evidence-based clinicalpractice guidelines. Chest 2006; 129(Suppl):238S–249SA comprehensive, evidence-based review of nonspecificcough agents; the few that have been shown to be effectiveare identified.Wheeze33. Pasterkamp H, Kraman SS, Wodicka GR. Respiratorysounds: advances beyond the stethoscope.Am J Respir Crit Care Med 1997; 156:974–987This state-of-the-art review provides an update on allaspects of respiratory sound.34. Forgacs P. The functional basis of pulmonarysounds. Chest 1978; 73:399–405This review of the functional basis of lung sounds providesdiscussion of the topic by one of the “fathers” inthe field and is easy to read and understand.35. Pratter MR, Hingston DM, Irwin RS. Diagnosis ofbronchial asthma by clinical evaluation: an unreliablemethod. Chest 1983; 84:42–47This prospective, descriptive study was one of the firstto show how unreliable clinical evaluation can be in thediagnsosis of asthma. It also reports the spectrum andfrequency of causes of expiratory wheezing encounteredin a pulmonary clinic practice.36. Shim CS, Williams MH. Relationship of wheezingto the severity of obstruction in asthma. ArchIntern Med 1983; 143:890–892This study shows that expiratory wheezing obtainedby patient history or detected by physical examinationis lacking in sensitivity and specificity in diagnosingasthma.37. Marini JJ, Pierson DJ, Hudson LD, et al. The significanceof wheezing in chronic airflow obstruction.Am Rev Respir Dis 1979; 120:1069–1072Although unforced expiratory wheezing heard on physicalexamination has been shown to be significantly correlatedwith the severity of obstruction in patients withasthma and COPD, the correlations were not strongenough to permit consistent prediction for clinicalpurposes.38. King DK, Thompson T, Johnson DC. Wheezingon maximal forced exhalation in the diagnosisof atypical asthma. Ann Intern Med 1989; 110:451–455Wheezing heard during forced expirations has not beenshown to correlate with the severity of obstruction in COPD,nor has it been shown to be useful in diagnosing asthma.39. Al-Bazzaz F, Grillo H, Kazemi H. Response to exercisein upper airway obstruction. Am Rev RespirDis 1975; 111:631–640This study provides experimental data associating degreeof upper airway obstruction with dyspnea and stridor.40. Geffin B, Grillo HC, Pontoppidan H. Stenosis followingtracheostomy for respiratory failure. JAMA1971; 216:1984–1988This study provides additional data associating degreeof upper airway obstruction with signs and symptoms.41. Irwin RS, Pratter MR, Holland PS, et al. Postnasaldrip causes cough and is associated with reversibleupper airway obstruction. Chest 1984; 85:346–352This study provides prospective data that expiratorywheezing indistinguishable from asthma can originatefrom the upper airway in the PNDS.42. Curley FJ, Irwin RS, Pratter MR, et al. Coughand the common cold. Am Rev Respir Dis 1988;138:305–311This prospective, randomized, controlled clinical trialprovides additional data that the expiratory wheezingin PNDS is associated with an extrathoracic, variableupper airway obstruction.43. Macklem PT. Airway obstruction and collateralventilation. Physiol Rev 1971; 51:368–436This article reviews the anatomy and physiology ofsmall airway diseases.44. Gelb AF, Klein E. The volume of isoflow and increasein maximal flow at 50% of forced vital capacity duringhelium-oxygen breathing as tests of small airwaysdysfunction. Chest 1977; 71:396–399This study describes, demonstrates, and explains howhelium-oxygen breathing during recording of flow-volumeloops can be used to localize the site(s) of airflowobstruction.45. Miller RD, Hyatt RE. Obstruction lesions of the larynxand trachea: clinical and physiologic characteristics.Mayo Clin Proc 1969; 44:145–161This is a classic publication that reveals the usefulness ofphysiologic testing in distinguishing variable extrathoracicfrom variable intrathoracic upper airway lesions.46. Kryger M, Bode F, Antic R. Diagnosis of obstructionof the upper and central airways. Am J Med1976; 61:85–93An excellent review article that covers the physiology,pathophysiology, and clinical aspects of diagnosing air-450 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4507/10/09 10:22:34 PM


way-obstructing lesions.47. Kaminsky DA, Irvin CG. Anatomic correlates ofreversible restrictive lung disease. Chest 1993;103:928–931This publication provides convincing data that diseasescan present with reversible restrictive lung disease thatis clinically indistinguishable from asthma.48. Hudgel D, Cooper D, Souhrada J. Reversiblerestrictive lung disease simulating asthma. AnnIntern Med 1976; 85:328–332This is another publication that reports that patients canhave a reversible restrictive lung disease that is like asthmaand responds to conventional asthma medications.49. Smyrnios NA, Irwin RS. Wheeze. In: Irwin RS, CurleyFJ, Grossman RF, eds. Diagnosis and treatmentof symptoms of the respiratory tract. Armonk, NY:Futura Publishing Company, 1997; 117–153This chapter is a comprehensive, extensively referencedreview of all aspects of wheeze that discusses in detailthe many diseases that comprise the differential diagnosisof wheeze.50. Smyrnios NA, Irwin RS, Curley FJ. Chronic coughwith a history of excessive mucus production: thespectrum and frequency of causes, key componentsof the diagnostic evaluation, and outcome ofspecific therapy. Chest 1995; 108:991–997This prospective, descriptive study reports the spectrumand frequency of causes of cough-phlegm syndromes.Contrary to what had been believed, chronic bronchitisis not as common a cause of this syndrome as PNDS,asthma, or GERD. Consequently, if a patient has wheezeand a cough-phlegm syndrome, the diagnosis of COPDshould not reflexively be made clinically.51. Crapo RO. <strong>Pulmonary</strong> function testing. N Engl JMed 1994; 331:25–30This review article provides guidelines for interpretingroutine pulmonary function testing in determiningwhether or not the patient has reversible or irreversibleairflow obstruction.52. Webb J, Clark TJH, Chilvers C. Time course ofresponse to prednisolone in chronic airflowobstruction. Thorax 1981; 36:18–21This study reveals that significant reversibility inairflow obstruction may occur after weeks of systemiccorticosteroids in patients who would have beenthought to have irreversible disease based solely uponlack of bronchodilator response in the pulmonary functionlaboratory.53. Brown RF, Dibenedetto R, Russell D, et al. Variantcystic fibrosis in an elderly man. South Med J 1986;79:1430–1432This report documents that CF should be considered a possibilityin the proper context regardless of the patient’s age.54. Highsmith WE, Burch LH, Zhou Z, et al. A novelmutation in the cystic fibrosis gene in patients withpulmonary disease but normal sweat chloride concentrations.N Engl J Med 1994; 331:974–980This article describes how to use new diagnostic methodsin diagnosing CF when standard, traditional techniqueshave failed.Hemoptysis55. Corey R, Hla KM. Major and massive hemoptysis:reassessment of conservative management. Am JMed Sci 1987; 294:301–309This article defines hemoptysis and reassesses the relativeroles of surgery vs observation in patients withmassive hemoptysis.56. Irwin RS, Curley FJ, Robinson KA. Managinghemoptysis. In: Rippe JM, Irwin RS, Fink MP, et al,eds. Intensive care medicine. 6th ed. Boston, MA:Little, Brown and Company, 2008; 588–598This chapter defines pseudohemoptysis and provides adifferential diagnosis of the same. It also comprehensivelyreviews with extensive referencing the entire subjectof hemoptysis.57. Johnston H, Reiza G. Changing spectrum ofhemoptysis: underlying causes in 148 patientsundergoing diagnostic flexible fiberoptic bronchoscopy.Arch Intern Med 1989; 149:1666–1668This is a representative article that deals with the spectraand frequencies of causes of hemoptysis from the late1970s to the mid-1980s.58. Nelson JE, Forman M. Hemoptysis in HIV-infectedpatients. Chest 1996; 110:737–743This is a retrospective study that provides informationon the spectra and frequency of causes of hemoptysis inHIV-positive patients.59. Bartter T, Irwin RS, Phillips DA, et al. <strong>Pulmonary</strong>artery pseudoaneurysm: a potential complicationof pulmonary artery catheterization. Arch InternMed 1988; 148:471–473This article summarizes what is known about rupture ofa pulmonary artery by the balloon flotation catheter andreports the complication of pseudoaneurysm and how torecognize and treat it.60. Schaefer OP, Irwin RS. Tracheo-artery fistula.J Intensive Care Med 1995; 10:64–75This review article comprehensively discusses all<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 451ACC-PULM-09-0302-027.indd 4517/10/09 10:22:34 PM


aspects of tracheoartery fistula as a complication oftracheostomy except for describing the surgical operationsto correct the complication.61. Lyons HA. Differential diagnosis of hemoptysisand its treatment. Basics of respiratory disease1976; 5:26–30This article reviews the spectra and frequencies ofcauses of hemoptysis and includes idiopathic or essentialhemoptysis as a cause.62. Rath GS, Schaff JT, Snider GL. Flexible fiberopticbronchoscopy: techniques and review of 100 bronchoscopies.Chest 1973; 63:689–693This study also includes idiopathic hemoptysis as a specificcause of hemoptysis.63. Barrett RJ, Tuttle WM. A study of essential hemoptysis.J Thorac Cardiovasc Surg 1960; 40:468–474The entity of essential (or idiopathic) hemoptysis is discussedin detail.64. Savale L, Parrot A, Khalil A, et al. Cryptogenichemoptysis: from a benign to a life-threateningpathologic vascular condition. Am J Respir CritCare Med 2007; 175:1181–1185From a cohort of 81 patients referred for crytogenic hemoptysis,an abnormal superficial vessel contiguous to the epitheliumof the bronchial mucosa was found as the cause ofhemoptysis (Dieulafoy disease) in 5 of 9 who underwentsurgery. This diagnosis must be kept in mind.65. Murray JF. Postnatal growth and development ofthe lung. In: Murray JF, ed. The normal lung: thebasis for diagnosis and treatment of pulmonarydisease. Philadelphia, PA: WB Saunders, 1976; 21Provides a discussion of the anatomy of the bronchialand pulmonary circulations that is useful to review tobetter understand the pathogenesis of hemoptysis.66. Schwarz MI, Sutarik JM, Nick JA, et al. <strong>Pulmonary</strong>capillaritis and diffuse alveolar hemorrhage: a primarymanifestation of polymyositis. Am J RespirCrit Care Med 1995; 151:2037–2040In addition to reporting pulmonary capillaritis and diffusealveolar hemorrhage as a primary manifestation ofpolymyositis, this article briefly reviews the topic of pulmonarycapillaritis.67. Pratt LW. Hemoptysis. Ann Otol Rhinol Laryngol63:296–309, 1954Although patients with lung infections often have bloodin their sputum, this article makes the point that hemoptysisthat lasts for more than the usual 24 h suggests thatan endobronchial lesion or coagulopathy are present.68. Thomas HM III, Irwin RS. Classification of diffuseintrapulmonary hemorrhage [editorial]. Chest1975; 68:483–484In addition to providing a differential diagnosis ofcauses of diffuse intrapulmonary hemorrhage, thiseditorial makes the point that the lack of hemoptysisdoes not rule out a substantial intrapulmonaryhemorrhage.69. Zeiss CR, Leach CL, Smith LJ, et al. A serial immunologicand histopathologic study of lung injuryinduced by trimellitic anhydride. Am Rev RespirDis 1988; 137:191–196This publication demonstrates that intrapulmonaryhemorrhage can be caused by inhaled TMA.70. Patterson R, Nugent KM, Harris KE, et al. Immunologichemorrhagic pneumonia caused by isocyanates.Am Rev Respir Dis 1990; 141:226–230This study shows that inhaled isocyantes can also causeintrapulmonary bleeding.71. Sisson JH, Thompson AB, Anderson JR, et al. Airwayinflammation predicts diffuse alveolar hemorrhageduring bone marrow transplantation inpatients with Hodgkin’s disease. Am Rev RespirDis 1992; 146:439–443Diffuse intrapulmonary hemorrhage is a complicationof bone marrow transplantation.72. Jackson CL, Diamond S. Hemorrhage from the trachea,bronchi, and lungs of nontuberculous origin.Am Rev Tuberc 1942; 46:126–137This study demonstrates that chest radiographicfindingscan be normal up to 30% of the time when hemoptysisis caused by a lower respiratory tract disease.73. Kim JH, Follett JV, Rice JR, et al. Endobronchial telangiectasiasand hemoptysis in scleroderma. Am JMed 1988; 84:173–174This article documents that bronchoscopy can establisha site of hemoptysis different from that suggested by thechest radiograph.74. Smiddy JF, Elliott RC. The evaluation of hemoptysiswith fiberoptic bronchoscopy. Chest 1973;64:158–162This is another study showing that bronchoscopy canfind a site of hemoptysis different than that suggestedby the chest radiograph.75. Saumench J, Escarrabill J, Padro L, et al. Valueof fiberoptic bronchoscopy and angiography fordiagnosis of the bleeding site in hemoptysis. AnnThorac Surg 1989; 48:272–274If flexible bronchoscopy is performed within 24 h ofactive bleeding, there is a high yield of being able tolocalize the specific site.76. Pursel SE, Lindskog GE. Hemoptysis. Am Rev452 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4527/10/09 10:22:35 PM


Respir Dis 84:329–336, 1961Even rigid bronchoscopy has an excellent chance oflocalizing the specific site of bleeding in hemoptysis ifthe procedure is performed early.77. McGuinness G, Beacher JR, Harkin TJ, et al. Hemoptysis:prospective high-resolution CT/ bronchoscopiccorrelation. Chest 1994; 105:1155–1162If there is a delay of 48 h in doing bronchoscopy, thefrequency of being able to identify the site of bleedingdecreases substantially.78. De Lassence A, Fleury-Feith J, Escudier E, et al.Alveolar hemorrhage: diagnostic criteria andresults in immunocompromised hosts. Am J RespirCrit Care Med 1995; 151:157–163Provides guidelines for interpreting the results of BALfluid (percentage of siderphages) in determining whetheror not the patient has recently had diffuse alveolar hemorrhage.79. Fraser RG, Pare JAP, Pare PD, et al. Diseases of theairways. In: Fraser RG, Pare JAP, Pare PD, et al, eds.Diagnosis of diseases of the chest (vol 3) 3rd ed.Toronto, Canada: WB Saunders, 1990; 2186–2206Discusses the performance of routine chest radiographywhen the patient is suspected to have bronchiectasis.80. Kalluri R, Petrides S, Wilson CB, et al. Anti-al(IV)collagen autoantibodies associated with lungadeno carcinoma presenting as the Goodpasturesyndrome. Ann Intern Med 1996; 124:651–653Discusses the different components of collagen and specificallythe collagen in basement membranes as theyrelate to the diagnosis of Goodpasture syndrome.81. Zimmerman SW, Varanasi UR, Hoff B. Goodpasture’swith normal renal function. Am J Med 1979;66:163–171Immunofluorescent staining can be diagnostic of Goodpasturesyndrome even when the patient has no evidenceof renal disease.82. Border WA, Baehler RW, Bhathena D, et al. IgAanti-basement membrane nephritis with pulmonaryhemorrhage. Ann Intern Med 1979; 191:21–25<strong>Pulmonary</strong>-renal syndrome mimicking Goodpasturesyndrome can be caused by IgA ephropathy.83. Clarke CP, Jackson KA, Moreland M, et al. Bronchoscopicuse of the neodymium-yttrium- aluminumgarnetlaser for lesions of the trachea and bronchus.Med J Aust 1989; 150:260–262When hemoptysis is the result of bronchogenic carcinoma,the use of laser therapy can successfully stop thebleeding.84. Conlan AA. Massive hemoptysis: review of 123cases. J Thorac Cardiovasc Surg 1983; 85:120–124Bronchoscopically directed iced saline solution lavge ofthe site of hemoptysis is reported to stop hemorrhage.85. Rabkin JE, Astafjev VI, Gothman LN, et al. Transcatheterembolization in the management of pulmonaryhemorrhage. Radiology 1987; 163:361–365<strong>Review</strong>s the role of interventional radiology in controllingintrapulmonary bleeding.86. Hickey NM, Peterson RA, Leech JA, et al. Percutaneousembolotherapy in life-threatening hemoptysis.Cardiovasc Intervent Radiol 1988; 11:270–273<strong>Review</strong>s the role ofinterventional radiology in controllingintrapulmonary bleeding and emphasizes the pointthat multiple procedures may be necessary.87. Stoll JF, Bettmann MA. Bronchial artery embolizationto control hemoptysis: a review. CardiovascIntervent Radiol 1988; 11:263–269<strong>Review</strong>s the role of interventional radiology in controllingintrapulmonary bleeding and gives rates of rebleeding.88. Uflacker R. Bronchial artery embolization in themanagement of hemoptysis: technical aspects andlong-term results. Radiology 1985; 157:637–644<strong>Review</strong>s the role of interventional radiology and temporaryballoon occlusion in controlling intrapulmonarybleeding.89. Jardin M, Remy J. Control of hemoptysis: systemicangiography and anastomoses of the internalmammary artery. Radiology 1988; 168:377–383<strong>Review</strong>s the role of interventional radiology and temporaryballoon occlusion in controlling bleeding.90. Crocco JA, Rooney JJ, Fankushen DS, et al. Massivehemoptysis. Arch Intern Med 1968; 121:495–498This older study recommends emergency surgery formassive hemoptysis over conservative therapy; morerecent studies do not support this recommendation.91. MacDonald JA, Fraser JF, Foot CL, et al. Successfuluse of recombinant Factor VII in massive hemoptysisdue to community-acquired pneumonia. Chest2006; 130:577–579Recombinant Factor VII may be a useful temporizingmeasure in unstable patients with life-threatening pulmonaryhemorrhage even when there is no coagulopathy.Dyspnea92. Curley FJ. Dyspnea. In: Irwin RS, Curley FJ, GrossmanRF, eds. Diagnosis and treatment of symptomsof the respiratory tract. Armonk, NY: FuturaPublishing Company, 1997; 55–111This extensively referenced chapter comprehensively<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 453ACC-PULM-09-0302-027.indd 4537/10/09 10:22:35 PM


and critically reviews all aspects of dyspnea.93. Fedullo AL, Swineburne AJ, McGuire-Dunn C.Complaints of breathlessness in the emergencydepartment: the experience at a community hospital.NY State J Med 1986; 86:4–6This study provides data on spectrum and frequency ofcauses of dyspnea evaluated in an emergency department.94. Pearson SB, Pearson EM, Mitchel JRA. The diagnosisand management of patients admitted to hospitalwith acute breathlessness. Postgrad Med 1981;57:419–424This study provides data on spectrum and frequency ofcauses of dyspnea evaluated in hospitalized patients.95. Mustchin CP, Tiwari I. Diagnosing the breathlesspatient. Lancet 1982; 2:907–908This study provides data on spectrum and frequency ofcauses of dyspnea evaluated in hospitalized patients.96. DePaso WJ, Winterbauer RH, Lusk JA, et al. Chronicdyspnea unexplained by history, physical examination,chest roentgenogram, and spirometry: analysisof a seven-year experience. Chest 1991; 100:1293–1299This study reports the spectrum and frequency ofcauses of dyspnea evaluated in the outpatient setting.97. Pratter MR, Curley FJ, Dubois J, et al. Cause andevaluation of chronic dyspnea in a pulmonarydisease clinic. Arch Intern Med 1989; 149:2277–2282This study reports the spectrum and frequency ofcauses of dyspnea evaluated in the outpatient setting98. Kikuchi Y, Okabe S, Tamura G, et al. Chemosensitivityand perception of dyspnea in patients witha history of near-fatal asthma. N Engl J Med 1994;330:1329–1334This study shows that asthmatics with a history ofnear-fatal attacks have a reduced response to hypoxiaand reduced sensation of dyspnea.99. Martinez FJ, Stanopoulos I, Acero R, et al. Gradedcomprehensive cardiopulmonary exercise testingin the evaluation of dyspnea unexplained by routineevaluation. Chest 1994; 105:168–174This study reports the spectrum and frequency ofcauses of dyspnea evaluated in the outpatient setting.The frequency of GERD as the cause was 2%.100. Lum LC. The syndrome of chronic habitualhyperventilation. In: Hill OW, ed. Modern trendsin psychosomatic medicine. London, UK: Buttersworth,1976; 196Of 640 patients treated with slow abdominal breathing,Lum reported that 70% were rendered asymptomatic,20 to 25% were improved, and 5 to 10% failed torespond.101. Reardon J, Awad E, Normandin E, et al. Theeffect of comprehensive outpatient pulmonaryrehabilitation on dyspnea. Chest 1994; 105:1046–1052Deconditioned COPD patients were randomized to anoninterventional group or a group that underwent12 3-h sessions over 6 weeks of an education and exerciseprogram. Dyspnea in the interventional groupsignificantly improved.102. Rogers RM, Donahoe M, Costantino J. Physiologiceffects of oral supplemental feeding inmalnourished patients with chronic obstructivepulmonary disease: a randomized control study.Am Rev Respir Dis 1992; 146:1511–1517This study demonstrates that malnourished COPDpatients had a reduction in dyspnea after a 3-monthrefeeding intervention.103. Woodcock AA, Gross ER, Geddes DM. Drugtreatment of breathlessness: contrasting effects ofdiazepam and promethazine in pink puffers. BMJ1981; 283:343–346This controlled clinical trial failed to show a reductionin dyspnea in patients with COPD when a benzodiazepinewas used.104. Eimer M, Cable T, Gal P. Effects of clorazepate onbreathlessness and exercise tolerance in patientswith chronic airflow obstruction. J Fam Pract1985; 21:359–362This controlled clinical trial failed to show a reductionin dyspnea in patients with COPD when a benzodiazepinewas used.105. Foral PA, Malesker MA, Huerta G, et al. Nebulizedopioids use in COPD. Chest 2004; 125:691–694106. Baydur A. Nebulized morphine: a convenient andsafe alternative to dyspnea relief? Chest 2004; 125363–365These two studies review the lack of efficacy of inhaledopioids for relief of dyspnea in COPD and interstitiallung disease.107. Lewith GT, Prescott P, Davis CL. Can a standardizedacupuncture technique palliate disabling454 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4547/10/09 10:22:35 PM


eathlessness? A single-blind, placebo-controlledcrossover study. Chest 2004; 125:1783–1790Although there have been very few studies assessingthe efficacy of acupuncture for dyspnea relief, thisstudy failed to demonstrate that acupuncture is efficacious.108. Ong K-C, Kor A-C, Chong W-F, et al. Effectsof inhaled furosemide on exertional dyspnea inchronic obstructive pulmonary disease. Am JRespir Crit Care Med 2004; 169:1028–1033This pilot study suggests that inhaled furosemide may helpalleviate dyspnea in patients with COPD. These results<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 455ACC-PULM-09-0302-027.indd 4557/10/09 10:22:35 PM


Notes456 Symptoms of Respiratory Disease (Irwin)ACC-PULM-09-0302-027.indd 4567/10/09 10:22:35 PM


Mechanical Ventilatory SupportCurtis N. Sessler, MD, FCCPObjectives:• <strong>Review</strong> the determinants of oxygenation and ventilation• Compare the different modes of mechanical ventilatorysupport• Examine the role and methods for noninvasive positivepressureventilation• Examine the value of monitoring and ventilator graphics inunderstanding patient-ventilator dyssynchrony• Describe the complications associated with mechanicalventilatory support, including auto-positive end-expiratorypressure (PEEP) and ventilator-associated lung injury• <strong>Review</strong> the importance of patient-focused ventilation, particularlydifferentiating ventilation strategies for obstructivelung disease and for ARDS• <strong>Review</strong> the rationale and results of studies supporting alung-protective ventilation strategy for acute lung injury/ARDS• Examine structured protocols on liberation from mechanicalventilation and the endotracheal tubeKey words: intrinsic positive end-expiratory pressure; mechanicalventilation; noninvasive ventilation; respiratoryfailure; weaningMechanical ventilatory support is used as a keycomponent in the management of both hypoxemicrespiratory failure and hypercapnic respiratory failure,topics that are discussed in more detail in otherchapters. The principle techniques for providingartificial ventilation have changed during the pastcentury. Negative pressure support (PS) achievedwidespread use in the polio epidemic in the 1920swith the Drinker iron lung, followed by introductionof other negative-pressure devices. However,positive-pressure ventilation, which typically isdelivered through an artificial tracheal airway,has achieved broad acceptance, whereas negativepressureventilation is rarely used in currentpractice. Although the patient-ventilator interfacefor positive-pressure ventilation is most often anendotracheal (ET) tube or tracheostomy tube, theless-invasive approach of using a tightly fittingfull-face (oronasal) or nasal mask has achievedwidespread use in the treatment of selected formsof respiratory failure. Advances in mechanicalventilator technology during the past 50 years haveincluded efficient demand valves for triggeredventilation delivery, computerized integrationof physiologic measurements and gas-deliveryparameters, accurate gas-blending systems, andadvanced monitoring and safety systems thathave resulted in more options for the clinicianto match mechanical ventilation (MV) to patientneeds, better patient-ventilator interactions, andgreater patient safety. Experience with MV in thetreatment of patients with ARDS illustrates thatproper MV influences not just gas exchange but alsosurvival and other important outcomes.Physiology of Ventilation and MVDelivery of a breath occurs when there is achange in transpulmonary pressure, which isthe pressure differential between the mouth andthe pleural space. During a spontaneous breath,respiratory muscle contraction causes chest wallexpansion and diaphragmatic descent that createsnegative pressure within the pleural space. Enteringthrough the trachea, inspired gas moves viabulk flow down the pressure gradient through theconducting bronchi and bronchioles to the respiratoryunits, where gases move by diffusion withinalveoli and respiratory bronchioles, and then bydiffusion through the alveolar-capillary membraneto the pulmonary capillaries. Negative-pressureventilators simulate the action of the respiratorymuscles by creating subatmospheric pressure surroundingthe chest, thus expanding the thoraciccavity in a manner that is physiologically similarto spontaneous breathing. The loss of effectivediaphragm movement in the negative-pressureventilation system, however, yields less-efficientalveolar ventilation and gas distribution.Similar to spontaneous breathing and negativepressure ventilation, positive-pressure ventilatorsgenerate transpulmonary pressure, favoring bulkflow of gases down the trachea and the conductingairways. However, this is created by positivepressure within the airway, forcing air down thetrachea, sufficient to overcome pleural pressure to<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 457ACC-PULM-09-0302-028.indd 4577/10/09 8:52:25 PM


achieve gas delivery to respiratory lung units. It isnoteworthy that pleural pressure is often actually apositive pressure at rest as a result of the resistiveforces associated with chest wall, abdominal, orpleural factors. Common examples of increasedpleural pressure include large pleural effusions orpneumothoraces, increased chest wall thicknessand/or rigidity from obesity or burn eschar, anddiaphragmatic elevation from pregnancy, ascites,obesity, or intraabdominal hypertension. Forexample, plateau airway pressures of 30 cm H 2Omay be associated with a transpulmonary pressureof only 20 cm H 2O in a patient who is obese andwho has a pleural pressure of 10 cm H 2O.Distribution of inspired gases within the thoraxdiffers between spontaneous breathing and controlledpositive-pressure ventilation, particularlywhen the patient is in the supine position. Withpositive-pressure ventilation, gases are deliveredpreferentially to zones of least resistance, generallythe more anterior lung units. In contrast,diaphragmatic descent in spontaneous ventilationmore effectively inflates the inferiorly anddependently positioned juxta-diaphragmatic lungunits, improving ventilation/perfusion matchingof these units. Techniques to enhance ventilationof these lung units via spontaneous breathingand by prone positioning often yield improvedoxygenation.Positive-Pressure VentilationAlthough the notion of forcing gas down thetrachea to inflate the gas exchange lung units withoxygen-enriched gas is a seemingly straightforwardconcept, there are many variables to be consideredfor safe and effective MV. The principles ofMV will be discussed in the context of the primaryderangements of respiratory failure, key parametersof positive-pressure ventilation, conventionalmodes of ventilation in widespread use, additionalcommercially available modes of ventilation, andspecialized modes and techniques.Oxygenation and VentilationA conceptual framework for understandingthe key determinants of oxygenation and ventilationduring MV is displayed in Figure 1. Althoughachieving acceptable oxygenation and ventilationis certainly inter-related, approaching the issuesseparately can be useful to emphasize key elements.Hypoxemia is a common component of respiratoryfailure, and the most straightforward solution isto increase the fraction of inspired oxygen (Fio 2).Adjustments in airway pressure can have animportant impact on oxygenation as well, particularlyfor patients who have extensive parenchymallung disease. Considerable ventilation/perfusionmismatching with a large component of underventilatedor shunt-like alveoli contributes greatly tohypoxemia and is often responsive to increases inairway pressure through alveolar recruitment andretention. PEEP is typically added at low levels (ie,5 cm H 2O) to help splint open alveoli even in theabsence of diffuse lung disease; however, PEEP isgenerally increased to recruit and retain alveolithat are prone to cyclic collapse as airway pressuresdecrease during exhalation.Although alveolar distention throughout therespiratory cycle is best accomplished by applyinggreater levels of PEEP, other adjustmentsthat increase mean airway pressure (MAP)—particularly by increasing the duration of inspiratorydistending pressure, ie, inspiratory time(Ti)—can improve alveolar ventilation and oxygenation.Excessive alveolar pressure can be detrimental,however, by the following actions: (1)overdistending functional alveoli, thus increasingdead space; (2) increasing intrathoracic pressurethereby reducing venous return, cardiac output,and oxygen delivery; (3) overstretching alveoli,thereby producing ventilator-associated lunginjury (VALI); and (4) causing alveolar rupture,thereby producing air-leak barotraumas. Ventilationcan be conceptualized by understandingthe parameters related to delivery of individualbreaths, the frequency of breath delivery, and howconventional modes of ventilation differ regardingbreath delivery.Breath Delivery and Principal Modes of MVVolume and Pressure-Targeted Breaths: As displayedin Figure 1, each breath the patient receives can bedefined by the 3 “Ts”: (1) the “target” for delivery, ie,volume or pressure; (2) the “trigger” or signal to initiatethe breath; and (3) the “termination” criteria thatsignals the end of the inspiratory phase and beginningof the expiratory phase. Breath characteristics can458 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4587/10/09 8:52:26 PM


OxygenationVentilationMAP FIO 2PEEP Target TriggerInspiratorypressureInspiratorytimeControlled(Mandatory)- Pressure orvolumetargeted- Timetriggered- Time orvolumeterminatedPressureVolumeAdjust maxflow rate &pattern,inspiratorytimeAssisted(Mandatory)- Pressure orvolumetargeted- Patient (flowor pressure)triggered- Time orvolumeterminatedBreath CharacteristicsTimed(Ventilatorinitiated)Patientinitiated- Flow- PressureCommon Breath TypesTermination(Cycling)TimeVolumeFlowSupported(Spontaneous)- Pressuretargeted- Patient (flowor pressure)triggered- FlowterminatedMode & rate:Select mode,set minimumrate, patientdeterminestotal rateAssist control:all breaths arecontrolled orassistedSIMV:Set ratebreaths arecontrolled orassisted,additionalbreaths aresupportedSpontaneous/ PSV(PressureSupportVentilation):No set rate.All breathsare supportedFigure 1. Mechanical ventilation parameters that influence oxygenation and ventilation. Oxygenation is controlled by adjustingFio 2and parameters that change MAP. Ventilation is determined by breath characteristics, ventilator mode, and respiratory rate,including minimum mandatory breaths and additional patient breaths. Each breath can be defined by the delivery target (pressureor volume) and the signals to initiate and to terminate the breath. There are three types of breaths as defined by initiation andtermination signals: ventilator-initiated mandatory breath, patient-initiated mandatory breath, and patient-initiated spontaneousbreath. The mode determines the type of breaths that can be mandatory and those that are in addition to the minimum set rate.be determined by the clinician (ie, mandatory controlledor assisted breath) or influenced by cliniciansettings and patient effort (ie, pressure- supportedspontaneous breath). These parameters differ in thecommon modes of ventilation (Table 1). An importantconcept is that mandatory breaths are “guaranteed,”both in terms of a minimum number of breaths eachminute, as well as delivery of a predetermined volumeor a predetermined pressure and Ti. In contrast,spontaneous breaths are dependent on patient effortfor both breath initiation as well as breath duration.Mandatory breaths can be volume or pressure-targeted,whereas spontaneous breaths areonly pressure-targeted in conventional modes. Thepatient interacts with the ventilator via sensorsthat are in-line pressure or flow transducers thatrespond to the patient’s spontaneous efforts. Datasuggest that flow-triggered sensors are more sensitiveto the patient’s efforts than are the demandsensors triggered by changes in pressure, unlessthe pressure transducer is positioned at the distalend of the ET tube. However, clinically available<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 459ACC-PULM-09-0302-028.indd 4597/10/09 8:52:26 PM


Table 1. Comparison of Principal Modes of VentilationMode of ventilation AC Mode SIMV Spontaneous/PSVMandatory breaths Yes Yes NoSpontaneous breaths* No Yes YesVolume or pressure Either Mandatory: eitherspontaneous: pressure onlyPressureAdvantages Stabilization mode Comfortable modeDisadvantagesRespiratory alkalosis, airtrapping with tachypneaLess comfortableUnreliable ventilation if there isapnea, weakness, deteriorationin lung mechanics*Spontaneous breaths are all initiated by the patient, and the termination of inspiration (and thus inspiratory time) is determinedby the patient.sensors are positioned inside the ventilators and,thus, the pressure or flow signal is dampened bythe dead space of the ventilator tubing that connectsthe ET tube to the ventilator. Depending onthe sensitivity and responsiveness of the ventilator,respiratory muscular efforts may not be sensedby the ventilator, causing poor ventilator/patientinteraction. This asynchrony increases the imposedwork of breathing (WOB) that the patient has togenerate to initiate a breath.Volume-Targeted Breaths—In volume-targetedbreaths, a tidal volume (Vt) is chosen, and the breathis delivered by applying an inspiratory flow rate(flow limitation) and pattern of delivery (squarewave, sine wave, or decelerating or ramp pattern).Typical flow rates vary from 30 to 80 L/min, withgreater flows resulting in shorter Ti and shorterinspiratory/expiratory (I:E) ratios but greater peakinspiratory pressure (PIP). Similarly, square waveflow results in shorter Ti and greater PIP, whereasthe decelerating pattern has the opposite effects. Collaborationwith a skilled respiratory therapist is importantto fine-tuning these parameters. The breathis terminated and exhalation permitted after the setVt has been delivered. In some volume-targetedmodes, the Ti is also set, and exhalation begins afterTi is concluded. The airway pressure that is generatedto deliver the breath varies, increasing to a peakpressure as the lung is inflated. Airway pressures aredependent on the ventilator settings as well as lungand chest wall mechanics. For example, greater PIPcan be result of faster flow rates or larger Vt but alsobronchospasm, ET obstruction, or tension pneumothorax,with no change in set flow or Vt.Pressure-Targeted Breaths—In contrast to volume-targetedbreaths, pressure-targeted (or pressure-limited)breaths do not guarantee the deliveryof specific Vt. The inspiratory pressure (Pi) and theTi will determine the Vt. The Pi is set by the clinicianin all conventional pressure-targeted modes,but Ti can be determined by the clinician (as ina mandatory breath) or by the patient (as witha spontaneous breath). If the Pi and Ti are heldconstant but lung mechanics change, the Vt maybe larger or smaller. For example, the resolution ofpulmonary edema would improve lung compliance,which would result in a larger Vt for the same Pi andTi, whereas new onset of bronchospasm would beassociated with a smaller Vt from increased airwayresistance.Breath Initiation (Trigger), Breath Termination,Breath Frequency, and Principle Modes of Ventilation—Thefrequency of breath delivery is dependenton the signal(s) to initiate the breath, the modeof ventilation, the minimum set rate for mandatorybreaths, and patient effort. The influence of ventilatormode on respiratory frequency and type ofbreaths is compared in Table 1 and illustrated inhypothetical cases in Table 2. For ventilator-initiated(or -controlled) mandatory breaths, the breath isinitiated when a clinician-determined time intervalhas been reached. For example, if a frequency of10 breaths/min is set, a new breath will be initiatedevery 6 s. Among the conventional modes ofventilation (Table 1), the frequency of mandatorybreaths can be set with the assist control (AC) modeor synchronized intermittent mandatory ventilation(SIMV) mode. Mandatory breaths can also be initiatedby the patient (ie, patient-initiated [or -assisted]mandatory breath; see Fig 1). This feature promotespatient comfort by recognizing a patient’s inspiratoryeffort and delivering a mandatory breath thathas the same clinician-determined characteristics asif ventilator-initiated.460 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4607/10/09 8:52:26 PM


Table 2. Hypothetical Examples of Varying the Drive to Breath, the Number of Patient Inspiratory Efforts, and Clinician-Selected Parameters on Breath Types and Characteristics in PrincipalModes of Ventilation*Scenarios ParametersA. Baseline(12 inspiratoryefforts/min)Mode setFrequencyPressure orvolumeMandatorybreathsSpontaneousbreathsB. Apnea MandatorybreathsC. 30 patient inspiratoryefforts/minD. ↓set f to 2/min; thereare 12 patient inspiratoryefforts/minSpontaneousbreathsMandatorybreathsSpontaneousbreathsMandatorybreathsSpontaneousbreathsAssist Control 10/min Assist Control 10/min SIMV 10/min SIMV 10/min SpontaneousPressure Volume Mandatory: pressure,Supported: pressure12 total, mix ofcontrolled andassisted12 total, mix ofcontrolled and assisted10 total, mix of controlledand assistedMandatory: volume,Supported: pressure10 total, mix ofcontrolled and assisted0 0 2 2 1210 controlled0 assisted10 controlled0 assisted10 controlled0 assisted10 controlled0 assisted0 0 0 0 030 total, most assistedsince tachypneic30 total, most assistedsince tachypneic10 total, 0-10 controlled,the rest assisted10 total, 0-10 controlled,the rest assisted0 0 20 20 3012 total,2 controlled;10 assisted12 total,2 controlled;10 assisted2 total probablyall controlled2 total probablyall controlled0 0 10 10 12Supported:pressureE. <strong>Pulmonary</strong> edema develops Tidal volume Decreases No change Decreases No change DecreasesPeak airway No change Increases No change Increases No changepressure0000*Five examples of ventilator modes are presented. The rate is set at 10 breaths/min for all except the spontaneous mode. Five scenarios are presented in which different numbersof patient inspiratory efforts are made (A, B, and C), there is a reduction in the set rate to 2 breaths/min (D), or pulmonary edema develops (E). For the first four scenarios, theexpected effect on mandatory and spontaneous breaths is depicted. In scenario E, the effect on Vt and peak airway pressure is illustrated.ACC-PULM-09-0302-028.indd 4617/10/09 8:52:27 PM


The AC and SIMV modes differ in the numberof mandatory breaths permitted. With theAC mode, a minimum frequency (for example,10 breaths/min) is set; however, all additionalbreaths that the patient triggers are also mandatorybreaths. In contrast, with SIMV, the number ofbreaths that are mandatory is always determinedby the clinician-set minimum frequency. Any additionbreaths are treated as “spontaneous” breathsthat may be pressure “supported” (see scenarios Cand D in Table 2 for examples). For SIMV, ventilatormanufacturers apply different algorithms thatdetermine whether an inspiratory effort triggersan assisted mandatory breath or a spontaneouspressure-supported breath. For example, withsome ventilator models, the delivery of a mandatorybreath is followed by a “refractory period,”during which time a patient’s inspiratory effortwill trigger a spontaneous breath but not a mandatorybreath. Finally, in the spontaneous mode, thepatient determines both the onset and terminationof each breath, thus controlling both the frequencyof breathing and the Ti (which generally variesfrom breath to breath). These spontaneous breathsare usually pressure supported. The spontaneousmode with all breaths receiving PS is widelyreferred to as pressure support ventilation (PSV).In the spontaneous mode, Vt typically varies frombreath to breath and will also change with alterationsin lung mechanics.Summary and Comparison of Principle Modes ofVentilation: AC, SIMV, and Spontaneous Modes: Thethree principle modes of ventilation are comparedin Table 1. Among these modes, an internationalprospective surveillance study performed in 1998of 5,000 mechanically ventilated adults showedthat AC was the most commonly applied mode,with nearly two thirds of patients with acute exacerbationof COPD and 60% of patients with ARDSreceiving AC ventilation. Only 10 to 15% of patientswith COPD or ARDS received SIMV with or withoutPS, whereas nearly 10% of patients with COPD hadPSV, and 10 to 15% of patients with ARDS receivedpressure-controlled ventilation (PCV). The effectof hypothetical changes in respiratory frequencyor respiratory drive and changes in lung mechanicson ventilatory parameters for the three principlemodes are displayed in Table 2.AC Ventilation—AC ventilation may be volumeor pressure targeted, and all breaths have the samesettings (Vt for volume targeted, Pi and Ti for pressuretargeted). The clinician sets the minimal numberof breaths per minute. In addition, inspiratoryefforts trigger the ventilator to supply additionalmandatory breaths. Usually, AC ventilation is usedwhen the operator desires to minimize the amountof WOB that the patient has to perform. AC ventilationcannot be used as a weaning mode becauseall breaths the patient takes are mandatory breathsthat are associated with minimal WOB regardlessof the minimum frequency setting (see scenario Cin Table 2). Because all breaths are mandatorybreaths, hyperventilation in this setting can bedetrimental because of the resulting high minuteventilation and associated respiratory alkalosisbut also in the potential for “stacking” of breathsthat could result in an increase in airway pressurecaused by auto-PEEP and dynamic hyperinflation(scenario D in Table 2).SIMV—Intermittent mandatory ventilation(IMV) was originally developed as a bridge betweencontrolled mandatory ventilation and spontaneousbreathing. With controlled mandatory ventilation,all breaths are ventilator-initiated mandatorybreaths, and patient inspiratory efforts do not resultin any airflow or ventilation, an uncomfortablesensation. With IMV, the patient’s inspiratory effortpromotes delivery of a spontaneous breath from areservoir bag that contains gas with the same Fio 2as for mandatory breaths. The size of the breath isdetermined by patient effort and lung mechanics.Most modern ventilators use “synchronized” IMV,which permits the patient to initiate a mandatorybreath (ie, an “assisted” mandatory breath), or initiatea spontaneous breath. Current ventilators allowsupport of the spontaneous breaths with a predeterminedpressure (ie, a PS breath). Thus with SIMV,there is a combination of controlled mandatory,assisted mandatory, and pressure-supported spontaneousbreaths. Whether an assisted or supportedbreath is produced is dependent on the timing ofthe inspiratory effort. Over the course of a minute,the total number of mandatory breaths, includingboth controlled and assisted breaths, cannot exceedthe preset frequency. Although the original intentionof SIMV was to speed the transition from fullventilatory support to spontaneous breathing bygradually decreasing the number of mandatorybreaths, clinical trials actually demonstrate SIMVto be inferior to PSV for weaning patients off462 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4627/10/09 8:52:27 PM


ventilator support. Patient comfort can be impairedin this mode as well, with the patient receiving threedifferent breath types.Spontaneous Ventilation/PSV—Spontaneousventilation requires patient effort for the initiationof all breaths. Additionally, the termination of allbreaths is determined by a reduction in inspiratoryeffort that is reflected in a decrease in flow below apreset percentage of peak flow. This has been calledflow-cycled termination. If PEEP is applied, thencontinuous positive airway pressure (CPAP) is presentbecause the patient’s airway pressure is positiveduring inspiration and expiration. Virtually all cliniciansnow augment spontaneous breaths with PS,usually in the 5 to 25 cm H 2O range, often referred toas PSV. With PSV, flow commences when a pressure ora flow sensor is triggered. The duration of the addedPS is determined by the spontaneous inspiratoryflow, related to patient inspiratory effort. PSV cyclesoff when the spontaneous inspiratory flow decreasesto a preset level (usually approximately 25% of thepeak flow). The amount of pressure can be adjustedto achieve comfortable breathing, with acceptableVt and respiratory rate (frequency, or f ). Researchsuggests the PSV is the most comfortable mode ofventilation for the alert patient. Additionally, theamount of PS can be reduced, thus placing a greaterrequirement on the patient for WOB during transitionfrom ventilatory support to spontaneous breathing.In the past, PSV was a popular approach toweaning. The patient was placed on spontaneousmode and the PS level sequentially lowered, increasingthe amount of patient WOB. The patient withadequate respiratory muscle strength and reserveexperiences minimal change in the breathing pattern(f, Vt, and thoracoabdominal synchrony) asthe PS level is reduced. However, if the patient isnot ready to assume all the required WOB, f willincrease, and Vt will decrease as the PS level issequentially diminished. In current approaches toliberation of the ventilator, PSV (5 to 8 cm H 2O) isoften used to overcome the added WOB that accompaniesthe resistance of breathing through the ETtube during a spontaneous breathing trial. Becausemandatory breaths are not provided in this mode,reliance on the patient for initiating each breathas well as having sufficient respiratory drive andrespiratory muscle strength to achieve adequateminute ventilation is an important limitation toconsider if the patient is likely to become apneicor weak. Further, as with any pressure-targetedmode, deterioration in lung mechanics can resultin hypoventilation despite previously effective PSsettings.Some Additional Commercially Available Modesof VentilationPCV: PCV is a controlled mode of ventilation(no spontaneous breaths) that previously waspopular for the ventilatory management of ARDS,particularly with extended duration of the inspiratoryphase in excess of inspiratory phase, socalledpressure controlled inverse ratio ventilation orPC-IRV. Recall that the I/E ratio is normally 1:2to 1:4. PC-IRV uses I/E ratios of 1:1, with ratiosof 3:1 occasionally used. PCV is similar to pressure-targetedAC or SIMV but without assisted orspontaneous breaths. With PCV, the ventilator rapidlyachieves the clinician-set Pi, which is usuallydelivered in a decelerating waveform. The operatorsets the duration of inspiration (and thus the Vtvaries) by varying the Ti (time cycled) or by settingthe I/E ratio. Inspiratory flow from the ventilator tothe patient will continue until the preset Ti or the I/Eratio has occurred. In the time-cycled mode, if thepreset pressure is achieved early in the inspiratorycycle, an inspiratory hold will occur. The increase inMAP seen with PC-IRV can be effective for alveolarrecruitment and retention and has been associatedwith better oxygenation. Outcome studies have notdemonstrated greater survival or short duration ofMV with PC-IRV. The major disadvantages of PCVare the need for heavy patient sedation and thevariability of delivered Vt and minute ventilationwith changes in respiratory mechanics.Bilevel Ventilation and Airway Pressure ReleaseVentilation (APRV): Bilevel ventilation is, inessence, a modification of SIMV in which mandatorybreaths are pressure-targeted and PS is used toaugment spontaneous breaths. The unique featureof bilevel ventilation is that spontaneous breathscan be performed throughout the respiratory cycle.Thus, in contrast to other pressure-targeted modes,including SIMV, AC, and PCV, the patient can takespontaneous breaths even during a prolongedlung inflation. Mandatory breaths are delivered byincreasing airway pressure to a new greater level ofpressure for a set time, and exhalation occurs whenairway pressure is decreased to the PEEP level.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 463ACC-PULM-09-0302-028.indd 4637/10/09 8:52:27 PM


The frequency is set as with any mandatory breathsand like SIMV or AC can be initiated by the ventilator(ie, controlled) or initiated by the patient (ie,assisted).PS can be applied to augment spontaneousbreaths, but as a safety feature, the amount of PSis added relative to the PEEP level. For example, ifPEEP is 10 cm H 2O and the Pi is 15 cm H 2O, thenthe plateau airway pressure will be 25 cm H 2O.An increase to 20 cm H 2O PS would increase thepeak pressure of spontaneous breaths taken duringthe expiratory phase to 30 cm H 2O (ie, 10 cmH 2O PEEP 20 cm H 2O PS)—this sets the upperlimit for PS. Spontaneous breaths taken while at aninspiratory plateau pressure of 25 cm H 2O wouldonly be augmented to the 30 cm H 2O limit; thus, aneffective PS pressure of only 5 cm H 2O is provided.Bilevel ventilation is a potentially attractive modefor ARDS, combining the advantages of ventilationwith pressure-targeted prolonged inspiratorybreaths, yet by permitting spontaneous ventilationsat any time in the respiratory cycle, reducingthe requirement for sedation and neuromuscularblockade.The term APRV has different connotationsdepending on locale. Bilevel ventilation with I/Eratio of 1:1 has been called APRV in some Europeanstudies; however, in the United States, the termAPRV is generally applied to pressure-targetedventilation with a very long Ti and a very brief (ie, 0.8 s) expiratory time. Typically, exhalation to aset PEEP of zero is often performed; however, theactual PEEP is probably considerably greater as aresult of intrinsic or auto-PEEP that develops fromincomplete exhalation. In European clinical studies,bilevel ventilation with 1:1 I/E ratio and no PS forspontaneous breaths was associated with improvedgas exchange and patient comfort. There is a paucityof outcomes studies that demonstrate improvedoutcomes for APRV with short expiratory time. Animportant concern for the very brief expiratory timeis the abrupt development of life-threatening auto-PEEP caused by increased expiratory resistancefrom retained secretions or bronchospasm.Proportional-Assist Ventilation (PAV): PAV wasdesigned to deliver positive-pressure ventilationin proportion to respiratory resistance and elastance.PAV attempts to incorporate a feedback loopfrom the spontaneous breathing pattern to betterproportion subsequent positive-pressure breaths.This requires real-time, ongoing measurement ofthese parameters with rapid adjustments of PAVto keep pace with changes in patient effort and resistance,which can be significant even when thereare no clinical changes in the patient’s status. Conceptually,this form of ventilation should optimizepatient-ventilator interaction. However, a recentlypublished study examining WOB with PAV vs PSVfailed to show superiority of PAV over PSV duringquiet breathing or when dead space was added tothe circuit.Noninvasive Positive-PressureVentilation (NPPV)NPPV is widely used in a variety of settingsof respiratory failure. NPPV is delivered througha tight-fitting full-face (oronasal) or nasal mask. Itcan be delivered via a pressure-targeted or volumetargetedmode delivered with a conventional ICUventilator or a portable ventilator specificallydesigned for bilevel positive airway PS. The mask isconnected to the ventilator via a single inspiratoryand expiratory line fitted with a special exhalationvalve or by dual lines such as used with conventionalMV. The operator sets the inspiratory positiveairway pressure as well as the expiratory positiveairway pressure. These setting are analogous to PSand PEEP, respectively. This bilevel ventilator canbe used in a spontaneous mode or in a spontaneous/timedmode that is the equivalent of SIMVplus PSV. Being a pressure-targeted mode of ventilation,bilevel NPPV does not guarantee deliveryof a constant Vt or minute ventilation. A limitationto NPPV is that the monitoring systems in conventionalventilators may be too sensitive to air leaksthat occur around the mask. Recently, bilevel pressuredevices have incorporated more sophisticatedalarm systems to circumvent this problem.NPPV has been used in many forms of respiratoryfailure. Among the many proposed indications,acute hypercapnic respiratory failure in thepatient with COPD has the strongest evidence basefor recommendation. Several multicenter randomizedcontrolled trials (RCTs) have demonstratedlower intubation rates, shorter duration of MV,and improved survival. Additional indicationswith demonstration of improved outcomes includetreatment of cardiogenic pulmonary edema, feverand pulmonary infiltrates in immunocompromised464 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4647/10/09 8:52:27 PM


hosts, postextubation respiratory distress causedby upper-airway obstruction, and persistentwea ning failure with extubation to NPPV. It isimportant to recognize that patients who requireemergent ET intubation are poor candidates forNPPV, such as those with overt respiratory distressor cardiopulmonary arrest. Additionally, suitablecandidates for NPPV are those who are hemodynamicallystable, have sufficiently alert mentalstatus (Glasgow coma scale score 10), lack offacial trauma or deformity, and adequate clearanceof respiratory secretions.Correction of profound hypoxemia is often notpossible with NPPV, even with additional sourcesof supplemental oxygen. During NPPV, one mustavoid air leaks, eye irritation, and nasal abrasions.Careful attention by respiratory therapists andbedside nurses, including the establishment of agood mask fit and careful titration of airway pressuresto tolerance, will help with the success ofNPPV. It is particularly important to continuouslyre-evaluate tolerance of NPPV and to identify in atimely fashion impending failure of NPPV. Expertsrecommend that if NPPV has not resulted in clearimprovement within 2 h of onset, ET intubationand conventional ventilation should be stronglyconsidered. Signs of improvement include reducedrespiratory distress (ie, less tachypnea and use ofaccessory muscles, better thoracoabdominal coordination),or physiologic improvement as determinedby arterial blood gas analysis or breathingpattern analysis. It is worth considering that manystudies are modest in size, patients enrolled representa small percentage of patients with acuterespiratory failure at those centers, and blindingsubjects and investigators to the intervention isimpossible. Finally, the sobering results of a multicenterRCT that demonstrated a greater mortalityrate for non-COPD patients who experiencedextubation failure and were randomized to NPPV(vs usual care) emphasize the importance of closeobservation and rapid reintubation for a patientwho is not responding to NPPV.Monitoring During MVAirway pressures, volumes, flow, and othervariables can be monitored during ventilation.Monitoring incorporates continuous measurementof key parameters, display of numericaland graphic data to enhance visual inspection,establishment of thresholds that define acceptablelimits for key parameters, and use of visual andaudible alarms that alert clinicians of breaches inestablished thresholds. Key safety measurementsinclude low ventilation or low pressure alarms thatmight indicate ventilator disconnection or loss ofairway, apnea alarms, and high airway pressurealarms. Elevation of PIP is important to detect andshould prompt investigation of potential causes.The following equation provides clues as to reasonsfor elevated PIP:PIP flow resistance plateau pressure(Pplat)where Pplat Vt/compliance PEEP. Thus,PIP elevation might be caused by increased airwayresistance, in which case the Pplat would beunchanged. Common causes of increased airwayresistance include bronchospasm and ET tube narrowingby secretions or a patient’s biting of thetube. Elevation of both PIP and Pplat is typicallyproduced by parenchymal lung disease (pulmonaryedema, pneumonia, auto-PEEP); pleuraldisease (pneumothorax); chest wall abnormality(obesity); or increased abdominal pressure (ascites,pregnancy).Close examination of graphic display of airwaypressure, flow, and volume vs time can provideimportant clues as to patient-ventilator asynchrony,the likelihood that auto-PEEP is present, and themode and settings of positive-pressure ventilation.Some of the recognized forms of patient—ventilator asynchrony include ineffective triggering,double triggering, auto triggering, and flowasynchrony. Ineffective triggering is considerablymore common than the others, with double triggeringsecond in frequency in prospective studies.An example of ineffective triggering is displayedin Figure 2, with the arrow corresponding to theevent. Ineffective triggering occurs when an inspiratoryeffort by the patient does not create changein airway pressure or flow of sufficient magnitudeto initiate a positive pressure breath. The characteristicfindings seen on the graphic display includea transient peak in flow and concomitant dip inairway pressure (that is the result of the patient’sinspiratory effort) that does not result in a triggeredpositive pressure breath. The most common causeof ineffective triggering is auto-PEEP, often in the<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 465ACC-PULM-09-0302-028.indd 4657/10/09 8:52:28 PM


Figure 2. Ineffective triggering patient-ventilator dyssynchrony. Graphic display of flow and pressure over time demonstrates twoconventional breaths followed by an ineffective trigger in which patient inspiratory effort (arrows) does not trigger a breath.setting of COPD. Worsening airflow obstruction,tachypnea, high trigger sensitivity, and excessivePS can exacerbate auto-PEEP. Double-triggering isidentified by the presence of two positive pressurebreaths separated by a very brief expiratory phase(less than one half of the preceding Ti; Fig 3). Oftenthe patient’s ventilatory demand is high, and theset ventilator time is too brief, leading to a decreasein airway pressure early in exhalation that triggersan immediate second positive pressure breath.Complications of Positive-PressureVentilationComplications of MV and artificial airways areimportant issues to address proactively to prevent,recognize in a timely fashion, and manage appropriately.Some of the more common complicationsare listed in Table 3. Several complications requireadditional discussion.<strong>Pulmonary</strong> BarotraumaAir leaks typically develop from rupture ofalveoli, particularly alveoli that abut fixed structuressuch as bronchi and vessels. With continuedleaking, such as with positive-pressure ventilation,air then tracks in a path of least resistance, oftenalong the bronchovascular bundles toward thehila of the lung, rupturing into mediastinum, andthen through the thoracic inlet to produce subcutaneousemphysema. Eventually, air rupturesinto the pleural space, producing a pneumothorax.It is important to recognize lesser forms ofbarotrauma, such as subcutaneous emphysemaor pneumomediastinum, so that adjustments canbe made to reduce ventilatory pressures and haltthe progression.Dynamic Hyperinflation and Auto-PEEPAvoidance of overinflation of the lung relies ondelivery of appropriately sized tidal breaths deliveredwith appropriate transpulmonary pressureplus allowance of full exhalation before initiatingthe next breath. Failure to adhere to these conceptsincreases the likelihood of developing dynamichyperinflation from overinflation and presenceof auto-PEEP (also called intrinsic PEEP or occultPEEP) from failure to fully deflate (ie, “stacking”breaths). Both are potentially dangerous conditionsthat might contribute to respiratory compromiseFigure 3. Double-triggering patient-ventilator dyssynchrony. Graphic display of flow and pressure over time demonstrates threeconventional breaths followed by a double-triggered breath (arrows). See text for definition.466 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4667/10/09 8:52:28 PM


Table 3. Complications of MV and Artificial AirwaysOxygen toxicity<strong>Pulmonary</strong> barotrauma: air leakPneumothoraxPneumomediastinum/pneumopericardium/subcutaneousemphysema<strong>Pulmonary</strong> interstitial emphysema and air cystsSystemic air embolismDynamic hyperinflation and auto-PEEP or intrinsic PEEPHemodynamic compromiseAlveolar overdistentionAltered diaphragm/chest wall position and function,impairing ventilationVALIVolutraumaAtelectraumaBiotraumaMultiple organ dysfunctionVentilator-associated pneumoniaTracheomalaciaTrauma and dysfunction of vocal cordsTrauma to lip, teeth, oral cavity, nose, pharynxUnplanned extubationEndotracheal tube malpositionBronchial intubationEsophageal intubationTracheostomy bleeding, stoma problems, decannulationGI bleeding from stress ulceration/gastritisand hemodynamic compromise. Hyperinflationcan overdistend alveoli, increasing the likelihoodof rupture as well as compressing the alveolar capillaries,thus converting functional alveoli to deadspace. Additionally, hyperinflation can place thediaphragm in a flattened, mechanically disadvantagedposition that contributes to smaller Vt, whichin turn promotes tachypnea, thus exacerbatingair trapping by further reducing exhalation time.Auto-PEEP is also a common cause of ineffectivetriggering patient-ventilator asynchrony becausethe patient must overcome the auto-PEEP withdecrease in intrathoracic pressure of sufficientmagnitude to be sensed by the ventilator. Whenextreme, hypotension can progress to frank shockand even to cardiopulmonary arrest, typicallywith pulseless electrical activity. Mechanismsinclude reduced venous return, plus a componentof altered right and left heart chamber filling, andincreased right ventricular afterload from theincreased intrathoracic pressure.A common scenario is the patient with airflowobstruction who is tachypneic in the setting of positive-pressureventilation, thus causing insufficientexhalation time and progressive hyperinflationand auto-PEEP. Another situation is the use of veryshort exhalation time (ie, 1s), such as with APRV,in which even mild bronchospasm or respiratorysecretion retention can impair exhalation, producingincomplete emptying. Unexpected auto-PEEPcan occur in patients without obvious risk factors.Finally, manual ventilation with an Ambu bag isoften performed too rapidly and with uncontrolledVts, thus causing transient hyperinflation. Amanual rate of 12 breaths/min is generally recommended.The development of serious hyperinflationcan be subtle and can be recognized by wheezingexhalation persisting up to the next positive pressurebreath, respiratory efforts (chest movement)without ventilator triggering, characteristic findingsof ineffective trigger on ventilator graphics,hyperinflation on chest radiograph, or a decrease inthe Vt if the patient is receiving positive-pressureventilation in a pressure-targeted mode. Auto-PEEP is not directly measured during conventionalventilation, thus the designation occult PEEP. Onecan temporarily occlude the exhalation port bypressing a button on most modern ventilators,creating a no-flow state that allows equilibrationof the pressure in the ventilator tubing (where thepressure is sensed) with the pressure deep withinthe lung. Total PEEP is measured, and one subtractsthe extrinsic (or set) PEEP, yielding an estimationof auto-PEEP.Often the presentation is more dramatic withhypotension and decreasing oxygen saturation.Life-threatening hyperinflation can be immediatelytreated by disconnecting the ET or tracheostomytube from the ventilator, thus allowing passiveexhalation—that may take 15 s—and preventdelivery of additional breaths. This exhalation isoften followed by rapid and dramatic improvementsin BP and oxygenation. More definitivetreatment includes bronchodilators, sedation toreduce spontaneous respirations, and ventilatoradjustment to increase expiratory time (reducerespiratory rate, lower Vt, and perhaps increaseinspiratory flow rate). If the patient is breathingspontaneously and has auto-PEEP-inducedineffective trigger breaths, extrinsic PEEP canbe increased to a level several cm H 2O less thanmeasured auto-PEEP to reduce the magnitudeof pressure decrease the patient must generate<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 467ACC-PULM-09-0302-028.indd 4677/10/09 8:52:29 PM


to trigger a breath. The magnitude of inspiratoryeffort by the patient can be measured by the useof an esophageal balloon that displays changes inpleural pressure during these efforts.Ventilator-Associated Lung InjuryThere is considerable experimental evidencefrom animal models and accumulating evidencefrom patients that MV can be injurious, producingalveolar trauma and a systemic inflammatoryresponse. It is hypothesized that excessive alveolarvolume coupled with increased transalveolarpressure results in shearing forces that disruptthe fragile alveolar architecture. These deleteriouseffects are thought to be the worst when there islocal inhomogeneity of ventilation. This has beencalled volutrauma because experiments in animalmodels identify alveolar distention, rather than thedistending pressure per se, as the most importantparameter. Additionally, the cyclical inflation anddeflation of alveoli produces damage and inflammation,or so-called atelectrauma. The subsequentmechanical damage to lung tissue activates inflammatorymediators that act as a pulmonary sourcefor multiple-organ failure in patients with acutelung injury (ALI) or ARDS and, to a lesser extent, inpatients who do not satisfy criteria for ALI/ARDS.Figure 4 is an illustrated depiction of the majorpathophysiologic and altered gas exchange eventsdocumented in ALI, with superimposed similareffects of MV on the lung that are derived fromanimal model research and human studies. Thesimilarities are striking. Initially shown in animalmodels, but more recently in patients with ALI, theapplication of low Vts, ie, 6 mL/kg of predictedbody weight in humans, is associated with reducedinflammation, less lung injury, shorter duration ofMV, and improved survival. Further, many studiesdemonstrate that the application of PEEP at a levelabove the LIP prevents the cyclic alveolar collapseand reduced lung injury and inflammation.Pathophysiology ofAcute Lung InjuryAdverse Effects ofMechanical VentilationSystemicInflammationProtein-rich<strong>Pulmonary</strong>Edema<strong>Pulmonary</strong>InflammationAlveolarCapillaryDamageVALIAlveolarOverdistentionDamage &Cyclic AlveolarCollapsePersistentInflammation &DisorganizedFibrotic RepairSurfactantDysfunctionOxygenToxicityFibrosisImpaired GasExchange &Reduced LungComplianceAlveolarCollapseBarotrauma,AlveolarOverdistensionFigure 4. Hypothetical depiction of the pathophysiology of ALI and adverse effects of MV as related to components of lunginjury.468 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4687/10/09 8:52:29 PM


Patient-Focused MVThe goals of MV include improving hypoxemiaand hypercapnia, relief of suffering and distresssuch as dyspnea, providing respiratory supportduring circumstances in which the patient experiencesan impaired drive to breathe (ie, general anesthesia),and applying evidence-based strategiesto promote survival and timely recovery. Lungprotectivestrategies and permissive hypercapniasupersede goals of normalization of gas exchange.One can conceptualize stages of respiratory failureand MV support as stabilization, maintenance, andrecovery.Initiating MVMV may be required when there is a failure ofoxygenation or ventilation. Life-threatening respiratorydistress or cardiopulmonary arrest requiresimmediate attention and emergent ET intubation,followed by initiation of MV. NPPV should beconsidered for many patients with progressiverespiratory insufficiency, particularly those withrespiratory failure as the result of COPD exacerbationor cardiogenic pulmonary edema who havesuitably alert mental status and hemodynamicstability. Caveats regarding patient selection andearly monitoring and management for the responseto NPPV are discussed.A strategy for providing full ventilatory supportthat will provide adequate ventilation andoxygenation is usually sufficient. However, somepatients with respiratory failure have ARDSor have significant airflow obstruction from anexacerbation of asthma or COPD, conditions thatrequire specific ventilatory strategies that differsubstantially based on difference in pulmonaryphysiology. Ventilatory strategies for severe airflowobstruction and for ALI/ARDS are compared inTable 4.MV During Airflow ObstructionStatus asthmaticus as well as acute exacerbationof COPD is associated with marked limitationof expiratory airflow. Such a state increases thelikelihood of developing dynamic hyperinflation(increased end-inspiratory pressure and volume)and auto-PEEP (increased lung volume andintrathoracic pressure at the end of exhalation) asdiscussed previously. Additionally, such hyperinflationincreases the likelihood of developingbarotrauma. Ventilatory management includesadjusting the ventilator to prevent development ofdynamic hyperinflation and auto-PEEP in additionto the early recognition and timely managementof worsening of airflow obstruction. Key tenets ofmanagement include providing sufficient sedation(and possibly neuromuscular blockade) to controltachypnea, reducing the set respiratory rate (ie, f) toprovide an expiratory phase of sufficient durationto allow near-complete emptying, and using smallVt. Of note, reducing the respiratory rate from 20breaths/min to 10 breaths/min will more thandouble the expiratory time. Examination of theflow vs time graphics and examining the patient’schest for continued wheezing as one progressivelyincreases the expiratory time can help identifywhen flow is near completion. Minute ventilationdecreases significantly with these measures anddevelopment of hypercapnia is commonplace.Interestingly, because reductions in Vt and f canreduce the severity of alveolar hyperinflationand dead space ventilation, the increase in Pco 2can be blunted as the number of perfused alveoliincreases. Many clinicians have become comfortablewith permissive hypercapnia with a pH inthe 7.2 to 7.1 range, although a safe lower limit isnot well defined. Follow-up of patients who wereallowed to remain hypercapnic showed no longtermdeleterious effects. Permissive hypercapniais contraindicated in patients with intracranialhypertension (in whom hypercapnia could inducecerebral vasodilation and worsen the intracranialpressure) and in patients with an irritable myocardium(in whom hypercapnia may induce cardiacarrhythmias). Some clinicians administer IVbicarbonate in an attempt to maintain a pH 7.2 orso, but others do not. If a pressure-targeted modeis used, care must be taken to monitor Vt, for asthe airflow obstruction improves, Vt is likely toincrease significantly.MV for ALI/ARDsConsiderable experimental evidence from variousanimal models has demonstrated that largerVts and higher plateau or transpulmonary pressuresare associated with ALI. CT examination of<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 469ACC-PULM-09-0302-028.indd 4697/10/09 8:52:29 PM


Table 4. Comparison of Ventilatory Strategies for Severe Airflow Obstruction and for ALI/ARDS Using ARDSNet Guidelines or AlternativeStrategies Supported by Results of RCTsALI/ARDSVariablesSevere AirflowObstructionARDSNet StrategyAlternative StrategyFio 2To achieve Spo 2of 88–93%or Pao 2of 55–80 mm HgTo achieve Spo 2of 88–93% orPao 2of 55–80 mm Hg*To achieve Spo 2of 88–93% or Pao 2of55–80 mm HgMode of ventilation Varies AC mode AC modePressure or volume Either Volume PressureTarget Vt 8 mL/kg PBW † 6 mL/kg PBW 6 mL/kg PBWTotal rate, breaths/min Low (10–12) 25 (to 35 if needed) 25 (to 35 if needed)I/E target 1:5 1:2–1:1 1:3–1:1PEEP Minimal ‡ Modest; use PEEP-Fio 2* More aggressive; use PEEP-Fio 2*Target Pplat 30 cm H 2O 30 cm H 2O 40 cm H 2ORecruitment maneuvers No No Yes, after ventilator disconnectionSedation/NMBYes, sufficient to preventtachypnea; avoid NMB ifpossiblePermissive hypercapnia Ok; Pco 2may be very high Ok; pH 7.3 Ok; pH 7.3YesYes (less sedation if bilevel mode isused)*See Table 5.†Male PBW (in kilograms) 50 2.3 (height in inches 60); Female PBW (in kilograms) 45.5 2.3 (height in inches 60).‡Consider adding PEEP if significant auto-PEEP is present or there is WOB from ineffective triggering dyssynchrony. PEEPshould be set to about 2 cm H 2O less than auto-PEEP.the lungs of patients with ARDS reveals a smallvolume of normal or atelectatic lung, supportingthe notion of “baby”-sized lungs in ARDS. Someearlier RCTs demonstrated improved outcomes inARDS when patients were ventilated with lowerVts (and other ventilatory adjustments) comparedwith conventional volumes, but in some trials, nodifference was found.In 2000, publication of the result of the ARDSNetwork pivotal RCT that documented significantlylower rates of mortality among patients whoreceived Vt of 6 mL/kg predicted body weightcompared with those who received 12 mL/kg hasset the standard for MV for ALI/ARDS. There werealso more ventilator-free and organ failure-freedays but no difference in the frequency of barotraumanor in oxygenation. The key componentsof the National Heart, Lung and Blood InstituteARDS Clinical Network (ARDSNet) approach toventilating patients with ALI/ARDS are summarizedin Table 4. The ARDSNet approach to MV forARDS has been widely endorsed as an evidencebasedcornerstone of management of critically illpatients. However, several groups of investigators,including ARDSNet investigators, have demonstratedin subsequent real-life studies years afterpublication of ARDSNet results that clinicians arenot using low Vts. In one study in which threecenters were examined, only 16% of patients hadVt 8 mL/kg predicted body weight (PBW), andvirtually none were 6 mL/kg. Barriers to adoptioninclude lack of willingness to relinquish ventilatorcontrol, recognition of ARDS, and concernsfor contraindications, discomfort, and impairedgas exchange. Improved compliance with low Vtventilation was strongly associated with use of awritten protocol in a one study.On the basis of an earlier study by Amato andcoworkers, more recent smaller studies by Ranieriet al and Villar et al, as well as other evidence, investigatorshave addressed some of the finer points ofMV for ARDS/ALI, including the use of greaterPEEP, use of a pressure-targeted rather than a volume-targetedmode, and the addition of recruitmentmaneuvers. There are now three large-scale ( 500subjects in each) RCTs that address using a greaterPEEP level—stratified by oxygenation—compareddirectly with the original ARDSNet study parametersand PEEP levels. In all studies, PEEP averageapproximately 8 to 10 cm H 2O in traditionalARDSNet ventilation and 10 to 16 cm H 2O withthe greater PEEP strategy. All groups received Vtsof 6 mL/kg PBW. In the Assessment of Low TidalVolume and Increased End-Expiratory Volume to470 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4707/10/09 8:52:29 PM


Obviate Lung Injury trial, Brower and the ARDSNetinvestigators demonstrated better oxygenation butno difference in rate of mortality, ventilator-freedays, or barotrauma.In 2008, Meade and colleagues compared pressure-targetedventilation with allowable Pplatup to 40 cm H 2O, greater PEEP, and recruitmentmaneuvers to ARDSNet ventilation. The experimentalapproach resulted in better oxygenation,including less mortality caused by refractoryhypoxemia, and less frequent use of rescue therapybut no difference in rate of mortality, duration ofMV, or incidence of barotrauma. Also, in 2008,Mercat and coworkers compared an “increasedrecruitment” strategy with sufficient PEEP appliedto increase Pplat to 28 to 30 cm H 2O vs ARDSNetventilation. The increased recruitment grouphad better oxygenation, fewer cases of refractoryhypoxemia, less need for rescue therapy, andmore organ failure-free and ventilator-free days.No difference in barotrauma or rate of mortalitywas noted, although there was a trend (p 0.06)for lower mortality with more severe hypoxemia.The approach used by Mercat et al probably causedalveolar overdistention in patients with milderlung injury and lower levels of PEEP and shouldprobably be avoided in this subgroup.We conclude that greater PEEP levels, such asguided by PEEP-Fio 2in Table 5, should be considered,particularly if more severe hypoxemiais present. Recent work suggests that esophagealpressure monitoring can help individualize settingPEEP that improves gas exchange. Evidence for useof recruitment maneuvers is less compelling, withonly transient improvement in oxygenation andpotential for transient hemodynamic compromisein one large RCT. Use of a recruitment maneuver(typically 40 cm H 2O inflation held for 40 s) seemsmost likely to help when performed after ventilatordisconnection and loss of alveolar distention.Adjunctive and Alternative Therapy forRefractory Hypoxemia and ARDSAs many as one third of patients with ARDSreceive adjunctive therapy, such as inhaled nitricoxide and prone positioning for refractory hypoxemia.Additionally, novel ventilatory approaches,including high-frequency oscillatory ventilation(HFOV), may be considered. Inhaled nitric oxide(INO) is a vasodilator that, when administeredin very small doses (1 to 20 ppm) by inhalation,can increase perfusion to aerated alveoli withoutcausing systemic vasodilatation. As a result aboutthree fourths of ARDS patients will experience a 10% increase in Pao 2. Although INO can occasionallyhave dramatic effects in selected cases,its effects are often relative short-lived (48 h), andlarge RCTs have failed to show benefits in rates ofmortality or duration of MV.Prone positioning is accomplished by transferringthe mechanically ventilated patient for supineto prone position where they remain for 8 to12 hor longer. Prone positioning promotes a more-evendistribution of ventilation relative to perfusion,removes the weight of the heart from compressinglung tissue, and aids in respiratory secretion clearance.Oxygenation is improved in approximatelythree fourths of ARDS patients, but Pao 2revertsto near-baseline values when patients resume thesupine position. Outcomes studies demonstrateno benefit in rate of mortality or duration of MV,and the use of greater doses of sedation and neuromuscularblockade, risk for tube or catheter dislodgment,and potential for pressure sores must beconsidered. HFOV has been investigated as the ultimateform of lung-protective ventilation becausevery small Vts are delivered 180 to 300 times/min.In contrast to high-frequency jet ventilation, HFOVventilation is delivered with active inspiration andactive exhalation. Its management can be complexand subjective. Troubleshooting of patient deteriorationis more complex than with conventionalventilation, as is ventilator adjustment. HFOV canlead to impressive improvement in gas exchangein selected cases, but although HFOV has beendemonstrated in a multicenter RCT to be equivalentto convention ventilation for ARDS, it has not beencompared with low Vt ventilation. More researchis needed to confirm its role.Weaning and Liberation FromVentilationWeaning and Liberation: Evolving ConceptsTraditionally, separation of the patient fromMV has primarily been a process of progressivelyreducing the proportion of ventilationthat is supported artificially, ie, weaning from the<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 471ACC-PULM-09-0302-028.indd 4717/10/09 8:52:30 PM


Table 5. PEEP-FIO 2Values for Table 4*Variables 0.3 0.4 0.4 0.5 0.5 0.6 0.7 0.7 0.8 0.9 1.0 1.0ARDSNet protocolPEEP, cm H 2O 5 5 8 8 10 10 10 14 14 14 18 24Alternative approach †PEEP, cm H 2O 5–10 10–18 18–20 20 20 20–22 22Fio 2*PEEP is increased in increments of 2 cm H 2O; Fio 2increased by 0.1.†From Meade et al. JAMA 2008; 299:637–645.ventilation, while increasing the WOB that thepatient undertakes until it is judged that the patientis capable of breathing independently, at whichtime the ET tube is removed. Different approacheswere used, most commonly progressive reducingin PS in the PSV mode, decreasing the mandatorybreath rate in SIMV mode, or increasing the durationof spontaneous breathing trials in AC.As a result of clinical trial results, clinicians nowfocus on identifying the patient who, as judged bysuccessfully passing a number of conditions, has ahigh likelihood of breathing independently off theventilator and tolerating removal of the ET tube,without the process of progressively reducing support.Historically, measures of respiratory musclestrength, endurance, and gas exchange such as negativeinspiratory force, vital capacity, minute ventilation,maximum voluntary ventilation, respiratoryrate, or f/Vt, were relied on. Effective programsto liberate the patient from MV incorporate evidence-basedstructured techniques, thus improvingfrequency and consistency of measurements andactions, are multidisciplinary and team based,incorporate criteria to proceed to the next step (thuseliminating unnecessary delays), integrate clinicaljudgment, and have been demonstrated to improveoutcomes. The concepts and criteria are based onthe pathophysiology of respiratory failure and alsorecognize the importance of nonpulmonary organdysfunction in readiness to proceed.Protocol Based Liberation from MVIn 1996, Ely and colleagues developed a multistepprotocol to identify resolution of respiratoryfailure and in a RCT demonstrated faster weaning,shorter duration of ventilation, and fewercomplications (unplanned extubation, prolongedventilation, reintubation, and tracheostomy).This protocol incorporated daily screening of allmechanically ventilated patients by respiratorytherapists for hemodynamic stability, cessationof sedative infusions, adequate oxygenation, lowPEEP level, adequate cough, and adequate respiratorymuscle strength as judged by f/Vt 105while breathing without support. Patients whosuccessfully passed this screening protocol underwenta 2-h spontaneous breathing trial (SBT) with5 cm H 2O CPAP or T tube. Those who passed weredeemed likely (85% likelihood) to be successfullyextubated. This protocol has served as a basic templatefor many weaning protocols (specific criteria)and structured approaches (daily performance byrespiratory therapists and nurses). Most, but notall, published reports of such protocols have demonstratedshorter duration of ventilation, shorterICU length of stay (LOS), cost savings, and/orother benefits. It is worth noting, however, thatstrict adherence to overly conservative criteria canactually delay weaning and/or extubation.Evidence-Based Consensus Recommendations: InDecember 2001, a collective task force composedof representatives from the American College ofChest Physicians, American Association for RespiratoryCare, and Society of Critical Care <strong>Medicine</strong>published evidence-based guidelines for weaningand discontinuation of MVS. Table 6 summarizeskey recommendations regarding all aspects of theprocess. They offer specific recommendations forassessment of discontinuation potential (ie, screening)as follows: (1) evidence for some reversal ofthe underlying cause for respiratory failure; (2) adequateoxygenation (Pao 2/Fio 2 150 to 200, PEEP 5 to 8 cm H 2O, Fio 2 0.4 to 0.5), and pH 7.25;472 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4727/10/09 8:52:30 PM


Table 6. Components of Protocols for Discontinuation ofVentilationParametersMedical stabilityMental statusOxygenationLung mechanicsVentilatoryenduranceSecretion clearanceAirway patencyMiscellaneousfactorsMeasuresPresence of shock, use ofvasopressors, pHUse of continuous sedativeinfusion, sedation scale criteriaFio 2, PEEP, Pao 2/Fio 2ratiof/Vt, pHSBTCough strength, sputum volume/suction frequencyCuff-leak testCondition improving?(3) hemodynamic stability (no active myocardialischemia or clinically significant hypotension); and(4) capability to initiate an inspiratory effort. Theserecommendations serve as a solid basis for currentrecommendations; however, newer data allow us toreaddress these and other parameters.Components of Protocols to AssessDiscontinuation of VentilationThere are many factors that can influence thesuccess a patient has for tolerating withdrawalof ventilatory support and removal of the ETtube. Some of the parameters and criteria that canaffect this process are listed in Table 6, and severaldeserve additional comment. There is considerableevidence that mental status is an important determinantof timely recovery from MV because overlysedated patients may have an impaired driveto breathe and an inability to protect the airwayonce the ET tube is removed. Level of consciousis assessed in protocols indirectly by the use ofcontinuous infusion of sedatives or directly withthe use of a sedation scale.Shorter duration of MV has been demonstratedin numerous prospective studies that use strategiesto reduce oversedation, including daily interruptionof sedation, intermittent (rather than continuous)sedative and opioid administration, and matchingpatient characteristics to drugs. In one study, thecombination of a daily awakening trial plus a SBTwas superior to SBT alone. The combination of poorcough, high sputum volume, and poor mental statusis a particularly worrisome combination, even if thepatient is capable of passing an SBT. In one study,80% of patients who met this combination of findingswere reintubated. Adequacy of oxygenationafter extubation is important to predict. However,some protocol criteria may inadvertently delay extubationif the criteria are too restrictive. For example,some protocols require Fio 2 0.4, PEEP 5 cmH 2O, and/or Pao 2/Fio 2 200, yet many patientswho have Po 2 95 mm Hg on Fio 2of 0.5 will haveadequate oxygenation postextubation. In fact, oxygenationcriteria (Pao 2/Fio 2 180 mm Hg) was themost common weaning criterion never passed ina large cohort of patients who were successfullyextubated. In 1991, Yang and Tobin found f/Vt tobe a reliable measure of lung mechanics and goodpredictor of successful discontinuation. However,some patients, particularly smaller individuals andthose with parenchymal lung disease, may have amore rapid, shallow pattern of breathing. As notedpreviously, investigators have demonstrated thatincorporating f/Vt 105 as a screening criteriondelayed weaning. Some protocols use greater thresholds(ie, 125) or omit f/Vt from their protocol.SBT: The SBT is widely considered to be thepivotal test, incorporating the assessment of lungmechanics and endurance. The patient is placed onlow-level ventilatory support intended to overcomethe added WOB of breathing through the ET tube,or without any support for a period of 30 to 120 minand observed for deterioration in vital signs or ventilatoryfailure. This is however, considerable variabilityin several parameters with different criteriafor the mode and settings of ventilatory support,duration of SBT, and termination (failure) criteria.Although T tube, “flow-by,” CPAP, or PSV is oftenused as low-level ventilatory support, automatictube compensation (ATC) is emerging as a form of“smart” PSV. ATC is designed to provide enough PSto overcome the resistance of breathing through anartificial airway. This pressure is recalculated every5 ms and is based on the tube size and type (ET vstracheostomy) plus instantaneous flow measurement.Thus, greater pressure is provided if the tubehas a smaller diameter or if flow is faster. In onestudy, ATC was superior to CPAP for overall success(SBT completed and extubation for 48 h).Airway Patency and Postextubation Stridor:Although clinicians have long recognized the importanceof postextubation stridor in extubation failure,recent research has helped clarify the importance<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 473ACC-PULM-09-0302-028.indd 4737/10/09 8:52:30 PM


and validated tools for assessment. Postextubationstridor is observed in 5 to 30% of patients, dependingon the patient population. Those who developstridor tend to be sicker, have longer duration intubation,and are more likely to have had a difficultintubation and/or self-extubation. By deflating theET tube cuff and detecting airflow around the tube(either by measurement using comparison of volumeexhaled through the tube before and after cuffdeflation, or simply hearing a leak), one increasesthe ability to predict postextubation stridor. “Cuffleak” 18% was identified as being a good predictor(85% sensitive, 72% specific) of postextubationstridor in one study, although thresholds of 15 to25% have been identified by others. Administrationof methylprednisolone (MP) (40 mg q6h 4 doses)to patients with poor cuff-leak ( 24%) was effectivein reducing likelihood of postextubation stridor (6%vs 30% with no MP) in one study. In another study,giving MP at a dose of 20 mg q4h 4 doses beforeextubation was effective in reducing postextubationstridor for unselected patients.Recommendations for Weaning and LiberationAll endotracheally intubated mechanicallyventilated patients should undergo daily evaluationto determine their potential for discontinuationof ventilation. This should be a structured,evidence-based protocol that incorporates keycomponents in a multidisciplinary team-basedapproach that promotes efficiency. Interestingly,computerized weaning that is integrated withinmechanical ventilators is being explored as a meansof standardizing and streamlining the process,with shorter duration of MV and shorter ICU LOSdemonstrated in one study. The author’s recommendationsfor specific components and caveatsrelated to these parameters are listed in Table 7. Itis worth acknowledging that knowledge regardingweaning and extubation continuous to evolve, thatsome recommendations are based on opinion andlower-level evidence, and that clinical skills play animportant and complementary role. An importantcomponent of the weaning process is to recognizethat the patient who fails this evaluation repeatedlyand/or who fails extubation should undergoevaluation for possible causes of failure, such asthose listed in Table 8. Finally it is worth consideringextubation followed by immediate institutionof NPPV for the patient with COPD who hasfailed SBT. In a small RCT, NPPV was superior toconventional management for patients who failedthree SBTs, in regards to duration of MV, ICU LOS,and hospital LOS. However, one must carefullyconsider reasons for SBT failure and troubleshootpotential causes before using this approach.TracheostomyTracheostomy has been available for manyyears as a surgically placed artificial airway thatpermits longer-term placement compared with anET tube. Tracheostomy is considered to be morestable, more comfortable, and is associated withlower doses of sedative and opioid medications.Bedside percutaneous dilatational tracheostomy(PDT) is achieving widespread acceptance as analternative to traditional approaches in which anopen surgical technique is performed in the operatingroom. Meta-analyses of prospective studiesindicate that PDT is associated with less infection,scarring, and expense, with a trend for fewer complicationsbut greater rates of obstruction/decannulation.The optimal timing of tracheostomy remainsunsettled, although a small meta-analysis indicatedshorter duration of MV and shorter ICU LOS with“early” tracheostomy. Some experts favor early tracheostomyfor patients with Acute Physiology andChronic Health Evaluation II 25 or combinationof impaired mental status and poor oxygenationbecause these groups tend to have long courses ofrespiratory failure. Discontinuation of MV in thepatient with a tracheostomy is generally by traditionalweaning. This is typically accomplished byprogressively increasing the duration of periodswithout ventilatory support, often using oxygendelivered directly via the tracheostomy tube, andomitting the ventilator circuit.Summary and ConclusionsMV is a life-saving modality for treatingrespiratory failure and to providing gas exchangesupport. There are many subtleties in effectivemanagement that include adopting a patientfocusedapproach, using evidence-based strategiesfor MV as well as for discontinuation of ventilation,and avoiding complications related to MV and theartificial airway.474 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4747/10/09 8:52:31 PM


Table 7. Recommendations for Evaluation of the Ability To Tolerate Discontinuation of Ventilation and Extubation*Screening Criteria1. Is patient hemodynamically stable (not in shock, no more than minimal vasopressor requirement)?2. Is underlying cause of respiratory failure improving?3. Is the patient alert or have easy arousal (opens eyes to voice) and cognition (follows simple commands)?Perform daily interruption of sedation in patients in whom this is feasible.4. Is cough strength moderate or better and suctioningfrequency no more than every 2 h?5. Is Pao 2/Fio 2ratio 150 and PEEP 8 cm H 2O?6. Is f/Vt ratio 125? Consider eliminating f/Vt ratioCuff-leak test1. Is air movement around ET tube with cuff deflated 15%? If cuff leak is 25%, consider administeringMP (see text)SBT1. Does patient tolerate breathing on minimal support for 60 min? Consider using ATC as mode of ventilatorysupport during SBT, if available*All intubated patients receiving MV should undergo screening. Those who satisfy the criteria but fail mentalstatus testing (screening criterion 3) should have sedation interrupted (ie, discontinue all IV sedative andopioid drugs) and repeat testing for screening criteria. Patients who satisfy the screening criteria shouldundergo a cuff-leak test. If the cuff leak is 15%, MP should be administered (four doses of 40 mg IVevery 6 h or 20 mg IV every 4 h). Patients who satisfy the screening criteria and who pass the cuff leak testshould undergo SBT testing; if patients pass the SBT, they should be extubated.Table 8. Modifiable Factors Related to Failure of Weaning From MVUnresolved precipitating processReversible airflow obstructionIncreased endotracheal tube resistanceExcessive respiratory secretionsImpaired mental statusRespiratory depressant medicationsExcessive sedative and opioid medicationsMetabolic alkalosisElectrolyte imbalanceHemodynamic instabilityIschemic heart diseaseInfection and sepsisMalnutritionOverfeedingNeuromuscular disordersPsychological factorsAnnotated BibliographyAcute Respiratory Distress Syndrome Network. Ventilationwith lower tidal volumes as compared with traditionaltidal volumes for acute lung injury and the acuterespiratory distress syndrome. N Engl J Med 2000;342:1301–1308Large, multicenter, randomized sentinel trial with 861 patientsshowed that patients receiving MV with low Vt (6.2 mL/kgideal weight, Pplat 30 cm H 2O) had significantly lower ratesof mortality (p 0.007) than patients receiving MV withconventional Vt (12 mL/kg, Pplat 30cm H 2O) as well asfewer days receiving MV.Acute Respiratory Distress Syndrome Network. Higherversus lower positive end-expiratory pressures in patientswith the ARDS. N Engl J Med 2004; 351:327–336Greater PEEP levels improved oxygenation but did notimprove survival in this controlled, multi-institutional, randomizedtrial.Adhikari NK, Burns KE, Friedrich JO, et al. Effect ofnitric oxide on oxygenation and mortality in acutelung injury: systematic review and meta-analysis. BMJ2007; 334:779Meta-analysis: no mortality benefit, better oxygenation, butmore renal failure with INO.Amato MBP, Barbas CSV, Medeiros DM, et al. Beneficialeffects of the “open lung approach” with low distendingpressures in ARDS. Am J Respir Crit Care Med 1995;152:1835–1844A controlled trial that suggests that maintaining alveolarpatency while avoiding alveolar overdistention results inimproved outcome for patients with ARDS. This was an impetusto the ARDS Network trial of low Vt ventilation.Brochard L, Rauss A, Benito S, et al. Comparison of threemethods of gradual withdrawal from ventilatory supportduring weaning from mechanical ventilation. Am JRespir Crit Care Med 1994; 150:896–903<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 475ACC-PULM-09-0302-028.indd 4757/10/09 8:52:31 PM


This study concludes that SIMV is not a preferred weaningmode.Brower RG, Morris A, MacIntyre N, et al. Effects ofrecruitment maneuvers in patients with acute lung injuryand acute respiratory distress syndrome ventilated withhigh positive end-expiratory pressure. Crit Care Med2003; 31:2592–2597As studied, recruitment maneuvers did not accomplish muchin this ARDSNet substudy.Cheng KC, Hou CC, Huang HC, et al. Intravenous injectionof methylprednisolone reduces the incidence ofpostextubation stridor in intensive care unit patients.Crit Care Med 2006; 34:1345–1350Four doses of MP, 40 mg q6h, significantly reduced postextubationstridor.Cohen JD, Shapiro M, Grozovski E, et al. Extubationoutcome following a spontaneous breathing trial withautomatic tube compensation versus continuous positiveairway pressure. Crit Care Med 2006; 34:682–686Automatic tube compensation (“smart” PSV) during SBTwas better than CPAP for overall weaning success.Dellinger RP, Zimmerman JL, Taylor RW, et al. Effects ofinhaled nitric oxide in patients with acute respiratorydistress syndrome: results of a randomized phase II trial.Inhaled Nitric Oxide in ARDS Study Group. Crit CareMed 1998; 26:15–23Multicenter RCT: better oxygenation, not much else, with INO.Derdak S, Mehta S, Stewart TE, et al. High-frequencyoscillatory ventilation for acute respiratory distress syndromein adults: a randomized, controlled trial. Am JRespir Crit Care Med 2002; 166:801–808HFOV was equivalent to conventional but greater Vt ventilationfor ARDS. This study needs repeating with a low Vtstrategy.Ely EW, Baker AM, Dunagan DP, et al. Effect on thedura tion of mechanical ventilation of identifyingpatients capable of breathing spontaneously. N EnglJ Med 1996; 335:1864–1869Classic RCT demonstrating benefit of identifying whenpatients are ready to be liberated from the ventilator.Esteban A, Frutos F, Tobin MJ, et al. A comparison of fourmethods of weaning patients from mechanical ventilation:Spanish Lung Failure Collaborative Group. N EnglJ Med 1995; 332:345–350SIMV is not a preferred weaning mode.Esteban A, Frutos-Vivar F, Ferguson ND, et al. Noninvasivepositive-pressure ventilation for respiratory failureafter extubation. N Engl J Med 2004; 350:2452–246RCT demonstrated a greater rate of mortality with NPPV afterextubation failure. Patients with COPD were excluded.Ferrer M, Esquinas A, Leon M, et al. Noninvasive ventilationin severe hypoxemic respiratory failure: a randomizedclinical trial. Am Rev Respir Crit Care Med 2003;168:1438–1444NPPV can reduce the rate of mortality in some hypoxemicrespiratory failure groups.Gattinoni L, Tognoni G, Pesenti A, et al. Effect of proneposition on the survival of patients with acute respiratoryfailure. N Engl J Med 2001; 345:568–573Randomized, unblinded, controlled trial of 304 patients failedto show improvement in 10-day or 6-month mortality. Thestudy was criticized for using only 6 h/d of prone positioningin many patients.Girard TD, Kress JP, Fuchs BD, et al. Efficacy and safetyof a paired sedation and ventilator weaning protocol formechanically ventilated patients in intensive care (Awakeningand Breathing Controlled trial): a randomisedcontrolled trial. Lancet 2008; 371:126–134Combination of daily interruption of sedation and SBT betterthan SBT alone.Griffiths J, Barber VS, Morgan L, et al. Systematic reviewand meta-analysis of studies of the timing of tracheostomyin adult patients undergoing artificial ventilation.BMJ 2005; 330:1243Meta-analysis of timing of tracheostomy.Huang C-J, Lin H-C. Association between adrenal insufficiencyand ventilator weaning. Am J Respir Crit CareMed 2006; 173:276–280Treatment of patients who were found to have adrenal insufficiencywith hydrocortisone resulted in significantly shorterduration of weaning and improved weaning success.International consensus conferences in intensive caremedicine: ventilator-associated lung injury in ARDS: thisofficial conference report was cosponsored by theAmerican Thoracic Society, The European Society ofIntensive Care <strong>Medicine</strong>, and The Societé de Réanimationde Langue Française, and was approved by the ATS<strong>Board</strong> of Directors, July 1999. Am J Respir Crit Care Med1999; 160:2118–2124Detailed review of experimental evidence, risk factors, monitoring,and prevention of VALI.Kress JP, Pohlman AS, O’Connor MF, et al. Daily interruptionof sedative infusions in critically ill patientsundergoing mechanical ventilation. N Engl J Med 2000;342:1471–1477Daily interruption of continuous IV sedation significantlyshortened length of MV and days in the ICU.Lightowler J, Wedzicka J, Elliot M, et al. Non-invasivepositive pressure ventilation to treat respiratory failureresulting from exacerbations of chronic obstructive476 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4767/10/09 8:52:31 PM


pulmonary disease: Cochrane systematic reviews andmeta-analysis. BMJ 2003; 326:184–189This Cochrane analysis solidifies the role of NPPV inCOPD.MacIntyre NR, Cook DJ, Ely EW, et al. Evidence-basedguidelines for weaning and discontinuing ventilatorsupport. Chest 2001; 120:375S–395SProvides basis for approaches to weaning.Meade MO, Cook DJ, Guyatt GH, et al. Ventilation strategyusing low tidal volumes, recruitment maneuvers,and high positive end-expiratory pressure for acute lunginjury and acute respiratory distress syndrome: a randomizedcontrolled trial. JAMA 2008; 299:637–645Recent large Canadian RCT that showed that pressure-targeted,liberal PEEP, recruitment, and more liberal Pplat whilekeeping Vt at 6 mL/kg PBW was as good as traditional ARD-SNet ventilation for mortality and duration of MV but hadbetter oxygenation.Mercat A, Richard JC, Vielle B, et al. Positive end-expiratorypressure setting in adults with acute lung injury andacute respiratory distress syndrome: a randomized controlledtrial. JAMA 2008; 299:646–655Recent large RCT demonstrating more ventilator-free days inaddition to better oxygenation and less requirement for “rescuetherapy” for hypoxemia; no clear mortality benefit was demonstrated.Nava S, Ambrosino N, Clini E, et al. Noninvasive mechanicalventilation in the weaning of patients withrespiratory failure due to chronic obstructive pulmonarydisease: a randomized, controlled trial. Ann InternMed 1998; 128:721–728This sentinel study demonstrated the superiority of NPPVover MV through an ET tube in COPD patients with hypercapnicrespiratory failure.Nava S, Carbone G, DiBattista N, et al. Noninvasive ventilationin cardiogenic pulmonary edema: a multicenterrandomized trial. Am Rev Respir Crit Care Med 2003;168:1432–1437This study established that NPPV could improve oxygenationand reduce the need for intubation in patients with cardiogenicpulmonary edema. However, the rate of mortality was notaffected nor was it worse with bilevel positive airway pressurevs CPAP.Putensen C, Mutz NJ, Putensen-Himmer G, et al. Spontaneousbreathing during ventilatory support improvesventilation-perfusion distributions in patients with acuterespiratory distress syndrome. Am J Respir Crit CareMed 1999; 159:1241–1248Improved gas exchange and hemodynamics with bilevel ventilation(called APRV using 1:1 I:E ratio).Ranieri VM, Suter PM, Tortorella C, et al. Effect ofmechanical ventilation on inflammatory mediators inpatients with acute respiratory distress syndrome: arandomized controlled trial. JAMA 1999; 282:54–61Combination of low Vt and high PEEP was associated withdecreases in inflammatory mediators in BAL fluid and blood,compared with conventional ventilation in ARDSRubenfeld GD, Cooper C, Carter G, et al. Barriers toproviding lung-protective ventilation to patients withacute lung injury. Crit Care Med 2004; 32:1289–1293We do not follow low Vt ventilation recommendations andhere’s why.Sessler CN. Mechanical ventilation of patients with acutelung injury. Crit Care Clin 1998; 14:707–729<strong>Review</strong> of the topic is presented.Sessler CN, Varney K. Patient-focused sedation andanalgesia in the ICU. Chest 2008; 133:552–565Comprehensive recent review of sedation and analgesia inpatients on MV.Talmor D, Sarge T, Malhotra A, et al. Mechanical ventilationguided by esophageal pressure in acute lunginjury. N Engl J Med 2008; 359:2095–2104In a pilot RCT, guidance of setting PEEP by measured esophagealpressure as a surrogate for pleural pressure to estimatetransalveolar pressure led to better oxygenation and lungcompliance without increasing barotrauma. Although thereare technical concerns regarding accuracy of esophageal pressure,the concept of individualized setting of PEEP (ratherthan by a table) is attractive.Tanios MA, Nevins ML, Hendra KP, et al. A randomized,controlled trial of the role of weaning: Predictorsin clinical decision making. Crit Care Med 2006;34:2530–2535Addition of f/Vt (rapid shallow breathing index) to weaningprotocol slowed weaning by 1 day, although duration of MVwas identical as extubation failure.Taylor RW, Zimmerman JL, Dellinger RP, et al. Lowdoseinhaled nitric oxide in patients with acute lunginjury: a randomized controlled trial. JAMA 2004;291:1603–1609Five ppm of INO improved oxygenation for 24 h, but no outcomesbenefits were seen.Umoh NJ, Fan E, Mendez-Tellez PA, et al. Patient andintensive care unit organizational factors associated withlow tidal volume ventilation in acute lung injury. CritCare Med 2008; 36:1463–1468Use of a written protocol was highly associated with improvedcompliance with low Vt ventilation.Varelmann D, Wrigge H, Zinserling I, et al. Proportionalassist vs pressure support ventilation in patients with<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 477ACC-PULM-09-0302-028.indd 4777/10/09 8:52:31 PM


acute respiratory failure: cardiorespiratory responsesto artificially increased ventilatory demand. Crit CareMed 2005; 33:1968–1975PAV, even when combined with an automatic tube compensationprogram, did not reduce WOB significantly better thanPSV.Villar J, Kacmarek RM, Perez-Mendez L, et al. A highpositive end-expiratory pressure, low tidal volume ventilatorystrategy improves outcome in persistent acuterespiratory distress syndrome: a randomized, controlledtrial. Crit Care Med 2006; 34:1311–1318RCT of 108 patients showing better outcomes with low Vt andhigh PEEP.Vitacca M, Ambrosino N, Clini E, et al. Physiologicresponse to pressure support ventilation delivered beforeand after extubation in patients not capable of totallyspontaneous autonomous breathing. Am J Respir CritCare Med 2001; 164:638–641This study demonstrates that NPPV decreased diaphragmenergy expenditure to the same degree as applying PS in intubatedpatients.Walsh TS, Dodds S, McArdle F. Evaluation of simplecriteria to predict successful weaning from mechanicalventilation in intensive care patients. Br J Anaesth 2004;92:793–799Patients who are successfully weaning often do not have Pao 2:FIO 2> 180 mm Hg. Other parameters addressed as well.Ware LB, Matthay MA. The acute respiratory distresssyndrome. N Engl J Med 2000; 342:1334–1349This article reviews ARDS.Wheeler AP, Bernard GR. Acute lung injury and theacute respiratory distress syndrome: a clinical review.Lancet 2007; 369:1553–1564This article reviews ARDS.Yang KL, Tobin MJ. A prospective study of indexes predictingthe outcome of trials of weaning from mechanicalventilation. N Engl J Med 1991; 324:1445–1450This classic article defines the rapid shallow breathing index.Young M. Tidal volumes used in acute lung injury: whythe persistent gap between intended and actual clinicalbehavior? Crit Care Med 2006; 34:543–544Clinicians do not follow low Vt recommendations reliably.478 Mechanical Ventilatory Support (Sessler)ACC-PULM-09-0302-028.indd 4787/10/09 8:52:31 PM


Unusual and Uncommon <strong>Pulmonary</strong> DisordersJay H. Ryu, MD, FCCPObjectives:• Discuss several unusual and uncommon pulmonarydisorders• Describe clinical manifestations and facilitate recognitionof these disorders• Discuss management options for these disordersKey words: Birt−Hogg−Dubé syndrome; constrictivebron chiolitis; IgG4 sclerosing disease; pulmonary alveolarproteinosis; pulmonary amyloidosis; spontaneous pneumomediastinumRare (orphan) disease is defined as a disease orcondition affecting 200,000 persons in the UnitedStates. 1a An estimated 25 million people in theUnited States have 1 of the 6,000 rare diseases.The failure to diagnose a rare or unusual diseaseis generally attributable to the failure to considerthe diagnosis in the first place.The diagnosis of unusual pulmonary disordersis facilitated by considering a broad differentialdiagnosis at the outset and recognizing characteristicclinical context and imaging features associatedwith these disorders. Histopathology can behelpful but needs to be correlated with the clinicaland radiologic features in formulating the finalclinicopathologic diagnosis. Rational treatmentand accurate assessment of prognosis depend onan accurate diagnosis.IgG4-Related Sclerosing DiseaseIn 2001, Hamano et al 1a reported the findingsof high serum IgG4 concentrations and denselymphoplasmacytic infiltrates containing IgG4-positive plasma cells in the pancreas in 20 patientswith sclerosing pancreatitis (also called autoimmunepancreatitis). Subsequently, similar IgG4-relatedlesions were identified in other organs such as thebile duct, salivary gland, lacrimal gland, liver,kidney, retroperitoneum (retroperitoneal fibrosis),aorta (inflammatory aneurysm), as well as the lung.It is now known that IgG4-related sclerosing diseasecan be localized to one or two organs or presentwith systemic multisystem disease. 2,3IgG4-related sclerosing disease manifestssimilar histopathologic findings in different organsand is characterized by diffuse lymphoplasmacyticinfiltrate consisting of extensive IgG4-positiveplasma cells and T lymphocytes along with fibroticchanges. 2 Other histopathologic features mayinclude eosinophilic infiltration and obliterativeangiitis.Quantitatively, IgG4 is the smallest subclass ofIgG and accounts for only 3 to 6% of total IgG innormal serum. The IgG subclasses exhibit differencesin their effector functions. Involvement ofimmunologic mechanisms appears likely in IgG4-related sclerosing disease, but exactly what roleIgG4 plays in the pathogenesis and pathophysiologyof this disorder remains unclear. 2 The targetantigens in this inflammatory disorder and thetrigger for the serum IgG4 elevation also are notknown.IgG4 sclerosing disease occurs in adults, morecommonly men than women, with a median ageof 60 years (range, 17 to 80 years). 2,3 Manifestationsof IgG4-related sclerosing disease commonlyinclude pancreatic involvement, but extrapancreaticlesions without pancreatic involvement canoccur. 2,3 Systemic symptoms including fever alsomay be seen.Intrathoracic manifestations of IgG4-relatedsclerosing disease may include focal parenchymallesions (eg, inflammatory pseudotumor, plasma-cellgranuloma); interstitial infiltrates (organizing pneumonia,nonspecific interstitial pneumonia patterns);mediastinal/hilar lymphadenopathy; and mediastinalfibrosis. 2,4 Malignancy, eg, lung cancer, may besuspected on initial presentation. Radiologically,the manifestations include solitary nodule or mass,round-shaped ground-glass opacities, bilateralinterstitial infiltrates consisting of reticular andground-glass opacities, and thickening of bronchovascularbundles and interlobular septa. On18F-fluorodeoxyglucose PET scanning, abnormal<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 479ACC-PULM-09-0302-029.indd 4797/10/09 8:52:52 PM


18F-fluorodeoxyglucose uptake is observed in pancreaticand extrapancreatic lesions.Serum IgG4 levels and immunostaining of thebiopsy specimen with anti-IgG4 antibody are usefulin making the diagnosis. 2 Although the serum IgG4level was initially reported to be increased in nearlyall patients with autoimmune pancreatitis, 1a theprevalence of increased IgG4 levels appears to belower in patients presenting with various extrapancreaticlesions of IgG4-related sclerosing disease.In patients with intrathoracic abnormalitiesassociated with extrathoracic inflammatory orfibrotic lesions, particularly pancreatitis, the diagnosisof IgG4-related sclerosing disease should beconsidered. Even in the absence of extrathoracicmanifestations, some unusual cases of intrathoraciclesions such as mediastinal fibrosis and focal inflammatorylesions may be IgG4 related in origin.IgG4-related sclerosing disease responds wellto corticosteroid therapy. 2–4 The starting dose ofprednisone is typically 30 to 60 mg/d for 1 to 2weeks followed by gradual tapering of the doseover the subsequent several weeks to a low maintenancedose, eg, 10 mg/d. In the absence of recurrentdisease, corticosteroid therapy is eventuallydiscontinued.<strong>Pulmonary</strong> Alveolar Proteinosis<strong>Pulmonary</strong> alveolar proteinosis (PAP), alsoknown as pulmonary alveolar phopholipoproteinosis,is a rare parenchymal lung disease that is characterizedby accumulation of lipoproteinaceousmaterial in the alveoli. Most cases (90%) are idiopathic;the remaining cases consist of congenitaland secondary forms of PAP. 5–7 The congenital formis caused by mutations in the genes encoding surfactantproteins or the receptor for granulocytemacrophagecolony-stimulating factor (GM-CSF).Secondary PAP occurs in association with hematologicor immunodeficiency disorders, inhalationof chemicals or inorganic dusts (eg, silica), or certainpulmonary infections. PAP results fromimpaired surfactant metabolism by macrophagescaused by alterations in GM-CSF signaling. In mostcases of idiopathic PAP, anti-GM-CSF anti bodiesare detectable in the serum and BAL fluid.The median age at the time of diagnosis ofidiopathic PAP is 39 years; most patients are 20 to60 years of age. 5 Clinical features are nonspecific.Although most patients present with insidiousonset of exertional dyspnea and cough, somepatients may be asymptomatic. 5,6 Less commonsymptoms include fever, fatigue, weight loss, chestpain, and hemoptysis. Inspiratory crackles areheard in 20 to 50% of patients; digital clubbing isuncommon. 5,6Mildly increased levels of serum lactate dehydrogenaseare common. Increased levels of lungsurfactant protein A and D and mucin-like glycoprotein(KL-6) in the serum have also been foundin patients with PAP; however, these findings arenonspecific and not limited to this disease. 5,6Patients affected by PAP may have periodic acid-Schiff (PAS)-positive material or increased levelsof surfactant protein A in their sputum, but thesealso are nonspecific findings. Serologic testing foranti-GM-CSF autoantibodies is not a routinelyavailable assay.Chest radiography usually reveals bilateral,patchy airspace infiltrates. 8 Less commonly, aninterstitial pattern may be seen. Infiltrates often aremore prominent in the perihilar regions, and this“bat-wing” pattern may be mistaken for pulmonaryedema. High-resolution CT (HRCT) of thechest demonstrates ground-glass and/or consolidativeinfiltrates in patchy or diffuse distributions.Sharp demarcation of the infiltrates from surroundingnormal lung tissue is commonly observed.Reticular opacities or interlobular septal thickeningare present within the airspace infiltrates, creatinga “crazy-paving” pattern. 8,9 This pattern is characteristic,but not specific, for PAP, and it can be seenin diffuse alveolar damage superimposed on usualinterstitial pneumonia (“exacerbation” of idiopathicpulmonary fibrosis), acute interstitial pneumonia,ARDS, cardiogenic pulmonary edema, anddrug-induced lung disease. 8,9 <strong>Pulmonary</strong> functiontesting demonstrates a restrictive defect with areduced diffusing capacity. 6,10In an appropriate clinical setting, the diagnosisof PAP may be made by BAL fluid, which typicallyyields a milky effluent. 5,10 Under light microscopy,this fluid shows large amounts of PAS-positivelipoproteinaceous material. Transbronchial biopsyalso provides the diagnosis in most cases and demonstratesaccumulation of granular, PAS-positive,lipoproteinaceous material within the alveolarspaces with preserved alveolar architecture,although thickening of the alveolar septa and480 Unusual and Uncommon <strong>Pulmonary</strong> Disorders (Ryu)ACC-PULM-09-0302-029.indd 4807/10/09 8:52:53 PM


interstitial fibrosis may occur in some cases. 10 Inthe absence of a superimposed infection, very fewinflammatory cells are seen in the lung tissue. Surgicalbiopsy is now less frequently required toconfirm a PAP diagnosis.The treatment standard for PAP is whole-lunglavage, during which repeated instillation anddrainage of the lung with aliquots of isotonic salinesolution (up to a total volume of 20 to 30 L) is performeduntil the run off is clear. 5,6,10 This procedureis performed via a bronchoscope passed througha double-lumen endotracheal tube with the patientunder general anesthesia. Chest percussion is performedduring the lavage procedure to facilitateremoval of the lipoproteinaceous material. Themain indication for treatment is limiting exertionaldyspnea associated with hypoxemia. Approximately85% of patients show improvement afterlung lavage. The median duration of clinical benefitis approximately 15 months. Eventually, 60 to70% of patients require repeat lavage. Lung lavageis associated with improved survival. 5,6 Consistentwith its proposed importance in surfactant clearance,preliminary experience with GM-CSF administrationin PAP has shown favorable responses in40 to 50% of patients, with some patients experiencinga complete response. 10,11 GM-CSF has beenadministered subcutaneously (dose range, 5 to 20µg/kg/d) or via aerosolization (dose range, 250 to500 µg bid every other week). 6,12The clinical course of patients with PAP is variable;spontaneous resolution can occur as well asprogression to respiratory failure. 5,6 Five- and 10-year survival rates in PAP are 75% and 68%, respectively.5,6 Death caused by PAP is uncommon andusually related to respiratory failure or uncontrolledinfection. Superimposed infection, especiallywith opportunistic pathogens such asNocardia, may occur and disseminate. Lung transplantationis an option in those individuals withprogressive disease that is not adequately respondingto medical therapy. Recurrent PAP after doublelungtransplantation has been reported. 13<strong>Pulmonary</strong> AmyloidosisAmyloidosis is a heterogeneous group ofacquired or inherited diseases characterized bydeposition of amyloid in the extracellular tissuespace. Amyloid, a fibrillar, insoluble, proteinaceousmaterial, can deposit in the tracheobronchial treeor pulmonary parenchyma in either localized ordiffuse pattern. 14 This pulmonary involvement canoccur as part of a systemic process or be isolatedto the lung. 14,15The most frequent types of amyloidosis are theAL (primary) and AA (secondary) types. AL amyloidosiscan occur alone or in association withmultiple myeloma or another plasma-cell dyscrasia.AA amyloidosis occurs with chronic inflammatoryor infectious disorders. The frequency ofpulmonary involvement is low with secondary andfamilial amyloidosis, whereas pulmonary involvementby AL (primary) systemic amyloidosis isrelatively common.Patients with pulmonary amyloidosis with orwithout previous amyloidosis may present to pulmonologistsfor an evaluation of respiratory symptomsor abnormal findings on chest imaging.Laryngeal and tracheobronchial forms of pulmonaryamyloidosis may be seen as focal nodules/plaques or diffuse submucosal infiltration andusually occur as localized amyloidosis. Diffuseendobronchial involvement is recognizable bronchoscopicallyas shiny, pale plaques with scatteredfocal stenoses. Symptoms associated with airwayinvolvement depend on the extent of luminal compromiseand include dyspnea, cough, hemoptysis,wheeze, atelectasis, and recurrent pneumonias.<strong>Pulmonary</strong> parenchymal involvement manifestsas single or multiple nodules or diffuse parenchymalinfiltrates. The nodular form is oftendetected incidentally on chest radiography. Thesenodules may grow slowly, cavitate, or calcify. Adiffuse interstitial pattern is usually seen withsystemic AL amyloidosis. HRCT reveals bilateralnodular and/or reticular pattern with septal thickening.16 Pleural effusions may occur as the resultof pleural involvement or from heart failure causedby cardiac amyloidosis. Other pulmonary manifestationsare mediastinal and/or hilar adenopathy,mediastinal masses, and obstructive sleep apnearesulting from macroglossia.The diagnosis of pulmonary amyloidosis isachieved by obtaining biopsy material that revealsthe characteristic apple-green birefringence withpolarized microscopy after staining with Congored. 14,15 Tracheobronchial and diffuse parenchymalforms of pulmonary amyloidosis can be diagnosedby bronchoscopy, although the endoscopist must<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 481ACC-PULM-09-0302-029.indd 4817/10/09 8:52:53 PM


e prepared for potential patient bleeding. Lungnodules are diagnosed by needle aspiration biopsyor surgical resection. Diagnosis of pulmonaryamyloidosis should lead to an evaluation to identifythe underlying cause or condition leading toamyloid deposition, if it is not already known, aswell as the extent of involvement.The treatment of amyloidosis varies with differenttypes of pulmonary involvement and theunderlying cause of amyloidosis. Treatment ofdiffuse parenchymal amyloidosis associated withAL (primary) amyloidosis is directed against theunderlying plasma cell dyscrasia. This is also truefor hilar/mediastinal adenopathy. The progno -sis is poor for patients with primary systemicamyloidosis who have pulmonary involvement.Symptomatic tracheobronchial involvement withlocalized stenosis can be treated with bronchoscopiclaser or surgery. Recently, external beamtherapy has been reported to be effective in thetreatment of tracheobronchial amyloidosis. 17 Theuse of bevacizumab, an antivascular endothelialgrowth factor antibody, has been reported to beeffective in managing refractory pleural effusionspatients with primary (AL) systemic amyloidosis. 18Treatment of AL systemic amyloidosis involvessystemic chemotherapy and peripheral stem-celltransplantation.Birt-Hogg-Dubé SyndromeBirt-Hogg-Dubé (BHD) syndrome is a rare,inheritable disorder (autosomal dominant) that isassociated with cystic lung disease and pneumothoraces.19 Recent reports suggest that BHD maybe an underdiagnosed cause of pneumothorax.BHD was first described in 1977 and is characterizedby the cutaneous triad of fibrofolliculomas(hamartoma of the hair follicle), trichodiscomas,and skin tags, along with a propensity for renaltumors. It is caused by germline mutations in theBHD (FLCN) gene that lies on chromosome 17 andencodes a tumor-suppressor protein, folliculin. 20,21Folliculin is highly expressed in a variety of tissues,including the skin, kidney, and lung.Characteristic skin lesions of BHD typicallyappear as firm, dome-shaped papules in adulthoodduring the third or fourth decades of life and occurpredominantly on the face, scalp, neck, and upperchest. 19 Renal tumors associated with BHD syndromeinclude oncocytic hybrid tumor, chromophobe renalcellcarcinoma, clear-cell carcinoma, and papillaryrenal-cell carcinoma. 19Patients with BHD syndrome who present withcystic lung disease or a history of pneumothoracestend to be middle aged without a previous diagnosisof the underling genetic syndrome. 21,22Approximately 90% of adult patients with BHDwill have lung cysts of varying degrees on CT. 21Approximately 25% of patients with BHD willexperience pneumothorax, which will typicallyoccur during third to sixth decades of life. 21,22 Somegermline mutations of the BHD gene can causepulmonary manifestations (cystic lung disease andpneumothorax) in the absence of skin or kidneylesions. 20 Histopathologic features in the lung arenonspecific and include intraparenchymal air-filledspaces surrounded by normal parenchyma or athin fibrous wall. 22Cystic lung disease seen in BHD syndromeneeds to be distinguished from other lung diseasescharacterized by multifocal or diffuse cysticchanges, including lymphangioleiomyomatosis,pulmonary Langerhans-cell histiocytosis, lymphocyticinterstitial pneumonitis, and Pneumocystispneumonia. 23 <strong>Pulmonary</strong> lymphangioleiomyomatosis,whether occurring in a sporadic form orassociated with tuberous sclerosis complex, affectswomen of childbearing age almost exclusively. Itcan be associated with renal tumors (angiomyolipomas)and may at times be difficult to distinguishfrom BHD. 24 Lymphocytic interstitial pneumoniaand Pneumocystis pneumonia both cause parenchymalchanges aside from cysts, such as groundglassopacities, consolidation, nodules, andreticular opacities. 23 In addition, both of these disordersusually are symptomatic and are associatedwith specific clinical contexts. <strong>Pulmonary</strong> Langerhans-cellhistiocytosis encountered in adults isusually a smoking-related interstitial lung diseaseand is characterized by irregular cystic lesions andnodules that predominantly affect the upper andmid lung zones with relative sparing of the bases. 25Rarely, metastatic neoplasms, such as adenocarcinomasand low-grade sarcomas, can present withcystic lung lesions. 23The diagnosis of BHD has obvious implicationson family members and the need for their evaluation.Patients with BHD who experience pneumothoraxare at high risk of recurrence and should be482 Unusual and Uncommon <strong>Pulmonary</strong> Disorders (Ryu)ACC-PULM-09-0302-029.indd 4827/10/09 8:52:53 PM


managed accordingly, ie, consider pleurodesis.These patients also need to be screened for renalneoplasms. 21Constrictive BronchiolitisConstrictive bronchiolitis (also called obliterativebronchiolitis or bronchiolitis obliterans) is aform of obstructive lung disease that results frombronchiolar scarring and narrowing. 26,27 Bronchiolesare small airways (internal diameter 2 mmthat do not contain cartilage in their walls. Membranousand terminal bronchioles are purely airconducting,and respiratory bronchioles ( 0.5 mmin diameter) contain alveoli in their walls. Constrictivebronchiolitis can be seen in a variety of clinicalsettings, most recognizably in organ transplantrecipients, ie, obliterative bronchiolitis syndrome.26,27 Outside of organ transplant recipients,constrictive bronchiolitis may present as obstructivelung disease of obscure etiology (eg, no smokinghistory) or abnormal CT findings suggestive ofsmall airways disease with air trapping.The types of constrictive bronchiolitis can beclassified by etiology into following groups 26 :• Allograft recipients;• Postinfectious (viruses, mycoplasma, bacteria,fungi);• Connective tissue diseases (rheumatoid arthritis,systemic lupus erythematosus, etc);• Inhalational injury (eg, “popcorn lung” causedby diacetyl 28 );• Ingested toxins (eg, Sauropus androgynus);• Chronic hypersensitivity pneumonitis;• Drugs (such as penicillamine, lomustine, cocaine);• Other associations, including inflammatory boweldiseases, microcarcinoid tumorlets, paraneoplasticpemphigus, menstrual associations; and• Cryptogenic constrictive bronchiolitis.Patients with constrictive bronchiolitis typicallypresent with slowly worsening exertionaldyspnea, sometimes accompanied by a chroniccough. 26,29 On auscultation, the lungs usually soundclear without adventitious sounds. Wheezing maybe heard in some patients. Evidence of expiratoryslowing and hyperinflation will be observed inmore severe cases. <strong>Pulmonary</strong> function testing willdemonstrate airflow obstruction, usually withoutsignificant reversibility.Chest radiography typically shows hyperinflationin the absence of any parenchymal infiltrates.HRCT reveals a mosaic pattern with patchy areasof air trapping. 26,29–31 Scattered changes of bronchiectasismay be seen in some patients.The treatment of constrictive bronchiolitisvaries with the underling cause. 26,29 Causative exposureor underlying disease needs to be controlled.In general, corticosteroids administered systemicallyor by inhaled route are relatively ineffectivein reversing the airflow obstruction. 26,29 In organtransplant recipients, administration of macrolides,eg, azithromycin, 250 mg three times per week, hasdemonstrated some beneficial effects. 32,33Spontaneous PneumomediastinumPneumomediastinum is defined as the presenceof free air in the mediastinum. Pneumomediastinumcan be spontaneous, traumatic, or iatrogenic(surgical and endoscopic procedures). Spontaneouspneumomediastinum (SP) is defined as the presenceof air in the mediastinum without precedingtraumatic or procedural injury and is subdividedinto primary (no apparent lung disease) and secondary(known lung disease) categories.Potential sources of air into the mediastinumin cases of SP include airways, lung, or esophagus.The proposed pathophysiology involves the developmentalveolar rupture with dissection of airalong the interstitium and bronchovascular tissuesheath following a centripetal pattern toward thehilum and the mediastinum. Rarely, pneumomediastinummay be caused by a mediastinal infectionwith a gas-producing organism.Preexisting lung diseases are identified inapproximately one half of patients with SP andinclude asthma, chronic obstructive lung disease,and interstitial lung diseases. 34,35 Precipitatingevents preceding SP may include coughing,sneezing, vomiting, exercise or athletic activities,inhalational drug use, Valsalva maneuver duringchildbirth, and the playing of wind instruments. 34These activities are associated with increase alveolarpressure. In some patients, no precipitatingevent can be identified. 35,36SP is typically encountered in young men andwomen in their third and fourth decades of life butcan be encountered in subjects over a wide rangeof ages. 34–36 Presenting clinical manifestations<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 483ACC-PULM-09-0302-029.indd 4837/10/09 8:52:53 PM


usually consist of chest pain and/or dyspnea. Othercommon symptoms may include dysphagia, odynophagia,neck pain and swelling, and dysphonia.In some patients, SP may be detected radiologicallyin the absence of any relevant symptoms.Crepitations from subcutaneous emphysemamay be present in the neck or chest wall. 34–36 TheHamman sign (a mediastinal crunching or clickingsound that is synchronous with the heart beat) isuncommon. The use of plain chest radiography(posteroanterior and lateral views) is usually sufficientto reveal the presence of mediastinal air. 35However, chest radiographic findings may be normalup to 30% of the time. 34,35,37 Thus, CT is thediagnostic standard for pneumomediastinum.The presence of pneumomediastinum is commonlyviewed as an ominous finding, with potentiallycatastrophic complications. However, SP isgenerally not associated with identifiable pathologiccause such as esophageal or tracheobronchialrupture. 34–38 In the absence of worrisome symptomssuch as severe dysphagia, fever, leukocytosis, orhemodynamic instability, extensive diagnosticevaluation is unnecessary and conservative managementsuffices. Resolution of pneumomediastinumand symptoms occurs over the following 1 to2 weeks. Associated pneumothorax may developin 10 to 30% of patients, but these patients do notrequire drainage in most cases. Recurrence ofpneumomediastinum is unusual. 34–38References1. Office of Rare Diseases, National Institutes ofHealth. Disease information from NORD. Availableat http://www.rarediseases.org/info/about.html. Accessed June 15, 20091a. Hamano H, Kawa S, Horiuchi A, et al. High serumIgG4 concentrations in patients with sclerosingpancreatitis. N Engl J Med 2001; 344:732–7382. Kamisawa T, Okamoto A. IgG4-related sclerosingdisease. World J Gastroenterol 2008; 14:3948–39553. Pickartz T, Mayerle J, Lerch MM. Autoimmunepancreatitis. Nat Clin Pract Gastroenterol Hepatol2007; 4:314–3234. Neild GH, Rodriguez-Justo M, Wall C, et al. Hyper-IgG4 disease: report and characterisation of a newdisease. BMC Med 2006; 4:235. Seymour JF, Presneill JJ. <strong>Pulmonary</strong> alveolar proteinosis:progress in the first 44 years. Am J RespirCrit Care Med 2002; 166:215–2356. Trapnell BC, Whitsett JA, Nakata K. <strong>Pulmonary</strong>alveolar proteinosis. N Engl J Med 2003; 349:2527–25397. Inoue Y, Trapnell BC, Tazawa R, et al. Characteristicsof a large cohort of patients with autoimmunepulmonary alveolar proteinosis in Japan. Am JRespir Crit Care Med 2008; 177:752–7628. Frazier AA, Franks TJ, Cooke EO, et al. From thearchives of the AFIP: pulmonary alveolar proteinosis.Radiographics 2008; 28:883–899; quiz 9159. Johkoh T, Itoh H, Muller NL, et al. Crazy-pavingappearance at thin-section CT: spectrum of diseaseand pathologic findings. Radiology 1999; 211:155–16010. Ioachimescu OC, Kavuru MS. <strong>Pulmonary</strong> alveolarproteinosis. Chron Respir Dis 2006; 3:149–15911. Venkateshiah SB, Yan TD, Bonfield TL, et al. Anopen-label trial of granulocyte macrophage colonystimulating factor therapy for moderate symptomaticpulmonary alveolar proteinosis. Chest 2006;130:227–23712. Wylam ME, Ten R, Prakash UBS, et al. Aerosolgranulocyte-macrophage colony-stimulating factorfor pulmonary alveolar proteinosis. Eur RespirJ 2006; 27:585–59313. Parker LA, Novotny DB. Recurrent alveolar proteinosisfollowing double lung transplantation.Chest 1997; 111:1457–145814. Lachmann HJ, Hawkins PN. Amyloidosis and thelung. Chron Respir Dis 2006; 3:203–21415. Utz JP, Swensen SJ, Gertz MA. <strong>Pulmonary</strong> amyloidosis:the Mayo Clinic experience from 1980 to1993. Ann Intern Med 1996; 124:407–41316. Pickford HA, Swensen SJ, Utz JP. Thoracic crosssectionalimaging of amyloidosis. AJR Am J Roentgenol1997; 168:351–35517. Neben-Wittich MA, Foote RL, Kalra S. Externalbeam radiation therapy for tracheobronchial amyloidosis.Chest 2007; 132:262–26718. Hoyer RJ, Leung N, Witzig TE, et al. Treatmentof diuretic refractory pleural effusions with bevacizumabin four patients with primary systemicamyloidosis. Am J Hematol 2007; 82:409–41319. Schmidt LS, Nickerson ML, Warren MB, et al.Germline BHD-mutation spectrum and phenotypeanalysis of a large cohort of families withBirt–Hogg–Dubé syndrome. Am J Hum Genet2005; 76:1023–1033484 Unusual and Uncommon <strong>Pulmonary</strong> Disorders (Ryu)ACC-PULM-09-0302-029.indd 4847/10/09 8:52:54 PM


20. Gunji Y, Akiyoshi T, Sato T, et al. Mutations of theBirt–Hogg–Dubé gene in patients with multiplelung cysts and recurrent pneumothorax. J MedGenet 2007; 44:588–59321. Toro JR, Pautler SE, Stewart L, et al. Lung cysts,spontaneous pneumothorax, and genetic associationsin 89 families with Birt–Hogg–Dubésyndrome. Am J Respir Crit Care Med 2007;175:1044–105322. Ayo DS, Aughenbaugh GL, Yi ES, et al. Cystic lungdisease in Birt–Hogg–Dubé syndrome. Chest 2007;132:679–68423. Ryu JH, Swensen SJ. Cystic and cavitary lungdiseases: focal and diffuse. Mayo Clin Proc 2003;78:744–75224. Ryu JH, Moss J, Beck GJ, et al. The NHLBI lymphangioleiomyomatosisregistry: characteristics of 230patients at enrollment. Am J Respir Crit Care Med2006; 173:105–11125. Vassallo R, Ryu JH. <strong>Pulmonary</strong> Langerhans’ cellhistiocytosis. Clin Chest Med 2004; 25:561–57126. Ryu JH, Myers JL, Swensen SJ. Bronchiolar disorders.Am J Respir Crit Care Med 2003; 168:1277–129227. Visscher DW, Myers JL. Bronchiolitis: the pathologist’sperspective. Proc Am Thorac Soc 2006; 3:41–4728. van Rooy FGBGJ, Rooyackers JM, Prokop M,et al. Bronchiolitis obliterans syndrome in chemicalworkers producing diacetyl for food flavorings.Am J Respir Crit Care Med 2007; 176:498–50429. Cordier JF. Challenges in pulmonary fibrosis: 2.Bronchiolocentric fibrosis. Thorax 2007; 62:638–64930. de Jong PA, Dodd JD, Coxson HO, et al. Bronchiolitisobliterans following lung transplantation: earlydetection using computed tomographic scanning.Thorax 2006; 61:799–80431. Kang E-Y, Woo OH, Shin BK, et al. Bronchiolitis:classification, computed tomographic and histopathologicfeatures, and radiologic approach.J Comp Assist Tomogr 2009; 33:32–4132. Gottlieb J, Szangolies J, Koehnlein T, et al. Longtermazithromycin for bronchiolitis obliterans syndromeafter lung transplantation. Transplantation2008; 85:36–4133. Verleden GM, Vanaudenaerde BM, Dupont LJ,et al. Azithromycin reduces airway neutrophiliaand interleukin-8 in patients with bronchiolitisobliterans syndrome. Am J Respir Crit Care Med2006; 174:566–57034. Caceres M, Ali SZ, Braud R, et al. Spontaneouspneumomediastinum: a comparative study andreview of the literature. Ann Thorac Surg 2008;86:962–96635. Macia I, Moya J, Ramos R, et al. Spontaneous pneumomediastinum:41 cases. Eur J Cardiothorac Surg2007; 31:1110–111436. Newcomb AE, Clarke CP. Spontaneous pneumomediastinum:a benign curiosity or a significantproblem? Chest 2005; 128:3298–330237. Takada K, Matsumoto S, Hiramatsu T, et al. Managementof spontaneous pneumomediastinumbased on clinical experience of 25 cases. RespirMed 2008; 102:1329–133438. Jougon JB, Ballester M, Delcambre F, et al. Assessmentof spontaneous pneumomediastinum: experiencewith 12 patients. Ann Thorac Surg 2003;75:1711–1714<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 485ACC-PULM-09-0302-029.indd 4857/10/09 8:52:54 PM


Notes486 Unusual and Uncommon <strong>Pulmonary</strong> Disorders (Ryu)ACC-PULM-09-0302-029.indd 4867/10/09 8:52:54 PM


Mediastinal and Other NeoplasmsJay H. Ryu, MD, FCCPObjectives:• Discuss the anatomic compartments of the mediastinum• Discuss the diagnostic approach to mediastinal lesions andmanagement options• Discuss uncommon intrathoracic neoplasmsKey words: carcinoid; germ-cell tumors; hamartoma; lymphadenopathy;lymphomas; mediastinal tumors; mediastinum;neurogenic tumors; salivary gland type carcinomaMediastinumThe mediastinum is the intrathoracic compartmentlocated between the two pleural cavities. Itincludes those structures bounded by the thoracicinlet, diaphragm, sternum, vertebral bodies, andpleura. Several different schemes exist in subdividingthe mediastinum into compartments. 1 For thisdiscussion, we will use the three-compartmentmodel: anterior, middle, and posterior. The anteriorcompartment refers to the retrosternal space thatis anterior to the heart and the great vessels; itincludes the thymus and lymph nodes as well asadipose and connective tissue. The posterior compartmentextends from the posterior heart borderand trachea to the posterior aspect of the vertebralbodies. It includes the descending aorta, azygousveins, esophagus, autonomic ganglia and nerves,thoracic duct, and lymph nodes. In between is themiddle compartment, which contains the heart,ascending aorta and aortic arch, brachiocephalicvessels, main pulmonary arteries and veins, venacavae, lymph nodes, trachea, and main bronchi.Mediastinal TumorsThe differential diagnosis of a mediastinal massis facilitated by identifying its location within themediastinal compartments (Table 1). Thus, CTscanning is the most important imaging procedurein the diagnosis of mediastinal lesions. CT also canbe used to define the density of the mass (fat, fluid,Table 1. Mediastinal Masses According to CompartmentsLocationAnteriorMiddlePosteriorDisorderThymoma and other thymic tumorsLymphomasGerm cell tumorsThyroid goiter and other thyroid tumorsMetastasesMiscellaneous: parathyroid tumors,mesenchymal neoplasms (eg, lipoma,leiomyoma, lymphangioma), andMorgagni herniaLymphomasMetastasesNonmalignant lymphadenopathy(eg, infections, sarcoidosis, silicosis,and Castleman disease)Mediastinal cystsVascular lesionsMiscellaneous (eg, lymphangiomas andhernias)Neurogenic neoplasmsEsophageal lesionsMiscellaneous (eg, lateral meningocele,extramedullary hematopoiesis,descending aortic aneurysm, andBochdalek hernia)air, calcification, heterogeneous attenuation) andeffect on adjacent structures (compression, infiltration).Sometimes, a confident diagnosis can beestablished based on CT features, eg, teratoma,mediastinal goiter, and pericardial cyst. MRI typicallyadds little additional information beyond thatprovided by CT scanning, but it may be useful inpatients with contraindication to the use of iodinatedIV contrast or when additional characterizationof musculoskeletal or neurovascular anatomyis needed, eg, neurogenic tumor with intraspinalextension. 2In general, the diagnostic approach to mediastinalmasses includes consideration of the followingparameters: location; radiologic characteristics;clinical context (ie, age, symptoms, concurrent andprevious illnesses); and tempo (ie, assess fromevolution of clinical manifestations and comparisonto previous imaging studies). Approximately25 to 30% of mediastinal masses occurring in adults<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 487ACC-PULM-09-0302-030.indd 4877/10/09 8:53:14 PM


are malignant, with a higher risk in those massesappearing in the anterior mediastinum and inchildren. 3 Symptomatic patients are also morelikely to have a malignancy (50%) compared withthose without symptoms (10 to 20%). 3 Patients withmediastinal masses may experience localizedsymptoms (related to compression or invasion ofadjacent mediastinal structures, metastatic sites),constitutional symptoms, or symptoms related toparaneoplastic syndromes. Overall, the most frequentmediastinal lesions are thymoma, neurogenicneoplasms, lymphomas, and foregut cysts.Anterior Mediastinal LesionsThymomaThymoma is defined as a low-grade epithelialneoplasm of the thymus, and it is typically locatedin the superior aspect of the anterior mediastinum.It is the most common mediastinal neoplasm inmost published case series. 3 Grossly, most thymomasare encapsulated firm masses but may havenecrotic, hemorrhagic, or cystic components. Invasivethymomas tend to infiltrate the surroundingmediastinal structures. Microscopically, thymomasare composed of a mixture of neoplastic thymicepithelial cells and nonneoplastic T lymphocytes. 4,5Accordingly, thymomas are subclassified as predominantlyepithelial, predominantly lymphocytic,or mixed lymphoepithelial types. 4,5 Thepredominantly epithelial cell type category displaysa greater rate of recurrence and tumor-relateddeaths than other subtypes. Thymomas are alsosubclassified based on the type of neoplastic cellspresent; polygonal-cell type, spindle-cell type, andmixed-cell type. Precise histologic subclassificationcan be difficult because histologic heterogeneity iscommon within thymomas. Several schemes existfor classification of thymomas.Thymoma usually occurs in adults, particularlyduring the fourth and fifth decades of life. 3,5,6One half to two thirds of patients present withsymptoms such as cough, chest pain, dyspnea,dysphagia, constitutional symptoms, or symptomsrelated to one or more of the parathymic syndromes.These parathymic syndromes occur in 30to 50% of patients with thymoma and includemyasthenia gravis (most common), pure redcell aplasia, hypogammaglobulinemia Goodsyndrome, and connective tissue diseases. 7 Theremaining cases are discovered as asymptomaticmediastinal mass on routine chest radiography. 3,5Thymoma appears on chest radiography as arounded, well-circumscribed, anterior mediastinalmass. 3 On CT, thymoma is typically homogeneous,discrete soft-tissue mass, but at times, it mayexhibit heterogeneous attenuation related to hemorrhage,necrosis, or cystic changes.The diagnosis of thymoma may be obtained byultrasound or CT-guided core needle biopsy orsurgical excision. 3,6 Local invasive disease may beobserved in approximately one third of cases, butlymph node and hematogenous spread are relativelyrare. Anatomic staging of thymoma is basedon assessment of capsular invasion at the timeof surgery and on microscopic examination.Thymomas can be classified into three categoriesfor staging: encapsulated, invasive, or metastatic.Optimal treatment for thymoma is completesurgical excision. 5,6,8 Adjunctive chemotherapy andradiation therapy are used for locally invasive ormetastatic disease. 6 In patients with thymomaassociated with myasthenia gravis who are 55 to60 years of age, thymectomy yields improvementin the neurologic disease, although the benefit ofthymectomy is not immediate, with remission rategradually increasing during the following 5 to 10years. 7 Thymectomy in patients with pure red cellaplasia results in 40 to 50% remission rate, whereasthose with hypogammaglobulinemia do not benefitfrom thymic resection. The overall 5-year survivalrate associated with encapsulated thymomasis 75%, whereas patients with invasive thymomashave a 5-year survival rate of 50%. 5,8Other Thymic TumorsThymic carcinoma is a rare epithelial malignancywith malignant histologic features (highor low grade) and typically presents as a large,poorly defined, infiltrative anterior mediastinalmass. 9,10 Thymic carcinoma typically occurs inthe fifth and sixth decades of life and has predominancein male patients. 9,10 Most patientspresent with respiratory and/or systemic symptoms.Regional lymph node and pulmonarymetastases as well as pleural and/or pericardialeffusions frequently are noted at presentation.Surgical resection is the preferred treatment,488 Mediastinal and Other Neoplasms (Ryu)ACC-PULM-09-0302-030.indd 4887/10/09 8:53:15 PM


when feasible. 10 Response to adjuvant cisplatinbasedchemotherapy and radiotherapy aregenerally poor, with 5-year survival rates ofapproximately 30%. 10Thymic carcinoid is a rare, malignant neuroendocrineneoplasm that is associated with similarclinical and radiologic presentation as for thymiccarcinoma. 11,12 Approximately one third of patientsexperience ectopic hormone production, includingCushing syndrome (most common), multipleendocrine neoplasia syndrome type I, syndromeof inappropriate antidiuretic hormone secretion,and carcinoid syndrome. Surgical resection, whenfeasible, is the treatment of choice. 11,12Thymolipoma is a rare benign neoplasm thatis composed of mature thymic and adipose tissueas demonstrated on a CT scan. It usually affectsyoung adults. Treatment is surgical resection, andthe prognosis is excellent. 8,12 Thymic cyst is anotherrare mediastinal lesion and appears as a unilocularor multilocular lesion, often with calcification, onCT. It may be acquired or congenital. 12 Surgicalresection is the recommended treatment.LymphomasLymphomas represent 4 to 5% of the estimatednew cases of cancer overall and 3% of cancer deathsin the United States. 13 Lymphoma involves themediastinum usually as part of widespread disease;in 10% of patients, lymphoma is limited tothe mediastinum. 14 Primary mediastinal lymphomasoccur typically in the anterior mediastinumand more commonly presents as Hodgkin disease(HD) than non-Hodgkin lymphoma (NHL). 14,15Patients with primary mediastinal lymphoma areusually young adults who may or may not besymptomatic (local or constitutional symptoms) atthe time of diagnosis.Radiologic studies 14,15 generally demonstrate alobulated mass or bulky adenopathy in the anteriormediastinum. The diagnosis of lymphoma requiresan adequate biopsy specimen obtained by surgicalexcision of an enlarged lymph node or extranodalmass. Core needle biopsy may also yield adequatetissue. Treatment of mediastinal HD consists ofradiotherapy and/or chemotherapy; both treatmentmodalities usually are instituted in the treatmentof NHL. Prognosis is better for those withmediastinal HD compared with NHL. 3,14Germ-Cell TumorsMediastinal germ-cell tumors consist of adiverse group of benign and malignant neoplasmsthat are usually located in the anterior mediastinum.Germ-cell tumors are uncommon neoplasmsthat most frequently present as testicular tumors,but approximately 5% of these tumors can primarilyarise in extragonadal locations such as themediastinum. 16 Germ-cell tumors account for 15%of anterior mediastinal tumors in adults. 3,16–18 Likeprimary gonadal germ-cell tumors, they are mostcommon in young men (usually in the third decadeof life). Germ-cell tumors are generally classifiedinto the following three groups: teratomas, seminomas,and nonseminomatous tumors. 3,16–18Teratomas (benign or mature) account for 60 to70% of mediastinal germ-cell tumors and containvarying amounts of tissues derived from at leasttwo of the three primitive germ layers: ectoderm,endoderm, and mesoderm. Teratoma typicallyaffects children and young adults. The majority ofpatients with benign teratomas are asymptomatic,and the tumor is discovered incidentally. 3,16 Characteristicbut uncommon symptoms include expectorationof hair (trichoptysis) or sebaceous debrisattributable to a communication between thetumor and the airways. Radiologically, mediastinalteratomas are well-circumscribed round or lobularmasses that may contain fat (75%), calcification(20 to 50%), or cystic changes (80%). 3,16 Occasionally,teeth or bone may be recognized and are verysuggestive of the diagnosis. Malignant degenerationof this tumor can occasionally occur. Treatmentof mediastinal teratoma is surgical excision performedthrough a median sternotomy or posterolateralthoracotomy. 16,18 Thoracoscopic resection hasalso been reported. 18Primary mediastinal seminoma is the secondmost common form of mediastinal germ-cell tumorbut is the most common malignant germ-cell tumorin this location. 3,16 Mediastinal seminomas typicallyare found in men in the third and fourth decadesof life who are usually symptomatic, most commonlywith chest pain. Increased serum level of-human chorionic gonadotrophin (-hCG)is noted in a minority of patients, and serum fetoprotein (AFP) level is usually within normalrange. Radiologically, mediastinal seminomamanifests as a large, lobular, well-circumscribed,<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 489ACC-PULM-09-0302-030.indd 4897/10/09 8:53:15 PM


homogeneous mass. Fine-needle aspiration biopsyis often adequate to make the diagnosis. Pureseminomas are very radiosensitive, but if nonseminomatouscomponents, eg, embryonal carcinoma,are present, these tumors tend be moreaggressive and do not readily respond to radiationtherapy. Patients with locally advanced diseasemay be treated with chemotherapy, followed byresection of residual tumor. At the time of diagnosis,60 to 70% of patients have metastatic diseasethat requires cisplatin-based combination chemotherapywith or without radiation. 3,16Nonseminomatous germ-cell tumors includechoriocarcinoma, embryonal carcinoma, endodermalsinus tumor, teratocarcinoma, and mixed germcellneoplasms. These rare tumors are commonlygrouped together because of therapeutic and prognosticsimilarities, including rapid growth and pooroutlook. These tumors typically occur in youngmen (usually in the third and fourth decades of life),and they are commonly associated with symptomsthat include chest pain, hemoptysis, cough, fever,weight loss, gynecomastia, or superior vena cavalsyndrome. 3,18 Serum tumor markers, AFP and-hCG, are increased in approximately 80% and30% of patients, respectively. 16 Radiologically, thesetumors present as large, irregularly shaped heterogeneousmasses with areas of necrosis, hemorrhage,or cystic changes, along with evidence invasion intoadjacent structures and metastases. 3 Treatment usuallyinvolves cisplatin-based combination chemotherapy,followed by surgical resection of residualneoplasm. 3,18 Serial monitoring of serum AFP and-hCG is helpful in assessing the response totherapy and detecting early recurrence.ThyroidIntrathoracic goiter can present as an anteriormediastinal mass and usually results from extensionof a cervical thyroid goiter into the mediastinum.An intrathoracic goiter can cause symptomsby compressing the trachea or esophagus but, morecommonly, symptoms are absent. Although mostintrathoracic goiters are benign, 10 to 15% maycontain thyroid cancer or lymphoma. 3,19 On chestradiography, intrathoracic goiter is seen as a massassociated with tracheal narrowing or deviation oras superior mediastinal widening. CT demonstratesa lobular, discrete mass with heterogeneousattenuation. Caution needs to be exercised in theadministration of iodinated radiocontrast dyebecause iodine may induce hyperthyroidism.Symptomatic goiters are surgically excised, usuallythrough a cervical incision. 20,21Other Anterior Mediastinal MassesAlthough rare in occurrence, lesions arisingfrom parathyroid tissue can produce an anteriormediastinal mass, including parathyroid cysts,carcinomas, and adenomas. These tumors tend tobe small and are usually detected by CT. Other rarecauses of anterior mediastinal masses includemesenchymal tumors (such as lipoma, leiomyoma,and lymphangioma) and Morgagni hernia. Occasionally,metastatic disease may present as ananterior mediastinal mass without a precedingcancer diagnosis.Middle Mediastinal MassesThe differential diagnosis of middle mediastinalmasses is broad and includes lesions of manycauses.Mediastinal LymphadenopathyThere are many neoplastic and nonneoplasticdiseases that can cause mediastinal and/or hilarlymphadenopathy. 3 Metastatic mediastinal/hilarlymphadenopathy may be seen with lung cancer,lymphomas, and cancers of extrapulmonaryorigin. 15 Infections, particularly fungi and mycobacteria,are also commonly associated with mediastinaland/or hilar lymphadenopathy. There aremany other disorders that cause mediastinallymphadenopathy and include sarcoidosis, drugs(eg, phenytoin), silicosis, amyloidosis, and Castlemandisease. The diagnostic approach to mediastinallymphadenopathy requires integration of theradiologic findings, clinical context, and tempo ofthe disease process.Mediastinal CystsPrimary mediastinal cysts are benign lesionsusually found in the middle mediastinum andinclude bronchogenic cyst, enteric cyst, and pericardialcyst. 3,22,23 Bronchogenic cyst is the most490 Mediastinal and Other Neoplasms (Ryu)ACC-PULM-09-0302-030.indd 4907/10/09 8:53:15 PM


common. Bronchogenic cysts are lined with ciliatedrespiratory epithelium and contain serous fluid,mucus, blood, or purulent material. Bronchogeniccysts typically are found in adults. 24,25 Patients arecommonly asymptomatic at presentation, althoughsymptoms can be produced from airway compressionor communication of the cyst with the airway.These cysts are most commonly located in thesubcarinal or paratracheal regions but are occasionallyin the lung. Radiologically, bronchogenic cystsappear as homogeneous, nonenhancing masses ofvariable attenuation depending on the compositionof the fluid. 23 When water attenuation is seen, itcan be highly suggestive of the diagnosis. Definitivediagnosis can be obtained by CT-guided,bronchoscopic, or endoscopic aspiration of the cystfluid. 24,25 Most bronchogenic cysts are surgicallyexcised, but fluid aspiration and observation areother management options. 24,25Enteric cysts are lined by enteric or stratifiedsquamous epithelium and are usually seen in themiddle or posterior mediastinum. 24 The majorityof enteric cysts are encountered in children. Theymay cause symptoms by compression of the tracheaor esophagus, as well as mucosal ulceration,hemorrhage, or cyst rupture if the cyst containsgastric or pancreatic mucosa. 3 Surgical excision isthe treatment of choice. 24 Prognosis after completeexcision is excellent. 24Pericardial cysts are lined by mesothelial cellsand appear as well-circumscribed lesions abuttingthe heart (most commonly in the area of the rightcardiophrenic angle), the diaphragm, and theanterior chest wall. 23 CT demonstrates a unilocularlesion with characteristic water attenuationand a thin cystic wall. Most pericardial cysts arecongenital but they can be acquired lesions. Thesecysts are usually asymptomatic but can occasionallycause cardiac compromise. 26 These lesions donot require intervention unless symptoms ordiagnostic uncertainty exists. 22Vascular LesionsVascular lesions, such as aneurysms and fistulas,constitute approximately 5 to 10% of allmediastinal masses and may radiologicallyresemble neoplasms. The diagnosis is usuallyestablished by contrast-enhanced CT, MRI, and/orangiography.Other Middle Mediastinal MassesLymphangiomas are rare congenital or acquireddisorders involving the lymphatic vessels. 27Although most commonly found in the neck, theycan also occur in the mediastinum and appear assolitary or multilobulated cystic masses. Theselesions are considered benign in nature but canslowly enlarge and infiltrate adjacent structures. 27,28Total resection is optimal, if feasible. Mediastinallipomatosis is a benign condition characterized byabnormal adipose tissue accumulation in the mediastinum.It is associated with endogenous andexogenous steroid excess, as well as obesity. CTreveals diagnostic features. 29Posterior MediastinumNeurogenic TumorsThe most common cause of posterior mediastinalmasses is neurogenic neoplasm, whichaccounts for 75% of primary posterior mediastinalneoplasms. 30 Neurogenic tumors may arise fromthe peripheral, autonomic, or paraganglionic nervoussystems. Ninety percent of neurogenic neoplasmsoccur in the paravertebral gutters. Themajority (70 to 80%) of neurogenic neoplasms arebenign; the occurrence of multiple lesions suggeststhe diagnosis of neurofibromatosis and the associatedrisk of malignant transformation. 3Approximately one half of neurogenic tumorsare first detected on chest radiography in asymptomaticpatients. 3 CT helps to define the location,characteristics, and margins of the mediastinal mass.Approximately 10% of mediastinal neurogenictumors extend through the neural foramen into thespinal canal, creating a dumbbell shape (ie, “dumbbelltumor”). In this setting, MRI is more sensitivethan CT in delineating the extension through neuralforamen and spinal canal involvement. 2Peripheral nerve sheath tumors are the mostcommon type (40 to 65%) of mediastinal neurogenictumors and include schwannoma (alsocalled neurilemmoma), neurofibroma, and malignantperipheral nerve sheath neoplasm. 3,30 Theyare most commonly found in asymptomatic adultsin the third and fourth decades of life. The vastmajority of schwannoma and neurofibromas arebenign and slow growing. They usually arise from<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 491ACC-PULM-09-0302-030.indd 4917/10/09 8:53:15 PM


a posterior spinal nerve root. Radiologically,schwannomas and neurofibromas are well-circumscribed,round, or lobular posterior mediastinalmasses. 3 Associated abnormalities may be seen inthe ribs, vertebral bodies, and neural foramina asthe result of pressure and erosion caused by theenlarging mass. The treatment of choice is surgicalresection. Combined thoracic and neurosurgicalapproach may be needed for dumbbell tumors. 31Malignant peripheral nerve sheath neoplasmsare rare spindle-cell sarcomas ( 5% of nervesheath tumors) and include malignant neurofibrosarcomas,malignant schwannomas, and neurogenicfibrosarcomas. Approximately one half ofthese are associated with neurofibromatosis. 30 Painand enlarging mass are common presenting manifestationsof this tumor, which appears radiologicallyas a sharply marginated, round, heterogeneousmass on CT. Evidence of local invasion and/ormetastases may also be revealed. Complete surgicalresection is the optimal treatment; in patientswith unresectable tumors, adjuvant chemotherapyand radiation are options. 3,30,32Neoplasms of the sympathetic ganglia arisefrom nerve cells and include ganglioneuroma,ganglioneuroblastoma, and neuroblastoma.These tumors are more common in the pediatricpopulation, and most are malignant. 3,30 Ganglioneuroblastomaand neuroblastoma are aggressivemalignancies that usually affect children in the firstdecade of life. Treatment for limited disease issurgical resection. Management of advanced diseaseincludes chemotherapy and radiation. Ganglioneuromasare benign neoplasms and occur inolder children and young adults. One half ofpatients are symptomatic from local effects of thetumor or intraspinal extension. CT demonstratesa well-circumscribed paraspinal mass that may beassociated with skeletal displacement or erosion.MRI may be needed to exclude intraspinal extension.3 Complete surgical resection is the treatmentof choice and may necessitate a combined thoracicand neurosurgical approach.Mediastinal paraganglionic tumors are veryrare and may arise from either sympathetic orparasympathetic cells. 33 They can occur in all threecompartments of the mediastinum. Paravertebralparagangliomas occur in young adults and may befunctional, ie, secrete catecholamines. Catecholamine-secretingparagangliomas may be associatedwith neurofibromatosis 1, von Hippel-Lindau disease,Carney triad, and multiple endocrine neoplasiatype 2. 34 Mediastinal paragangliomas generallybehave as low-grade, indolent neoplasms. 30,33 Thetreatment of choice is surgical excision.Other Posterior Mediastinal MassesHiatal hernias are common and may manifestradiologically as a retrocardiac mass. Lateralthoracic meningoceles are rare and consist ofredundant meninges that protrude through theneural foramen. 35 CT scans demonstrate a wellcircumscribed,paraspinous cystic lesion (homogeneouswater attenuation) that is continuous withthe thecal sac. Symptomatic lesions are treated withsurgical excision.Extramedullary hematopoiesis refers to theproliferation of hematopoietic cells outside of thebone marrow in conditions of compromised tissueoxygenation. Most cases are associated with hemoglobinopathiesmanifesting severe, chronic anemiasuch as thalassemia major, hereditary spherocytosis,hemolytic anemia, or sickle-cell anemia. 36,37Myeloproliferative disorders and other chronicillnesses associated with persistent anemia mayalso lead to extramedullary hematopoiesis.Extramedullary hematopoiesis can be seen in theposterior mediastinum as one or multiple, inhomogeneouslycontrast-enhancing paravertebralmasses. 36,37 The diagnosis can be confirmed withpercutaneous fine-needle aspiration or thoracoscopicbiopsy, if needed. Extramedullary hematopoiesisusually regresses with treatment of anemiaand the underlying condition or with radiationtherapy.Thoracic duct cyst is a rare lesion that can presentin the posterior mediastinum and occasionallyin the neck (left supraclavicular fossa). It may ariseas a congenital anomaly or related to inflammatoryor traumatic injuries. Symptomatic worsening orincrease in size of the lesion after meals is a diagnosticclue. 38 The diagnosis can be established byneedle aspiration of the chylous fluid within thecyst or lymphangiography.Uncommon Thoracic TumorsThere are many types of uncommon to raretumors that may be encountered in the thorax. 39–41492 Mediastinal and Other Neoplasms (Ryu)ACC-PULM-09-0302-030.indd 4927/10/09 8:53:16 PM


Among these, more frequently encountered tumorsinclude the carcinoid tumor, salivary gland-typecarcinomas (mucoepidermoid carcinoma andadenoid cystic carcinoma), and hamartoma.CarcinoidBronchial carcinoid tumors are malignant neoplasmswith neuroendocrine differentiation andaccount for 1 to 2% of all thoracic tumors. 42,43 Thereis no clear association with smoking. Carcinoidtumors have been histologically subclassified intotypical (more common) and atypical types. 39,42Atypical tumors have greater mitotic activity andnecrosis.Patients with typical bronchial carcinoids usuallypresent in the fifth decade of life, whereasthose with atypical carcinoids present later. 39,42The majority of carcinoid tumors are central orperihilar in location. Patient often present withsymptoms related to bronchial obstruction (eg,recurrent pneumonias or wheezing) or vascularityof the tumor (eg, hemoptysis). 43 Bronchialcarcinoids are occasionally associated carcinoidsyndrome caused by release of serotonin andother vasoactive substances into the systemiccirculation. 43 Rarely, biopsy or manipulation of abronchial carcinoid results in acute carcinoidsyndrome. Bronchial carcinoids can also causeCushing syndrome and acromegaly as a result ofthe ectopic production of adrenocorticotrophichormone and growth hormone-releasing factor,respectively. 42,43Radiologically, bronchial carcinoids maymanifest obstructive pneumonia, discrete ovoidhilar mass or, occasionally, a peripheral nodule. CThelps identify the location and extent of the tumoras well as the presence or absence of mediastinallymphadenopathy. MRI appears to be more sensitivethan CT in detecting hepatic metastases (themost common metastatic site). Somatostatin receptorscintigraphy using a radiolabeled somatostatinanalog such as 111 In-octreotide can sometimes behelpful in localizing the primary tumor and instaging. 43Carcinoids are generally attached to the bronchusby a broad base but can be polypoid andcreate a ball-valve effect. Caution should be usedin biopsy of these vascular tumors (typically pinkto red mass). Typical carcinoid tumors are usuallyindolent in behavior, and hilar/mediastinal nodalmetastases are uncommon. 42,43 However, atypicalcarcinoids have an aggressive course, and metastasisto mediastinal lymph nodes are seen in 20 to60% of cases. Atypical carcinoids are associatedwith a greater recurrence rate compared with thetypical type. The 5-year survival rate associatedwith atypical carcinoid is 40 to 60%, compared with85 to 100% for the typical carcioid. 42,44Surgical resection of the tumor with completemediastinal lymphadenectomy is the treatment ofchoice for localized bronchial carcinoid tumors. 42,43Surgical resection of liver metastases may be ofbenefit in patients with limited hepatic disease. 43Hepatic artery occlusion or embolization is anoption for patients who are not candidates forhepatic resection. 43Octreotide, a somatostatin analog, and interferonare used in the treatment of metastatic carcinoidtumors. 42 Cytotoxic chemotherapy has had alimited therapeutic effect. 43 Radiolabeled somatostatinanalogs and biological agents targeting thevascular endothelial growth factor pathway arebeing investigated. 42Salivary Gland-Type CarcinomasMucoepidermoid carcinoma and adenoidcystic carcinoma are the most common forms ofsalivary gland-type carcinomas that can occur inthe thorax, accounting for 0.1 to 0.2% of all thoracictumors. 45 Both are considered low-grademalignancies, and lymph node spread is rare touncommon. Neither tumor has a clear associationwith smoking.Mucoepidermoid carcinoma is derived fromminor salivary gland tissue in the proximal tracheobronchialtree. 46 This tumor usually appearsas an endobronchial mass in the trachea or majorbronchi. 45,46 Histologically, mucoepidermoid carcinomaconsists of three components: mucussecreting,squamous, and intermediate cells. 46<strong>Pulmonary</strong> mucoepidermoid carcinoma occurs ina wide range of patients, from children to elderly.Patients usually present with cough, hemoptysis,wheezing, or obstructive pneumonia. CT mayidentify an endobronchial lesion or findings ofobstructive pneumonia. Surgical resection is thetreatment of choice and complete resection portendsan excellent prognosis.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 493ACC-PULM-09-0302-030.indd 4937/10/09 8:53:16 PM


Adenoid cystic carcinomas (previously calledcylindroma) also arise in the trachea or a majorbronchus in most cases. It occurs in adults, mostcommonly in the fifth decade of life. 45,47 Dyspnea,cough wheeze, and hemoptysis are the most commonpresenting symptoms. Surgical resection isthe treatment of choice; however, this tumor has apropensity to recur locally and metastasize. 45,47 Laterecurrence as long as 15 to 20 years after initialresection has been reported. Median survival afterresection is approximately 6 years. 45,47 Radiationand palliative laser therapy are used for unresectabletumors.HamartomaHamartoma is defined as a benign tumor-likemalformation that is composed of abnormal anddisorganized mixture of tissue elements. <strong>Pulmonary</strong>hamartoma typically consists of a combinationof cartilage, connective tissue, smooth muscle,fat, and respiratory epithelium. 48 Hamartoma is themost common benign neoplasm to occur in thelung, accounting for approximately 75% of allbenign lung tumors and 4% of solitary pulmonarynodules. 48<strong>Pulmonary</strong> hamartomas occur in all patient agegroups; the greatest incidence is in the fourth toseventh decades of life. 48,49 The majority of pulmonaryhamartomas are located in the periphery ofthe lung and are not associated with symptoms. 48,49Radiologically, hamartoma appears as a solitary,round, or lobulated opacity with smooth margins,generally measuring 1 to 3 cm in size. Calcificationis detected in approximately 25% of hamartomas,appearing as small flecks throughout the lesion;the classic appearance of “popcorn” calcificationis infrequent. Areas of fat density are seen by CTin up to one half of cases. Hamartomas growslowly, if at all. Diagnostic confirmation, if needed,requires tissue diagnosis that can be obtained bytransthoracic needle aspiration biopsy, bronchoscopy,or thoracoscopy.References1. Aquino SL, Duncan G, Taber KH, et al. Reconciliationof the anatomic, surgical, and radiographicclassifications of the mediastinum. J ComputrAssist Tomogr 2001; 25:489–4922. Thompson BH, Stanford W. MR imaging of pulmonaryand mediastinal malignancies. Magn ResonImaging Clin N Am 2000; 8:729–7393. Duwe BV, Sterman DH, Musani AI. Tumors of themediastinum. Chest 2005; 128:2893–29094. Suster S. Diagnosis of thymoma. J Clin Pathol 2006;59:1238–12445. Thomas CR, Wright CD, Loehrer PJ. Thymoma:state of the art. J Clin Oncol 1999; 17:2280–22896. Casey EM, Kiel PJ, Loehrer PJ Sr. Clinical managementof thymoma patients. Hematol-Oncol Clin NAm 2008; 22:457–4737. Tormoehlen LM, Pascuzzi RM. Thymoma, myastheniagravis, and other paraneoplastic syndromes.Hematol-Oncol Clin N Am 2008; 22:509–5268. Wright CD, Kessler KA. Surgical treatment of thymictumors. Semin Thorac Cardiovasc Surg 2005;17:20–269. Moran CA, Suster S. Thymic carcinoma: currentconcepts and histologic features. Hematol-OncolClin N Am 2008; 22:393–40710. Eng TY, Fuller CD, Jagirdar J, et al. Thymic carcinoma:state of the art review. Int J Radiat OncolBiol Phys 2004; 59:654–66411. Detterbeck FC, Parsons AM. Thymic tumors. AnnThorac Surg 2004; 77:1860–186912. Strollo DC, Rosado-de-Christenson ML. Tumors ofthe thymus. J Thorac Imaging 1999; 14:152–17113. Jemal A, Siegel R, Ward E, et al. Cancer statistics,2008. CA Cancer J Clin 2008; 58:71–9614. Ryu J, Habermann T. <strong>Pulmonary</strong> lymphoma: primaryand systemic disease. Semin Respir Crit CareMed 1997; 18:341–35215. Sharma A, Fidias P, Hayman LA, et al. Patternsof lymphadenopathy in thoracic malignancies.Radiographics 2004; 24:419–43416. Wood DE. Mediastinal germ cell tumors. SeminThorac Cardiovasc Surg 2000; 12:278–28917. Strollo DC, Rosado-de-Christenson ML. Primarymediastinal malignant germ cell neoplasms: imagingfeatures. Chest Surg Clin N Am 2002; 12:645–65818. Takeda S, Miyoshi S, Ohta M, et al. Primarygerm cell tumors in the mediastinum: a 50-yearexperience at a single Japanese institution. Cancer2003; 97:367–37619. Strollo DC, Rosado de Christenson ML, Jett JR.Primary mediastinal tumors. Part 1: tumors of theanterior mediastinum. Chest 1997; 112:511–52220. de Perrot M, Fadel E, Mercier O, et al. Surgicalmanagement of mediastinal goiters: when is a494 Mediastinal and Other Neoplasms (Ryu)ACC-PULM-09-0302-030.indd 4947/10/09 8:53:16 PM


sternotomy required? Thorac Cardiovasc Surg2007; 55:39–4321. Netterville JL, Coleman SC, Smith JC, et al. Managementof substernal goiter. Laryngoscope 1998;108:1611–161722. Zambudio AR, Lanzas JT, Calvo MJR, et al. Nonneoplasticmediastinal cysts. Eur J CardiothoracSurg 2002; 22:712–71623. Jeung M-Y, Gasser B, Gangi A, et al. Imaging ofcystic masses of the mediastinum. Radiographics2002; 22:S79–S9324. Cioffi U, Bonavina L, De Simone M, et al. Presentationand surgical management of bronchogenicand esophageal duplication cysts in adults. Chest1998; 113:1492–149625. Patel SR, Meeker DP, Biscotti CV, et al. Presentationand management of bronchogenic cysts in theadult. Chest 1994; 106:79–8526. Patel J, Park C, Michaels J, et al. Pericardial cyst:case reports and a literature review. Echocardiography2004; 21:269–27227. Faul JL, Berry GJ, Colby TV, et al. Thoracic lymphangiomas,lymphangiectasis, lymphangiomatosis,and lymphatic dysplasia syndrome. Am J RespirCrit Care Med 2000; 161:1037–104628. Park JG, Aubry M-C, Godfrey JA, et al. Mediastinallymphangioma: Mayo Clinic experience of 25cases. Mayo Clin Proc 2006; 81:1197–120329. Nguyen KQ, Hoeffel C, Le LH, et al. Mediastinallipomatosis. South Med J 1998; 91:1169–117230. Reeder LB. Neurogenic tumors of the mediastinum.Semin Thorac Cardiovasc Surg 2000; 12:261–26731. Shadmehr MB, Gaissert HA, Wain JC, et al. Thesurgical approach to “dumbbell tumors” of themediastinum. Ann Thorac Surg 2003; 76:1650–165432. Wright CD, Mathisen DJ. Mediastinal tumors: diagnosisand treatment. World J Surg 2001; 25:204–20933. Suster S, Moran CA. Neuroendocrine neoplasmsof the mediastinum. Am J Clin Pathol 2001;115(suppl):S17–2734. Young WF Jr. Paragangliomas: clinical overview.Ann New York Acad Sci 2006; 1073:21–2935. Chen KM, Bird L, Barnes P, et al. Lateral meningocelesyndrome: vertical transmission and expansionof the phenotype. Am J Med Genet 2005;133A:115–12136. Koch CA, Li C-Y, Mesa RA, et al. Nonhepatosplenicextramedullary hematopoiesis: associateddiseases, pathology, clinical course, and treatment.Mayo Clin Proc 2003; 78:1223–123337. Berkmen YM, Zalta BA. Case 126: extramedullaryhematopoiesis. Radiology 2007; 245:905–90838. Dool JJ, de Bree R, van den Berg R, et al. Thoracicduct cyst: sclerotherapy as alternative for surgicaltreatment. Head Neck 2007; 29:292–29539. Brambilla E, Travis WD, Colby TV, et al. The newWorld Health Organization classification of lungtumours. Eur Respir J 2001; 18:1059–106840. Chan AL, Shelton DK, Yoneda KY. Unusual primarylung neoplasms. Curr Opin Pulm Med 2001;7:234–24141. Gimenez A, Franquet T, Prats R, et al. Unusual primarylung tumors: a radiologic-pathologic overview.Radiographics 2002; 22:601–61942. Kulke MH. Clinical presentation and managementof carcinoid tumors. Hematol-Oncol Clin N Am2007; 21:433–455; vii-viii43. Gustafsson BI, Kidd M, Chan A, et al. Bronchopulmonaryneuroendocrine tumors. Cancer 2008;113:5–2144. Thomas CF Jr., Tazelaar HD, Jett JR. Typical andatypical pulmonary carcinoids: outcome in patientspresenting with regional lymph node involvement.Chest 2001; 119:1143–115045. Molina JR, Aubry MC, Lewis JE, et al. Primary salivarygland-type lung cancer: spectrum of clinicalpresentation, histopathologic and prognostic factors.Cancer 2007; 110:2253–225946. Liu X, Adams AL. Mucoepidermoid carcinoma ofthe bronchus: a review. Arch Pathol Lab Med 2007;131:1400–140447. Gaissert HA, Grillo HC, Shadmehr MB, et al. Longtermsurvival after resection of primary adenoidcystic and squamous cell carcinoma of the tracheaand carina. Ann Thorac Surg 2004; 78:1889–1896;discussion 1896–188748. Whyte RI, Donington JS. Hamartomas of thelung. Semin Thorac Cardiovasc Surg 2003; 15:301–30449. Gjevre JA, Myers JL, Prakash UB. <strong>Pulmonary</strong> hamartomas.Mayo Clin Proc 1996; 71:14–20<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 495ACC-PULM-09-0302-030.indd 4957/10/09 8:53:16 PM


Notes496 Mediastinal and Other Neoplasms (Ryu)ACC-PULM-09-0302-030.indd 4967/10/09 8:53:16 PM


Pathology of Airway Disease and Organizing PneumoniaHenry D. Tazelaar, MD, FCCPObjectives:• <strong>Review</strong> the morphologic characteristics of COPD• Summarize the pathology of disease affecting small andlarge airways• <strong>Review</strong> the histopathologic features of asthma• <strong>Review</strong> the complex constellation of histologic abnormalitiesin patients with allergic bronchopulmonary fungal disease,including bronchocentric granulomatosis, and mucoidimpaction of bronchi• <strong>Review</strong> and contrast the histologic characteristics oforganizing pneumonia and constrictive (obliterative)bronchiolitisKey words: allergic bronchopulmonary fungal disease; obstructiveairway disease; obliterative bronchiolitis; organizingpneumonia; pathologyObstructive <strong>Pulmonary</strong> DiseaseEmphysemaEmphysema is defined as “a condition of thelung characterized by abnormal, permanent enlargementof airspaces distal to the terminal bronchiole,accompanied by destruction of their walls, and withoutobvious fibrosis.” 1 Four major types of emphysemaare found in four different clinical settings 2 :Centriacinar:• Most common form of emphysema.• Most common correlate of severe airflowobstruction in smokers. 3• Also seen in coal workers, ie, simple coal worker’spneumonoconiosis.• Upper lobes most often are affected.• Destruction of alveoli affects the alveoli immediatelyadjacent to second- and third-orderrespiratory bronchioles. Lung at the peripheryof the lobule appears normal.• In severe cases, there is overlap with panacinar.Panacinar:• α 1-Protease inhibitor deficiency (ZZ), IV drugabuse, Ritalin, Swyer-James syndrome, aging lung.• Involves destruction of entire acinus.• Lower lobes most affected initially.Distal Acinar:• Least common form of emphysema.• Not usually associated with airflow obstruction.• Acini adjacent to pleura and interlobular septamost destroyed.• Upper lobes most often involved.• May contribute to spontaneous pneumothoracesand bullae formation in tall, asthenic maleadolescents.Paracicatricial Airspace Enlargement (So-CalledIrregular or “Scar” Emphysema):• Airspace enlargement next to foci of fibrosis.• May not have clinical correlate depending onunderlying disease, eg, Langerhans’ cell histiocytosis.Blebs and Bullae• Bleb: Air within the visceral pleural layers, soa form of interstitial “emphysema.” (The termemphysema really being a misnomer here by theaforementioned definition.) 4• Bullae: Emphysematous space 1 cm in diameterin the distended state. 4Chronic BronchitisThe clinical definition of chronic bronchitis isproductive cough of unknown cause occurring onmost days for 3 or more months for at least 2successive years. The pathologic definition usuallyrefers to a nonspecific combination of goblet-cellmetaplasia of bronchial luminal lining cells, thickenedbasement membrane, submucosal glandhyperplasia, smooth muscle hypertrophy, andchronic inflammation.Obviously, pathology is not the way to makethe diagnosis! The Reid index is used to measurethe thickness of mucous gland layer/thickness ofbronchial wall (normal 0.4). The submucosal<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 497ACC-PULM-09-0302-031.indd 4977/10/09 8:54:10 PM


gland hyperplasia appears to contribute the mostto mucus hypersecretion, but goblet cell metaplasialikely also contributes. The pathogenesis of chronicbronchitis appears related to inflammatory mechanismsin which neutrophils appear to play a keyrole. 5Significance of Small Airways and ObstructiveLung DiseaseSmall airways disease probably is a term thatshould be abandoned, as it has meant many differentthings since it was first introduced and doesnot refer to a specific disease. 4 The thickness ofsmall airways in patients who smoke (caused byfibrosis, inflammation, and mucus), however, doesappear to correlate with emphysema. 6AsthmaAs with chronic bronchitis, pathology is notdiagnostic in patients with asthma. 4 However,there are characteristic macroscopic and microscopicfindings. Macroscopic findings in patientsdying of status asthmaticus are mucous pluggingof airways, air trapping; in late/longstanding disease:saccular bronchiectasis, especially in theupper lobe.Microscopic findings include thickened basementmembranes; intraluminal mucous plugswith Charcot−Leyden crystals; Curschmann spirals(ie, coiled mucus plugs); creola bodies (ie,small groups of shed epithelial cells); goblet-cellmetaplasia; submucosal gland hyperplasia;inflammation, including eosinophils, mast cells,and mononuclear cells; and smooth-musclehypertrophy.Miscellaneous Diseases AffectingLarge and Small AirwaysAllergic Bronchopulmonary Fungal DiseaseAllergic bronchopulmonary fungal diseasesyndrome affects asthmatic patients and patientswith cystic fibrosis and results from an allergicreaction to fungi. Although Aspergillus sp are themost notorious fungi to cause this allergic phenomenon(allergic bronchopulmonary aspergillosis),multiple other fungi can cause the disease.Diagnostic criteria include the presence of six,including the first five, of the following 7,8 :• History of asthma or cystic fibrosis (10 to 15%of cases);• Immediate skin reactivity to fungal (usuallyAspergillus sp) extract;• Precipitating antibodies to fungi (usuallyAspergillus sp);• Elevated total serum IgE ( 1,000 ng/mL);• Elevated IgE and/or IgG to fungi, usuallyAspergillus fumigatus antigens;• Central bronchiectasis;• Recurrent pulmonary infiltrates;• Peripheral eosinophilia ( 500 μL).Pathology includes a complex constellation offindings in various combinations: bronchiectasis,mucoid impaction of bronchi, bronchocentricgranulomatosis (BCG), eosinophilic bronchiolitis,chronic eosinophilic pneumonia, bronchiectasis,and mucoid impaction of bronchi. 4,9 Macroscopicfindings include dilated large airways filled withtenacious mucous plugs and casts of variablecolor from green−grey to yellow−white. Microscopicfindings include lamellated mucus as seenin other allergic syndromes such as allergic fungalsinusitis rich in eosinophils in various stages ofdegeneration and Charcot−Leyden crystals, raredegenerating fungal hyphae (ie, not as many asone would see in an infection), and bronchocentricgranulomatosis, which are extremely rare(most cases of possible BCG are not BCG but trueinfection).In BCG, bronchioles appear replaced by necrotizinggranulomas centrally located within theairway. Fungal hyphae may be identified in thenecrotizing granuloma, but they should be rare andnoninvasive. Airways adjacent to those with BCGmay contain an exudate not unlike that describedpreviously in the context of mucoid impaction ofbronchi.Acute BronchiolitisThere are many causes of acute bronchitis, 9most significantly infection and aspiration (may beassociated with foreign material, eg, food). Thepathology includes ulceration of mucosa, inflammation,usually including neutrophils, and intraluminalexudate.498 Pathology of Airway Disease and Organizing Pneumonia (Tazelaar)ACC-PULM-09-0302-031.indd 4987/10/09 8:54:11 PM


Chronic BronchiolitisChronic bronchiolitis 10 is a nonspecific pathologicterm used to describe the presence of chronicinflammation and/or fibrosis around and in wall ofbronchiole and the presence of peribronchiolar metaplasia.Usually, it is part of the spectrum of pathologyfound in a variety of primary airway diseases andas a component of many classically “interstitial”diseases, eg, extrinsic allergic alveolitis.Respiratory (Smokers’) BronchiolitisRespiratory bronchiolitis is an incidental findingin smokers. 11 It may be responsible for asymptomaticsmall nodules on imaging in smokers andis defined as the presence of pigmented macrophageswithin respiratory bronchioles and surroundingalveoli. Mild chronic inflammation andfibrosis of bronchiole may be found. Histology isthe same as in respiratory bronchiolitis-interstitiallung disease.Follicular BronchiolitisFollicular bronchiolitis 12 is a form of lymphoidhyperplasia associated with connective tissuesdiseases (especially Sjögren syndrome and rheumatoidarthritis) and immunodeficiencies. It isdescribed as chronic inflammation with germinalcenters around bronchioles.Diffuse PanbronchiolitisDiffuse panbronchiolitis is a form of bronchiolitisassociated with sinusitis. Predominantly seenin Japanese adults, it is associated with humanleukocyte antigen BW54. Macroscopic findingsinclude bronchiectasis and hyperinflation. Microscopicfindings are not entirely specific but includeacute and chronic inflammation of respiratorybronchioles, follicular bronchiolitis, interstitialfoamy macrophages, and intraluminal exudatecontaining neutrophils.Constrictive (Obliterative) BronchiolitisThis entity can be found in multiple clinicalsettings, including lung transplantation (chronicrejection), bone marrow transplantation (graft vshost disease), drug toxicity, connective tissue disease,and idiopathic disease. Common exposuresinclude diacetyl (popcorn worker’s lung), sulfurmustard gas, and infection (eg, postadenoviral).Pathology includes bronchiolar and peribronchiolarfibrosis with narrowing and eventual obliterationof the lumen. Constrictive bronchiolitis maybe preceded by acute and chronic bronchiolitis, buthistology is similar regardless of etiology.Organizing Pneumonia (BronchiolitisObliterans Organizing Pneumonia)Organizing pneumonia (OP) 13,14 is a myxoidappearingintraluminal (alveolar, alveolar duct, andterminal and respiratory bronchioles) fibroblasticproliferation (“fibrosis”) associated with mildchronic interstitial inflammation. The term OP isnow preferred over bronchiolitis obliterans organizingpneumonia (BOOP) because the bronchiolitis obliteranscomponent (which could be called proliferativeobliterative bronchiolitis) is often difficult toappreciate pathologically 47 and because it is rare forpatients to present with “bronchiolitis.”OP may appear in clinical settings or as a histologiccomponent of the following: idiopathic(cryptogenic OP); connective tissue disease; drugreaction (eg, bleomycin, amiodarone); resolvingbacterial and viral infection; extrinsic allergicalveolitis; nonspecific interstitial pneumonia (maybe seen focally in 50% of cases); radiation; reactionto tumor; infarct; Wegener granulomatosis; andchronic eosinophilic pneumonia.Pathology is the same regardless of clinicalcategory: polypoid plugs of proliferating fibroblastsin alveoli and alveolar ducts (strictly speaking,this would have constituted the OP of BOOP)and terminal respiratory bronchioles (the BO ofBOOP), mild interstitial pneumonia confined toareas of organization, intra-alveolar foamy macrophages(from a failure to clear secretions as theresult of small airway obstruction), and no collagen(dense) fibrosis.Relationship Between Proliferative ObliterativeBronchiolitis in OP and Constrictive ObliterativeBronchiolitisThe exact relationship between the two diseasesis uncertain. In most cases, they appear to be<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 499ACC-PULM-09-0302-031.indd 4997/10/09 8:54:11 PM


separate and distinct entities with different clinical,radiographic, and morphologic features.In certain circumstances, such as lung transplantrecipients, silo filler’s disease and, occasionally,rheumatoid arthritis, there is evidence tosuggest that intraluminal organization localized tomembranous and respiratory bronchioles cancause obstructive airways disease and can “progress”to a lesion that is indistinguishable from thelate stage of constrictive bronchiolitis.[Note to participants regarding references:There are some primary sources quoted, but thetextbooks have additional photographs worthstudying.]References1. Snider G, Keinerman J, Thurlbeck W. The definitionof emphysema. Report of a National Heart, Lung,and Blood Institute, Division of Lung Diseasesworkshop. Am Rev Respir Dis 1985; 132:182−1852. Travis W, Colby TV, Koss MN, et al. Non-neoplasticdisorders of the lower respiratory tract. In: Atlas ofnontumor pathology. Washington, DC: AmericanRegistry of Pathology and the Armed Forces Instituteof Pathology, 20023. West W, Nagai A, Hodgkin JE, et al. The NationalInstitutes of Health Intermittent Positive PressureBreathing trial—pathology studies. III. The diagnosisof emphysema. Am Rev Respir Dis 1987;135:123−1294. Churg A, Myers J, Tazelaar H, et al. Thurlbeck’spathology of the lung. 3rd ed. New York, NY: ThiemeMedical Publishers, 20055. Barnes P. Chronic obstructive pulmonary disease.N Engl J Med 2000; 343:269−2806. Hogg J, Chu F, Utokaparch S, et al. The natureof small-airway obstruction in chronic obstructivepulmonary disease. N Engl J Med 2004;350:2645−26537. Greenberger PA, Patterson R. Diagnosis and managementof allergic bronchopulmonary aspergillosis.Ann Allergy 1986; 56:444−4488. Cockrill BA, Hales CA. Allergic bronchopulmonaryaspergillosis. Annu Rev Med 1999; 50:303−3169. Leslie KO, Wick MR. Practical pulmonary pathology:a diagnostic approach. Philadelphia, PA:Churchill Livingstone, 2005; 81310. Visscher D, Myers J. Bronchiolitis: the pathologist’sperspective. Proc Am Thorac Soc 2006; 3:41−4711. Fraig M, Shreesha U, Savici D, et al. Respiratorybronchiolitis: a clinicopathologic study in currentsmokers, ex-smokers, and never-smokers. Am JSurg Pathol 2002; 26:647−65312. Aerni MR, Vassallo R, Myers JL, et al. Follicularbronchiolitis in surgical lung biopsies: clinicalimplications in 12 patients. Respir Med 2008;102:307−31213. Katzenstein A, Askin F. Surgical pathology of nonneoplasticlung disease. 4th ed. Philadelphia, PA:W.B. Saunders, 2006; 50314. Elicker B, Pereira CA, Webb R, et al. High-resolutioncomputed tomography patterns of diffuseinterstitial lung disease with clinical andpathological correlation. J Bras Pneumol 2008; 34:715−744500 Pathology of Airway Disease and Organizing Pneumonia (Tazelaar)ACC-PULM-09-0302-031.indd 5007/10/09 8:54:11 PM


Pathology of Diffuse and Neoplastic DiseaseHenry D. Tazelaar, MD, FCCPObjectives:• <strong>Review</strong> morphologic characteristics and potential etiologiesof diffuse alveolar damage• <strong>Review</strong> and contrast morphologic characteristics of the“idiopathic” interstitial pneumonias• <strong>Review</strong> histopathologic features of diffuse lung diseasewith specific histologic features, including Langerhanscell histiocytosis, hypersensitivity pneumonia, sarcoidosis,lymphangioleiomyomatosis, pulmonary alveolar proteinosis,Wegener granulomatosis, hot tub lung, and pulmonaryrejection• <strong>Review</strong> certain infectious pathogens that can be identifiedon the basis of their morphologic characteristics in tissuesections• Discuss the morphology of neuroendocrine tumors andother lung cancersKey words: granulomatous disease; infection; interstitial lungdisease; neoplasia; transplantDiffuse Lung DiseaseDiffuse Alveolar Damage/Acute InterstitialPneumoniaDiffuse alveolar damage (DAD) is the mostcommon pathologic correlate of ARDS. 1,2 Causesinclude pulmonary edema; septic shock; oxygentoxicity; drugs; radiation; infection (influenza;Pneumocystis carinii pneumonia); and trauma. DADis associated with connective tissue disease eg,acute lupus pneumonitis, and its idiopathic variantis represented by acute interstitial pneumonia andHamman-Rich syndrome. 3,4Microscopic Pathology:• The disease is temporally uniform (homogenous);microscopically, the lungs look the samefrom field to field. There are two phases: acuteand organizing (often overlapping caused byattempted repair at the same time there isongoing injury). Acute (exudative) includesinterstitial edema, type I pneumocyte sloughing,and the formation of hyaline membranes;organizing includes proliferating/reactive typeII pneumocytes and interstitial fibroblasts withfocal airspace organization. (These areas maylook like typical organizing pneumonia.) Acuteand organizing thrombi within vessels are common.Metaplastic bronchiolar epithelium maybe very atypical and mimic carcinoma.Differential Diagnosis:• Cryptogenic organizing pneumonia;More subacute course;Process is patchy around bronchioles;Hyaline membranes are not seen; andOrganization is only intraalveolar and intraluminal,not interstitial.“Idiopathic” Interstitial PneumoniasIdiopathic interstitial pneumonias comprise aheterogeneous group of pneumonias with acute,eg, acute interstitial pneumonia, and more chronicpresentations. The American Thoracic Society/European Respiratory Society endorses the classificationscheme shown in Table 1. 4 Althoughcauses for other interstitial pneumonias are notalways known, eg, sarcoidosis, the idiopathic interstitialpneumonias are characterized histologicallyby varying combinations of interstitial fibrosis,organization, and chronic inflammation in both oreither the interstitium and alveolar spaces withoutthe presence of granulomas or specific cell types,eg, Langerhans cells.Usual Interstitial PneumoniaClinical findings of usual interstitial pneumonia(UIP) 5−7 : idiopathic disease clinical diagnosisof idiopathic pulmonary fibrosis (IPF).• Common in connective tissue disease, as amanifestation of drug reaction and rarelyas manifestation of chronic hypersensitivitypneumonitis. 8 Gross findings include lowerlobe and peripheral predominant fibrosis and a<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 501ACC-PULM-09-0302-032.indd 5017/10/09 8:54:34 PM


Table 1. Clinical and Pathologic Classification of Idiopathic Interstitial Pneumonias* †Clinical DiagnosisAIP, Hamman-rich disease idiopathic diffuse alveolardamageIPF, cryptogenic fibrosing alveolitisDIPRB-ILDNonspecific interstitial pneumoniaCryptogenic organizing pneumonia (formerly idiopathicbronchiolitis obliterans organizing pneumonia)Lymphocytic interstitial pneumoniaPathologic PatternDADUIPDesquamative interstitial pneumoniaRespiratory bronchiolitisNonspecific interstitial pneumonia: cellular and fibroticvariantsOrganizing pneumoniaLymphocytic interstitial pneumonia*The American Thoracic Society/European Respiratory Society endorses some separate terms to describe the disease clinicallyand pathologically because some pathologic patterns may be associated with known disease, ie, they are not idiopathic,eg, usual interstitial pneumonia occurring in the setting of rheumatoid arthritis is still UIP, but it is not IPF.†From ATS/ERS. 4honeycomb appearance. Microscopic findingsinclude the following:Geographic heterogeneity: looks different fromfield to field, eg, variegated fibrosis worse insubpleural and paraseptal regions. Areas ofnormal lung (usually centrilobular) are foundbetween areas of fibrosis.Temporal heterogeneity (refers to type of “fibrosis”):dense collagen and fibromyxoid fibro -blastic foci (look similar to small flat areas oforganizing pneumonia) are observed.Inflammatory infiltrate usually consists ofchronic inflammatory cells. Germinal centersand dense lymphoid infiltrates are associatedwith UIP in connective tissue diseases,eg, rheumatoid arthritis. 9 A total of 70% ofpatients are smokers; therefore, pigmentedalveolar macrophages, as seen in respiratorybronchiolitis (RB) and desquamative interstitialpneumonia (DIP), are common.Honeycomb changes are most advanced at thebases and periphery.Work by some investigators 10,11 but not all 12 hasshown a correlation between survival and thenumber of “fibroblast foci,” a finding that suggeststhe critical pathway to end-stage fibrosis is not“alveolitis” (as has been supposed for many years)but ongoing epithelial damage and repair.Differential Diagnosis: (1) DIP (also see Table 2)diffuse macrophage accumulation and temporallyuniform; (2) cryptogenic organizing pneumonia;injury is temporally uniform, areas of recent organizationare more pronounced and intraalveolar,and areas of dense collagen deposition are absent;Table 2. Idiopathic Interstitial Pneumonias: Contrasting Pathologic FeaturesVariables NSIP UIP DIP/RB-ILD AIPTemporal appearance Uniform Variegated Uniform UniformInflammation Prominent Patchy Scant ScantCollagen Patchy fibrosis Prominent Patchy (none to Minimalminimal in RB-ILD)Fibroblast proliferation No Typical Rare LateOrganizing pneumonia-like foci, % 50 No No FocalHoneycomb change No Yes No RareIntraalveolar macrophages Focal Focal Diffuse (DIP), patchy No(RB-ILD)Hyaline membranesNo; in “accelerated”formNo No Focal502 Pathology of Diffuse and Neoplastic Disease (Tazelaar)ACC-PULM-09-0302-032.indd 5027/10/09 8:54:34 PM


(3) nonspecific interstitial pneumonia; fibrosis isgeographically and temporally more uniform; (4)accelerated idiopathic pulmonary fibrosis; foci ofdiffuse alveolar damage are present. 13−15Accelerated Idiopathic <strong>Pulmonary</strong> Fibrosis 13–15Underlying UIP can be recognized clinically(previous history of slowly progressive interstitialfibrosis with characteristic features of IPF), radiographically,and/or on the basis of findings at thetime of surgical lung biopsy. Wedge lung biopsywill show underlying UIP with DAD or, likely asthe result of sampling, either just DAD or just UIP.In some cases, only organizing pneumonia isseen.Nonspecific Interstitial Pneumonia/FibrosisFindings for nonspecific interstitial pneumonia/fibrosis16,17 include geographic homogeneity(looks same from field to field), fibrosis and interstitialthickening with inflammation, and is NOTvariegated. Some cases lack fibrosis altogether.Fibrosis, when present, is not worse in subpleuraland paraseptal regions, but uniform. Honeycombappearance is absent, with rare exceptions. Thereis also temporal homogeneity: refers to typeof fibrosis. Dense collagen may be present but,when present, is the only type present, and fibromyxoidfibroblastic foci are absent, with rareexceptions.DIPDIP 18,19 is closely related to respiratory bronchiolitis-interstitiallung disease (RB-ILD; see nextsection). A total of ∼90% of patients are cigarettesmokers. DIP is much more uniform at low magnificationand lacks the variegated appearance ofUIP. Alveolar septa are thickened by an inflammatoryinfiltrate that often includes mononuclearcells and occasional germinal centers. The moststriking (and definitional) feature is the presenceof numerous lightly pigmented macrophageswithin most of the distal air spaces. Collagenfibrosis occurs and may be associated with a honeycombappearance.RB-ILDRB-ILD 20−23 is a diagnosis of exclusion, as thehistology of RB may be seen in any active or pastsmoker. Microscopic findings include RB (pigmentedalveolar macrophages in and aroundrespiratory bronchioles) and a lacks of significantfibrosis, interstitial inflammation, or germinalcenters.Differential Diagnosis: RB-ILD and DIP representpoints along a spectrum, and some cases fallsomewhere in between; therefore, their separationmay not always be possible. Such cases maybe diagnosed as smoking-related interstitial lungdisease.Lymphoid Interstitial PneumoniaLymphoid interstitial pneumonia (LIP) 24−26 is adistinct clinicopathologic syndrome that can occurin a variety of settings (Table 3). Microscopic findingsinclude exquisitely interstitial and bronchovascularinfiltrate of small lymphocytes andplasma cells. Immunophenotypic studies havedemonstrated that T cells predominate in thealveolar septal infiltrates, whereas B lymphocytespredominate in the peribronchiolar zones withpolyclonal plasma cells. Poorly formed non-necrotizinggranulomas with multinucleated giant cellsare present in a minority of cases. There is polyclonallight-chain expression in the plasma cellcomponent in HIV and non-HIV–associatedcases.Table 3. Clinical Features of Lymphocytic Interstitial PneumoniaVariables HIV-Negative HIV-PositiveFemale/male ratio 2.2:1Mean age, yr 51 13 to 77Symptoms, %Dyspnea 69 80Cough 62 60Fever 22 60Weight loss 22 100Adenopathy 95Autoimmune diseases, % Common UncommonSjögren syndrome 20 YesMortality, % 30 75Mean survival, mo 20 10.4<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 503ACC-PULM-09-0302-032.indd 5037/10/09 8:54:35 PM


Langerhans Cell HistiocytosisMicroscopic findings of Langerhans cell histiocytosis1,3,5 include (1) bronchiolocentric nodularto stellate lesions; (2) variable fibrosis; and (3)a diagnostic cell, ie, Langerhans cell with convolutednuclei (kidney-bean shaped, grooved,resemble buttocks). They usually are admixedwith some eosinophils and smoker’s macrophages.Immunohistochemical findings includeS100 protein, CDla, Langerin, HLR-DR.On electron microscopy, Birbeck granules (pentilaminarstructure with a “tennis racket” morphology)appear. “Cystic” change is seenradiographically: Langerhans cell histiocytosisresults from dilated bronchioles, paracicatricialairspace enlargement (so-called scar emphysema),and necrosis in the center of the lesions.As lesions age, they become less cellular andmore fibrotic and may be confused with thechronic interstitial pneumonias. The processappears to start as a polyclonal one, but clonalitymay develop. 27Hypersensitivity Pneumonitis/ExtrinsicAllergic AlveolitisMicroscopic findings of hypersensitivitypneumonitis/extrinsic allergic alveolitis 1,3,5include a classic triad (seen in ∼80% of patientswho undergo biopsy with a clinical diagnosis):(1) Cellular interstitial pneumonia composedmainly of lymphocytes, plasma cells, and macrophagesoccasionally admixed with eosinophils(only 10% of cases) and neutrophils; (2) chronicbronchiolitis with organizing pneumonia; and (3)vague poorly formed nonnecrotizing granulomasor giant cells. Fibrosis may develop and be indistinguishablefrom UIP. 28 well-proven cases clinicallyhave been reported to resemble nonspecificinterstitial pneumonia, cryptogenic organizingpneumonia, and UIP. 8,29Hot Tub LungHot tub lung 30−32 may represent an infectionor a hypersensitivity reaction to Mycobacteriumavium-intracellulare organisms. Microscopic findingsinclude non-necrotizing granulomatousinflammation (although the granulomas oftenare larger and more exuberant than those seenin classic hypersensitivity pneumonitis) andpatchy, chronic interstitial pneumonia thatresembles that found in hypersensitivity pneumonitis.SarcoidosisThe diagnosis of sarcoidosis 1,3,5 usually requiresthe pathologic presence of granulomatous inflammation.Transbronchial biopsy has become themethod of choice and can be expected to yield adiagnosis in nearly 80% of patients, including 70%with stage 1 disease. The likelihood of obtaining adiagnostic biopsy is related to the number ofspecimens obtained, and the best results require aminimum of 4 specimens in patients with stage 2or 3 disease and as many as 10 in patients withstage 1 disease.Microscopic findings (berylliosis looks identical):• Granulomatous inflammation without bronchiolitisor cellular interstitial pneumonia;• Granulomas;Compact, well-circumscribed collections ofepithelioid histiocytes and multinucleatedgiant cells surrounded by a rim of lymphocytesand collagen fibrosis;Central necrosis (∼20% of cases);Distribution along lymphatic routes (bronchovascularbundles, interlobular septa, and visceralpleura);• Multinucleated giant cells may contain a varietyof cytoplasmic inclusions which should notbe confused with foreign material (however,these are not specific for sarcoid);Schaumann (conchoidal) bodies (blue, calcified);Asteroid bodies: star-shaped protein;Sheet-like and needle-shaped birefringent crystallineinclusions (often endogenous calciumoxalate);• Granulomatous vasculitis (∼70% of cases);• Nodular sarcoidosis (coalescence of the granulomascombined with dense collagen fibrosis);• Necrotizing sarcoidosis-sarcoid background oflymphangitic nonnecrotizing granulomatousinflammation associated with broad foci ofparenchymal necrosis and vasculitis.504 Pathology of Diffuse and Neoplastic Disease (Tazelaar)ACC-PULM-09-0302-032.indd 5047/10/09 8:54:35 PM


Table 4. Differential Diagnosis of Diffuse Lung Disease With Small Granulomas*Hypersensitivity Pneumonitis Sarcoidosis Hot Tub LungInterstitial pneumonia − − Chronic bronchiolitis − − −GranulomasWell formed − Single giant cells − − −Necrosis − Organizing pneumonia − − −Cultures − − Mycobacterium aviumintracellulare* rare; occasional; prominent feature.Differential diagnosis of disease with smallgranulomas is discussed in Table 4.<strong>Pulmonary</strong> Alveolar Proteinosis<strong>Pulmonary</strong> alveolar proteinosis (PAP) 1,3,5,33,34 isdue to a defect in alveolar macrophage functioncaused by granulocyte/macrophage colony-stimulatingfactor antibodies. The etiologies of pulmonaryalveolar proteinosis are listed in Table 5.Microscopic findings include the accumulation ofgranular eosinophilic material in alveoli (positivewith stains for periodic acid Schiff), minimal (ifany) interstitial pneumonia, and late fibrosis.Electron microscopy shows lamellar bodies ofsurfactant.LymphangioleiomyomatosisThe gross findings of lymphangioleiomyomatosis1,3,5 include randomly distributed cysticTable 5. Classification of PAPPrimary PAPSecondary PAPImmunocompromised patients, hematologic disorders(polycythemia vera, leukemias, idiopathic thrombocytopenicpurpura, multiple myeloma),AIDSInfections, Nocardia, Pneumocystis mycobacteriaEnvironmental inhalants, dusts (silica, aluminum), toxins(solvents, insecticides, cleansers)Neoplasmsairspaces with thin walls involving upper andlower lobes. The main pathology includes a disorderlyproliferation of smooth muscle cells, cystic“emphysematous” spaces (common) that usuallycontain smooth-muscle bundles within at least aportion of their walls, and hemosiderin ladenmacrophages that may be present within air spacesas the result of pulmonary hemorrhage. Immunohistochemistryof smooth muscle cells is positivefor HMB45, actin, desmin, melan A, and estrogenand progesterone receptor proteinOther pathology includes micronodular pneumocytehyperplasia, or circumscribed proliferationsof cytologically bland type 2 pneumocytes ina pattern resembling bronchioloalveolar adenocarcinoma.35 Evidence suggests that lymphangioleiomyomatosiscells may either metastasize ordevelop from progenitor cells capable of migratingto the lung. 36Eosinophilic PneumoniaThere are two types of eosinophilic pneumonia,chronic and acute. Histologic feature ofchronic eosinophilic pneumonia 1,3,5 are constantregardless of clinical context (Table 6). The majorabnormality is filling of distal air spaces by a cellularexudate of eosinophils and histiocytes,including multinucleated giant cells. Relativeproportions of the two cell types are variable,alveolar septa are thickened and contain a similarinflammatory infiltrate, and OP is a frequent associatedfinding.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 505ACC-PULM-09-0302-032.indd 5057/10/09 8:54:35 PM


Table 6. Associations and Causes of Chronic EosinophilicPneumonia PathologyAllergic bronchopulmonary fungal diseaseSimple pulmonary eosinophiliaSystemic infection (parasites, fungi)Connective tissue disease, especially rheumatoid arthritisRadiation (ipsilateral or contralateral)Cigarette smoking, especially at initiation or with change ofbrands*Drugs esp. salicylates, antibiotics †Churg Strauss syndromeHodgkin lymphomaInflammatory bowel diseaseLung cancer*From Uchiyama H, Suda T, Nakamura Y, et al. Alterationsin smoking habits are associated with acute eosinophilicpneumonia. Chest 2008; 133:1174−1180.†From Allen JN. Drug-induced eosinophilic lung disease.Clin Chest Med 2004; 25:77−88.In acute eosinophilic pneumonia, there is diffusealveolar damage with prominent tissueeosinophilia. 1,3,5Wegener GranulomatosisClassic pathology includes the following:• Necrotizing granulomatous inflammation withgeographic borders (appears more like parenchymalnecrosis than a true granuloma sincethe foci usually lack significant numbers of epithelioidhistiocytes);Centrally in foci of necrosis there is amorphouseosinophilic to basophilic debris withremnants of necrotic neutrophils and otherinflammatory cells (so-called dirty necrosis);Early lesions described as “palisading granulomas”and microabscesses;Palisading granulomas are tiny granulomascomposed of a single layer of palisading epithelioidhistiocytes that either radiate arounda central point or surround a central eosinophilicstructure resembling a collagen bundle.As the palisaded granulomas enlarge,they become more microabscess-like;• Necrotizing segmental (involving portion ofthe wall) vasculitis of arteries and veins; and• Randomly dispersed, darkly staining multinucleatedgiant cells common variants.• Common variants include: (1) presentationas a solitary pulmonary nodule limited to thelung; (2) hemorrhage with or without capillaritis(vasculitis involving capillaries recognizedby presence of neutrophils in alveolar walls);(3) bronchocentric; (4) organizing pneumonia -like; and (5) eosinophilic pneumonia-like.• c-ANCA titers are positive in approximately90% of patients with active generalized disease,60 to 85% of patients with active limited disease,and 30 to 35% of patients with quiescent disease.Positive p-antineutrophil cytoplasmic antibodytiters generally representing autoantibodiesdirected against myeloperoxidase are less specificbut have been reported to be positive insome patients with Wegener granulomatosis.PneumoconiosesAsbestos-related reactions (Table 7): the pneumoconioses5,37 is “asbestosis.” Macroscopically,firm, fibrotic lungs with areas of honeycomb changecan be found, whereas microscopically, markedinterstitial fibrosis with minimal inflammatoryinfiltrate are found, as well as UIP-like reactions.The presence of asbestos bodies, fibrosis, and exposurehistory are needed for a definitive diagnosis.Asbestos bodies are iron-encrusted fibers (onetype of ferruginous body, which is a more genericterm) that typically are beaded and dumbbell shapedwith a thin translucent core. There are no generallyaccepted criteria defining how many asbestos bodiesmust be identified in any given case for a diagnosisof asbestosis, but the presence of even a singleasbestos body in a routine tissue section usuallysignifies “above-background” asbestos exposure.Table 7. <strong>Pulmonary</strong> Complications of Asbestos Exposure*Pleural diseaseEffusionFibrosisPlaquesPseudoneoplasms (rounded atelectasis)Parenchymal lung diseaseAsbestosisAsbestos airways diseaseNeoplasmsMalignant mesotheliomaBronchogenic carcinoma (especially in cigarette smokers)*From Churg and Green. 37506 Pathology of Diffuse and Neoplastic Disease (Tazelaar)ACC-PULM-09-0302-032.indd 5067/10/09 8:54:35 PM


Table 8. Morphologic Classification of SilicosisSimple (nodular) silicosisSilicotic nodules 1 cmUpper lung zonesComplicated silicosisConglomerate nodules 1 cmUpper and middle lung zonesAcute silicosis (silicoproteinosis)Alveolar proteinosisSilicotuberculosisSilicosisGross findings of silicosis (Table 8) include firm,discrete, rounded lesions with variable amountsof black pigment and nodules in a lymphatic distribution(ie, around bronchovascular bundles, insubpleural and paraseptal areas).Microscopic findings include:• Discrete foci of concentric layers of hyalinizedcollagen;Abundant dust-filled histiocytes; andBirefringent particles (usually); often a mix ofsilica and silicates);When necrosis is present, complication bytuberculosis should be considered. Nodules mayexpand and eventually coalesce to form largermasses; when these masses measure 1 cm indiameter, they are referred to as conglomeratenodules (complicated silicosis).<strong>Pulmonary</strong> InfectionsSee Tables 9−12 for a listing of pulmonary infections.1,538, 39Pathology of Lung TransplantationPrimary graft failure in lung transplantation isdescribed as the immediate deterioration of pulmonaryfunction after successful revascularizationof the graft. Diffuse alveolar damage is foundmicroscopically in the most severe cases.RejectionThe grading scheme is strictly pathologic anddoes not rely on clinical parameters (eg, for thediagnosis of obliterative bronchiolitis syndrome):• Grade 0, negative for rejection: normal pulmonaryparenchyma without evidence of mononuclearinfiltration or alveolar hemorrhage;• Grade 1, minimal acute rejection: infrequentperivascular mononuclear infiltrates venules areusually involved before arterioles. The mononuclearcell cuff is usually more than two toTable 9. Comparison of <strong>Pulmonary</strong> Fungal Pathogens*Variables Histo Cocci Blasto Crypto Asper Mucor CandidaTissue reactionNecrotizing granulomas Yes Yes Yes Yes/no No No NoVascular invasion/infarct No No No No Yes Yes NoAC inflammation/abscess No No Yes No No No YesMorphologyBudding yeast Yes No Yes Yes No No YesLavage size, microns 3 3 30–60 8–15 4–7 3–6Shape Oval No Round Irregular Hyphae Hyphae DimorphicSpherules/endospores No Yes No No No No NoThick cell wall No Yes Yes No No No NoMucin-positive capsule No No No Yes No No NoHyphae No Rare No No Yes Yes RareSize, microns 3–5 10–15Septate Yes NoBranching 45° 90°Pseudohyphae No No No No No No Yes*Data are from Katzenstein and Askin. 1 AC = acute; Histo Histoplasma; Cocci Coccidioides; Blasto Blastomyces;Crypto Cryptococcosis; Asper Aspergillus; Mucor Zygomycetes.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 507ACC-PULM-09-0302-032.indd 5077/10/09 8:54:36 PM


Table 10. Comparison Between Actinomyces and NocardiaVariables Actinomyces NocardiaSulfur granules Yes NoSpecial stainsGram Yes YesGMS Yes YesAcid fast* No Yes*Modified acid fast (Fite stain) substituting weak acid fordecolorization.Table 11. Staining Characteristics of PneumocystisStains Frothy Exudate Cysts SporozoitesHE Yes No NoPapanicolaou Yes No NoUltraviolet fluorescence Yes YesGMS Yes NoGram-Weigert Yes NoCresyl-Echt violet Yes NoPAS-Gridley Yes NoToluidine blue Yes YesWright stain No YesGiemsa No YesMethylene blue No YesMonoclonal antibodies (immunofluorescence/immunoperoxidase)YesYesTable 12. Morphology of Selected Viral PathogensVariables CMV Herpes AdenovirusRespiratory SyncytialVirusTissue responseInterstitial pneumonia Yes No No NoNecrotizing bronchopneumonia Yes/no Yes Yes YesBronchiolitis Rare Yes Yes YesTracheobronchitis No Yes Yes YesDAD Yes/no Uncommon Yes UncommonGiant-cell pneumonitis No No No YesInclusionsIntranuclear Yes Yes Yes NoCowdry A No Yes Yes NoGround glass No Yes No NoCytoplasmic Yes No No No“Smudge” cells No No Yes Nothree cells thick and is composed of small round,plasmacytoid, and transformed lymphocytes;• Grade 2, mild acute rejection: frequentperivascular inflammatory cell infiltratessurrounding venules and arterioles. Infiltratesare readily recognizable and are usually fiveto six cells in thickness in the vascular adventitiand no obvious infiltration is found byinflammatory cells into the adjacent alveolarseptae or airspaces;508 Pathology of Diffuse and Neoplastic Disease (Tazelaar)ACC-PULM-09-0302-032.indd 5087/10/09 8:54:36 PM


• Grade 3, moderate acute rejection: all of thehistologic features of mild acute rejection.Inflammatory cell infiltrate extends into perivascularand peribronchiolar alveolar septaeand airspaces;• Grade 4, Severe acute rejection: diffuse perivascular,interstitial, and airspace infiltratesof inflammatory cells, including neutrophils.Alveolar pneumocyte damage usually is associatedwith necrosis, airspace hemorrhage,and sometimes a necrotizing vasculitis. Antibody-mediatedrejection and pulmonary hemorrhage,capillaritis, and complement depositionare present.Neoplasia 5,40Squamous Cell CarcinomaGross findings vary from small to large,obstructive lesions that commonly are cavitated.Microscopic findings include keratinization andintercellular bridges. Variants include papillary,clear cell, small cell, and basaloid. Immunohistochemistryis thyroid transcription factor 1–, cytokeratin5/6, and p63.AdenocarcinomaAdenocarcinomas are more commonly peripheral.Microscopically, they are gland forming (majority)or contain mucin by special stains (solid variant).They can be acinar, papillary, solid, and bronchioloalveolar.Immunohistochemistry is chromogranin/synaptophysin: 10 to 20%, cytokeratin 7 andcytokeratin 20±, and thyroid transcription factor 1.k-ras mutations can be present.Bronchioloalveolar CarcinomaIn bronchioloalveolar carcinoma there are nodulesor areas of consolidation, more commonlyperipheral. Mucinous subtypes usually replaceentire lobe and show preservation of underlyingarchitecture with large pools of mucus within airspaces.Microscopic findings include nonmucinous,mucinous, and mixed variants. The epithelium iswell differentiated, uniform, and grows along intactalveolar walls, and there is no invasion into underlyingstroma. Multifocality with microsatellite lesionsis common. The differential diagnosis: atypicaladenomatous hyperplasia, (1) cytologic atypia is lessmarked, and (2) typically 0.5 cm.Small Cell CarcinomaA total of 70% of cases present as perihilar mass.Extensive lymph node metastases are common.Small cell carcinoma typically is peribronchial;endobronchial lesions are uncommon. Microscopically,small cell carcinomas are round-to-fusiformnuclei, nuclear molding, faint or absent nucleoli,and have scant cytoplasm. Extensive necrosis maybe found, and 5% of small cell carcinoma may becombined with a nonsmall cell component. Immunohistochemically,20% of cases may be negativeclassic for neuroendocrine markers.Large Cell CarcinomaLarge cell carcinomas are central or peripheraland are typically large. They appear microscopicallyas having a “sheets and nests” growth patternwith extensive necrosis, large nuclei with prominentnucleoli, and no evidence of squamous orglandular differentiation by light microscope. Variantsinclude large cell neuroendocrine carcinoma,basaloid, lymphoepithelioma-like, and clear cellcarcinoma. On electron microscopy, 80% showTable 13. Working Formulation for Classification and Gradingof <strong>Pulmonary</strong> Allograft Rejection*A. Acute rejection (perivascular, interstitial and alveolarinflammation)Grade 0 – NoneGrade 1 – MinimalGrade 2 – MildGrade 3 – ModerateGrade 4 – SevereB. Airway inflammationGrade 0 – NoneGrade 1R – Low gradeGrade 2R – High gradeGrade X – Not gradableC. Chronic airway rejection – bronchiolitis obliterans0 – Absent1 – PresentD. Chronic vascular rejection – accelerated graft vascularsclerosisAntibody-mediated rejection*From Stewart et al. 39<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 509ACC-PULM-09-0302-032.indd 5097/10/09 8:54:36 PM


glandular differentiation; 10% show squamousdifferentiation.Carcinoid TumorletA carcinoid tumorlet is usually an incidentalmicroscopic finding. Multiple tumorlets maymimic metastatic tumor, eg, breast cancer. 41,42 Theyrarely are associated with syndrome of airflowobstruction. 41 Microscopically, nests of neuroendocrinecells ( 0.5 cm) are found embedded infibrotic stroma. They are usually adjacent to thebronchiole.Neuroendocrine Cell HyperplasiaNeuroendocrine cell hyperplasia 43 is definedas increased neuroendocrine cells within bronchiolarepithelium.Typical and Atypical CarcinoidsMicroscopic, neuroendocrine cells with organoid,trabecular, insular, palisading ribbon, androsette-like architecture are found. They haveround-to-oval nuclei with finely granular chromatinand inconspicuous nucleoli. Sromal changesinclude bone, cartilage, dense fibrosis, and amyloid.Atypical carcinoids have 2–10 mitoses per 10high powered fields and/or focal necrosis.Large Cell Neuroendocrine CarcinomaMicroscopic findings include organoid, palisading,trabecular patterns; large, polygonal nucleiand low nuclear/cytoplasmic ratio with frequentnucleoli; and high mitotic rate ( 10 mitoses /10high powered fields). Necrosis can be prominent.Immunohistochemistry: chromogranin 80%,synaptophysin 40%, and bombesin 40%. Bydefinition, one of the aforementioned neuroendocrinemarkers must be positive.Lymphomatoid Granulomatosis-LymphomaLymphomatoid granulomatosis-lymphomawas described at a time when the existence ofextranodal lymphomas was not accepted, but thisdisease is a malignant lymphoma. 44,45 It appearswith a nodular consolidation; nodules may havecentral necrosis. Microscopically, nodules or diffuseinfiltrates of lymphoid cells appear; centralnecrosis and cavitation can be seen in largernodules; prominent vascular invasion is usuallyseen. The cell population is heterogeneous, butlarge atypical cells are malignant. Most cells arepositive for T-cell markers (CD4, CD8), but themajority of the malignant cells are B cells. Theyshow either immunoglobulin gene rearrangements(B-cell) or T-cell receptor rearrangements(in the few that represent T-cell lymphomas).Epstein−Barr virus DNA often is detected by insitu hybridization.References[Note to participants regarding references: there aresome primary sources quoted, but the textbookshave additional photographs worth studying.]1. Katzenstein A, Askin F. Surgical pathology of nonneoplasticlung disease. 4th ed. Philadelphia, PA:Saunders Elsevier, 20062. Parambil JG, Myers JL, Aubry MC, et al. Causes andprognosis of diffuse alveolar damage diagnosedon surgical lung biopsy. Chest 2007; 132:50–573. Travis W, Colby T, Koss M, et al. Non-neoplasticdisorders of the lower respiratory tract. In: Atlas ofnontumor pathology. Washington, DC: AmericanRegistry of Pathology and the Armed Forces Instituteof Pathology, 20024. American Thoracic Society/European RespiratorySociety International Multidisciplinary ConsensusClassification of the Idiopathic InterstitialPneumonias. This joint statement of the AmericanThoracic Society (ATS), and the European RespiratorySociety (ERS) was adopted by the ATS boardof directors, June 2001 and by the ERS ExecutiveCommittee, June 2001. Am J Respir Crit Care Med2002; 165:277–3045. Churg A, Myers J, Tazelaar H, et al. Thurlbeck’spathology of the lung. 3d ed. New York, NY:Thieme Medical Publishers, 20056. Katzenstein A, Askin F. Surgical pathology of nonneoplasticlung disease. 4th ed. Philadelphia, PA:W.B. Saunders, 2006; 5037. Katzenstein AL, Mukhopadhyay S, Myers JL.Diagnosis of usual interstitial pneumonia and distinctionfrom other fibrosing interstitial lung diseases.Hum Pathol 2008; 39:1275–1294510 Pathology of Diffuse and Neoplastic Disease (Tazelaar)ACC-PULM-09-0302-032.indd 5107/10/09 8:54:36 PM


8. Trahan S, Hanak V, Ryu JH, et al. Role of surgicallung biopsy in separating chronic hypersensitivitypneumonia from usual interstitial pneumonia/idiopathic pulmonary fibrosis: analysis of 31 biopsiesfrom 15 patients. Chest 2008; 134:126–1329. Atkins SR, Turesson C, Myers J, et al. Morphologicand quantitative assessment of CD20 B cell infiltratesin rheumatoid arthritis-associated nonspecificinterstitial pneumonia and usual interstitialpneumonia. Arthritis Rheum 2006; 54:635–64110. King T, Jr., Schwarz M, Brown K, et al. Idiopathicpulmonary fibrosis: relationship between histopathologicfeatures and mortality. Am J Respir CritCare Med 2001; 164:1025–103211. King T, Jr., Tooze J, Schwarz M, et al. Predictingsurvival in idiopathic pulmonary fibrosis: scoringsystem and survival model. Am J Respir Crit CareMed 2001; 164:1171–118112. Hanak V, Ryu JH, de Carvalho E, et al. Profusionof fibroblast foci in patients with idiopathic pulmonaryfibrosis does not predict outcome. Respir Med2008; 102:852–85613. Churg A, Müller NL, Silva CIS, et al. Acute exacerbation(acute lung injury of unknown cause) inUIP and other forms of fibrotic interstitial pneumonias.Am J Surg Pathol 2007; 31:277–28414. Parambil J, Myers J, Ryu J. Histopathologic featuresand outcome of patients with acute exacerbationof idiopathic pulmonary fibrosis undergoingsurgical lung biopsy. Chest 2005; 128:3310–331515. Kim D, Park JH, Park BK, et al. Acute exacerbationof idiopathic pulmonary fibrosis: frequency andclinical features. Eur Respir J 2006; 27:143–15016. Katzenstein A, Fiorelli R. Nonspecific interstitialpneumonia/fibrosis. Histologic features and clinicalsignificance. Am J Surg Pathol 1994; 18:136–14717. Travis WD, Hunninghake G, King TE Jr., et al. Idiopathicnonspecific interstitial pneumonia: report ofan American Thoracic Society project. Am J RespirCrit Care Med 2008; 177:1338–134718. Aubry M, Wright J, Myers J. The pathology ofsmoking-related lung diseases. Clin Chest Med2000; 21:11–35, vii19. Ryu J, Colby T, Hartman T, et al. Smoking-relatedinterstitial lung diseases: a concise review. EurRespir J 2001; 17:122–13220. Fraig M, Shreesha U, Savici D, et al. Respiratorybronchiolitis: a clinicopathologic study in currentsmokers, ex-smokers, and never-smokers. Am JSurg Pathol 2002; 26:647–65321. Vassallo R, Jensen EA, Colby TV, et al. The overlapbetween respiratory bronchiolitis and desquamativeinterstitial pneumonia in pulmonaryLangerhans cell histiocytosis: high-resolution CT,histologic, and functional correlations. Chest 2003;124:1199–120522. Myers J, Veal CF Jr., Shin MS, et al. Respiratorybronchiolitis causing interstitial lung disease. Aclinicopathologic study of six cases. Am Rev RespirDis 1987; 135:880–88423. Yousem S, Colby T, Gaensler E. Respiratory bronchiolitis-associatedinterstitial lung disease and itsrelationship to desquamative interstitial pneumonia.Mayo Clin Proc 1989; 64:1373–138024. Saldana M, Mones J. Lymphoid interstitial pneumoniain HIV infected individuals. Prog SurgPathol 1991; 12:181–21525. Koss M, Hockhholzer L, Langloss JM, et al. Lymphoidinterstitial pneumonia: clinicopathologicaland immunopathological findings in 18 cases.Pathology 1987; 19:178–18526. Cha S, Fessler MB, Cool CD, et al. Lymphoid interstitialpneumonia: clinical features, associationsand prognosis. Eur Respir J 2006; 28:364–36927. Yousem S, Colby TV, Chen YY, et al. <strong>Pulmonary</strong>Langerhans’ cell histiocytosis: molecular analysisof clonality. Am J Surg Pathol 2001; 25:630–63628. Churg A, Muller NL, Flint J, et al. Chronic hypersensitivitypneumonitis. Am J Surg Pathol 2006;30:201–20829. Ohtani Y, Saiki S, Kitaichi M, et al. Chronic birdfancier’s lung: histopathological and clinical correlation.An application of the 2002 ATS/ERS consensusclassification of the idiopathic interstitialpneumonias. Thorax 2005; 60:665–67130. Khoor A, Leslie KO, Tazelaar HD, et al. Diffusepulmonary disease caused by nonturberculousmycobacteria in immunocompetent people (hottub lung). Am J Clin Pathol 2001; 115:755–76231. Hanak V, Kalraa S. Hot tub lung: presenting featuresand clinical course of 21 patients. Respir Med2006; 100:610–61532. Aksamit T. Hot tub lung: infection, inflammation,or both? Semin Respir Infect 2003; 18:33–3933. Seymour J, Presneill J. <strong>Pulmonary</strong> alveolar proteinosis:progress in the first 44 years. Am J RespirCrit Care Med 2002; 166:215–23534. Trapnell B, Whitsett J, Nakata K. <strong>Pulmonary</strong> alveolarproteinosis. N Engl J Med 2003; 349:2527–2539<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 511ACC-PULM-09-0302-032.indd 5117/10/09 8:54:37 PM


35. Muir T, Leslie KO, Popper H, et al. Micronodularpneumocyte hyperplasia. Am J Surg Pathol 1998;22:465–47236. Karbowniczek M, Astrinidis A, Balsara B, et al.Recurrent lymphangiomyomatosis after transplantation:genetic analyses reveal a metastaticmechanism. Am J Respir Crit Care Med 2003;167:976–98237. Churg A, Green F. Pathology of occupationallung disease. 2nd ed. Baltimore, MD: Williams &Wilkins, 199838. Stewart S. Pathology of lung transplantation.Semin Diagn Pathol 1992; 9:210–21939. Stewart S, Fishbein MC, Snell GI, et al. Revision ofthe 1996 working formulation for the standardizationof nomenclature in the diagnosis of lung rejection.J Heart Lung Transplant 2007; 26:1229–124240. World Health Organization classification oftumours. In: Travis WD, et al., eds. Pathology andgenetics of tumours of the lung, pleura, thymusand heart. Lyon: IARC Press, 200441. Aubry MC, Thomas CF, Jett JR, et al. Significanceof multiple carcinoid tumors and tumorlets in surgicallung specimens: analysis of 28 patients. Chest2007; 131:1635–164342. Darvishian F, Ginsberg MS, Klimstra DS, et al. Carcinoidtumorlets simulate pulmonary metastasesin women with breast cancer. Hum Pathol 2006;37:839–84443. Armas OA, White DA, Erlandson RA, et al. Diffuseidiopathic pulmonary neuroendocrine cell proliferationpresenting as interstitial lung disease. AmJ Surg Pathol 1995; 19:963–97044. Liebow A, Carrington C, Friedman P. Lymphomatoidgranulomatosis. Hum Pathol 1972; 3:457–55845. Myers J, Kurtin PJ, Katzenstein AL, et al. Lymphomatoidgranulomatosis. Evidence of immunophenotypicdiversity and relationship to Epstein-Barrvirus infection. Am J Surg Pathol 1995; 19:1300–1312512 Pathology of Diffuse and Neoplastic Disease (Tazelaar)ACC-PULM-09-0302-032.indd 5127/10/09 8:54:37 PM


Pleural DiseaseSteven A. Sahn, MD, FCCPObjectives:• Understand the pathogenesis of pleural fluid (PF) formationin health and disease• Appreciate the clinical presentation, radiographic features,and the most common causes of pleural effusions• Appreciate the diagnostic value of PF analysis• Understand the management of patients with pleuraleffusions, especially malignant and parapneumoniceffusions• Understand the pathogenesis, causes, clinical features,and management of patients with spontaneous pneumothoraxKey words: effusions; empyema; exudates; malignant pleuralpleural disease; pneumothorax; transudatesPathogenesis of Pleural EffusionsNormal Pleural FluidIn the normal pleural space, there is approximately0.1 to 0.2 mL/kg body weight of pleural fluid (PF).Normal PF is clear, with 500 nucleated cells/L;the differential is approximately 2% neutrophils,0% basophils, 7 to 11% lymphocytes, 61 to 77%macrophages, and 9 to 30% mesothelial. PF pHhas been reported in the range of 7.60 to 7.64, andthere is a biocarbonate gradient (PF minus serumequals 8 mEq/L). 1,2AnatomyThe parietal pleura is supplied by systemiccapillaries (intercostal vessels) that are situatedclose to mesothelial surface. The visceral pleura issupplied by the bronchial circulation that is situateda greater distance from mesothelial surface. 3 Theformation of PF is a function of parietal pleura. Anultrafiltrate of parietal pleural capillaries increasesinterstitial pressure and promotes movement offluid into pleural space between the mesothelial celljunctions. PF is resorbed from the pleural spacewith movement through the stomata of parietalTable 1. Mechanisms of PF Formation in Disease*Mechanism Clinical Example ClassificationIncreased Pmv Congestive heart TransudatefailureDecreased peri-Pmv Atelectasis TransudateDecreased P Hypoalbuminemia TransudateIncreased capillarypermeability Pneumonia ExudateImpaired lymphaticdrainageYellow nailsyndrome ExudatePeritoneal-pleuralcommunicationHepatichydrothorax TransudateThoracic duct Chylothorax ExudaterupturePEEVO Duropleural fistula Transudate*PEEVO pleural effusion of extravascular origin;Pmv microvascular pressure; P oncotic pressure.pleura. The PF then moves into lymphatic lacunaeand subsequently lymphatic ducts and well-formedlymphatic channels and ultimately into mediastinallymph nodes. A clinically-relevant pleural effusiondevelops when PF formation exceeds resorption. 1The mechanisms for PF formation and clinicalexamples are shown in Table 1.Assessment of the Patient with aPleural EffusionSymptoms of Pleural EffusionsThe patient may present with symptoms suchas pleuritic chest pain (lupus pleuritis) or withoutsymptoms, eg, most patients with a benign asbestospleural effusion (BAPE). When patients with apleural effusion are symptomatic, dyspnea andchest pain are the most common findings. Dyspneamay be caused by a large or massive pleural effusionin a patient with normal lungs, a moderateeffusion in patients with some underlying lungdisease, and a small-moderate effusion in patientswith severe lung disease. A large effusion causesipsilateral mediastinal shift, depression of the<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 513ACC-PULM-09-0302-033.indd 5137/10/09 10:23:09 PM


ipsilateral diaphragm, outward movement of theipsilateral chest wall, and lung compression whenthere is no endobronchial lesion or fixed mediastinum.Dyspnea is perceived by the patient with alarge-to-massive pleural effusion because of itseffect on the previously mentioned structures withinput from neurogenic receptors in the lung andchest wall.Patients with a pleural effusion may presentwith pleuritic chest pain, which is associated withpleural inflammation. Pleuritic chest pain has beendescribed as having a “stitch in the side” or a “stabbing”or “shooting” pain that may be exacerbatedby deep inspiration, cough, or sneezing. Anymaneuver resulting in chest wall splinting, such asmanual pressure over the chest wall, will minimizethe pain. However, a splinting maneuver will notdifferentiate between other causes of pleuritic-likechest pain, such as rib fractures or from inflammationof the pleura per se.Value of PF AnalysisThe clinician should recognize that PF analysis(PFA) in isolation will establish a definitivediagnosis, such as malignancy or empyema, inonly a minority of patients. However, the numberand specific definitive diagnoses will vary withthe population being studied. In a prospectivestudy of 129 patients with pleural effusion, thoracentesisprovided a definitive diagnosis in only18% of patients and a presumptive diagnosis in55% of patients. In the remaining 27% of patients,the PF findings were unhelpful diagnosticallybecause the values were compatible with two ormore clinical possibilities. 4 However, in a numberof these patients, the findings were useful inexcluding other possible diagnoses, such as infection.Therefore, history and physical examination,radiologic evaluation, and ancillary blood testsare crucial in formulating a pretest diagnosticprobability. PFA is a valuable test that may notprovide a definitive diagnosis but can offer aconfident clinical diagnosis if there is a thoughtfulprethoracentesis evaluation. However, the moreknowledgeable the clinician is concerning PFA,the more likely undiagnosed pleural effusionswill decrease. The clinician who is more informedabout PFA should be able to “make a confidentclinical diagnosis” in virtually all patients with apleural effusion. 5 Diagnoses that can be established“definitively” by PF analysis are shown inTable 2. 5,6Virtually all patients with a newly discoveredpleural effusion should have a thoracentesis performedto obtain a diagnosis to guide management.Exceptions would be a secure clinicaldiagnosis, such as typical congestive heart failure(CHF) and a very small volume of fluid as withviral pleurisy. Observation is warranted in theaforementioned examples; however, if the clinicalTable 2. Diagnoses That Can Be Established “Definitively”by PFA*DiseaseDiagnostic PF TestsEmpyemaObservation (pus, putrid odor)MalignancyPositive cytology; high salivaryamylase level in absence ofesophageal ruptureLupus pleuritis Presence of LE cells; ANA 1:320;and pleural fluid/serum 1.0Tuberculous effusion Positive acid-fast bacilli stain orcultureEsophageal rupture High amylase (salivary), low PFpH; presence of food particlesor squamous epithelial cells oncytologyFungal effusion Positive potassium hydroxide stainor cultureChylothoraxTriglyceride 110 mg/dL (highlikelihood); presence of chylomicronsdefinitiveHemothoraxHematocrit (pleural fluid/blood 0.5)UrinothoraxCreatinine (pleural fluid/serum 1.0)Peritoneal dialysis Protein ( 0.5 g/dL) and pleuralfluid/ serum glucose ratio 2.0Rheumatoid pleurisy pH, 7.00, glucose 30 mg/dL andLDH 1,000 IU/L (highly likelyif empyema excluded); characteristiccytology definitiveDuropleural fistula Presence of 2-transferrinGlycinothorax Glycine in PF (complication oftransurethral surgery withglycine bladder infusion)EVM of a CVC Glucose PF/S 1.0 with glucoseinfusion; PF protein 0.5 g/dLwith saline infusion; white fluidwith high triglycerides with lipidinfusionBiliopleural fistula Green fluid; bilirubin PF/S 1.0*LDH lactate dehydrogenase; EVM of CVC extravascularmigration of central venous catheter.514 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5147/10/09 10:23:10 PM


situation changes or deteriorates, a thoracentesisshould be performed promptly. Approximately30 mL of PF is all that is necessary for a completePFA. In 2009, most clinicians use ultrasound guidancefor thoracentesis for safety purposes; theuse of ultrasonography virtually excludes thepossibility of a pneumothorax caused by needlepuncture.Observation of PFSome diagnoses can be established at the bedsideby visual inspection of the fluid (Table 3). Forexample, if pus is aspirated from the pleural space,the diagnosis of empyema is established and, ifthe pus has a putrid odor, anaerobic organismsare causative. A chylothorax can be suspected ifthere is milky fluid; however, this appearancecould also be caused by empyema or a cholesteroleffusion.Table 3. Observations of Pleural Effusions Helpful in Diagnosis*ObservationsColor of fluidPale yellow (straw)Red (bloody)White (milky)BrownBlackYellow-greenColor of enteral tubeFeeding or centralVenous line infusateGreenish tintCharacter of fluidPusViscousDebrisTurbidAnchovy pasteSatin-like sheenLooks like waterLooks like urineOdor of fluidPutridAmmoniaSuggested DiagnosisTransudate, some exudatesMalignancy, BAPE, PCIS, orpulmonary infarction if traumaexcludedChylothorax or cholesteroleffusion; EVM of CVC withlipid infusionLong-standing bloody effusion;rupture of amoebic liverabscessAspergillus niger infectionRheumatoid pleurisyFeeding tube has entered pleuralEVM of CVCBiliopleural fistulaEmpyemaMesotheliomaRheumatoid pleurisyInflammatory exudatesRupture of amebic liver abscessCholesterol effusionDuropleural fistulaUrinothoraxAnaerobic empyemaUrinothorax*See Table 2 for abbreviations not used in the text.The finding of 80% lymphocytes limitsthe possibilities of the exudative pleural effusion(Table 4). The most common of these diagnosesworldwide is a tuberculous pleural effusion.Although these diagnoses do not always have 80% lymphocytes, when this degree of lymphocytosisis present, one of these diagnoses is virtuallyalways present. 6PF eosinophilia is defined as a PF eosino -phil count of 10% of the total nucleated cells(Table 5). Interleukin-5 appears to be an importantchemotactic factor attracting bone marrowproducedeosinophils into the pleural space. Inpatients who require thoracotomy for spontaneouspneumothorax, eosinophilic pleuritis is commonlyencountered within hours of the pneumothorax.Although eosinophils appear to move rapidly intothe pleural space after a pneumothorax, after ahemothorax, eosinophils do not appear in PF for 7to 14 days. 7 Interestingly, PF eosinophilia is associatedwith delayed peripheral blood eosinophiliaafter a traumatic hemothorax that does not resolveuntil all eosinophils have left the pleural space.Table 4. Lymphocyte-Predominant ( 80%) Exudative PleuralEffusions*DiseaseTuberculous effusionChylothoraxLymphomaYellow nail syndromeChronic rheumatoidpleurisySarcoidosisPost-CABG effusions( 2 mo) including PCISAcute lung rejectionCommentusually 90 to 95% lymphocytes;acutely, may beneutrophil-predominant2,000 to 20,000 lymphocytes/L; most common cause isNHLOften 100% of nucleated cellsare lymphocytes; diagnosticyield on cytology or pleuralbiopsy higher with NHLA cause of benign persistenteffusion; protein discordantexudateOften associated with anunexpandable lungUsually 90% lymphocytes;effusion in 2% of sarcoidpatients; discordant exudate(by protein only)82 to 99% lymphocytes; unilateralon left or bilaterallyPF may be first manifestationof rejection*Consistently but not always 80%; other exudates rarelyhave 80% lymphocytes. NHL non-Hodgkin lymphoma.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 515ACC-PULM-09-0302-033.indd 5157/10/09 10:23:10 PM


Table 5. PFE*DiseasePneumothoraxHemothoraxBenign asbestospleural effusion<strong>Pulmonary</strong> embolismParasitic diseaseFungal diseaseDrug-inducedCarcinomaChurg-StrausssyndromeLymphomaTuberculous pleurisyPF Glucose and pHCommentLung and pleural tissue eosinophiliawithin a few hours afterpneumothoraxMay take 1–2 wk to develop afterblood enters pleural space30% incidence of PFE; up to 50%eosinophils in PFAssociated with infarction andhemorrhagic effusionParagonimiasis, hydatid disease,amebiasis, ascariasisHistoplasmosis,coccidioidomycosisDantrolene, bromocriptine,nitrofurantoin, valproic acid,and others5 to 8% with PFEPleural effusions in 30% withmarked PFE and peripheraleosinophiliaHodgkin disease (rare)Rare*PF eosinophils/total nucleated PF cells of 10%.PF glucose and pH should be measured on allexudative effusions because: (1) they can narrowthe differential diagnosis; (2) in a parapneumoniceffusion, they provide information helpful in managementstrategies; and (3) in malignant effusions,it provides information relating to extent of pleuralinvolvement with tumor, ease of diagnosis, prognosis,and management. 8 The finding of a pleuraleffusion with a pH 7.30 signifies substantialaccumulation of hydrogen ions in the pleuralspace. There is a direct relationship between thepH of PF and glucose; if the pH is low, the glucoseis low, and when the pH is normal, the glucose isnormal. This correlation suggests that the pathophysiologicprocesses responsible for this biochemicalphenomenon are interrelated (Table 6).The mechanism responsible for a low pH andglucose in complicated parapneumonic effusionsand empyema and esophageal rupture (an aerobicempyema) is an increased rate of glucose utilizationby neutrophil phagocytosis and bacterialmetabolism with the accumulation of the endproductsof glycolysis, CO 2, and lactic acid, causingthe pH to decrease. 9 In contrast, in malignant pleuraleffusions and chronic rheumatoid pleural effusions,the mechanism of low pH and glucose isrelated to an abnormal pleural membrane thatinhibits the efflux of CO 2and lactate from the pleuralspace rather than increased PF metabolism. 10Transudates vs ExudatesChemical analysis helps to determine wheth -er the fluid is an exudate, which results frominfection, inflammation, malignancy, or impairedzlymphatic drainage from the pleural space. Atransudate, which is caused by imbalances inhydrostatic and oncotic pressures, is associatedwith a normal pleura. Light and colleagues 11 haveproposed a diagnostic rule for categorizing pleuraleffusion as an exudate. The rule defines an exudativeeffusion if any one of the following criteria arefulfilled: (1) a PF lactate dehydrogenase (LDH)level 0.67, ie, the upper limit of normal for thelaboratory serum LDH value; (2) a PF/serum proteinratio 0.5; and (3) a PF/serum LDH ratio 0.6. The use of a three-test combination with“and/or” rule maximizes diagnostic sensitivity fordetecting exudative pleural effusions but lowersspecificity.A high diagnostic sensitivity is desirable whenscreening for conditions, such as exudative effusions,that have important clinical implications. Ametaanalysis of 1,444 patients 11 determined that allof the following tests have statistically similarTable 6. Diagnoses Associated With Pleural Fluid Acidosis (pH 7.30) and Low Glucose Concentration (PF/Serum 0.5)*Diagnosis pH Incidence (%) Glucose (mg/dL)CPPE and4.50–7.29 (~100) 0–40empyemaEsophageal rupture 5.50–7.00 (~100) 0–59Chronic tuberculous 6.90–7.05 (100) 0–30empyemaChronic rheumatoid 7.00 (~100) 0–30pleurisyMalignancy 6.95–7.29 (30) 30–59Tuberculous7.00–7.29 (20) 30–59effusionLupus pleuritis 7.00–7.29 (20) 30–59*CPPE complicated parapneumonic effusion.516 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5167/10/09 10:23:10 PM


diagnostic accuracy compared with the Light criteria:PF protein concentration 3.0 g/dL; PF/serum protein ratio 0.5; PF LDH greater thantwo-thirds the upper limits of normal of serumLDH; PF/serum LDH ratio 0.6; PF cholesterol 45, 54, 55, or 60 mg/dL; PF/serum cholesterolratio 0.3; and albumin gradient 1.2 g/dL. 12In a receiver operating characteristic analysisof 200 consecutive patients with pleural effusions,a PF LDH of 163 IU/L (serum upper limits of normal,200; ratio 0.82) was the best cutoff point foran exudate (area under the curve, 0.89), followedby PF/serum total protein ratio of 0.5 (area underthe curve, 0.86). 13 If serum tests are not available,single tests or test combinations that rely only onPF results can be used with an excellent diagnosticaccuracy. A Bayesian approach with application oflikelihood ratios to pretest estimates of the probabilityof an exudative effusion improves diagnosticaccuracy. For example, a PF/serum protein ratioof 0.70 has a much greater likelihood of being anexudate than a ratio of 0.51 (Tables 7, 8). 14TransudatesCHFCHF is the most common transudate and mostcommon cause of pleural effusions in the elderly.CHF PF is a consequence of pulmonary venoushypertension; a pulmonary capillary wedge pressure 24 mm Hg in the acute setting is stronglyassociated with bilateral pleural effusions. 15 Thesepatients have the classic signs and symptoms ofCHF; fine crackles usually are present, most prominentlyin the bases.The chest radiograph most commonly showsbilateral effusions (R L) with cardiomegaly andevidence of interstitial and alveolar edema. KerleyB lines may be observed. An isolated right effusion,8%, and isolated left effusion, 4%, may occur. Theseverity of pulmonary edema correlates directlywith the volume of PF. 16 Unless CHF develops afteran acute myocardial infarction, the absence of cardiomegalyin a patient with bilateral effusionsshould raise the possibility of another cause of thebilateral effusions.PFA reveals a serous fluid with 500 mononuclearcells/L. Serum and PF glucose areequivalent and range from pH 7.45 to 7.55. BothTable 7. Causes of Transudative Pleural Effusions*DiseaseCHFHepatic hydrothoraxNephrotic syndromePeritoneal dialysis(CAPD)HypoalbuminemiaUrinothoraxAtelectasisConstrictivepericarditisTrapped lungSuperior vena cavalobstructionDuro-pleural fistulaVentriculopleuralshuntVentriculoperitonealshuntEVM of CVCCommentDiuresis can result in exudate byprotein and LDHOccurs frequently without clinicalascitesSmall, subpulmonic, and bilateraleffusionsSmall, right-sided effusion mostcommon; acute, massive, righthydrothorax can occur shortlyafter initiating dialysisAnasarca virtually alwayspresent; serum albumin level, 1.5 g/dLCaused by ipsilateral obstructiveuropathySmall effusion caused bydecreased peri-PMVBilateral effusions due topulmonary and systemichypertensionA result of remote pleuralinflammationProbably caused by acute systemicvenous hypertensionCSF leak into pleural space afterdisk surgery or trauma30% incidence of symptomaticeffusionsEffusion can form with intrathoracicmigration of catheterMost common with left subclavianlines; transudate withsaline solution or glucoseinfusion*CAPD continuous ambulatory peritoneal dialysis; CSF cerebrospinal fluid. See Table 2 for other abbreviations notused in the text.the PF protein and LDH may reach “exudative”levels after diuresis, which can be confirmed 16 withuse of the serum- PF albumin gradient ( 1.2 supportsa transudate). 17The diagnosis of pleural effusions from CHF ispresumptive with the compatible clinical presentation.However, an elevated brain natriureticpeptide and compatible echocardiogram demonstratinga low left ventricular ejection fraction ordiastolic dysfunction supports the diagnosis. 18Treatment includes reducing microvascularpressure with diuretics, angiotensin-convertingenzyme inhibitors, spironolactone, and digitalis.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 517ACC-PULM-09-0302-033.indd 5177/10/09 10:23:11 PM


Table 8. Causes of Exudative Pleural EffusionsInfectionBacterial pneumoniaTuberculousParasitesFungal diseaseAtypical pneumoniasNocardia, ActinomycesSubphrenic abscessRupture of amebic liver abscessHepatic abscessSplenic abscessViral hepatitisSpontaneous esophageal ruptureMalignancyCarcinomaLymphomaLeukemiaMesotheliomaSarcoma (angio, Kaposi)Inflammatory and othersPancreatic disease (acute; pancreaticopleural fistula)BAPE<strong>Pulmonary</strong> embolismRadiation therapyUremic pleuritisSarcoidosisPCISHemothoraxARDSConnective tissue disease/vasculitisLupus pleuritisRheumatoid pleurisyMixed connective tissue diseaseChurg-Strauss syndromeWegener granulomatosisFamilial Mediterranean feverLymphatic abnormalitiesChylothoraxYellow nail syndromeLAMNoonan syndromeLymphangiectasiaMovement of fluid from abdomen to pleural spaceAcute pancreatitisPancreaticopleural fistulaMeigs syndromeCarcinomaLymphomaChylous ascitesLung entrapmentCholesterol effusion (TB; rheumatoid pleurisy)CPE/empyemaCABG SurgeryEndocrine dysfunctionHypothyroidismOvarian hyperstimulation syndromeIatrogenicEsophageal instrumentationEsophageal sclerotherapyCentral venous catheter misplacement/migrationDrug inducedEnteral feeding tube misplacement into pleural spaceLeakage of percutaneous biliary drainage tubeCABG surgeryIf medical management is not completely effective,a unilateral pleurodesis for these medically refractiveeffusions is a reasonable option; however,bilateral pleurodesis is probably contraindicatedbecause of the risk of the patient not being able toclear the heart failure fluid into the pleuralspaces.Hepatic HydrothoraxHepatic hydrothorax is defined as a pleuraleffusion in the setting of cirrhosis of the liver withportal hypertension and typically hypoalbuminemia;other causes of a transudative effusion mustbe excluded. 19 Although clinical ascites is usuallypresent, up to 16% of patients develop an hepatichydrothorax in the absence of clinical ascites. 20 Thepathogenesis of these effusions is attributable totransdiaphragmatic movement of the ascitic fluidinto the pleural space through congenital diaphragmaticdefects that are rendered patent by theincreased peritoneal pleural pressure.The patients typically have stigmata of cirrhosisand present with the insidious onset of dyspneawith exertion. The chest radiograph usually demonstratesa small-to-moderate right effusion 70%of the time; an isolated left pleural effusion andbilateral effusions have each been reported inapproximately 15% of patients. 21 The heart is typicallynormal in size and massive effusions, occupyingthe entire hemithorax, occur in only 5% ofpatients. PFA demonstrates a serous or sanguinousfluid with 500 mononuclear cells/L. PF serumand glucose are equivalent and the pH ranges from7.45 to 7.55.518 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5187/10/09 10:23:11 PM


The diagnosis is presumptive in the properclinical setting with evidence of a low PF total protein(commonly 1.0 g/dL) and an LDH clearly inthe transudative range. The clinical diagnosis canbe confirmed by demonstrating radionuclide passagefrom the ascitic fluid to the pleural space.Medical treatment entails sodium restriction,a loop diuretic, and spironolactone. Approximately25% of patients have a refractory hepatic hydrothorax.22 If medical treatment fails, transjugularintrahepatic portosystemic shunt should beconsidered to manage the refractory ascites. 23Video-assisted thoracoscopic surgery to patchdiaphragmatic defects and perform talc poudragehas been used but is associated with a high rate ofmorbidity and mortality in these very ill patients.Chemical pleurodesis via chest tube has a low successrate. Furthermore, chest tube drainage in theseindividuals is contraindicated because it promotesinfection, depletes protein and lymphocytes, andincreases the risk of renal failure and often createsa prolonged fluid leak after removal of the chesttube. Liver transplantation is the best treatment fordecompensated hepatic cirrhosis and, therefore,for patients with hepatic hydrothorax.Clinical manifestations that should suggest thatthe patient may have spontaneous bacterial pleuritisare increased temperature, new-onset dyspnea,abdominal pain, and encephalopathy. 24 Inthis situation, an immediate thoracentesis withbedside inoculation of blood culture bottles shouldbe performed.AtelectasisAtelectasis is a common cause of pleural effusionsin the ICU (40% of patients on admission and60% during medical ICU stay) 25 ; pleural effusionsare also observed postoperatively, especially afterupper-abdominal surgery, thoracotomy, and coronaryartery bypass grafting (CABG) surgery. Atelectasismay also be observed after pulmonaryembolism without an infarction, an endobronchialobstruction from lung cancer, acute pancreatitis,splenic infarction, and subphrenic or hepaticabscess and the absence of a sympathetic effusion.The pleural effusion from atelectasis developsfrom decreased perimicrovascular pressure thatfavors fluid movement from the parietal pleuralinterstitium into the pleural space until the pressuregradient returns to normal.Clinically, patients with an atelectatic pleuraleffusion may be asymptomatic or complain ofdyspnea or chest pain. These patients may have anincreased alveolar-arterial (a-a) oxygen (O 2) gradient.The chest radiograph typically shows smallunilateral or bilateral effusions with minimal evidenceof volume loss with a normal heart size andclassic plate-like densities.PF analysis shows a serous fluid with 500mononuclear cells/L with a pH 7.45 and aglucose level similar to serum glucose. The diagnosisis presumptive based on the patient’s presentation.Treatment depends on the underlyingdisease and chest physiotherapy typically resultsin rapid resolution in the post-surgical patients.Nephrotic SyndromeEffusions develop in 20 to 25% of patients withnephrotic syndrome. The development of theseeffusions is associated with the degree of hypoalbuminemia.The effusions form because of decreasedoncotic pressure, which at times may be associatedwith an increase in perimicrovascular pressure.Individuals with nephrotic syndrome and pleuraleffusions are typically asymptomatic because theeffusions are typically small. Patients become dyspneicwhen they develop volume overload. Theseeffusions tend to occur when the albumin concentrationis reduced to 1.8 g/dL. 26 The chest radiographtypically shows small bilateral effusions that arefrequently subpulmonic and a normal heart size. OnPFA, the fluid is serous with 500 mononuclearcells/L, a pH 7.45, and a PF and serum glucosethat are equivalent. The diagnosis is presumptive inthe appropriate clinical setting with a low serumalbumin. These patients, who are in a hypercoagulablestate because of a loss of clotting inhibitors inthe urine, may develop renal vein thrombosis andsubsequent pulmonary embolism. <strong>Pulmonary</strong> embolismshould always be considered in patients withnephrotic syndrome who develop pleuritic chestpain when the pleural effusion is significantly largeron one side orthe PF is hemorrhagic, exudative, orneutrophil predominant. 27 Management entails treatingthe underlying disease and preventing volumeoverload.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 519ACC-PULM-09-0302-033.indd 5197/10/09 10:23:11 PM


Trapped LungTrapped lung, the end stage of an infectious orinflammatory pleural effusion, presents as a persistenttransudative effusion months to years afterthe acute pleural process. It is a relatively uncommoncause of a pleural effusion, representing 5%of effusions in the Medical University of SouthCarolina database of 1,000 patients. Trappedlung is one of the three forms of an unexpandablelung; the other two causes of an unexpandablelung are an endobronchial lesion, usually secondaryto carcinoma of the lung, and chronic atelectasis,which may eventually resolve over timewithout intervention. Trapped lung is caused byvisceral pleural restriction when a fibrous membranecovers a portion of the visceral pleura,preventing lung expansion to the chest wall.The increased negative intrapleural pressurefavors PF formation until a new steady state isestablished. 28Patients with trapped lung may present withan asymptomatic pleural effusion seen on a routinechest radiograph or with dyspnea on exertionmonths to years after the acute pleural injury.Common causes of trapped lung include a complicatedparapneumonic effusion or empyema,hemothorax, tuberculous pleural effusion, uremicpleuritis, rheumatoid pleurisy, and after CABGsurgery. During a therapeutic thoracentesis,patients often develop severe, anterior, substernalchest pain as a result of the decrease in pleuralpressure, which is quickly resolved by allowingair to enter through the pleural catheter. 29 Thepleural effusion in trapped lung after thoracentesisreaccumulates relatively rapidly during aperiod of 72 h and does not result in relief ofdyspnea.The chest radiograph typically shows a smallto-moderateunilateral effusion without significantmediastinal shift. 29 PFA shows a serous fluidwith a low nucleated cell count, typically 500/L,and a mononuclear predominance characterizedby a lymphocytosis 50% of the total nucleatedcells. On a number of occasions, the protein is inthe range expected for an exudate. We still believethat the fluid should be considered a transudateand that the pleural protein concentration dependson the plasma protein concentration, the proteinreflection coefficient, solvent filtration into thepleural space (wash-down effect), and bulk flowof fluid via the pleural lymphatics. Finding atransudate by these criteria in effusions of vascularorigin almost always indicates a normalprotein reflection coefficient but also increasedfiltration and increased lymphatic bulk flowconditions, narrowing the differential diagnosisconsiderably.Although the PF of trapped is by definition ofvascular origin and the absence of inflammationand malignancy implies a normal protein reflectioncoefficient, solvent filtration may not be increased,and lymphatic bulk flow may be impaired. On thebasis of these considerations, we do not interpretan isolated increase in PF protein concentration asunequivocal evidence of an abnormal proteinreflection coefficient that would imply activeinflammation. 29 The diagnosis of a trapped lung isestablished by finding an initial negative pleuralliquid pressure on manometry that decreased precipitouslyafter removal of PF. These patientstypically have a pleural elastance of 15 cm of H 2Oper liter of fluid removed. Failure of the lung toexpand on chest radiograph after removal of allfluid, in the absence of bronchial obstruction, alsosubstantiates the diagnosis.Treatment of trapped lung is reassurance if thepatient is asymptomatic and decortication ifthe patient is symptomatic and the underlying lungis normal. Decortication for a trapped lung may besuccessful years after the diagnosis.HypoalbuminemiaThe precise incidence of hypoalbuminemia isunknown. However, it is commonly seen in hospitalizedpatients with AIDS, patients with otherchronic illnesses, and those in medical ICUs. Thepathogenesis of these effusions is decreasedoncotic pressure that generally requires a serumalbumin of 1.8 g/dL. It is highly unlikely apatient with hypoalbuminemia presents with anisolated pleural effusion without concomitantanasarca.The patient typically has small bilateral pleuraleffusions and is generally not dyspneic at rest orwith minimal activity; however, those patients withsignificant underlying pulmonary disease or heartdisease may be breathless on exertion. The chestradiograph typically shows small-to-moderate520 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5207/10/09 10:23:12 PM


ilateral effusions with a normal heart size. OnPFA, the clinician will find paucicellular levels (ie, 500 total nucleated cells/L) that are predominantlymononuclear. PF pH is 7.45 to 7.55, and thePF glucose is equivalent to the serum glucose. LDHand protein are clearly in the transudative range.Hypoalbuminemia-induced pleural effusions area clinical diagnosis in a patient with a very lowserum albumin and when other possible causes ofthe effusions have been excluded. The treatmentis to maximize nutrition and prevent protein losseither through the GI track or kidneys.Continuous Ambulatory Peritoneal DialysisSmall right-sided, or less commonly, bilateralpleural effusions develop in approximately 2% ofpatients on continuous ambulatory peritonealdialysis (CAPD). Acute, massive, right-sided pleuraleffusions may occur, usually in women in1 month (range 1 day to 2 years) after startingperitoneal dialysis. 30 The pathogenesis of theseeffusions is peritoneal to pleural movement of fluidthrough diaphragmatic defects that becomeenlarged from acute increases in peritoneal pressureduring dialysis. Multiparity and peritonitisare risk factors for the development of pleural effusionswith CAPD.The majority of these patients are asymptomatic,but some with enlarging effusions haveinsidious onset of dyspnea with exertion. Patientswith acute massive effusions may present withacute respiratory failure and are often diagnosedas uncontrolled congestive heart failure. 31The chest radiograph typically shows a smallright-sided or bilateral effusions or a massiveeffusion with an acute hydrothorax. PFA showsa serous fluid with a very low total protein (usually 0.5 g/dL) and glucose in the range of 200to 2,000 mg/dL. The LDH ranges from 6 to15 IU/L and the fluid contains 100 mononuclearcells/L. Diagnosis is established with a highdegree of clinical confidence in the patient onCAPD who has a PF total protein value of 0.5g/dL and a PF/serum glucose ratio 2. Treatmentis observation for patients who are asymptomaticwith small pleural effusions. Patientswith a large refractory symptomatic effusion orwith an acute massive right effusion should bechanged to hemodialysis. There have been reportsof video-assisted thoracoscopic surgery pleurodesisused to successfully manage patients onCAPD. 32UrinothoraxThe precise incidence of urinothorax isunknown; however, approximately 40 cases havebeen reported in the literature. It is highly likelythat many cases are underdiagnosed and underreported.The most common cause of urinothoraxis obstructive uropathy, where urine exudes fromthe capsule of the kidney and moves into the mediastinumor directly into the pleural space via thediaphragm due to a pressure gradient. Reportedcauses of urinothorax include bladder and prostatecancer, posterior urethral valves, renal cysts, nephrolithiasis,surgical ureteral manipulation, blunttrauma to the kidney, renal transplantation, ilealconduit with ureteral obstruction, bladder laceration,and pregnancy. 33,34The most common presenting symptom of aurinothorax is dyspnea after surgery or trauma.Occasionally a small asymptomatic urinothorax isfound on a routine postoperative chest radiograph.The interval between the precipitating event andurinothorax has been reported to be from 8 h to2 months, with most patients diagnosed within 48 hof the event. 33The chest radiograph typically shows a smallto-moderatepleural effusion ipsilateral to theobstructed kidney; however, there are reports ofbilateral and contralateral effusions. 33,34 In mostinstances, there is no other abnormality on thechest radiograph.On PFA, the fluid looks like and smells likeurine. It has between 23 and 1,500 mononuclearcells/L and typically has a protein 1.0 g/dL;the pH ranges from 7.00 to 8.00 and PF and serumglucose concentrations are equivalent. 35 The diagnosisis established by finding a PF/serum creatininedifference of 1.0; the range reported in theliterature is 1.1 to 15.7. The PF/serum creatininedifference needs to be measured before relief of theobstruction because the latter will cause a rapidreversal in the PF/serum creatinine difference.Urinothorax is the only cause of a low pH ( 7.30)transudate with a normal arterial pH. 35 Treatmentis relief of the urinary obstruction, which results ina rapid reversal of the PF/serum creatinine<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 521ACC-PULM-09-0302-033.indd 5217/10/09 10:23:12 PM


difference, an increase in PF pH, and rapid resolutionof the pleural effusion.ExudatesParapneumonic EffusionsA parapneumonic effusion is defined as apleural effusion associated with pneumonia orlung abscess. An uncomplicated parapneumoniceffusion is a small-to-moderate effusion that resolveswith antibiotics only without pleural spacesequelae. A complicated parapneumonic effusionrequires pleural space drainage for resolution ofpleural sepsis. An empyema, or pus in the pleuralspace, represents the end stage of a complicatedparapneumonic effusion and always requires pleuralspace drainage. A parapneumonic effusion isthe most common cause of an exudative effusion.A total of 40 to 57% of patients with bacterial pneumoniawill develop parapneumonic effusions.However, only a small percentage (10 to 15%) havecomplicated effusions, and empyema develops inonly 5 to 10%. 36The stages of a parapneumonic effusion are asfollows: (1) exudative (capillary leak) with an estimatedtime period of 5 to 7 days; (2) fibrinopurulentor bacterial invasion and fibrin formationstage, which tends to occur after 7 days up to2 weeks; and (3) the organizational or empyemastage, which generally occurs within 2 to 4 weeksof onset of the pleural effusion. 36,37 The most commonorganisms causing empyema are anaerobicbacteria, staphylococcus, Gram-negative aerobes,and the pneumococcus. At some time during thefibrinopurulent stage, the clinician loses the abilityto treat the patient with antibiotics alone, anddrainage is required.The patient initially presents with symptomstypical of pneumonia: fever, pleuritic chest pain,cough with purulent sputum, and increasing dyspneawith exertion. The presence of these symptomscannot determine whether the patient has acomplicated or uncomplicated parapneumoniceffusion, and therefore chest imaging, clinical factors,and PFA are often helpful in this determination.36,37 An effusion 40% of the hemithorax onchest radiograph, the presence of an intrapleuralair-fluid level, the presence of loculation or multiloculations,marked pleural thickening 5 mm, orenhancement of the pleural membranes on chestCT scan increase the likelihood that the patient hasa complicated parapneumonic effusion or empyema.An anaerobic infection, prolonged pneumoniahistory, failure to respond to antibiotic therapy,virulence of the underlying bacterial pathogen, andhypoalbuminemia also suggest the presence of acomplicated parapneumonic effusion. 36,37On PFA, the finding of purulent fluid establishesthe presence of an empyema and alwaysrequires pleural space drainage. A positive bacterialculture, low PF pH of 7.20 to 7.30, a low glucoseconcentration 60 mg/dL, or a PF to serum ratio 0.5, or a high LDH 1000 IU/L are strong indicatorsfor the need for pleural space drainage. Itappears that the pH is the most sensitive of theseparameters. 36,38The diagnosis of a parapneumonic effusion isrelatively straightforward. In the clinical setting ofpneumonia, PFA shows a neutrophil-predominant,concordant exudate (by both LDH and protein criterion)supporting the diagnosis of a parapneumoniceffusion. Uncomplicated parapneumonic effusionsrequire antibiotics only without dose escalation.Complicated parapneumonic effusions requireappropriate antibiotics and pleural space drainage,which may be performed with a small-bore chesttube inserted under ultrasound guidance. 36,37,39,40 Theuse of fibrinolytic agents is controversial and clearlydepends on the stage of the parapneumonic effusions.In the early fibrinolytic stage, fibrinolyticsmay shorten the clinical course. In establishedempyemas, the Multicenter Intrapleural Streptokinasetrial showed that streptokinase was of nobenefit in decreasing the rate of mortality or needfor surgery in a large number of patients. 41 Empyemectomyand decortication by thoracoscopy orthoracotomy is necessary in patients with establishedempyemas to resolve pleural sepsis and toprevent severe pleural fibrosis or development ofan unexpandable lung. In patients who are toodebilitated to undergo surgery, open drainage isan option. 36,40Viral and Mycoplasma PneumoniaPleural effusions caused by viral infectionstend to be small and self-limited and may developin up to 20% of patients. However, because of thesmall volume of PF and spontaneous resolution in522 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5227/10/09 10:23:12 PM


most patients; PF is usually not sampled. Themajority of institutions are not equipped to performviral cultures and, therefore, these effusionsare typically not recognized. However, a pleuraleffusion develops in approximately 40% of patientswith adenovirus 3 and 7. Adenovirus infection isendemic throughout the year and occurs in all agegroups, although it is more common in school-agedchildren. 42The PF is a paucicellular exudate with a predominanceof mononuclear cells. In the acutephase, there may be an increased number of neutrophils.The diagnosis is established by documentingincreasing complement fixation titers orculturing the virus from the fluid.Hantavirus, which has been reported predominantlyfrom the Four Corners area of theUnited States (ie, New Mexico, Arizona, Utah, andColorado) is the virus transmitted to humans viacontact with deer mice. Pleural effusions are causedby either cardiac dysfunction or increased vascularpermeability. The fluid can be either transudate inthe initial phase with cardiac dysfunction or anexudate with vascular leak. Chest radiographshows a normal-sized heart and bilateral infiltrateswith blood tests documenting thrombocytopenia,a left shift in the myeloid series, and enlargedimmunoblastoid lymphocytes. 43Patients with Mycoplasma pneumoniae infectioncan develop bronchopneumonia involving one ormultiple lobes. A small pleural effusion occurs in5 to 20% of patients. 44 The diagnosis is establishedby finding specific antibody titers or by isolatingM pneumoniae from the PF. Positive polymerasechain reaction results from PF samples are associatedwith the abnormality observed on chestradiographs. 45 Macrolides are the drug of choice,and no specific treatment is necessary for the pleuraleffusion.Malignant Pleural Effusions (Carcinoma,Lymphoma )Malignant pleural effusions are the secondmost common cause of exudates. Lung and breastcarcinoma represent approximately 60% of allmalignant pleural effusions, with lymphomarepresenting approximately 10%. 46,47 Malignantpleural effusions in lung cancer result from ipsilateraltumor invasion into the pulmonary arterywith embolization to the visceral pleural surface.Malignant cells migrate into the pleural space andadhere to the parietal pleural surface. Bilateralmalignant pleural effusions in lung cancer developfrom liver metastasis and the same pathologicprocess in the contralateral side. 47 With breastcancer, an ipsilateral malignant effusion occurswhen there is metastasis through the chest walllymphatics or hematogenous spread via the liver.In Hodgkin lymphoma, the effusion is predominantlycaused by a lymphatic obstruction, whereasin non-Hodgkin lymphoma there is more commonlydirect pleural invasion. 47 Non-Hodgkinlymphoma is one of the most common causes ofa chylothorax.Bilateral pleural effusions in the setting ofmalignancy are associated with mediastinal nodeinvolvement or bilateral parenchymal metastasis.Bilateral effusions also may occur years after mediastinalradiation secondary to either constrictivepericarditis or mediastinal lymphatic fibrosis.Lymphangitic carcinomatosis is seen with adenocarcinomaof the lung, breast, stomach, pancreas,prostate, and thyroid. A paramalignant effusion isdefined as PF developing in the setting of a knownmalignancy but malignant cells are not found inthe PF. Lymphatic obstruction appears to be themost important mechanism of a paramalignanteffusion and is also associated with larger effusions.47 In most series of patients with malignanteffusions, 5% are transudates. Transudates maybe caused by early lymphatic obstruction or endobronchialobstruction resulting in atelectasis orconcomitant congestive heart failure. The courseof a paramalignant effusion and a malignant effusionhas not been compared.Patients with a malignant effusion generallypresent with the insidious onset of exertionaldyspnea and cough, typically with a known malignancy.47 Chest radiographs may show a largeto-massivepleural effusion with contralateralmediastinal shift or bilateral effusions with anormal heart size or an apparent large effusionwith ipsilateral shift. Any unilateral effusion inan elderly patient should suggest a possiblemalignancy. A massive pleural effusion (occupyingthe entire hemithorax) is most commonly tothe cause of malignancy. Bilateral malignanteffusions are most common with a nonlung primary.With an apparent large effusion without<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 523ACC-PULM-09-0302-033.indd 5237/10/09 10:23:13 PM


contralateral shift, the most likely diagnosis islung cancer with obstruction of the ipsilateralmainstem bronchus. The classic triad of lymphangiticcarcinoma is ipsilateral mediastinaladenopathy, Kerley B-lines, and a pleuraleffusion.PFA is variable in a malignant pleural effusion.The fluid may be serous or grossly hemorrhagicand is typically a concordant exudate.Approximately 10 to 14% of patients with amalignant pleural effusion will have an elevatedPF amylase (salivary) concentration that is virtuallydiagnostic of adenocarcinoma of the lung inan individual who clearly does not have esophagealperforation. A low PF pH and glucose in amalignant pleural effusion is associated withpoorer survival, a greater yield on initial PF cytology,and poorer response to chemical pleurodesisthan those with normal pH malignant pleuraleffusions. 7,46–48A malignant pleural effusion is diagnosed byfinding malignant cells either in PF or on pleuralbiopsy. The yield of cytology and pleural biopsydepends on the stage of the disease, with the moreadvanced disease being associated with greatersensitivities; therefore, the sensitivity of cytologyranges from 45 to 90% and percutaneous pleuralbiopsy from 30 to 50% with thoracoscopy beingthe most sensitive diagnostic test with yields of95 to 100% depending on the expertise of theoperator. 46Treatment of patients with malignant pleuraleffusion is palliative; outpatient therapeutic thoracentesisfor patients with far-advanced disease andcontralateral mediastinal shift is appropriate.Chemical pleurodesis with talc poudrage or slurryfor refractory effusions with dyspnea relieved bythoracentesis is an additional option. An indwellingcatheter for failed pleurodesis or lung entrapmentis an outpatient procedure, and the patientcan manage his or her effusion at home. 49 Pleurectomyis rarely performed for failed pleurodesisunless there is an expected survival 6 months.All patients with a suspected malignant effusionshould undergo pleural manometry to determinewhether the lung is completely expandable, makingthe patient a candidate for pleurodesis. If thelung is not fully expandable, pleurodesis is futile.In these patients, an indwelling catheter should beplaced for palliation. 49Malignant MesotheliomaMalignant mesothelioma is the most commonprimary malignant tumor of the pleura that arisesfrom the mesothelial surfaces of the pleura. Mesotheliomamay also occur in the peritoneal cavityand pericardium. The estimated incidence of mesotheliomain the United States is 2,200 cases/year,and it appears to be on the rise. The predominantcause of a malignant mesothelioma is exposure toasbestos, with approximately 40% of malignantmesothelioma patients having a documentedasbestos exposure. The lifetime risk of mesotheliomais estimated to be as high as 18 to 30% ofasbestos workers. 50 However, there is no directcorrelation to the amount and duration of asbestosexposure. The latency period from asbestos exposureto the development of mesothelioma variesfrom 20 to 50 years.The World Health Organization classificationof pleural tumors recognizes four major histologicsubtypes: epithelioid, sarcomatoid, desmo plastic,and biphasic. 50 The epithelioid variant ismost common. Immunohistochemistry has essentiallyreplaced electron microscopy as the goldstandard for diagnosis. Pathologists typically usea panel of markers to confirm the diagnosis. Calretinin,CK5-6, Wilms’ tumor-1 antigen, and D2-40positivity support the diagnosis of malignantmesothelioma. 51The majority of patients with malignant mesotheliomapresent in the fifth to seventh decades oflife. The most frequent presenting symptom isnonpleuritic chest pain (60%) followed by dyspnea(25%) and cough (20%). Some patients are asymptomaticat diagnosis with a unilateral pleural effusionfound on a routine chest radiograph. Thedisease is usually unilateral.The initial radiographic manifestation is frequentlya large unilateral pleural effusion withcontralateral shift; the majority of the lesions areright-sided. Mesothelioma also may present as apleural mass or with diffuse pleural thickening.A small percentage of patients will show evidenceof asbestosis. In the later stages of disease, thepatient develops ipsilateral mediastinal shiftas the lung is encompassed by the tumor. CTwill detect invasion of the chest wall, ribs, andmediastinal structures and can be used forstaging.524 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5247/10/09 10:23:13 PM


There are no specific biomarkers in PF for thediagnosis of malignant mesothelioma. The effusionsare typically exudative with protein concentration 4 g/dL and are lymphocyte predominant.PF LDH levels often are 600 IU/L. Patients withadvanced disease present with extensive pleuralinvolvement and a low PF pH and glucose, portendinga poor prognosis as well as a poor responseto pleurodesis. Other poor prognostic factors at thetime of diagnosis include thrombocytosis, leukocytosis,anemia, fever, sarcomatoid or mixed histology, 65 years of age, poor performance status,and male gender. 50,52Recent chemotherapy trials have demonstratedsome evidence of antitumor activity withanthracyclines, platinum derivatives, and antimetabolites.Pleurodesis may be effective early in thecourse if lung entrapment is not present. Pleurectomyhas been shown to be more successful thantalc pleurodesis in reducing the recurrence of thepleural effusion. There are some long-term survivorsafter extrapleural pneumonectomy, whenthere is a component of a multimodality treatmentsuggesting that this procedure may alter thenatural history of the disease in appropriatelyselected patients who are in early stages. 53 Otherapproaches to treatment are still in clinical trialsincluding immunotherapy, targeted therapy, andgene therapy. 50<strong>Pulmonary</strong> EmbolismPleural effusions occur in 40 to 50% of patientswith angiogram-documented pulmonary embolism.A pleural effusion results from ischemia,leading to increased capillary permeability and theleak of protein-rich fluid into the pleural space. Theeffusion can also be the result of atelectasis secondaryto chest pain. A large, hemorrhagic pulmonaryinfarction can rarely result in a hemothorax in apatient who is treated with heparin.The presence of a pleural effusion does not alterthe symptoms and signs of pulmonary embolism.Ipsilateral pleuritic chest pain, acute onset of dyspnea,tachypnea, and tachycardia are the typicalfindings. The patient may be afebrile or have a lowgradefever, but rarely exceeding 101.5F (38.6C).Hemoptysis may occur with a pulmonary infarction.Virtually all patients with a pleural effusion frompulmonary embolism have ipsilateral chest pain.The chest radiograph typically shows a smallunilateral pleural effusion virtually always lessthan one third of the hemithorax. 53 The effusion ispresent on the initial presentation in approximately90% of cases. A radiographic infarction is seen inapproximately one half of the patients and is usuallyin the lower lobe. A pleural effusion in associationwith a radiographic infarction tends to belarger than if no radiographic infarct is present butits volume is still less than onethird of the hemithorax.55,56PFA in pulmonary embolism is highly variable.54,57 One small study 55 showed that approximately20% of patients with a pulmonary embolismhad a transudative effusion. An exudative effusionmay be hemorrhagic with neutrophil or mononuclearcell predominance depending on the timingof the thoracentesis in relationship to the acutepulmonary embolus. The PF and serum glucosesare equivalent, and the pH is typically 7.30. Theclassic bloody, neutrophil-predominant exudate ispresent in only approximately one third of patients.The diagnosis of pulmonary embolism is bestestablished by CT angiography; a ventilation/perfusion scan may be useful in situations in whichthe patient cannot receive IV contrast or the chestradiograph shows significant airspace disease orsevere emphysema. A positive lower-extremityDoppler study and elevated d-dimer with a smallipsilateral neutrophil-predominant exudate makesthe diagnosis highly likely.Immediate anticoagulation with heparinshould be instituted when the diagnosis is highlylikely or definitively established. A hemorrhagiceffusion or minimal hemoptysis is not a contraindicationto anticoagulation. There are rare reportsof hemothorax developing during heparin therapyin the setting of a large hemorrhagic infarction. Nospecific treatment is required for the pleural effusionof pulmonary embolism. A small effusionwithout radiographic infarction tends to peak involume by 72 h and resolves in 7 to 10 days.Patients with a radiographic infarction and a largerpleural effusion usually require 14 to 21 days forresolution. 54–57Tuberculous Pleural EffusionA tuberculous pleural effusion is the mostcommon form of extrapulmonary tuberculosis,<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 525ACC-PULM-09-0302-033.indd 5257/10/09 10:23:13 PM


occurring in up to 10% of untreated purified proteinderivative (PPD) converters. These effusionsdevelop when a subpleural focus of tuberculosisruptures into the pleural space followed by a cellmediatedimmune response to the tuberculousantigens. A tuberculous pleural effusion can occurin the following two settings: (1) postprimaryphase (3 to 6 months after the initial exposure) and(2) r-activation any time thereafter. 58–60These patients are typically symptomatic atpresentation, exhibiting nonproductive cough(80%), chest pain (75%), which is usually pleuritic,and fever. Symptoms tend to be more acute inpostprimary disease than with reactivation. Theage of patients presenting with a tuberculous effusiontends to be increasing because more reactivationdisease has been observed. 58The PPD skin test is positive in 69 to 100% ofpatients. A false-negative PPD results from thefollowing: (1) circulating mononuclear cells thatsuppress sensitized T lymphocytes in the peripheralblood and skin but not in the pleural space;or (2) sequestration of PPD-reactive lympho cytesin the pleural space. Chest radiograph typicallyshows a unilateral small-to-moderate pleuraleffusion, which is more commonly on the right.A tuberculous pleural effusion is rarely massive.Bilateral pleural effusions are associated withhematogenous disease in 10% of HIV-positivepatients and 5% in non-HIV–positive patients.Parenchymal infiltrates are noted in 33% onstandard chest radiograph, whereas parenchymallesions are seen more commonly (75%) onchest CT scan. Infiltrates are less commonly seenwith postprimary disease. The effusion is ipsilateralto the infiltrate and is loculated in 30%of patients. 58,59PFA reveals serous fluid that may be serosanguineousin 10% but is never frankly bloody. Theeffusion is always an exudate with total protein 5.0 g/dL in 77% of patients. The total nucleatedcell count ranges from 2,000 to 8,000 L and is classically90 to 95% lymphocytes; 90% of patients have 60% lymphocytes. Neutrophils are predominantwith an acute presentation. PF eosinophilia or 5%mesothelial cells make the diagnosis of tuberculouspleural effusion unlikely. 61 The PF glucose is 60 mg/dL and the pH 7.30 in approximately20% of patients. The PF pH is virtually never 7.40. 58Table 9. Diagnosis of Tuberculous Pleural Effusion*Tests Sensitivity, %Percutaneous pleural biopsyHistology 63–85Culture 55–80Pleural fluid culture 13–70Sputum4–50 (4% with isolatedeffusion on CXR)Pleural biopsy acid-fast bacilli 5–18smearPleural fluid acid-fast bacilli smear 5N-PCR 51VATS pleural biopsy ~100*CXR chest radiograph; N-PCR nested polymerase chainreaction; VATS video-assisted thoracoscopic surgery.A combination of histology and culture of thepleural biopsy tissue, culture of PF, and sputumculture establishes the diagnosis in up to 86% ofcases (Table 9). 58 A high PF adenosine deaminase(ADA) level of 40 to 60 U/L supports the diagnosisif rheumatoid arthritis and empyema are unlikely.An ADA level 40 U/L has a high negative predictivevalue for tuberculous pleural effusions. AnADA 50 IU/L with a lymphocyte/neutrophil ratio 75% makes a tuberculous effusion highlylikely. 62Treatment is the same as for pulmonary tuberculosis63 ; however, because the organism load islow with a tuberculous pleural effusion, 6 monthsof isoniazid and rifampin appears to be effectivetherapy. 64 A tuberculous effusion usually resolvesin 4 to 6 weeks with treatment or spontaneouslybut can persist for up to 16 weeks. Untreated, therecurrence rate is 65% within 5 years, mostly pulmonarywithout a pleural effusion. 65 Therefore,a patient with an undiagnosed lymphocytepredominantexudate and positive PPD skin testneeds to be treated with isoniazid prophylaxis ifactive disease has been excluded. 65 Corticosteroidsare effective in resolving severe symptoms and incausing more rapid resolution of the pleural effusionbut has no effect on residual pleural thickening. 66Pancreatic Disease (Acute Pancreatitis, ChronicPancreatic Effusion [Pancreatopleural Fistula])Approximately 3 to 17% of patients with acutepancreatitis will develop a left pleural effusion. 67526 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5267/10/09 10:23:13 PM


Pleural effusions are much less common in patientswith chronic pancreatitis. Those with acute pancreatitisdevelop a pleural effusion from transdiaphragmaticpassage of amylase-rich PF into thepleural space with resultant increased capillarypermeability from inflammatory mediators. ThePF to serum amylase ratio is 1 and is explainedby the rapid renal clearance of amylase and pooramylase clearance from the pleural space. Achronic, large recurrent pleural effusion is associatedwith chronic pancreatitis and pseudocystformation. Pseudocysts develop in approximately10% of patients with acute pancreatitis, and 50%of those patients with a pseudocyst will develop apleural effusion. 68 The pleural effusion occurswhen a fistulous track develops from the pseudocystto the mediastinum, either pleural space, orthe pericardial sac. 69 Patients with acute pancreatitisvirtually always have abdominal symptoms andoccasionally pleuritic pain and dyspnea. The majorsymptom with a chronic pancreatic effusion isdyspnea attributable to the large pleural effusion;chest discomfort and cough are typical in thesepatients who do not have abdominal symptomsbut have a long history of alcohol abuse. 68,69The chest radiograph in acute pancreatitis is asmall left-sided pleural effusion in 60% of cases butis right-sided in 30% and bilateral in 10%. 67 In addition,there is usually an elevated left diaphragmand evidence of left basilar atelectasis. In the settingof a chronic pancreatic effusion, the effusionis large or massive in size and typically unilateralon the left but can be right-sided or bilateral. 68,69Pancreatic calcifications are typically observed onan abdominal film. The large pleural effusion usuallyrecurs rapidly after thoracentesis because thefluid is rapidly generated from the pseudocystalong the pressure gradient.PFA in acute pancreatitis reveals a turbid orhemorrhagic effusion that is neutrophil predominantand may have a total nucleated cell count ofup to 50,000/L. The PF and serum glucose concentrationsare equivalent, and the pH ranges from7.30 to 7.35. The PF/ serum amylase ratio is 1and above the upper limits of normal of serum.The initial PF amylase may be normal but willbe found to be elevated on a subsequent thoracentesis.67 PF from a chronic pancreatic effusion isserous or hemorrhagic in appearance and is also aneutrophil-predominant exudate. However, theseeffusions have very high PF amylase levels, often 100,000 IU/L. The reported pH range is 7.28 to7.50, with the vast majority being 7.30. The serumand PF glucose are equivalent. 68,69The diagnosis of an acute pancreatitis pleuraleffusion is established by finding a PF/ serum amylaseratio 1 or above the upper limits of normal ofserum in the appropriate clinical setting. A very highPF amylase, 50,000 to 100,000 IU/L, establishesthe diagnosis of a chronic pancreatic effusion withpseudocyst formation and fistulous track in analcoholic patient. Ultrasonography and CT can demonstratethe pseudocyst and the fistulous track.There is no specific treatment for the pleuraleffusion of acute pancreatitis, which tends toresolve in 2 to 3 weeks as pancreatic inflammationsubsides without pleural space sequelae. Patientswith a chronic pancreatic effusion may respond toconservative therapy, which includes pleural spacedrainage, bowel rest, and hyperalimentation; however,conservative therapy is only successful in 50%of patients. In those with refractory chronic pancreaticeffusions, percutaneous catheter drainageof somatostatin, the pseudocyst, or surgery areeffective (Tables 10–12). 68,69Rheumatoid PleurisyPleural effusions occur in rheumatoid arthritisin approximately 5% of patients; however, atautopsy, 50% of patients will have pleural fibrosisor an effusion (Table 13). 70 The pathogenesis forthese effusions is local immune complex-mediatedpleural inflammation. A chronic effusion may beseen with lung entrapment. Small fibrous plaques,rheumatoid nodules, or extensive fibrosis caninvolve the visceral pleura. There is an exfoliationof cells from the visceral pleural rheumatoid nodulesthat may produce an appearance of “debris”in the PF. 71 The low glucose and pH in rheumatoideffusions occur predominantly from an abnormalpleural membrane that inhibits glucose influx andhydrogen ion efflux from the pleural space. 72 Rheumatoidpleural effusions tend to occur in men (4:1male/female ratio) who have rheumatoid nodulesand moderate-to-severe active arthritis 5 to 10years after the onset of articular disease. 70 However,effusions can occur up to 3 years before orafter 20 years of the onset of rheumatoid arthritis. 70At presentation, patients complain of pleuritic<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 527ACC-PULM-09-0302-033.indd 5277/10/09 10:23:14 PM


Table 10. Differentiating Features of Pleural Effusions From Acute and Chronic PancreatitisClinical Features Acute Pancreatitis Chronic Pancreatitis (Pancreaticoplueral Fistula)Incidence 3–17% UncommonPresenting symptom Abdominal pain DyspneaSize of effusion Small Large to massiveSite Usually left sided Usually left sided, can be bilateral or right sidedPF amylase, IU/L 500–10,000 May be 200,000Pathogenesis Increased vascular permeability Direct fistulous communication from pancreatic pseudocystOutcome Specific treatment not required 50% require surgical intervention, percutaneous drainage orsomatostatinTable 11. Tests Useful in the Differential Diagnosis of Amylase-Rich Pleural EffusionsTests Acute Pancreatitis Chronic Pancreatic Effusion Esophageal Rupture MalignancyPleural fluid amylase Moderate Extremely elevated Minimal MinimalconcentrationPleural fluid /serum amylase 10:1 20:1 (mean 63,000) 5:1 3:1Pleural fluid amylase Pancreatic Pancreatic Salivary SalivaryisoenzymePleural fluid pH 7.30–7.39 7.28–7.39 5.50–7.00 7.05–7.40Serum lipase Elevated Normal to minimally increased Normal NormalTable 12. Causes of Amylase-Rich Pleural EffusionsDiagnosisPancreatitis, pancreaticopleural fistulaCarcinoma of the lung (usually adenocarcinoma)Adenocarcinoma of ovaryLymphomaEsophageal ruptureChronic lymphatic leukemiaPneumoniaRuptured ectopic pregnancyType of Amylase IsoenzymePancreaticSalivary (most common cause of salivary amylase-rich effusion)SalivaryMacroamylase/salivarySalivarySalivarySalivaryProbably salivarychest pain and dyspnea or may be asymptomatic,usually without fever. Patients with rheumatoidpleurisy typically have high serum rheumatoidfactors and presumably high levels of CCP antibodies.The chest radiograph typically shows asmall-to-moderate unilateral effusion with a normalcardiac silhouette. However, in one third ofpatients, another manifestation of rheumatoid lungdisease, such as rheumatoid nodules or interstitiallung disease, may be present. 70On PFA, the fluid may have several differentappearances, including a yellowish-green tint,turbidity, milky, or debris-like. 70–72 The milkyappearance suggests that the patient has longstandinglung entrapment and a cholesterol pleuraleffusion. The fluid typically contains a fewhundred mononuclear cells (mostly lymphocytes)if there is an unexpandable lung without activedisease, up to 15,000 mostly neutrophils/L inacute rheumatoid pleurisy or with a cholesteroleffusion with chronic inflammation. When theclassic triad of a glucose 30 mg/dL, a pH of 7.00,and LDH 1,000 IU/L, the clinical diagnosis ishighly likely as long as infection is excluded.Approximately 65% of patients on presentationwith rheumatoid pleurisy will have a PF glucose528 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5287/10/09 10:23:14 PM


Table 13. Differentiating Rheumatoid Pleurisy and Lupus Pleuritis*Characteristics Rheumatoid Pleurisy Lupus PleuritisFrequency over course of disease 5% 50–75%Onset 3–5 to 10 yr after onset of articular Any time; 5% as initial manifestation of SLEdiseaseClinicalMen; active RA; with or without Flare up of diseasepleuritic painChest radiograph Small unilateral effusion Unilateral or bilateral; with or withoutpericardial effusionAppearance Turbid, yellow-green, debris, milky Serous-serosanguinousPleural fluid analysisLow pH and glucose; increasedLDHNeutrophilic exudate; pH 7.30 and glucosePF/serum ratio 0.5 in 15–20% of patientsDiagnosis Characteristic cytology (see text) SLE cells present; PF ANA 1:320 and PF/serum ANA ratio 1 is suggestiveTreatmentAcute disease: corticosteroids; chronic Corticosteroidsdisease: unknownResolution 3–4 mo (up to 1 yr) 2–4 wkResidualPleural fibrosis; lung entrapment;cholesterol effusionNo sequelae or minimal pleural thickening*ANA antinuclear antibody; RA rheumatoid arthritis. 30 mg/dL, and 80% will have a glucose 50mg/dL. PF complement is decreased, immunecomplexes are present, and rheumatoid factor isusually 1:320. Cytology is diagnostic by findinglarge, elongated tadpole-shaped cells with multinucleatedgiant cells with a background of granularmaterial. 71 The cytologist should be advised thatrheumatoid pleurisy is a possibility. The diagnosisis presumptive with the characteristic PF triad inthe appropriate clinical setting; however, the cytologicappearance of the fluid is diagnostic. 70–75There are no controlled studies evaluating theeffectiveness of corticosteroids or nonsteroidalantiinflammatory drugs. There are anecdotalreports of responses to corticosteroids; however,the course is variable, and it is uncommon for theeffusion to resolve in 3 to 4 weeks. In fact, it usuallytakes several months but less than a year forresolution. Pleural thickening may be a residual insome patients evolving into a trapped lung thatmay require decortication. 76 A cholesterol effusionmay develop over the course of several years.Lupus PleuritisPleural effusion or pleuritic chest pain duringthe course of systemic lupus erythematosus (SLE)occurs in 50 to 75% of patients. 77 Lupus pleuritis isa presenting manifestation of SLE in 5% of patients.At autopsy, 67% of patients with SLE will havepleural adhesions, pleural thickening, and pleuraleffusions (Table 13). 77 Pleuritis and pleural effusionsresult from a local immune response withlow levels of complement and complement componentsfound in PF. In addition, PF anti-DNAcomplexes may be involved. A specific immunofluorescentpattern of nuclear staining of pleurallining cells with either anti-IgM, anti-IgG or anti-C3 has been documented. 77,78Lupus pleuritis is typically associated with aflare of SLE. Patients are virtually always symptomaticwith pleuritic chest pain (86 to 100%),pleural rub (71%), cough (65%), dyspnea (50%),and fever. The symptoms are similar in both nativeand drug-induced lupus. The chest radiographtypically shows small-to-moderate bilateral effusions;however, unilateral massive pleural effusionshave been reported. Alveolar infiltrates,atelectasis, and an increased cardiac silhouetteattributable to a pericardial effusion may benoted. 79PF analysis typically shows yellow or serosanguineousfluid. The total nucleated cell countranges from 100 to 20,000/L with either neutrophilsin the acute setting or mononuclear cellspredominating later. The glucose is 60 mg/dL<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 529ACC-PULM-09-0302-033.indd 5297/10/09 10:23:14 PM


and the pH is 7.30 in approximately 20% ofpatients. The presence of lupus erythematosus (LE)cells is diagnostic. A PF/serum antinuclear antibody(ANA) ratio of 1 and decreased complementlevels support the diagnosis. 78,79 However, LEcells are not found in the PF of all patients withlupus pleuritis and are usually associated with thepresence of serum LE cells, which may diminishthe value of pleural LE cell positivity. Moreover,the detection of LE cells is not straightforwardtechnically and may be subject to significantobserver variation.Pleuritic chest pain may respond to nonsteroidalantiinflammatory drugs but virtually alwaysrespond dramatically to corticosteroid therapy,although high doses may be needed for severepleuritis or large effusions. In refractory cases,immunosuppressive agents, such as azathioprine,added to corticosteroids are sometimes but notalways efficacious. Recurrent pleural effusionsusually respond to pleurodesis. Asymptomaticlupus effusions require no treatment, usuallyresolve spontaneously, and have no prognosticsignificance. However, persistent pleuritic painappears to be an adverse prognostic marker witha mean survival of 4 years. 77–79Postcardiac Injury SyndromePostcardiac injury syndrome (PCIS) occurs inup to 30% of patients after cardiac surgery and 1to 15% after a myocardial infarction. The pleuraleffusion is most likely an immunologic responseto cardiac injury. Patients with PCIS have increasedlevels of serum antimyocardial antibodies (AMA)compared with those without PCIS. PCIS may betriggered by a viral illness. 80PCIS is heralded by pleuritic chest pain in 90% of patients with fever, pericardial rub, dyspnea,and crackles (all 50%) 3 weeks (range2 days to 1 year) after myocardial injury. Thesepatients have an elevated erythrocyte sedimentationrate with a mean of 62 mm/h. Fifty percent ofthe patients have a leukocytosis. The chest radiographis abnormal in 95% of patients, with pleuraleffusion occurring in 83%, either left-sided or bilateral,and right-sided effusion only in 17%. There isaccompanying left lower-lobe infiltrates and anincreased cardiac silhouette, most likely from apericardial effusion. 81PFA demonstrates a serosanguineous or hemorrhagiceffusion in 70% of patients. The PF is a concordantexudate, with both the protein and LDHbeing in the exudative range. The nucleated cellcount ranges from 500 to 39,000/L, depending onthe timing of thoracentesis in relationship to theinitial pleural injury. Neutrophils are predominantearly. PF and serum glucose levels are equivalent,and the pH is 7.30. Low PF total hemolyticcomplement and complement components may befound and the PF to serum AMA is typically 1. 81The diagnosis is presumptive and is a diagnosisof exclusion. In the proper clinical setting, a PF/serum AMA ratio 1 and low PF complementlevels when adjusted for the serum and PF proteinconcentrations support the diagnosis. 82Patients usually require antiinflammatorytherapy for relief of symptoms, including prednisonein some cases. Approximately 50% of patientsrelapse and need additional antiinflammatorytherapy. Corticosteroids typically result in rapidresolution of symptoms; however, the natural historyof PCIS does not appear to be influenced bycorticosteroids because their withdrawal oftenresults in recurrence. Given the deleterious effectsof corticosteroids on wound healing and the propensityof PCIS effusions to recur after their withdrawal,it is prudent to reserve corticosteroids forpatients with moderate-to-severe persistent symptoms.Most patients have resolution without clinicallyimportant sequelae. 83Pleural Effusions After CABG Surgery (Not PCIS)Small left pleural effusions virtually alwaysoccur postoperatively as the result of atelectasisfrom left phrenic nerve trauma (Table 14). Less commonly,a larger, left hemorrhagic effusion developsafter internal mammary harvesting. These effusionsmay be asymptomatic or cause dyspnea and tendto persist for several months after surgery. 83–85The chest radiograph shows a small left-sidedeffusion with left lower-lobe atelectasis and anelevated left hemidiaphragm in the early postoperativeperiod associated with phrenic nerve injury.In patients with internal mammary harvesting, alarger effusion may develop and persist for severalmonths. A trapped lung may develop after 6months and typically is a small, unilateral effusionthat recurs rapidly after thoracentesis. 83530 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5307/10/09 10:23:14 PM


Table 14. Pleural Effusion Following CABG Surgery*Type of Pleural EffusionPerioperativeEarly (within 1 mo)Late (2–12 mo)Persistent ( 6 mo)PF after internal mammary artery harvestingis hemorrhagic with a low number of nucleatedcells, which are predominantly lymphocytic. Themajority of post-CABG effusions resolve overweeks to months; however, some patients mayrequire 1 to 3 therapeutic thoracenteses; less commonly,some progress to trapped lung with severerestriction requiring decortication. 29,82–84ChylothoraxDescriptionAtelectasisBloody effusions after IMAharvestingOthers (CHF, PE, parapneumonic,chylothorax)PCISPCISLymphocytic effusion ofuncertain originConstrictive pericarditisLung entrapmentTrapped lung*PE pulmonary embolism. From Heidecker J, Sahn SA.Clin Chest Med 2006; 27:267–283.When the thoracic duct or one of its majortributaries ruptures, chyle flows into the surroundingtissues. If the mediastinal pleura remain intact,chyle fills the mediastinum and forms a “chyloma”over the next several days before rupturing intothe pleural space, usually on the right at the baseof the pulmonary ligament. The thoracic duct,which has its origin in the cisterna chyli, is situatedin the midline anterior to the first and second lumbarvertebrae. The thoracic duct travels throughthe aortic hiatus of the diaphragm approximatelyat the level of the tenth to twelfth thoracic vertebraeto the right of the aorta. Approximately at the fifthto sixth thoracic vertebrae, the duct enters the leftposterior mediastinum and eventually joins thevenous circulation where the left subclavian andinternal jugular veins merge.Therefore, disruption of the thoracic ductbelow T5 to T6 causes a right-sided chylothorax,whereas injury to the duct above this level resultsin a left chylothorax. However, significantanatomic variation exists in the path of the thoracicduct in almost half of the population. 86 Chyle isnormally transported from the intestine to thelymphatic system into the blood. Long-chain triglycerides(containing 14 carbon units) andingested fats are transformed into chylomicronsand low-density lipoproteins in the intestines andform chyle that enters the intestinal lymphatic vessels.In contrast, medium-chain triglycerides (containing 12 carbon units) are directly absorbedinto the portal vein without entering intestinallymphatics. Approximately 60% of the dietary fatenters the lymphatics, and 1,500 to 2,500 mL ofchyle travels daily through these vessels. 86The most common cause of chylothorax ismalignancy, particularly non-Hodgkin lymphoma.Cardiothoracic surgery, especially esophagectomyand surgery for congenital heart disease, are themost common causes. 87 However, the incidence ofpostoperative chylothorax remains at 1%.Trauma, including innocuous hyperextension of thethoracic spine, has also been reported. Several miscellaneouscauses of chylothorax include lymphangioleiomyomatosis(LAM), with approximately25% of patients with LAM developing chylothoraxduring the course of their disease and 9% as aninitial presenting manifestation; other causesinclude lymphangiectasis, tuberculous lymphadenopathy,left subclavian vein thrombosis, filariasis,CABG surgery, chronic lymphocytic leukemia,Castleman disease (angiocentric or giant lymphnode hyperplasia), sarcoidosis, Kaposi sarcoma,yellow nail syndrome, Noonan syndrome, multiplemyeloma, Waldenström macroglobulinemia afterthoracic radiation, and goiter. 87Patients with chylothorax present with subacuteor insidious onset of dyspnea. The historymay be a clue to its cause. However, after investigation,a number of chylothoraces are termed idiopathic;these are most likely caused by innocuoushyperextension of the spine or an occult malignancy.Patients with chylothorax are usually notfebrile and do not have chest pain because chyledoes not tend to invoke an inflammatory response.Chyle is bacteriostatic and, therefore, empyema israre. Frequently, patients cough up chyle (chyloptysis).Sputum triglyceride concentrations havebeen reported to range from 662 to 2,600 mg/dL,which is greater than concurrent serum values.Chylomicrons have also been detected in sputum.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 531ACC-PULM-09-0302-033.indd 5317/10/09 10:23:15 PM


Table 15. Differentiating a Chylothorax and a Cholesterol EffusionCharacteristics Chylothorax Cholesterol EffusionIncidence Uncommon RareCauses Lymphoma, trauma, surgery, LAM Rheumatoid arthritis or empyema, TB (a form of lungentrapment)Onset Acute or subacute Insidious and chronicSymptoms Dyspnea None; dyspneaPFA Milky, serous, turbid or bloody; 80%lymphocytes, protein discordant exudate;triglycerides 110 mg/dL (highly likely);cholesterol 60 to 200 mg/dLMilky, satin-like sheen; neutrophil-predominantexudate; cholesterol 200 mg/dL; cholesterol/triglyceride ratio 1; triglyceride level may be 110 mg/dLDiagnosis Presence of chylomicrons Presence of cholesterol crystals; cholesterol/triglyceride ratio 1 and cholesterol 200 mg/dLTreatment Manage underlying disease; pleurodesis Observation; decorticationPFA shows fluid that is typically white andopalescent if fat is present; however, the fluid canbe clear and yellow in the adult who has not eatenfor 12 h or hemorrhagic if there is concomitanttrauma. The supernatant of a chylothorax fails toclear after centrifugation. The primary cells in chyleare T lymphocytes, which typically represent 80%of the cellular population. The total nucleated cellcount ranges from 400 to 6,800/L. PF protein hasbeen reported from 2.2 to 5.9 g/dL (Table 15).In contrast to a cholesterol pleural effusion, thecholesterol levels in chyle are substantially lowerand range from 65 to 220 mg/dL. 88 If the PF triglycerideis 110 mg/dL, there is a high likelihoodthat the patient has chylothorax. Conversely, if thetriglyceride level is 50 mg/dL, it is highlyunlikely that a chylothorax is present. 89 However,if the patient has been fasting for at least 12 h, thetriglyceride concentration may be falsely low. Ifchylomicrons are present, the diagnosis is establisheddefinitively. Chylous fluid has been reportedto have a pH from 7.40 to 7.80 and a glucose concentrationof 78 to 200 mg/dL. 88 Measurement ofthe PF/serum glucose ratio assists in differentiatinga chylous effusion (ratio 1.0) and pleuralattributed to extravascular migration of a centralvenous catheter with glucose containing parenteralnutrition with lipid into the pleural space (ratio 1.0) (Table 15). 90There are two major principles that should befollowed in managing a patient with a chylothorax.First, there should be a step-wise plan of care thatbegins with pleural space drainage if the patient issymptomatic and nutritional support with progressionof pleural synthesis if necessary, which can beaccomplished by various techniques. Second, prolongeddrainage of a chylothorax should beavoided to prevent immunosuppression and malnutrition.Patients with a traumatic chylothoraxare typically managed with chest tube drainage,bowel rest, and parenteral nutrition to minimizethe flow of chyle and maintenance of fluid andelectrolyte balance. Because most traumatic chylothoracesresolved within 10 to 14 days, protein lossand immunosuppression are usually minimal. Ifdrainage is persistent after 2 weeks, 1,500 mL for5 days, or if the patient develops significant weightloss or progressive loss of protein despite nutritionaltherapy, surgical intervention is indicated.There are recent reports of percutaneous catheterizationand embolization of the thoracic duct inpatients with traumatic chylothorax occurring aftersurgical procedures. Somatostatin and octreotidehave been reported to decrease chyle productionin postoperative chylothorax in small caseseries. 91LAMLAM, a rare lung disease of unknown cause, ischaracterized by the proliferation and infiltrationof the pulmonary interstitium with atypical musclecells. In patients with LAM, progressive respiratoryinsufficiency and pleural complications, specificallypneumothorax and chylothorax, are clinicalhallmarks. 92 Because the majority of initial episodesof pneumothorax or chylothorax occur before thediagnosis of LAM, their occurrence is often a sentinelevent that leads the clinician to consider thediagnosis of LAM.532 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5327/10/09 10:23:15 PM


The incidence of pneumothorax in LAM, witha reported range of 39 to 76%, is one of the highestamong diseases associated with secondary spontaneouspneumothorax. 91 Pneumothorax is commonlythe presenting event that leads to thediagnosis of LAM. In a large retrospective studyof pneumothorax from the LAM Foundation database,the prevalence of pneumothorax in 395 LAMpatients during the course of their disease was 66%(260 patients). 93 Approximately 80% of thesepatients developed at least one pneumothoraxbefore the diagnosis of LAM, with patients averaging2.6 pneumothoraces before diagnosis. The rateof pneumothorax recurrence is remarkable inpatients with LAM. In the LAM Foundation study,recurrence occurred in 140 (73% of 193 patients)who developed at least one pneumothorax. 93 Therecurrences were ipsilateral (71%) and contralateral(74%), occurring at an average of 21.7 and 30months after the initial pneumothorax, respectively,defining LAM as the disease with the highestrate of recurrence. 93 In patients with LAM, mostpneumothoraces (81%) occurred at rest or withminimal activity. Bilateral simultaneous pneumothoraceshave been reported in LAM; the LAMFoundation study identified 8 (4%) of 193 patientswho developed bilateral simultaneous pneumothoraxduring the course of their disease, withseveral patients experiencing recurrent bilateralsimultaneous pneumothoraces. 93The cardinal pathologic feature of LAM is aproliferation if immature smooth muscle cellsalong the peribronchial, perivascular, and perilymphaticstructures. 93 Compression and obstructionof these conduits result in the development ofairflow obstruction and pneumothorax, hemoptysisand alveolar hemorrhage, and chyloptysis andchylothorax, respectively. It has been suggestedthat bronchial obstruction by overgrowth of LAMcells is responsible for the obstructive pattern intrapped air and that the pathologic process ultimatelyleads to formation of diffuse, bilateral,thin-walled pulmonary cysts ranging in size froma few millimeters to a few centimeters in diameter. 94The classic CT features of LAM include reticulonodularshadows, cysts or bullae, and hyperinflation.95 Incidental small pneumothoracesoccasionally are discovered on CT scan performedfor other purposes. Patients with LAM shouldbe counseled on the symptoms associated withpneumothorax and given explicit instructions toseek medical attention when a pneumothorax issuspected.One of the major conclusions of the LAM Foundationstudy was that current experience withpneumothorax in LAM supports an early interventionalprocedure, either chemical pleurodesis orsurgery, following the first pneumothorax. 93 Therecommendation was put forth because of the highincidence of pneumothorax recurrence and associatedrate of morbidity, including a life-long averageof 1 month in the hospital for pneumothorax managementin LAM patients who develop an initialpneumothorax.However, interventional approaches for pneumothoraxin LAM may affect candidacy and outcomesof lung transplantations. There are studiesthat have evaluated the outcomes of lung transplantationin LAM patients who had a pleuralpleurodesis. However, a retrospective survey 96 of34 patients with LAM who underwent lung transplantationat 16 centers in the United States andEurope showed that 27 received single-lung transplant,6 received bilateral transplants, and onereceived a heart-lung transplant. Thirteen (38%) of34 patients had previous pleurectomy or pleurodesis.Also, 18 (53%) of 34 patients had extensivepleural adhesions that were judged to be of moderateseverity and severe intent. Moderate-to-severehemorrhage occurred in four patients, leading tointraoperative death in one patient and repeatthoracotomy in two patients. Overall, posttransplantationsurvival in this cohort of LAM patientswas similar to other chronic lung disease populations.It was concluded that although preoperativecomplications do occur in LAM patients who hadbilateral pleural procedures, lung transplantationremains an important option that improves longtermoutcomes. 96Chylothorax is a less common pleural complicationof LAM than pneumothorax, with a prevalenceof approximately 20 to 30% among allreported cases (see the section “Chylothorax”).Most chylothoraces are unilateral with no predilectionfor either hemithorax. Most, however, are largeenough to require intervention. Pneumothoracesand chylothoraces in patients with LAM do notappear to occur simultaneously. Some patientswith LAM develop chylous ascites before theirchylothoraces. Patients with chylothorax have a<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 533ACC-PULM-09-0302-033.indd 5337/10/09 10:23:15 PM


subacute onset of dyspnea, with the average ageof presentation being approximately 41 years ofage compared with the average age of diagnosisof LAM of 34. 97 The occurrence of chylothorax inthese patients did not correlate with the extent oflung involvement. Chylothorax in LAM most likelyresults from the obstruction of the lymphatic vesselby the infiltration of smooth muscle cells. Managementof chylothorax in LAM is no different thanfor other causes of chylothorax.Esophageal PerforationThe three distinct types of esophageal perforationare (1) traumatic (iatrogenic and barogenic),(2) inflammatory, and (3) neoplastic. 98 Iatrogeniccauses account for most esophageal perforations,with the most common causative procedure beingesophageal dilation which occurs 3 to 10 timesmore frequently than with diagnostic flexible fiberopticesophagoscopy. Spontaneous esophagealrupture, more appropriately termed barogenic rupture(Boerhaave syndrome), is a relatively rareoccurrence. Rupture is associated with vigorousvomiting. Mediastinitis and sepsis are responsiblefor the rates of high morbidity and mortality in thissyndrome. 99The pathogenesis of esophageal ruptureincludes the following: (1) the esophageal tearalways occurs longitudinally, (2) the tear alwaysoccurs in the lower half of the esophagus, (3) therupture pressure is approximately 5 PSI, and (4)the mucosa is the most resistant layer. 100 The loweresophagus is more susceptible to rupture becausethe upper esophagus is buttressed by striatedsmooth muscle fibers, whereas the lower esophaguscontains only unsupported smooth muscle.The left pleural space is predominantly involvedafter barogenic esophageal rupture because theesophagus deviates to the left to enter the esophagealhiatus, and therefore, the left lateral wall is indirect contact with the mediastinal pleura. However,perforation of the cervical esophagus usuallydoes not involve the pleural space.The most dramatic presentation of esophagealrupture is associated with barogenic perforation. 98This entity is seen most commonly in men in theirfourth-sixth decades of life with a history of alcoholism,dietary indiscretion, and malnutritionwho present with severe epigastric pain that ispleuritic and may be misdiagnosed as an acutemyocardial infarction, acute pancreatitis, perforatedpeptic ulcer, or empyema of the gallbladder.Pain typically follows severe vomiting and retchingand is persistent and severe in the upperepigastrium. Reflex spasm of the upper abdominalmuscles may focus the clinician on a primaryabdominal process. Fever is universally presentand occurs later than the chest pain. Dysphagiais invariably present. Pneumomediastinum is aconstant feature of esophageal rupture. Afteresophageal-mediastinal perforation, a “crunch”may be auscultated over the left heart synchronouswith the cardiac cycle.The presence or absence of radiographic findingsdepends on the time interval between theperforation and the initial radiographic examination,the site of the perforation, and the integrity ofthe mediastinal pleura. 101 Early chest radiographs,within minutes of an esophageal perforation, aretypically normal. Mediastinal widening with airfluidlevels indicates mediastinal infection and maytake several hours to appear radiographically.Mediastinal emphysema virtually never appearsbefore 1 h after perforation and never occurs inapproximately 40% of patients. With intrathoracicesophageal perforation, mediastinal changes aremore likely to occur. The presence and timing ofpleural changes are linked to the integrity of themediastinal parietal pleura. Rupture of the mediastinalparietal pleura occurs in most patients, withthe rapid development of a hydropneumothoraxor pyopneumothorax occurring. Most left-sidedpleural lesions occur because 70% of barogenicesophageal ruptures develop in the left posteriorlateral wall near the diaphragm. 102 Without pleuralrupture, mediastinal and subcutaneous emphysemaappear rapidly, and a small sympatheticpleural effusion attributed to mediastinitis developsinsidiously in approximately 75% of patients.When esophageal rupture is suspected, a contraststudy of the esophagus should be performedimmediately. The choice of contrast is limited to awater-soluble iodinated compound and bariumsulfate. Barium has the advantage of increasedradiographic density and better mucosal adherence.A disadvantage of barium is its ability toincite a foreign-body reaction with development534 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5347/10/09 10:23:16 PM


of pleural and mediastinal granulomas and possiblyfibrosis. Water-soluble contrast agents arerapidly absorbed and do not cause an inflammatoryreaction; however, the limitations are relatedto hypertonicity. Therefore, aspiration of thesecompounds into the tracheobronchial tree can createsignificant inflammation and precipitate pulmonaryedema. On chest CT, if air is found in themediastinum in the appropriate clinical setting, anesophageal perforation is highly likely. Thoracentesiscan establish the diagnosis once the mediastinalpleura have ruptured.PFA will depend on whether the mediastinumis still intact. If mediastinitis alone is present withoutrupture into the pleural space, the PF is predominantlya neutrophilic exudate with a pH 7.30, glucose concentration 60 mg/dL, andamylase levels less than the upper limits of normalserum amylase. After mediastinal parietal pleuralrupture, the patient develops an anaerobic empyema.The low PF pH and glucose are the result ofenhanced glycolysis from neutrophil phagocytosisand bacterial metabolism causing increased carbondioxide and lactic acid. 103 Therefore, finding a lowpH, typically between 5.00 and 7.00, in associationwith a high amylase level, which is salivary onisoenzyme analysis, establishes the diagnosis. ThePF at this stage should show high total proteinand LDH levels ( 1000 IU/L), the glucose concentrationshould be 60 mg/dL and may reach0 mg/dL. Furthermore, PF cytologic analysis ofhigh-speed cytocentrufication concentrates mayshow undigested food particles that confirmthe diagnosis, which may not be detected onGram stain and wet preparations. Squamous epithelialcells observed on cytologic examination alsoestablish the diagnosis.The patient with a barogenic esophageal rupturerepresents an emergency. On review 104 of 127patients with esophageal perforation, it wasreported that if primary closure was achievedwithin 24 h of rupture the outcome was excellent,with a 92% survival rate. After 24 h, survival wassubstantially worse, no specific treatment wasclearly superior, and complications were frequent.Immediate primary repair of barogenic esophagealrupture includes mediastinal and pleural spacedrainage and prompt treatment with antibioticsthat include anaerobic coverage.BAPEBAPEs are defined by exposure to asbestos,confirmation of the effusion by chest imaging orthoracentesis, absence of another disease relatedto the pleural effusion, and no development of amalignant tumor within 3 years. 105 The prevalenceof BAPE is related to dose, with 7%, 3.7%, and 0.2%effusions with severe, indirect, and peripheralexposure, respectively. The latency of these effusionswas shorter than for other asbestos-relateddisorders. BAPE was the most common asbestosrelatedabnormality during the first 20 years afterasbestos exposure. BAPE was most commonamong asbestos pipe coverers, less common inasbestos product or paper machine operators, andleast common among ship fitters, maintenancepersonnel, and welders. BAPE was the only manifestationwithin 10 years, and it was the mostcommon abnormality during the first 20 years afterasbestos exposure. 105Two thirds of patients report no symptomsduring their effusion. 105 When symptomatic, themost common symptom was pleuritic pain inapproximately 17% of patients and dyspnea in 9%of patients. Most BAPE effusions are small, unilateraleffusions with a small number presentingbilaterally. In patients with BAPE, pleural plaqueswere seen in 20% of patients and asbestosis observedin 10%. 105 Follow-up chest radiographs show ablunted costophrenic angle in approximately 90%of patients and diffuse pleural thickening in approximately50% of patients. Recurrent effusions developin approximately 30% of patients, sometimes ipsilateral,but more often contralaterally.PFA in 35 patients with BAPE showed a volumeof fluid up to 2,000 mL. The mean volume of theseeffusions was 460 mL. Approximately onehalf ofthe pleural effusions were hemorrhagic, and in 26%the exudate demonstrated PF eosinophilia. Othercells in the effusion were predominantly lymphocyteswith varying numbers of neutrophils andmesothelial cells. The PF glucose concentrationswere similar to blood glucose. 106,107An unusual variant of pleural fibrosis, calledrounded atelectasis, can result directly from a pleuraleffusion and often can be confused with possibletumor. Differentiation from a tumor can be problematic,but CT is of value in showing that the lesion<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 535ACC-PULM-09-0302-033.indd 5357/10/09 10:23:16 PM


simply represents peripheral atelectasis contiguousto the most significant pleural fibrosis and merelyrepresents a form of peripheral lobar collapse. 108Uremic Pleural EffusionAt autopsy, 20% of uremic patients have fibrinouspleuritis. 109,110 In patients undergoing decorticationfor restriction from uremic pleuritis, both thevisceral and parietal pleural surfaces are involvedwith a fibrotic peel covering both pleural surfaces.Pleural effusions from uremic pleuritis appear toresult from increased pleural and pulmonary capillarypermeability, possibly related to an immunecomplex injury.It is estimated that 3 to 15% of patients withuremia develop pleuritis. 111,112 The typical clinicalpresentation is a patient who has been on dialysisfor 1 to 2 years who presents with fever, chest pain,cough, and a transient pleural rub. Dyspnea isuncommon, and patients may be asymptomatic.The chest radiograph typically shows a unilateralsmall-to-moderate effusion, but at times, theeffusions can be bilateral and massive. 111,112 PFanalysis typically shows a serosanguineous tobloody effusion, although the PF can be serous. Thenucleated cell count is usually 1,500/L, with alymphocyte predominance averaging approximately70% of the total nucleated cells. The PF is aconcordant exudate by both protein and LDH, andthe PF and serum glucose concentrations are equivalent.The pH is typically 7.30 if there is not significantsystemic acidemia. 111,112 The PF/serumcreatinine ratio is 1 differentiating this effusionfrom a urinothorax. The diagnosis is presumptivebased on the PFA in the appropriate clinical setting.This effusion needs to be differentiated from a tuberculouspleural effusion because of the high risk oftuberculosis in this patient population.The treatment of a uremic pleural effusion iscontinued dialysis; the effusion resolves in 75% ofpatients during a 4- to 6-week period but mayrecur. Decortication has been successful for thosepatients who develop severe restriction from lungentrapment. 113,114Yellow Nail SyndromeYellow nail syndrome (YNS) was first describedby Samman and White 115 in 1964, who summarizeda series of 13 patients and referred to several otherreports from the 1920s and the early 1960s. Theirpatients had ankle edema and slow rates of nailgrowth ( 0.2 mm/week compared with thenormal rate of 0.5 to 1.2 mm/week). The authorssuggested that an abnormality of lymphatics mayexplain the pathogenesis of the syndrome. Twoyears later, Emerson 116 described the full triad ofslow-growing yellow nails, lymphedema, andpleural effusions, whereas Hiller and colleagues 117in 1972 reported that the presence of two of thethree symptoms were sufficient to establish thediagnosis.YNS is a rare disease, and its diagnosis is basedsolely on clinical criteria. YNS has been reportedfrom all five continents and includes all races. Themale/female ratio has been approximated at1:1.6. 118 The theory of impaired lymphatic drainageis supported by lymphangiographic findings inwhich most patients show a paucity of hypoplasticor dilated deficient lymphatics. 118 Electron microscopyhas shown, in a single patient, normal lymphaticssuggesting obstruction of lymph floweither in the major lymph vessels or at the lymphnodes. The age of onset varies from birth to as lateas 65 years of age with a median age of 40 years.Approximately 90% of the reported cases have hadyellow nails with the abnormal nails being theinitial symptom in 37%. Onycholysis was oftenpresent, and some patients have noted that thefingernails acted as a barometer of the syndromebecoming deeply yellow during exacerbation ofsymptoms and not as impressive during periodsof recession. Lymphedema of various degrees wasencountered in 80% of the reported cases and wasthe initial symptom in 34%. The edema was variablypitting and nonpitting. The edema could beconfined to the fingertips alone but was often morewidespread. 118Sixty-three percent of the published cases hadpleuropulmonary symptoms. In 29%, the initialsymptoms were related to the respiratory tract.Pleural effusions were found in 36% of all cases.Patients often had a history of 10 to 20 years ofrecurrent episodes of chronic bronchitis associatedwith bronchiectasis, chronic sinusitis, pneumoniaor pleuritis.Diseases associated with YNS have includederysipelas, cellulitis and lymphangitis, and thyroidabnormalities, including Hashimoto thyroiditis,536 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5367/10/09 10:23:16 PM


thyroid enlargement, hypothyroidism, and goiter.Of 97 patients, 6 (6%) developed malignancies.Hypogammaglobulinemia and nephrotic syndromehave also been reported in the setting ofYNS. 118The PF in YNS is a straw-colored clear exudatewith the total protein content generally being 3.0g/dL and often 4.0 g/dL. In most cases, the PFLDH is 0.67 of the upper limits of normal ofserum LDH but at times has been reported in theexudative range. The PF pH approximates 7.40;and the glucose is typically equivalent to the serumglucose, although there is a single report of a PFglucose of 10 mg/dL. The total cell count is 1,000/L with a predominance ( 80%) of lymphocytes.117,118With the full triad, the diagnosis of YNS is easilyestablished. However, signs and symptomsseldom appear simultaneously, and the full triadof the syndrome is not always present, giving riseto some differential diagnostic problems, especiallyconcerning the nail changes. 118Treatment can be symptomatic with no specifictherapy for YNS. However, there are reports oflocal steroid injection to the posterior nail foldsdemonstrating complete recovery, with normal nailgrowth in some patients. Other treatments thathave been tried with variable success include biotinand periodic nail debridement. Lymphedemaand pleural effusions, once present, seem to bepersistent. Pleurodesis has been successful in preventingthe recurrence of pleural effusions and hasbeen accomplished after pleurectomy, chemicalpleurodesis, and pleural abrasion. There have alsobeen reports of vitamin E 400 IU twice daily torestore the nails to normal growth and color andimprove lymphedema. No deaths attributable toYNS have been reported. 118AIDSA large study from a single US hospitalreported a 27% incidence of pleural effusions in222 hospitalized patients with AIDS. Infection wasthe cause of the effusions in two thirds of patients,with bacterial pneumonia being the most common(31%). 119 Pneumocystis jiroveci pneumonia resultedin pleural effusions in 15% of the patients, andMycobacterium tuberculosis was present in 8% of thepopulation. Noninfectious causes were found in34%, with hypoalbuminemia the most commoncause in 19% of patients. 119In most patients, no respiratory symptomswere noted; however, patients with pneumonia,tuberculous effusion, Kaposi sarcoma, and lymphomahad systemic symptoms and dyspnea.The chest radiographs in these patients mostcommonly revealed small unilateral or bilateralpleural effusions; the largest effusions wereseen with tuberculosis, Kaposi sarcoma, andlymphoma. 119PFA varies depending on the cause of the pleuraleffusion. In patients with hypoalbuminemiceffusions, the serum albumin was 1.5 g/dL.Pleural effusions in HIV patients with tuberculosisappeared not to be different compared with thosewithout HIV. P jiroveci effusions had a low numberof nucleated cells with an LDH discordant exudate,most likely reflecting the serum LDH. The effusionswith Kaposi sarcoma were hemorrhagic exudateswith negative cytology. 119Patients with AIDS and effusions had lowerCD4 counts and lower albumin concentrationsthan those AIDS patients without effusions. Allcases of pleural tuberculosis in patients with AIDSfrom 1988 to 1994 reported to the state of SouthCarolina were reviewed. Twenty-two (11%) of the202 AIDS patients with tuberculosis had pleuralinvolvement compared with 6% in the patientswithout AIDS. 120 Associated features of AIDS tuberculouspleurisy included a substantial weight lossand lower lobe infiltrates; there was no differencein PF characteristics between AIDS patients withCD4 counts 200/L compared with patients witha CD4 count 200/L. 120 Follow-up with chestradiography in 20 patients showed complete resolutionin 7, improvement in 10, and no improvementin 3.Hypothyroid Pleural EffusionA hypothyroid pleural effusion is a rare causeof an effusion and is defined as an effusion occurringsolely from hypothyroidism without concomitantcongestive heart failure, ascites, or othercause of pleural effusion. 121,122 The pathogenesis ofthe effusion is presumably increased capillarypermeability. On presentation, the patient usuallyhas obvious findings of hypothyroidism andmay have no respiratory symptoms because the<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 537ACC-PULM-09-0302-033.indd 5377/10/09 10:23:16 PM


effusions are small and respiratory drives may beblunted. There appears to be no correlation betweenthe development of the effusion and thethyroid-stimulating hormone concentration. 122 Thechest radiograph typically shows a small unilateralor small bilateral pleural effusions with a heart ofnormal size. PFA reveals a protein discordant exudatethat is serous or serosanguineous in appearance.There are a small number of mononuclearcells, and the PF and serum glucose are equivalent.The pH has been recorded as 7.30. 122 The diagnosisis presumptive and is one of exclusion; mosteffusions in a hypothyroid patient are caused byanother disease, such as pneumonia. Other effusionsare related to hypothyroid-induced congestiveheart failure, pericardial effusion or ascites.The classic hypothyroid effusion will resolve overweeks with thyroid replacement. 122Pleural Disease After Radiation TherapyFifty years ago, radiation pleuritis was reportedin 6% of patients treated for breast cancer withsupervoltage radiation. 123 Currently, radiationtherapy has been shown to cause pleuritis 6 weeksto 3 months after radiation therapy and is associatedwith a loculated exudative effusion. Late (ayear or longer) manifestations of radiation therapyinclude mediastinal fibrosis, superior venal cavalobstruction, and constrictive pericarditis. 124 Radiationpleuritis causes direct injury to the pleural andsubpleural capillaries causing increased capillarypermeability. The effects of late radiation fibrosisare manifested by impaired lymphatic drainagefrom the pleural space or imbalances in hydrostaticpressures. 124,125Many patients with radiation pleuritis areasymptomatic, but some manifest pleuritic chestpain or dyspnea 2 to 6 months after completionof radiation therapy, which is usually associatedwith radiation pneumonitis. Patients withremote radiation therapy may have dyspnea orsigns of superior vena caval obstruction or haveeffusions from constrictive pericarditis. Thechest radiograph of acute radiation pleuritisshows a small-to-moderate unilateral effusionwith a tendency toward loculation. With mediastinalfibrosis, there may be small to moderateunilateral or bilateral effusion with a normalcardiac silhouette.The PF analysis in radiation pleuritis is a nonspecific,lymphocyte-predominant exudate withreactive mesothelial cells. With the mediastinumfibrosis, the effusions are transudates caused byconstrictive pericarditis. 124The diagnosis should be considered in anypatient who has had previous radiation therapyand presents with either a transudative or exudativeeffusion. Effusions from radiation pleuritismay persist for months to years. Early use of corticosteroidsmay relieve symptoms and hastenresolution. With mediastinal fibrosis, malignancymust be excluded. Treatment is usually symptomatic;pericardectomy may be successful for thepatient with constrictive pericarditis.Drug-Induced Pleural DiseaseA large number of drugs have been documentedto cause pleural involvement (Table 16);however, the incidence is less than drugs implicatedin parenchymal lung disease. 126–129 In thepatient with an undiagnosed exudative effusion,a careful review of the their drug list must be done.Pleural involvement develops after a variableperiod of treatment ranging from a few days (nitrofurantoin)to several years (ergolines). Rarely,pleural involvement is noted after termination ofthe drug. The clinical imaging and pathologic patternof pleural involvement should conform toearlier observations with the drug. Other drugsshould be excluded as causative. Improvementshould follow discontinuation of the drug,and pleural involvement should recur after reexposure.Additional supporting information forthe diagnosis includes involvement of other serosalsurfaces, such as the pericardium. Eosinophilia inPF, blood, and BAL fluid suggest a hypersensitivityreaction, whereas an elevated ANA or antineutrophilcytoplasmic antibodies titer suggests adrug-induced autoimmune condition. High titersof ANA or LE cells in PF are associated with druginducedlupus pleuritis. The drugs associated withPF eosinophilia include clozapine, dantrolene,glicazide, isotretinoin, mesalazine, nitrofurantoin,propylthiouracil, sulfasalazine, and valproic acid.Eosinophilia is typically present in the blood, BALfluid or pleural tissue concomitant with PF eosinophilia;there may be concomitant pulmonaryinfiltrates as well.538 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5387/10/09 10:23:16 PM


Drugs that produce pleural thickening includeamiodarone, the ergolines, cyclophosphamide, andpractalol. 126 Drug-induced pneumothorax canoccur in patients with rheumatoid nodules, primaryor metastatic lung tumors (particularly germcell tumor or sarcoma), or lymphoma as the resultof treatment with cytotoxic chemotherapy.Pleuritic chest pain may be the most commonsymptom in patients with drug-induced as well asnative lupus, whereas more intense pain and frictionrubs have been described in patients withdrug-induced pneumonitis or organizing pneumoniafrom nitrofurantoin, carbamazepine, andmethacycline. Pleural effusions in association withdrug-induced pneumonitis include methotrexate,mitomycin, nitrofurantoin, angiotensin-convertingenzyme inhibitors, fenfluramine, amiodarone, andbleomycin. Furthermore, pleural effusions can beassociated with drug-induced increased capillarypermeability. Drugs that have caused ARDS withconcomitant pleural effusions include the following:granulocyte-macrophage colony-stimulatingfactor, cytosine arabinoside, gemcitabine, and ethchlorvynol.Increased pulmonary capillary permeabilityhas been noted with all-trans-retinoic acidand the ovarian hyperstimulation syndrome.PneumothoraxPneumothorax, or air in the pleural space, isclassified as spontaneous, traumatic, or iatrogenic(Table 17).Primary Spontaneous PneumothoraxThe incidence of pneumothorax occurs in 7 to18 per 100,000 patients/year for men and 1 to 6 per100,000 patients/year for women. Air can enter thepleural space from rupture of an apical pleuralbleb, which can be acquired from bronchial inflammationor be congenital. A pneumothorax may beprecipitated by changes in barometric pressure.Small, ipsilateral, hemorrhagic pleural effusionsare noted in approximately 10% of patients withprimary spontaenous pneumothorax (PSP). PSPmost commonly occurs in tall, thin boys and menages 10 to 30 years. Smoking increases the riskapproximately 20-fold in men and less in women.Ninety percent of PSPs occur at rest. Pleuritic chestpain and acute dyspnea are the most commonsymptoms, which typically resolve spontaneouslywithin 24 h. 130A pneumothorax can be diagnosed on a standardchest radiograph when there is visualizationof the visceral pleural line removed from the chestwall on an upright radiograph. An end-expiratorychest radiograph on lateral decubitus view withthe affected side superior may be helpful in problematiccases. The expiratory radiograph will makethe pneumothorax more apparent as the lung iscompressed and therefore increases its densityand the relative amount of air in the pleural space.An arterial blood gas shows mild-to-moderatehypoxemia and a respiratory alkalosis. The risk ofrecurrence after an initial PSP is 32 to 52%. 130The recurrence rate increases with each subse -quent pneumothorax. The recurrence rate doesnot appear to be affected by chest tube drainage(Table 18).Patients should be treated with supplementaloxygen, which results in a fourfold increase inpneumothorax gas resorption. Observation is warrantedwith an asymptomatic patient and absenceof progression over the course of 6 h. Simple aspirationis successful in 70% of patients with a moderatePSP. 131–133 A small-bore catheter can be left inplace if simple aspiration is not effective or placedin a patient with a continued air leak and continuedsymptoms. The management of PSP is shown inTable 18.Secondary Spontaneous PneumothoraxThe causes of secondary spontaneous pneumothoras(SSP) include the following: (1) diseases ofthe airways, (2) interstitial lung disease, (3) infection,(4) malignancy, and (5) others (see Table 18). 130The peak incidence of SSP occurs later than PSP.Hyperexpansion of the distal air spaces fromobstruction or inflammation in the airways canlead to alveolar rupture and retrograde movementof air along the bronchovascular bundle to themediastinum with rupture through the mediastinalparietal pleura or direct rupture through the visceralpleura from an inflammatory parenchymalprocess.Patients with SSP have more severe dyspneathan those with PSP because of their decreasedpulmonary reserve. Chest pain is less common andsevere than in PSP. However, these patients may<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 539ACC-PULM-09-0302-033.indd 5397/10/09 10:23:17 PM


Table 16. Drug-Induced Pleural Disease*Drug Presentation Chest Radiograph PFA CommentProcainamideNitrofurantoinDantroleneMethysergideProcarbazineMethotrexateBromocriptineAmiodaroneMitomycin1 yr (1 mo–8 yr) of treatmentprior (1.5 g/d) topolyarthralgias, fever,pleurisy; pleuritic painand effusion in 67%Acute: previous drugexposure; hours todays after drug; acuteonset of fever, dyspnea,and cough; bloodeosinophiliaChronic: insidious coughand dyspneaAfter 2 mo–3 yr oftreatment; pleuriticchest pain and fever;some asymptomatic;blood eosinophiliaAfter 2 mo–6 yr oftreatment; recurrentchest pain, dyspnea,and fever with pleuralrub and effusionsWith combinationchemotherapy andhours after procarbazine:fever, chills, cough,dyspnea, crackles andblood eosinophiliaAfterintermittent ormaintenancemethotrexate; chestpain, fever, cough,and dyspnea1 wk–3 yr after drug;pleurisy, dyspnea, andcoughDyspnea, cough, constitutionalsymptoms,pleuritic chest pain,pleural rub after 100 gof drug; pleural diseaseuncommonAsymptomatic ordyspnea, cough, chestpain after 36–40 mg/m 2total doseInfiltrates andeffusionAcute: lung infiltratesand effusionsInterstitial infiltratesusually withouteffusion; effusionsalways associatedwith infiltratesUnilateral effusionUnilateral or bilateralloculated effusionsand pleural thickeningBilateral interstitiallower lobe infiltrateswith unilateralor bilateraleffusionsEffusion withoutinfiltrates (lowdose); diffuse infiltrateswith smalleffusions (maintenance);thickeningof fissures withouteffusion (high dose)Interstitial infiltrates,effusion, and pleuralthickeningAlveolar or interstitialinfiltrates, pleuralthickening, andunilateral or bilateraleffusionsInterstitial infiltrates,pleural thickening,and small unilateraland bilateraleffusionsNonspecific exudateMay haveeosinophiliaNot reportedSerosanguineous;eosinophiliacommonSerous orserosanguineousexudate with lowcell countNot reportedNot reportedNonspecific exudatewith a low numberof cellsSerous orserosanguineouswith low cell count;foamy macrophagespresentNot reportedAlso exudative effusionwith hydralazine andquinidine; positiveANA and anti-histoneantibodySymptoms resolve rapidlyafter drug stoppedSequelae dependent onlength of drug therapySymptoms improve withindays of stopping drugbut CXR takes months toresolveSymptoms resolve quicklybut CXR resolves overmonths after drug isstoppedSymptoms and CXR clearrapidly after drug isstoppedSymptoms resolve daysto weeks after drugstoppedStructurally similar tomethysergide;withdrawal of drugresults in improvementof symptomsDose-related toxicity;stopping drug andsteroids improvessymptoms and CXRIsolated pleural diseasedoes not occur; alwaysassociated with interstitiallung disease(continued)540 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5407/10/09 10:23:17 PM


Table 16. (Continued)Drug Presentation Chest Radiograph PFA CommentBleomycinEffusions have occurredafter 150 U; symptomsdue to interstitial lungdiseaseSmall bilateraleffusions associatedwith interstitiallung diseaseNot reportedEffusions resolve wheninterstitial diseasetreated with steroidsValproic acidFever, chest pain orasymptomatic; peripheraleosinophilia oftenpresentUnilateral or bilateraleffusions; no parenchymalinfiltratesExudate with highLDH 1,000; IU/L40 to 84% eosinophiliaLymphoplasmacyticinfiltrate and eosinophilson pleural biopsy.Effusions resolve overseveral weeks after drugdiscontinued*Abbreviations as in Table 9.Table 17. Classification of PneumothoraxTable 18. Causes of SSPTypes of PneumothoraxCausesTypes of SSPCausesSpontaneouspneumothoraxTraumaticIatrogenicPrimary: absence of clinicallung diseaseSecondary: a manifestationof clinical lung diseasePenetrating or blunt thoracicinjuryBarotraumaInvasive proceduresInsertion of central venouscatheterThoracentesisPleural biopsyTransbronchial lung biopsyTransthoracic needle biopsyhave life-threatening hypoxemia, and hypotensionoccurs in 15%. The symptoms of SSP do not resolvespontaneously. A suspicion of pneumothorax shouldremain heightened in the COPD patient with dyspneaand unilateral chest pain. The presence ofunderlying lung disease often makes identificationof the visceral pleural line problematic. A supineradiograph can show pneumothorax gas juxtacardiacor in the costophrenic sulcus and may producea deep sulcus sign. CT scan may be necessary todiagnose a SSP. The arterial blood gas in patientswith a SSP shows significant hypoxemia and concomitanthypercapnia (Table 19). Virtually allpatients with SSP require chest tube drainagebecause the hypoxemia, hypercapnia, and respiratorydistress can lead to an untoward event. 133 Themanagement of SSP is discussed in Table 19.Diseases of the airwaysInterstitial lung diseaseInfectious diseaseMalignancyOthersTension PneumothoraxCOPDCystic fibrosisStatus asthmaticusLAMLangerhans cell histiocytosisSarcoidosisRheumatoid arthritis (usualinterstitial pneumonia)Radiation fibrosisIdiopathic pulmonary fibrosisPneumocystis pneumonia inAIDSNecrotizing Gram-negativepneumoniaAnaerobic pneumoniaStaphylococcal pneumoniaM tuberculosisSarcomaLung cancerCatamenial<strong>Pulmonary</strong> infarctionWegener granulomatosisMarfan syndromeEhler-Danlos syndromeBirt-Hogge-Dubé syndromeThe incidence of tension pneumothorax is 1 to2% in patients who develop spontaneous pneumothoraces.Tension pneumothorax is most commonwith traumatic pneumothorax and pneumothoraxoccurring in patients on positive pressure ventilation.A tension pneumothorax develops when<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 541ACC-PULM-09-0302-033.indd 5417/10/09 10:23:17 PM


Table 19. Management of Spontaneous PneumothoraxImmediate OptionsSupplementaloxygenObservationSimple aspirationChest tube drainageCommentsResults in fourfold increase inpneumothorax gas absorptionSmall PSP, asymptomatic, noprogression in 6 hSuccessful in 70% of patients withmoderate PSPsFor virtually all patients with SSPsAs soon as the patient is stabilized, a standard chesttube should be inserted.Iatrogenic PneumothoraxCauses of iatrogenic pneumothorax includesubclavian vein catheter insertion, thoracentesis,percutaneous pleural biopsy, transbronchial lungbiopsy, and mechanical ventilation. 130ReferencesTable 20. Prevention of Recurrence of SpontaneousPneumothoraxTherapeutic OptionsChemical pleurodesis viachest tubeVideo-assisted thoraco -scopic surgeryCommentsReduces recurrence rate; talcslurry (most effective) anddoxycycline usedStapling of large bleb with talcpoudrage, pleural abrasion,or partial pleurectomypleural pressure exceeds atmospheric pressureduring the entire respiratory cycle. It developswhen there is unidirectional flow of air from thelung into the pleural space, and air accumulates inthe pleural space because of a check-valve mechanism.Experimental data suggest that circulatorycollapse in tension pneumothorax is related todecreased tissue oxygen delivery from hypoxemiarather than impaired venous return. The patientpresents with severe respiratory distress, cyanosis,marked tachycardia, and hypotension. The ipsilateralhemithorax may be larger; however, at timesit may be problematic to differentiate between theinvolved and noninvolved hemithorax. The predominantsymptoms may be related to hemodynamicinstability rather than respiratory distress.The chest radiograph shows the characteristic findingsof contralateral mediastinal shift, depressionof the ipsilateral diaphragm, lung collapse, andwidening of the ipsilateral ribs.With a high index of suspicion, immediatedecompression of the involved hemithorax withinsertion of a large-bore needle into the secondaryintercostal space should be done without delay.With decompression, there is an immediatedecrease in heart rate and respiratory rate, restorationof the BP, and improvement in oxygenation.1. Staub ND, Wiener-Kronish JP, Albertine KH. Transportthrough the pleura: physiology of normal liquidand solute exchange in the pleural space. In:Chretien J, Bignon J, Hirsch A, eds. The pleurain health and disease. New York: Marcel Dekker,1985; 169–1932. Stauffer JL, Potts DE, Sahn SA. Cellular content ofthe normal rabbit pleural space. Acta Cytol 1978;22:570–5743. Albertine KH, Wiener-Kronish JP, Roos PJ, et al.Structure, blood supply, and lymphatic vesselsof the sheep’s visceral pleura. Am J Anat 1982;165:277–2944. Collins TR, Sahn SA. Thoracentesis: clinical value,complications, technical problems and patientexperience. Chest 1987; 91:817–8225. Sahn SA. Approach to patients with pleural diseases.In: Light RW, Lee YCG, eds. Textbook ofpleural diseases. 2nd ed. London: Hodder Arnold;2008, 201–2076. Heidecker JT, Kaplan AP, Sahn SA. Pleural fluidand peripheral eosinophilia from hemothorax:hypothesis for the pathogenesis of EPE in hemothoraxand pneumothorax. Am J Med Sci 2006;332:148–1527. Sahn SA, Good JT Jr. Pleural fluid pH and malignanteffusions: diagnostic, prognostic, and therapeuticimplications. Ann Intern Med 1988; 108:345–3498. Sahn SA, Reller LB, Taryle DA, et al. The contributionof leukocytes and bacteria to the low pH ofempyema fluid. Am Rev Respir Dis 1983; 128:811–8159. Good JT Jr., Taryle DA, Sahn SA. The pathogenesisof low glucose-low pH malignant effusions. AmRev Respir Dis 1985; 131:737–74110. Light RW, MacGregor I, Luchinger PC, et al. Pleuraleffusion, the diagnostic separation of transudatesand exudates. Ann Intern Med 1972; 77:507–513542 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5427/10/09 10:23:17 PM


11. Heffner JE, Brown LK, Barbieri C. Diagnostic valueof tests that discriminate between exudative andtransudative effusions. Chest 1997; 111:970–97912. Joseph J, Badrinath P, Basran G, et al. Is the pleuralfluid transudate or exudate? A re-visit of the diagnosticcriteria. Thorax 2001; 56:867–87013. Heffner JE, Sahn SA. Multilevel likelihood ratiosfor identifying exudative pleural effusions. Chest2002; 121:1916–192014. Wiener-Kronish JP, Matthay MA, Callen PW,et al. Relationship of pleural effusions to pulmonaryhemodynamics in patients with congestive heartfailure. Am Rev Respir Dis 1985; 132:1253–125615. Rabin CB, Blackman NS. Bilateral pleural effusion:its significance in association with a heart of normalsize. J Mt Sinai Hosp 1957; 24:45–6316. Romero-Candeira S, Fernandez C, Martin C. Influenceof diuretics on concentration of proteins andother components of pleural transudates in patientswith heart failure. Am J Med 2001; 110:681–68617. Broaddus VC. Diuresis and transudative effusionschangingthe rules of the game. Am J Med 2001;110:732–73518. Gegenhuber A, Mueller T, Dieplinger B, et al.Plasma B-type natriuretic peptide in patients withpleural effusions: preliminary observations. Chest2005; 128:1003–100919. Lazaridis KN, Frank JW, Krowka MJ, et al. Hepatichydrothorax: pathogenesis, diagnosis and management.Am J Med 1999; 107:262–26720. Xiol X, Castellote J, Cortes-Beut R, et al. Usefulnessand complications of thoracentesis in cirrhoticpatients. Am J Med 2001; 111:67–6921. Lieberman FL, Peters RL. Cirrhotic hydrothorax.Arch Intern Med 1970; 125:114–11722. Sese E, Xiol X, Castellote J, et al. Low complementlevel and opsonic activity in hepatic hydrothorax:its relationship with spontaneous bacterial empyema.J Clin Gastroenterol 2000; 36:75–7723. Chalasani N, Clark WS, Martin LG, et al. Determinatesof mortality in patients with advanced cirrhosisafter transjugular intrahepatic portosystemicshunting. Gastroenterology 2000; 118:138–14424. Chen TA, Lo GH, Lai KH. Risk factors for spontaneousbacterial empyema in cirrhotic patients withhydrothorax. J Chinese Med Assoc 2003; 66:579–58525. Mattison L, Coppage L, Alderman D, et al. Pleuraleffusions in the medical intensive care unit: prevalence,causes, and clinical implications. Chest 1997;111:1018–102326. Abrass CK. Clinical spectrum and complicationsof nephrotic syndrome. J Invest Med 1997; 45:143–15327. Llach F, Arieff AI, Massry SG. Renal vein thrombosisin nephrotic syndrome: a prospective studyof 36 adult patients. Ann Intern Med 1975; 83:8–1428. Doelken P, Sahn SA. Trapped lung. Semin RespirCrit Care Med 2001; 22:632–63629. Huggins JT, Sahn SA, Heidecker J, et al. Characteristicsof trapped lung: pleural fluid analysis,manometry, air-contrast chest computed tomography.Chest 2007; 131:206–21330. Rudnick MR, Coyle JF, Beck LH, et al. Acute massivehydrothorax complicating peritoneal dialysis:report of 2 cases and a review of the literature. ClinNephrol 1979; 12:38–4431. Lepage S, Bisson G, Verreault J, et al. Massivehydrothorax complicating peritoneal dialysis:isotopic investigation (peritoneopleural scintigraphy).Clin Nucl Med 1993; 18:498–50132. Jagasia M, Cole F, Stegman M, et al. Video-assistedtalc pleurodesis in the management of pleural effusionssecondary to continuous ambulatory peritonealdialysis: a report of three cases. Am J KidneyDis 1996; 28:772–77433. Salcedo J. Urinothorax: report of 4 cases and reviewof the literature. J Urol 1986; 135:805–80834. Stark DD, Shanes JG, Baron RL, et al. Biochemicalfeatures of urinothorax. Arch Intern Med 1982;142:1509–151135. Miller K, Wooten S, Sahn SA. Urinothorax: a causeof a low pH transudative pleural effusion. Am JMed 1988; 85:448–44936. Sahn SA. Diagnosis and management of parapneumoniceffusions and empyema. Clin Infect Dis2007; 45:1480–148637. Sahn SA. Clinical commentary: management ofcomplicated parapneumonic effusions. Am RevRespir Dis 1993; 148:813–81738. Heffner JE, Brown LK, Barbieri C, et al. Pleural fluidchemical analysis in parapneumonic effusions. AmJ Respir Crit Care Med 1995; 151:1700–170839. Colice GL, Curtis A, Deslauriers J, et al. Medicaland surgical treatment of parapneumonic effusions:an evidence-based guideline. Chest 2000;118:1158–117140. Rahman NM, Chapman SJ, Davies RJO. Theapproach to the patient with a parapneumoniceffusion. Clin Chest Med 2006; 27:253–266<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 543ACC-PULM-09-0302-033.indd 5437/10/09 10:23:18 PM


41. Maskell NA, Davies CWH, Nunn AJ, et al (MIST1 Trial). U.K. controlled trial of intrapleural streptokinasefor pleural infection. N Engl J Med 2005;352:865–87442. Hong JY, Lee HJ, Piedra PA, et al. Lower respiratorytract infections due to adenovirus in hospitalizedKorean children: epidemiology, clinical features,and prognosis. Clin Infect Dis 2001; 32:1423–142943. Bustamante EA, Levy H, Simpson SQ. Pleuralfluid characteristics in Hantavirus pulmonary syndrome.Chest 1997; 112:1133–113644. Waites KB, Talkington DF. Mycoplasma pneumoniaeand its role as a human pathogen. ClinMicrobiol Rev 2004; 17:697–72845. Narita M, Matsuzono Y, Itakura O, et al. Analysisof mycoplasmal pleural effusion by the polymerasechain reaction. Arch Dis Child 1988; 78:67–6946. Antony VB, Loddenkemper R, Astoul P, et al.Management of malignant pleural effusions. Am JRespir Crit Care Med 2000; 162:1987–200147. Tremblay A, Michaud G. Single-center experiencewith 250 tunneled pleural catheter insertions formalignant pleural effusions. Chest 2006; 129:362–36848. Heffner JE, Neitert PJ, Barbieri C. Pleural fluid pHas a predictor of survival for patients with malignanteffusions. Chest 2000; 117:79–8649. Sahn SA. Malignant pleural effusions. In: FishmanAP, Elias JA, Fishman JA, eds. <strong>Pulmonary</strong> diseasesand disorders. 4th ed. New York: McGraw-Hill;2008, 1505–151550. Sterman DH, Litzky LA, Albelda SM. Malignantmesothelioma and other primary pleural tumors.In: Fishman AP, Elias JA, Fishman JA, eds. <strong>Pulmonary</strong>diseases and disorders. 4th ed. New York:McGraw-Hill; 2008, 1536–155251. Curran D, Shamoud T, Therasse P, et al. Prognosticfactors in patients with pleural mesothelioma: theEuropean Organization for Research and Treatmentof Cancer Experience. J Clin Oncol 1998;16:145–15252. Ordonez NG. What are the current best immunohistochemicalmarkers for the diagnosis of epithelioidmesothelioma? A review and update. HumanPath 2007; 38:1–1653. Sugarbaker DJ, Flores RM, Jaklitsch MT, et al.Resection margins, extrapleural nodal status, andcell type determine postoperative long-term survivalin trimodality therapy of malignant pleuralmesothelioma: results of 183 patients. J ThoracCardivasc Surg 1999; 117:54–6554. Bynum LF, Wilson JE III. Radiographic features ofpleural effusions in pulmonary embolism. Am RevRespir Dis 1978; 117:829–83455. Bynum LF, Wilson JE III. Characteristics of pleuraleffusions associated with pulmonary embolism.Arch Intern Med 1976; 136:159–16256. Yap E, Anderson G, Donald J, et al. Pleural effusionin patients with pulmonary embolism. Respirology2008; 13:832–83657. Porcel JM, Madronero AB, Pardina M, et al. Analysisof pleural effusion in acute pulmonary embolism:radiological and pleural fluid data from 230patients. Respirology 2007; 12:234–23958. Gopi A, Madhvan S, Sharma S, et al. Diagnosis andtreatment of tuberculous pleural effusion in 2006.Chest 2007; 131:880–88959. Seibert AF, Haynes J Jr., Middleton R, et al. Tuberculouspleural effusion: twenty-year experience.Chest 1991; 99:883–88660. Valdes L, Alvarez D, San Jose E, et al. Tuberculouspleurisy: a study of 254 patients. Arch Intern Med1998; 158:2017–207161. Spriggs AI, Boddington MM. Absence of mesothelialcells from tuberculous pleural effusions. Thorax1960; 15:169–17162. Valdes L, Alvarez D, San Jose E, et al. Value of adenosinedeaminase in the diagnosis of tuberculouspleural effusions in young patients in the regionof high prevalence of tuberculosis. Thorax 1995;50:600–60363. Joint Tuberculosis Committee of the British ThoracicSociety. Chemotherapy and management oftuberculosis in the United Kingdom: recommendations1998. Thorax 1998; 53:536–54864. Butt AK, Moers D, Stead WW. Tuberculous pleuraleffusion: six-month therapy with isoniadizeand rifampin. Am Rev Respir Dis 1992; 145:1429–143265. Roper W, Waring JJ. Primary serofibrinous pleuraleffusion in military personnel. Am Rev Tuberculosis1955; 71:616–63466. Wyser C, Walzl G, Smedema J, et al. Corticosteroidsin the treatment of tuberculous pleurisy: a doubleblind,placebo-controlled, randomized study. Chest1996; 110:333–33867. Kaye MD. Pleuropulmonary complications of pancreatitis.Thorax 1968; 23:297–30668. Uchiyama T, Suziki T, Adachi A, et al. Pancreaticpleural effusion: case report and review of 113 casesin Japan. Am J Gastroenterol 1992; 87:378–391544 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5447/10/09 10:23:18 PM


69. Tauseef A, Nandakumar S, Le V, et al. Pancreaticopleuralfistula. Pancreas 2009; 38:e26–e3170. Walker WC, Wright V. <strong>Pulmonary</strong> lesions in rheumatoidarthritis. <strong>Medicine</strong> 1968; 47:501–52071. Nosanchuk JS, Naylor B. A unique cytologic picturein pleural fluid from patients with rheumatoidarthritis. Am J Clin Pathol 1968; 50:330–33572. Dodson WH, Hollingsworth JW. Pleural effusionin rheumatoid arthritis: impaired transport of glucose.N Engl J Med 1966; 275:1337–134273. Joseph J, Sahn SA. Connective tissue diseases andthe pleura. Chest 1993; 104:262–27074. Faurschou P, Francis D, Faarup P. Thoracoscopic,histological, and clinical findings in 9 cases ofrheumatoid pleural effusion. Thorax 1985; 40:371–37575. Sahn SA. Immunologic diseases of the pleura. ClinChest Med 1986; 6:103–11276. Brunk JR, Drash EC, Swineford O. Rheumatoidpleuritis successfully treated with decortication.Am J Med 1966; 251:545–55577. Harvery AM, Shulman LE, Tumulty PA, et al. Systemiclupus erythematosus: a review of the literatureand clinical analysis of 138 cases. <strong>Medicine</strong>1954; 33:291–43778. Pines A, Kaplinsky N, Olchovsky D, et al. Pleural-pulmonarymanifestations of systemic lupuserythematosus: clinical features of its subgroups—prognostic and therapeutic implications. chest1985; 88:129–13579. Good JT Jr., Antony VB, King TE Jr., et al. Lupuspleuritis: clinical features and pleural fluid characteristicswith special reference to antinuclear antibodytiters. Chest 1983; 84:714–71880. Engle MA, McCabe JC, Ebert PA, et al. The postpericardiotomysyndrome and antiheart antibodies.Circulation 1974; 49:401–40681. Stelzner TJ, King TE Jr., Antony VB, et al. The pleuropulmonarymanifestations of the postcardiacinjury syndrome. Chest 1983; 84:383–38782. Kim S, Sahn SA. Postcardiac injury syndrome: animmunologic pleural fluid analysis. Chest 1996;109:570–57283. Heidecker J, Sahn SA. The spectrum of pleuraleffusions following CABG surgery. Clin ChestMed; 2006; 27:267–28484. Lee YC, Vaz MA, Ely KA. Symptomatic persistentpost-coronary artery bypass graft pleural effusionsrequiring operative treatment: clinical andhistologic features. Chest 2001; 119:795–80085. Light RW. Pleural effusions after coronary arterybypass graft surgery. Curr Opin Pulm Med 2002;8:308–31186. Miller JJ. Anatomy of the thoracic duct and chylothorax.In: Shields T, Locicero J, Ponn R, et al, eds.General thoracic surgery. 6th ed. Philadelphia, PA:Lippincott, Williams & Wilkins; 2005, 879–88887. Doerr C, Allan M, Nichols FR, et al. Etiology ofchylothorax in 203 patients. Mayo Clin Proc 2005;80:867–87088. Agrawal V, Sahn SA. Lipid pleural effusions. AmJ Med Sci 2008; 335:16–2089. Staats BA, Ellefson RD, Budahn LL, et al. A lipoproteinprofile of chylous and nonchylous pleuraleffusions. Mayo Clin Proc 1980; 55:700–70490. Duntley P, Siever J, Korwes M, et al. Vascularinvasion by central venous catheters: clinical featuresand outcomes. Chest 1992; 101:1633–163891. Browse NL, Allan D, Wilson N. Management ofchylothorax. Br J Surg 1997; 84:1711–171692. Almoosa KF, McCormack FX, Sahn SA. Pleuraldisease in lymphangioleiomyomatosis. ClinChest Med 2006; 27:355–36893. Almoosa KF, Ryu J, Mendez J, et al. Managementof pneumothorax in the lymphangioleiomyomatosis:effects on recurrence and lung transplantationcomplications. Chest 2006; 129:1274–128194. Sobonya RE, Quan SF, Fleishman JS. <strong>Pulmonary</strong>lymphangioleiomyomatosis: quantitative analysisof lesions producing air-flow limitation. HumPathol 1985; 16:1122–112895. Chu SC, Horiba K, Usuki J, et al. Comprehensiveevaluation of 35 patients with lymphangioleiomyomatosis.Chest 1999; 115:1041–105296. Boehler A, Speich R, Russi EW, et al. Lungtransplantation for lymphangioleiomyomatosis.N Engl J Med 1996; 335:1275–128097. Ryu J, Doerr CH, Fisher SD, et al. Chylothoraxin lymphangioleiomyomatosis. Chest 2003;123:623–62798. Ferguson T. Esophageal perforations and mediastinalsepsis. In: Hardy J, ed. Critical surgicalillness. 2nd ed. Philadelphia, PA: WB Saunders;1980, 279–28999. Campbell T, Andrews J, Neptune W. Spontaneousrupture of the esophagus (Boerhaave syndrome):necessity of early diagnosis and treatment. JAMA1976; 235:526–528100. MacKenzie M. A manual of diseases of the throatand neck including the pharynx, larynx, trachea,<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 545ACC-PULM-09-0302-033.indd 5457/10/09 10:23:18 PM


oesophagus, nasal cavities and neck. London:Churchill Limitis; 1884101. Parkin G. The radiology of perforated esophagus.Clin Radiol 1973; 24:324–332102. Anderson R. Rupture of the esophagus. J ThoracSurg 1952; 24:369–388103. Good JT, Antony V, Reller L, et al. The pathogenesisof the low pleural fluid pH in esophagealrupture. Am Rev Respir Dis 1983; 127:702–704104. Bladergroen M, Lowe J, Postlethwait R. Diagnosisand recommended management of esophagealperforation and rupture. Ann Thorac Surg1986; 42:235–239105. Epler GR, McLoud TC, Gaensler EA. Prevalenceand incidence of benign asbestos pleural effusionin a working population. JAMA 1982; 247:617–622106. Hillerdal G, Ozesmi M. Benign asbestos pleuraleffusion: 73 exudates in 60 patients. Eur J RespirDis 1987; 71:113–121107. Mattson SB. Monosymptomatic exudative pleurisyin persons exposed to asbestos dust. Scand JRespir Dis 1975; 56:263–272108. Mintzer RA, Cugell DW. The association of asbestos-inducedpleural disease and rounded atelectasis.Chest 1982; 81:457–460109. Nidus BN, Matalon R, Cantaczino D, et al. Uremicpleuritis-a clinicopathological entity. N EnglJ Med 1969; 281:255–256110. Hopps HC, Wisler RW. Uremic pneumonitis. AmJ Pathol 1955; 31:261–273111. Berger HW, Ramnohan G, Neff MS, et al. Uremicpleural effusion: a study of 14 patients on chronicdialysis. Ann Intern Med 1975; 82:362–364112. Jarratt M, Sahn SA. Pleural effusions in patientson chronic hemodialysis. Chest 1995; 108:470–474113. Galen MA, Steinberg SM, Lowrie EG, et al. Hemorrhagicpleural effusion in patients undergoingchronic hemodialysis. Ann Intern Med 1975;82:359–361114. Rodelas R, Rakoswki TA, Argy WP, et al. Fibrosinguremic pleuritis during hemodialysis. JAMA1980; 243:2424–2425115. Samman PD, White WF. The yellow nail syndrome.Br J Dermatol 1964; 76:153–157116. Emerson PA. Yellow nails, lymphedema andpleural effusion. Thorax 1966; 21:247–253117. Hiller E, Rosenow EC III, Olsen AM. <strong>Pulmonary</strong>manifestations of yellow nail syndrome. Chest1972; 61:452–458118. Nordkild P, Kromann-Andersen H, Struve-ChristensenE. Yellow nail syndrome: the triad of yellownails, lymphedema, and pleural effusions.Acta Med Scand 1986; 219:221–227119. Joseph J, Strange C, Sahn SA. Pleural effusions inhospitalized patients with AIDS. Ann Intern Med1993; 118:856–859120. Frye MD, Pozsik CJ, Sahn SA. Tuberculous pleurisyis more common in AIDS than in non-AIDSpatients with tuberculosis. Chest 1997; 112:393–397121. Schneierson SJ, Katz M. Solitary pleural effusiondue to myxedema. JAMA 1958; 168:1003–1005122. Goettehrer A, Roa J, Stanford GG, et al. Hypothyroidismand pleural effusions. Chest 1990;98:1130–1132123. Bachman AL, Macken K. Pleural effusions followingsupervoltage radiation for breast carcinoma.Radiology 1959; 72:699–709124. Whitcomb ME, Schwarz MI. Pleural effusioncomplicating intensive mediastinal radiationtherapy. Am Rev Respir Dis 1971; 103:100–107125. Morrone N, Silva-Volpe VL, et al. Bilateral pleuraleffusion due to mediastinal fibrosis inducedby radiotherapy. Chest 1993; 104:1276–1278126. Rosenow E. Drug-induced bronchopulmonarypleural disease. J Allergy Clin Immunol 1987;87:778–787127. Huggins JT, Sahn SA. Drug-induced pleural disease.In: Camus P, Rosenow E, eds. Clin ChestMed 2004; 25:141–153128. Camus P. Drug-induced pleural diseases. In:Bourous D, ed. Pleural disease. New York: MarcelDekker; 2004, 317–352129. Sahn SA. Drug-induced pleural disease. In:Camus P, Rosenow E, eds. Drug-induced iatrogeniclung disease. London: Hodder Arnold;2009130. Sahn SA, Heffner JE. Spontaneous pneumothorax.N Engl J Med 2000; 342:868–874131. Harvey J, Prescott RJ. Simple aspiration versusintercostal tube drainage for spontaneous pneumothoraxin patients with normal lungs: BritishThoracic Society Research Committee. BMJ 1994;309:1338–1339132. Waki A, O’Sullivan RG, McCabe G. Simple aspirationversus intercostal tube drainage for primaryspontaneous pneumothorax in adults. CochraneDatabase Syst Rev 2007; 1:CD004479133. Baumann MH. Management of spontaneouspneumothorax. Clin Chest Med 2006; 27:369–381546 Pleural Disease (Sahn)ACC-PULM-09-0302-033.indd 5467/10/09 10:23:18 PM


Pleural PearlsSteven A. Sahn, MD, FCCPObjectives:• Understand the causes, pathophysiology, clinical presentation,diagnosis, and management of patients with trapped lung• Understand the clinical presentation, pathophysiology,diagnosis, and management of patients with yellow nailsyndrome• Understand the epidemiology, pathogenesis, clinical presentation,diagnosis, and management of patients withthoracic endometriosis• Understand the pathogenesis, clinical presentation, diagnosis,and management of patients with chronic tuberculousempyema• Understand the causes, pathogenesis, clinical presentation,diagnosis, and management of patients with urinothorax• Understand the causes, pathogenesis, diagnosis, and managementof patients with a duropleural fistula• Understand the causes, pathogenesis, diagnosis, and managementof patients with a cholesterol pleural effusion• Understand the pathogenesis, diagnosis, and managementof patients with spontaneous bacterial pleuritisKey words: cholesterol effusion; chronic tuberculous empyema;duropleural fistula; lung entrapment; spontaneousbacterial pleuritis; thoracic endometriosis; trapped lung;urinothorax; yellow nail syndromeTrapped LungTrapped lung is one of a small number of causesof a persistent benign pleural effusion. This conditiondevelops when a portion of the lung is coveredby fibrous tissue that prevents its expansion to thechest wall, leaving a persistent, fluid-filled space.Trapped lung is an uncommon consequence offibrinous or granulomatous pleuritis, in which afibrous membrane covers the visceral pleura whilethe lung is separated from the chest wall. Trappedlung should be considered a consequence of abnormalhealing in the pleural space with formation ofscar tissue on the visceral pleura while the lung ispartially collapsed. 1PathophysiologyThe diagnosis of trapped lung implies achronic, constant-volume pleural effusion and amechanical cause for the persistent fluid-filledpleural space. The pleural fluid in trapped lungtypically has low total protein and lactate dehydrogenase(LDH) concentrations consistent with atransudative fluid; however, depending on thetiming of the thoracentesis and the integrity of theparietal pleura, the protein concentration may bein the exudative range. 2 Transudative effusions ofvascular origin, such as those associated with heartfailure, are invariably caused by increased fluidfiltration across the capillary endothelium. Thisincreased fluid filtration in combination with anormal protein reflection coefficient results in a lowtotal protein concentration in the effusate. Fluidegress from the pleural space fluid in these conditionsis presumably by bulk flow via the parietalpleural lymphatics, preventing a secondary rise oftotal protein concentration in the effusion. Thereis currently no evidence that increased capillaryfluid filtration is a feature of trapped lung. Furthermore,bulk flow from the pleural space also maybe impaired in trapped lung as a result of theinvolvement of the parietal pleural lymphatics bya pleural peel and impairment of the pump functionof the lymphatics during ventilation. A paucityof information is known about the pathophysiologyof fluid persistence in trapped lung, and thefinding of an exudate by total protein criterionshould therefore not be interpreted as unequivocalevidence of an active pleural process. 2It is important to define the difference betweena trapped lung and lung entrapment, which isassociated with an active inflammatory or malignantprocess. A trapped lung is the end-stage of anunexpandable lung that is not caused by an endobronchiallesion or chronic atelectasis. All patientswith a trapped lung must begin with lung entrapmentconnoting an active process (infection,inflammation [rheumatoid pleurisy] or malignancy)that generates pleural fluid, in addition tothe fluid induced by an unexpandable lung. Themechanism of pleural fluid accumulation froman active inflammatory or infectious process ormalignant condition may be associated with lung<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 547ACC-PULM-09-0302-034.indd 5477/10/09 8:56:49 PM


entrapment and differs from the single mechanismoperative in trapped lung. 3In patients with trapped lung, there is no otherexplanation for the mechanical visceral pleuralrestriction of lung expansion or for the persistenceof the pleural effusion. In contrast, with intenseinflammation, as with empyema, there are otherpathophysiologic mechanisms that produce pleuralfluid in addition to the fluid induced by the hydrostaticimbalance. Lung entrapment, as can be presentwith malignancy, causes a pleural effusion bytwo mechanisms: a hydrostatic mechanism causedby the inability of the lung to expand to the chestwall because of visceral pleural tumor involvementand a capillary leak or lymphatic obstruction as aconsequence of malignant infiltration. In contrast,a pleural effusion from trapped lung is solelycaused by failure of lung expansion that results inhydrostatic imbalance. It follows therefore that atrapped lung would be a transudate and lungentrapment an exudate. A trapped lung, in reality,is the end stage of lung entrapment when theinflammatory process has resolved (Fig 1).PathophysiologyThe generation and removal of pleural fluid ina trapped lung are operative under conditionsfound in the normal pleural space with intacthydrostatic and oncotic pressure gradients. Thepleural fluid from a trapped lung exists becausethe forces promoting generation and removal ofpleural fluid are in equilibrium. The loss of lungvolume decreases pleural pressure and leads todecreased thoracic volume on the affected side.Moderate decreases in lung volume, such as withlobar atelectasis, lobectomy, or interstitial lungdisease, usually do not result in a pleural effusion;the fact that trapped lung, with similar or slightlymore volume loss, results in a pleural effusionrequires explanation.After lobectomy, the remaining lung expands,the diaphragm may be displaced upward, and thepleural space assumes its usual width with theremaining lung assuming the shape of the thoraciccavity. However, in trapped lung, the fibroelasticmembrane usually involves only the dependentlung, although the entire lung may be involved. Theaffected lung is not only prevented from expandingby the fibrous membrane but also is restricted toconforming to the shape of the thoracic cavity. Theunaffected lung expands normally during breathingand therefore will expand into any void left by theportion of lung that is unexpandable. For the effusioncaused by the trapped lung to persist when itis only partially unexpanded, the shape of theFigure 1. Flow chart of lung entrapment.548 Pleural Pearls (Sahn)ACC-PULM-09-0302-034.indd 5487/10/09 8:56:49 PM


trapped lung must be sufficiently different fromthat of the opposing parietal pleural surface andthe unaffected lung must be mechanically preventedfrom expanding into the space. Otherwise,the space would be obliterated during the reestablishmentof negative pleural pressures after resolutionof the inflammatory process. Conditions in thepleural effusion from a trapped lung are closer tothe conditions along lobar margins, with widerseparation of the pleural surfaces, a steeper pressuregradient, and more negative pressures.The initial mean pleural liquid pressure inpatients with trapped lung may be negative orminimally positive but rapidly and substantiallydecreases with fluid removal. A negative initialpleural pressure or minimally positive pleural pressureand an elastance (change in pleural pressureof 14.5 cm H 2O when 1 L of fluid is removed) inthe absence of an active pleural process or endobronchialobstruction establishes the diagnosis oftrapped lung.CausesA trapped lung can develop from an infection,a noninfectious inflammatory process, hemorrhage,or malignancy. A pleural effusion must havechronicity to allow for mature fibrous tissue toform on the visceral pleural surface while the pleuralsurfaces remain separated. Some conditionsassociated with the development of a trapped lunginclude coronary artery bypass grafting (internalmammary artery harvesting), infection (empyema),inflammation (rheumatoid pleurisy), andmalignancy (lung cancer metastatic to the pleura).Because of the extensive number of cardiac surgeriesperformed currently, especially with use ofinternal mammary artery vascularization for coronaryartery bypass grafting, cardiac surgery is oneof the most common causes for trapped lung. Inadequatemanagement of any of the previously mentionedconditions—whether therapy for theunderlying condition, inadequate pleural spacedrainage, or both—can lead to the development oflung entrapment and eventually a trapped lung.Clinical PresentationMost patients with trapped lung present withan asymptomatic, chronic, unilateral pleural effusiondiscovered on a routine chest radiograph or suspectedon physical examination. There is typicallya remote history of pleurisy, pneumonia, an inflammatorypleural process, coronary artery bypassgrafting surgery, 4 or chest trauma. The chestradiograph typically shows a small-to-moderateunilateral pleural effusion without significantmediastinal shift. In the asymptomatic patient, atrapped lung presents as a diagnostic dilemmarather than a condition requiring treatment. Somepatients with trapped lung, however, have varyingdegrees of dyspnea and restrictive physiology onpulmonary function testing and therefore shouldbe considered for decortication.DiagnosisThe sole cause for the persistent pleural effusionin trapped lung is hydrostatic imbalance. The presenceof malignancy, active infection, intense inflammation,a fibrinous exudate, or blood in the pleuralspace implies that other causes for fluid accumulationalso are present, excluding the diagnosis oftrapped lung. If trapped lung is suspected, a therapeuticthoracentesis with pleural manometry shouldbe performed. A sterile, lymphocyte-predominant,low-protein, low-LDH effusion without malignantcells is compatible, but not diagnostic, of trappedlung. The diagnosis is established with a significantdecrease ( 14.5 cm H 2O/L of fluid removed) inpleural liquid pressure, in linear fashion, duringfluid withdrawal. 5 The patient typically experiencessignificant, substernal chest pain from the increasein negative pleural pressure; relief via the allowanceof air through the catheter into the pleural spaceconfirms the diagnosis clinically. An air-contrast CTwith the affected side superior confirms the diagnosiswith visualization of a “thick visceral pleuralpeel.” The patient does not obtain relief of dyspneafrom the thoracentesis.After thoracentesis, reaccumulation of fluid tothe prethoracentesis volume generally occurswithin 48 h to several days. Complete expansionof the lung after thoracentesis is inconsistent withthe diagnosis of trapped lung. If the diagnosis isnot established and the lung has not expandedafter therapeutic thoracentesis, a chest CT scanshould be obtained to exclude severe parenchymaldisease; fiberoptic bronchoscopy should beconsidered to exclude endobronchial obstruction.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 549ACC-PULM-09-0302-034.indd 5497/10/09 8:56:50 PM


At thoracoscopy, a thin, resilient membrane isseen covering the lung surface and preventingreexpansion. On histologic examination, maturefibrosis is evident with few inflammatory cells. Ifchronicity and stability over time has been demonstrated,a diagnosis of trapped lung can beestablished with reasonable confidence. Thesepatients often have a history of a persistent pleuraleffusion for months to years. The effusionresolves completely after decortication if theunderlying lung is normal.ManagementA trapped lung can be prevented if there isappropriate management of the pleural space duringthe inflammatory process. For example, witha complicated parapneumonic effusion or empyema,appropriate antimicrobial therapy and timelydrainage of the pleural space will typically preventa fibrous peel from developing over the visceralpleural surface. Effusions after cardiac surgery andtuberculous pleuritis tend to resolve spontaneouslyor with the administration of drug therapy beforethe development of visceral pleural fibrosis. Whenan active pleural process has been excluded andthere is no further explanation other than amechanical cause for the persistent pleural effusion,a decision can be rendered regarding therapy.The asymptomatic patient with a small trappedlung does not appear to be at risk for a secondaryinfection or other complications and therefore willnot benefit from decortication. In a symptomaticpatient, the likelihood of improvement after decorticationand the feasibility of surgery should beassessed. If the trapped lung is severely diseasedreexpansion after decortication is problematic andtherefore surgery would not be expected to resultin symptomatic relief.The underlying lung is best assessed with achest CT scan. In contemplating surgery, the generalhealth of the patient and the mechanics ofsurgery need to be carefully considered. Ideally,the elastic membrane that covers the visceralpleura can be easily separated from the pleurawithout creating significant air leaks. However, attimes, a defined resection plane may not be present,increasing the likelihood of complications. ChestCT scan may be helpful by demonstrating extensionof fibrous tissue from the pleura surface intothe lung parenchyma. In the properly selectedcandidate, decortication can be achieved successfullyyears after the acute pleural injury.ConclusionTrapped lung should be included in the differentialdiagnosis of a stable, persistent pleuraleffusion. The diagnosis requires exclusion of anactive pleural process. With trapped lung, there isfailure of expansion of the lung after thoracostomyor thoracentesis with an initial negative or minimallypositive pleural pressure and a high pleuralspace elastance. The fibrous membrane coveringthe visceral pleura can be demonstrated with anair-contrast CT scan after removal of pleural fluid.Asymptomatic patients with trapped lung do notrequire therapy. Patients with dyspnea on exertionand restrictive physiology should be consideredfor decortication if the underlying lung is normal.Clinical Pearls• Trapped lung, a result of visceral pleural restriction,is a cause of a persistent, benign, constantvolumeeffusion that is not associated with activepleural infection or inflammation or malignancy.• Lung entrapment, a precursor of trapped lung,is associated with an active pleural process andan exudative effusion.• There is no substantial relief of dyspnea withtherapeutic thoracentesis in trapped lung, andsubsternal chest pain is a constant finding thatis a consequence of a significant decrease inpleural pressure.• A high pleural space elastance ( 14.5 cm H 2O/Lof fluid removed) confirms the diagnosis oftrapped lung.• A fibrous membrane covering the visceral pleuracan be detected on an air-contrast CT scan.• Asymptomatic patients with a trapped lungeffusion do not require treatment.• Decortication should be considered in patientswith exertional dyspnea and restrictive physiologyand normal underlying lung.References1. Doelken P, Sahn SA, Heidecker J. Trapped lung.Semin Respir Crit Care Med 2001; 22:631–635550 Pleural Pearls (Sahn)ACC-PULM-09-0302-034.indd 5507/10/09 8:56:50 PM


2. Huggins JT, Sahn SA, Heidecker J, et al. Characteristicsof trapped lung: pleural fluid analysis,manometry, and air-contrast chest CT. Chest 2007;131:206–2133. Lan R, Singh KL, Chuang M, et al. Elastance ofthe pleural space: a predictor for the outcome ofpleurodesis in patients with malignant effusion.Ann Intern Med 1997; 126:768–7744. Lee YCG, Vaz MAC, Ely KA, et al. Symptomaticpersistent post-coronary artery bypass graftpleural effusions requiring operative treatment:clinical and histologic features. Chest 2001; 119:795–8005. Huggins JT, Doelken P. Pleural manometry. ClinChest Med 2006; 27:229–240Yellow Nail SyndromeThe differential diagnosis of a persistent,benign pleural effusion present for years includeslymphatic abnormalities, such as yellow nailsyndrome (YNS), chylothorax, and lymphangiectasisand an unexpandable lung, such astrapped lung (postcoronary artery bypass graftingsurgery) and lung entrapment (cholesteroleffusion). Effusions that last several months,but not years, include benign asbestos pleuraleffusion, uremic pleurisy, rheumatoid pleurisy,postcardiac injury syndrome, and radiationpleuritis.Definition and DiagnosisYNS, first described more than 40 years ago,classically consists of the triad of yellow, slowgrowingnails, lymphedema, and respiratory tractdisease. 1,2 The latter includes pleural effusion,bronchiectasis, recurrent pneumonia, chronic bronchitis,and chronic sinusitis. However, the diagnosiscan be established when two components of thetriad are present.Epidemiology and PathophysiologyYNS is slightly more common in women andis a consequence of abnormal lymphatic vesselsthat result in impaired drainage and obstruction,leading to subungual edema, periorbital edema,lymphedema of the extremities, and pleural effusions.In addition, there is evidence that increasedmicrovascular permeability contributes to theclinical findings. Lymphangiograms show a paucityof hypoplastic and dilated lymphatics. 3 Theage of onset of symptoms ranges from birth to aslate as 65 years of age, with a median of 40 yearsof age. 4Clinical PresentationYellow or unsightly nails are the initial findingin only one-third of patients. These patients’ nailsgrow at approximately half the rate of the lowerlimits of normal nail growth. Other nail abnormalitiesseen in YNS include overcurvature, eithertransversely or longitudinally; thickening; onycholysis;cross-ridging; and loss of the lunulae andcuticles. Women frequently cover their unsightlynails with opaque nail polish that can obscure thefinding from the unsuspecting observer. Lymphedemais seen in approximately 80% of patientsduring the course of the disease and is the initialsymptom in approximately one-third of patients.Edema can be pitting or nonpitting and can beconfined to the fingertips or eyelids only. Pleuropulmonarysymptoms develop in approximately60% of patients during the course of the disease,whereas a third present with a respiratory manifestation.Pleural effusions have been noted in 36%of recorded cases. Patients often provide a historyof years of recurrent attacks of bronchitis (in anonsmoker), bronchiectasis without a knowncause, chronic sinusitis, or recurrent pneumonias.Chylothorax has also been reported. Symptomsseldom appear simultaneously, and the triad is notalways present.Pleural Fluid AnalysisThe pleural fluid may be unilateral or bilateral,with the volume varying from small to massive. Inmost patients, the fluid is straw colored but can besanguineous. The total protein concentrationranges between 3.5 and 4.5 g/dL, with an LDHconcentration usually, but not exclusively, in thetransudative range. The nucleated cell count isusually 2,000/L with a predominance ( 80%)of lymphocytes. The red blood cell count is usually 10,000/L. The glucose level is equivalent to thatin serum and the pH is approximately 7.40.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 551ACC-PULM-09-0302-034.indd 5517/10/09 8:56:51 PM


ManagementTherapy is usually symptomatic. Local steroidinjection and oral vitamin E have been reported tobe successful in treating the yellow nails. Afterthoracentesis, fluid recurs over weeks to months.Chemical and surgical pleurodesis have been successfulin those with symptomatic pleural effusions.5,6 After pleurectomy in a single patient, therewas remission of the yellow nails. Spontaneouspartial or complete resolution of the nail abnormalitiesoccurs in 30% of patients with occasionalrelapses. Lymphedema and pleural effusionsappear to be persistent, and spontaneous recoveryhas not been reported. There is a single reportof resolution of edema and chylothorax withoctreotide. 7Clinical Pearls• YNS should be considered in the differentialdiagnosis of a discordant exudative pleuraleffusion by protein criterion only with 80%lymphocytes that has been present for monthsor years.• A history of bronchiectasis, chronic bronchitis(in a nonsmoker), chronic sinusitis, or recurrentpneumonias in association with a chronic pleuraleffusion or peripheral edema suggests thediagnosis.• Slow nail growth is the most consistent nailfinding and is more common than yellow discolorationof nails; women often cover theirunsightly nails with opaque nail polish thatmay obscure the diagnosis from the unsuspectingobserver.• The triad of yellow, slow-growing nails, lymphedema,and respiratory tract disease is rarelypresent at the initial encounter; pleural effusionusually is a late manifestation that doesnot regress spontaneously.• The diagnosis can be established when at leasttwo components of the triad are present.• Pleurodesis is successful for treating the symptomaticpleural effusion.References1. Samman PD, White WF. The “yellow nail syndrome.”Br J Dermatol 1964; 76:153–1572. Emmerson PA. Yellow nails, lymphedema and pleuraleffusion. Thorax 1966; 21:247–2533. D’Alessandro A, Muzi G, Monaco A, et al. Yellownail syndrome: does protein leakage play a role? EurRespir J 2001; 17:149–1524. Norkild P, Kromann-Andersen H, Struve-ChristensenE. Yellow nail syndrome: the triad of yellownails, lymphedema and pleural effusions. Acta MedScand 1986; 219:221–2275. David I, Crawford FA, Hendrix GH, et al. Thoracicsurgical implications of the yellow nail syndrome. JThorac Cardiovasc Surg 1986; 91:788–7936. Sudduth D, Sahn SA. Pleurodesis for nonmalignantpleural effusions: recommendations. Chest 1992;102:1805–18607. Makrilakis K, Pavlatos S, Giannikopoulos G, et al.Successful octreotide treatment of chylous, pleuraleffusion and lymphedema in the yellow nailsyndrome. Ann Intern Med 2004; 141:246–247Thoracic EndometriosisEndometriosis is defined by the presence ofendometrial glands in stroma outside the uterinecavitary and musculature. The growth andmaintenance of these endometrial implants aredependent on the ovarian steroids; therefore, endometriosisoccurs only in women of reproductiveage or in those women who are receiving estrogenreplacement therapy. Endometriosis typicallyinvolves the pelvic structures, particularly theovaries, cul-de-sac, broad ligament, and uterosacralligaments. However, on rare occasions, endometrialtissue can occur in the thorax, abdomen,brain, and skin. 1EpidemiologyAlthough the prevalence of pelvic endometriosisis approximately 1%, thoracic endometriosis iseven less common. There were 110 cases of thoracicendometriosis reported in the English-languageliterature between 1966 and 1994. 2 These casesincluded catamenial pneumothorax, hemothorax,hemoptysis, chest pain, and pulmonary nodules.The term catamenial means “related to menses,”which is an important historical feature that suggeststhe diagnosis. The most common form ofthoracic endometriosis is catamenial pneumothorax.Catamenial pneumothorax occurred in 80552 Pleural Pearls (Sahn)ACC-PULM-09-0302-034.indd 5527/10/09 8:56:51 PM


(73%) of 110 cases, catamenial hemothorax in 15(14%), catamenial hemoptysis in 8 (7%), and lungnodules in 7 (6%). 2PathogenesisA number of hypotheses have been proposedto explain the pathogenesis of pelvic endometriosis.The most logical one is that retrograde menstruationresults in endometrial tissue traversingthe Fallopian tubes after failing to clear the peritonealcavity. Movement of endometrial tissue fromthe peritoneal to pleural cavity can occur eitherthrough congenital diaphragmatic defects, whichoccur more commonly in the right diaphragm, orthrough the pelvic veins, with embolization to thepulmonary or the pleural capillaries. 3,4Clinical PresentationAt presentation, the mean age of women withthoracic endometriosis is 35 years (range, 19 to 54years); the onset is typically several years after thepeak age of onset of pelvic endometriosis. Of the110 patients reported, 61 underwent laparoscopyor laparotomy; 51 (84%) had evidence of pelvicendometriosis. Pleural implants, however, werefound in 15% of patients who underwent thoracostomyor thoracotomy, whereas diaphragmaticdefects and/or parenchymal cysts or blebs wereobserved in 25% of patients.Patients with thoracic endometriosis typicallyhave symptoms within 24 to 48 h of the onset ofmenstruation; however, catamenial symptoms maynot occur with each menstrual cycle. Chest pain isthe most common symptom, occurring in 90% ofpatients; dyspnea occurs in about 30%. Catamenialpneumothorax is almost exclusively (95%) a rightsidedevent, and the pneumothorax is typicallysmall to moderate in size. 2,5DiagnosisThe diagnosis should be considered in awoman of reproductive age who presents with apneumothorax (nonsmoker), hemothorax, hemoptysis,or chest pain associated with menses. Theliterature is replete with references to multipleepisodes that are undiagnosed for years. Simplyhaving knowledge about this disease should leadto earlier diagnosis and timely, specific therapywith decreased morbidity.There have been isolated reports of the diagnosisbeing established by pleural fluid cytology,needle aspiration of lung lesions, cytology frombronchoscopic specimens, and visualization orbiopsy of pleural lesions. <strong>Pulmonary</strong> and pleuralnodules have been documented on chest CT within48 to 72 h of the onset of menses.ManagementThe goals of treatment are twofold: eradicationor suppression of thoracic endometrial tissue andprevention of reseeding from the pelvis. Initialtreatment usually consists of suppression of ectopicendometrium by inhibiting ovarian estrogen secretion.Therefore, oral contraceptives, progestins,danazol, or gonadotropin-releasing hormone analogshave all been used to suppress ovulation.Unfortunately, ovulation suppression appears tobe effective in less than half of the patients.The most effective form of therapy for patientswith catamenial pneumothorax or hemothorax ispleurodesis. Pleurodesis is accomplished by thoracoscopictalc poudrage, thoracoscopic pleuralabrasion, partial pleurectomy, and chemicalpleurodesis through a chest tube. However, evenwith a successful pleurodesis, patients may stilldevelop catamenial chest pain as long as endometrialtissue is present in the thorax. Recurrentsymptoms are presumably the result of cyclicalproliferation of the pleuropulmonary endometrialimplants in response to ovarian estrogens. Thesesymptoms can be relieved by hysterectomy withbilateral salpingo-oophorectomy but may recur ifestrogen replacement therapy is initiated and dormantthoracic endometrial tissue is reactivated.Plication of the diaphragmatic defects has beensuccessful in preventing pneumothorax in somepatients.Clinical Pearls• Thoracic endometriosis is a clinical diagnosisin women who develop right-sided pneumothorax,hemothorax, chest pain, or hemoptysisin association with the onset of menses.• Symptoms may be transient and may not occurwith each menstrual cycle.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 553ACC-PULM-09-0302-034.indd 5537/10/09 8:56:51 PM


• Most patients have pelvic endometriosis diagnosedseveral years earlier.• Suppression of ovulation prevents recurrencein 50% of patients.• Pleurodesis is effective in preventing pneumothoraxand hemothorax; however, catamenialchest pain may occur if endometrial tissuepersists in the thorax.References1. Olive DL, Schwartz LB. Endometriosis. N Engl JMed 1993; 328:1759–17692. Joseph J, Sahn SA. Thoracic endometriosis syndrome:new observations from an analysis of 110cases. Am J Med 1996; 100:164–1703. Vinatier D, Roazi G, Cosson M, et al. Theories ofendometriosis. Euro J Obstet Gynecol Reprod Biol2001; 96:21–344. Cowl CT, Dunn WF, Deschamps C. Visualization ofdiaphragmatic fenestration associated with catamenialpneumothorax. Ann Thorac Surg 1999; 68:1413–14145. Johnson MM. Catamenial pneumothorax and otherthoracic manifestations of endometriosis. Clin ChestMed 2004; 25:311–319Chronic Tuberculous EmpyemaDefinition and CausesChronic tuberculous empyema, an entity distinctfrom and much less common than tuberculouspleural effusion, represents chronic, active infectionof the pleural space. Chronic tuberculousempyema can occur in several settings: (1) progressionof the primary tuberculous effusion (usuallyassociated with a large to massive effusion) thathas led to lung entrapment; (2) direct extension ofinfection into the pleural space from thoraciclymph nodes or a subdiaphragmatic focus; (3)hematogenous spread; (4) after oleothorax or afterLucite ball plombage; (5) postpneumonectomy;and (6) after therapeutic pneumothorax. 1Clinical FeaturesIn patients with chronic tuberculous empyema,the inflammatory process may be present for yearswith a paucity of clinical symptoms. The prolongedasymptomatic course can be explained by themarked pleural thickening that virtually isolatesthe tubercle bacilli to the empyema cavity. Thesepatients often come to clinical attention at the timeof a routine chest radiograph or after the developmentof a bronchopleural fistula or empyemanecessitatis. 1The onset of a bronchopleural fistula, whichmay be dramatic, with acute fever, dyspnea, andproduction of copious, mucopurulent sputum, notonly heralds the disease but increases the risk ofspread of the tubercle bacillus through infectedsputum. Some patients may have insidious, constitutionalsymptoms, such as fatigue and weightloss, and also manifest low-grade fever and nightsweats. 2Empyema necessitatis occurs when the encapsulatedempyema erodes through the parietalpleura and discharges its contents. The most commonsite for empyema necessitatis is in the subcutaneoustissues of the chest wall; therefore, patientscan present with a chest wall mass. Subsequentcutaneous rupture can occur, resulting in a pleurocutaneousfistula. Other sites of empyema necessitatisinclude the esophagus, vertebral column,retroperitoneum, pericardium, flank, and groin.Before the development of antituberculous medications,Mycobacterium tuberculosis was the mostcommon cause of empyema necessitatis.Radiographic FindingsThe typical chest radiographic finding ofchronic tuberculous empyema is a moderate-tolarge,loculated pleural effusion with pleural calcificationand enlargement of the overlying ribs dueto the chronic infectious process. The diagnosis ofchronic tuberculous empyema is confirmed clinicallyby finding a thick, calcific, pleural rind andrib thickening surrounding loculated pleural fluidon CT imaging. Chest CT findings of a thick-walled,well-encapsulated, calcific pleural mass associatedwith an extrapleural mass are virtually diagnosticof tuberculous empyema necessitatis. 3 Direct communicationbetween the pleura and chest wallcollections is rarely demonstrated because of thethin nature of the tract, which is below the resolutionof the images obtained, or because the tract istangential to the imaging plane and is missed byvolume averaging. In addition to tuberculous554 Pleural Pearls (Sahn)ACC-PULM-09-0302-034.indd 5547/10/09 8:56:51 PM


empyema, the differential diagnosis of empyemanecessitatis includes bacterial empyema, lungabscess, blastomycosis, and actinomycosis.Pleural Fluid AnalysisThe definitive diagnosis of tuberculous empyemais established at thoracentesis by findingpurulent fluid that is smear positive for acid-factbacilli and subsequently cultures M tuberculosis.Tuberculous empyemas typically have nucleatedcell counts 100,000/L, with virtually all the cellsbeing neutrophils. The pleural fluid is acidic, withpH usually 7.20, and has a low glucose concentration,usually 20 mg/dL. Pleural fluid proteinconcentration is typically 5 g/dL, and theLDH is 1,000 IU/L. Anaerobic and aerobic culturesshould be performed because, on occasion,there is concomitant bacterial and mycobacterialinfection. 1ComplicationsComplications that have been associated withchronic tuberculous empyema include pleuralmalignancy (most commonly non-Hodgkin lymphoma),superior vena cava syndrome, and pleuralaspergillosis. 4TreatmentThe principles of treatment, pleural spacedrainage and antimicrobial chemotherapy, are thesame for both bacterial and tuberculous empyema.However, problems specific to chronic tuberculousempyema include the inability to reexpand thelung (lung entrapment) spontaneously with drainageand to achieve therapeutic drug levels in pleuralfluid, which can lead to drug resistance. 2,5Several patients have been described in whomchemotherapy for tuberculous empyema was complicatedby progressive, acquired drug resistance.It is believed the mycobacteria are exposed to subtherapeuticconcentrations of some or all the antituberculousdrugs as the result of poor penetrationof the agents into the empyema cavity because ofa severely fibrotic and calcific pleura; it is possiblethat the subtherapeutic drug concentrations resultin the emergence of resistant strains.There is a single report of successful nonsurgicaltreatment of tuberculous empyema withserial, space-emptying thoracenteses and 24months of isoniazid, rifampin, and ethambutol.Thoracentesis was repeated bimonthly for thefirst 2 months, monthly for 3 months, and lessfrequently as the fluid reaccumulated moreslowly. Twenty-four months of therapy was chosenbased on the rate of improvement of thepleural fluid by laboratory parameters. 6 Surgicaloptions for chronic tuberculous empyema includestandard decortication, decortication limited tothe parietal side of the peel’s sac, thoracoplasty,parietal wall collapse, open drainage, myoplasty,and omentopexy. 7−9Clinical Pearls• Much less common than tuberculous pleuraleffusion, tuberculous empyema representschronic, active infection in the pleural space.• Tuberculous empyema may be present foryears (up to 62 years) with a paucity of clinicalsymptoms.• Tuberculous empyema may be discovered onroutine chest radiograph or after the developmentof a bronchopleural fistula or empyemanecessitatis.• Chest CT is virtually diagnostic by finding athick, calcific pleural rind and rib thickeningsurrounding loculated pleural fluid.• Pleural fluid analysis shows purulent fluid thatis smear positive for acid-fast bacilli.• Anaerobic and aerobic cultures of the fluidshould be performed because, on occasion,there is concomitant bacterial and mycobacterialinfection.• Problems associated with treatment include theinability to reexpand the lung and to achievetherapeutic drug levels in pleural fluid.References1. Sahn SA, Iseman MD. Tuberculous empyema. SeminRespir Infect 1999; 14:82–872. Iseman MD, Madsen LA. Chronic tuberculousempyema with bronchopleural fistula resulting intreatment failure and progressive drug resistance.Chest 1991; 100:124–1273. Hulnick DH, Naidich DP, McCauley DI. Pleuraltuberculosis evaluated by computed tomography.Radiology 1983; 149:759–765<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 555ACC-PULM-09-0302-034.indd 5557/10/09 8:56:51 PM


4. Luchi K, Aozasa K, Yamamoto S, et al. Non-Hodgkin’s lymphoma of the pleural cavity developingfrom long-standing pyothorax: summaryof clinical and pathologic findings in thirty-sevencases. Jpn J Clin Oncol 1989; 19:249–2575. Elliott AM, Berning SEN, Iseman MD, et al. Thefailure of drug penetration and acquisition of drugresistance in chronic tuberculous empyema. TubercLung Dis 1995; 76:463–4676. Neihart RE, Hof DF. Successful non-surgical treatmentof tuberculous empyema in irreducible pleuralspace. Chest 1985; 88:792–7947. Massard G, Rouge C, Wihlm JM, et al. Decorticationis a valuable option for late empyema after collapsetherapy. Ann Thorac Surg 1995; 68:888–8958. Al-Kattan KM. Management of tuberculous empyema.Eur J Cardiothorac Surg 2000; 17:251–2549. Tatsumura T, Koyama S, Yamamoto K, et al. A newtechnique for one-stage radical eradication of longstandingchronic thoracic empyema. J Thorac CardiovascSurg 1990; 99:410–415UrinothoraxPleural effusions associated with renal diseaseoccur with nephrotic syndrome, salt and waterexcess leading to congestive heart failure, uremicpleurisy, peritoneal dialysis, perinephric abscess,and urinary tract obstruction.CausesAn effusion secondary to obstructive uropathyhas a unique pathogenesis and is termedurinothorax. With rare exception, hydronephrosiswith extravasation of fluid into the perirenalspace appears to be a prerequisite for the developmentof a urinothorax. Causes of urinothoraxreported in the literature include bladder andprostate cancer, posterior urethral valves, renalcysts, nephrolithiasis, surgical ureteral manipulation,blunt kidney trauma, renal transplant, ilealconduit with ureteral obstruction, and bladderlaceration. 1,2PathogenesisObstructive uropathy leads to perirenal fluidaccumulation and movement of urine transdiaphragmaticallyinto the ipsilateral pleural space. 2Removal of the obstructing lesion generally resultsin rapid resolution (within days) of the effusion.The pathogenesis of the low-pH transudate is notclearly established, although it is most likelyrelated to the low pH of the extravasated urinewith back-diffusion of hydrogen ions as thefluid passes from the retroperitoneal into the pleuralspace. 3 An alkaline urine at the time of extravasationcould result in a pleural fluid pH up to8.0.Clinical PresentationDepending on the size of the pleural effusionand the underlying pulmonary status, the patientmay present with dyspnea or be asymptomatic.Other symptoms would depend on the cause ofthe urinary tract obstruction. 2,3Chest RadiographThe chest radiograph should demonstrate asmall-to-moderate effusion ipsilateral to theobstructed kidney; however, there are reports ofbilateral and contralateral effusions.Pleural Fluid AnalysisThe pleural fluid looks and smells like urine.The fluid is a transudate that frequently has a pH 7.30. The total protein may be 1 g/dL if thereis no significant proteinuria. The pleural fluid/serum creatinine ratio is 1.0 as long as urinarytract obstruction persists. 1,3DiagnosisThe effusion is often discovered on a routinepostoperative chest radiograph. The diagnosisshould be suspected in the setting of obstructiveuropathy and can be confirmed by thoracentesisdemonstrating a pleural fluid/serum creatinineratio 1.0. The ratio of pleural fluid to serumcreatinine appears to be specific for the diagnosis,especially if measured within the first few daysof pleural fluid formation; an additional clue tothe diagnosis is the finding of a low-pH transudate.The latter is a unique combination as allother low-pH effusions (pH 7.30) have beenassociated with exudates (empyema, complicated556 Pleural Pearls (Sahn)ACC-PULM-09-0302-034.indd 5567/10/09 8:56:52 PM


parapneumonic effusion, esophageal rupture,carcinoma, rheumatoid pleurisy, tuberculosis, andlupus pleuritis).TreatmentRelief of obstruction results in rapid reversalof the pleural fluid/serum creatinine ratio andresolution of the effusion over several days withoutpleural space sequelae.Clinical Pearls• A urinothorax is typically an ipsilateral urinecollection in the pleural space due to obstructiveuropathy.• Urinothorax is the only single cause of a lowpHtransudate.• A pleural fluid/serum creatinine ratio 1.0 isdiagnostic of a urinothorax.• The pleural fluid/serum creatinine ratio 1.0and low pH will not be maintained once theurinary tract obstruction is relieved.• The effusion resolves rapidly over a few daysafter the relief of obstruction, and there are nopleural sequelae.References1. Stark BD, Schanes JG, Baron RL, et al. Biochemicalfeatures of urinothorax. Arch Intern Med 1982;142:1509–15112. Salcedo JR. Urinothorax: report of 4 cases and reviewof the literature. J Urol 1986; 135:805–8083. Miller KS, Wooten S, Sahn SA. Urinothorax: a causeof low pH transudative pleural effusions. Am J Med1988; 85:448–449Duropleural FistulaDefinitionA duropleural fistula (DPF) or subarachnoidpleural fistula represents communication betweenthe subarachnoid and pleural spaces.Pathophysiology and CausesThe mechanism of a DPF involves disruption ofthe dural membrane and parietal pleura, allowinga tract to develop between the subarachnoid andpleural spaces. Once the fistula develops, a pressuregradient is established, which allows cerebrospinalfluid (CSF) to flow from the positive-pressure subarachnoidspace to the negative-pressure pleuralspace.Most cases of DPF are secondary to blunt andpenetrating trauma, with a total of 23 casesreported since 1959. 1 The mechanism that resultsin a traumatic fistula can involve either a missiletraversing both the pleural and subarachnoidspaces or a vertebral fracture that tears the duraand parietal pleura. A fistulous tract usually occursbetween the upper thoracic subarachnoid spaceand the pleural space. With blunt trauma, the disruptionoccurs from extreme extension of the spine,resulting in the tearing of the relatively immobilenerve roots and dura. Significant chest wall compressionmay perforate the pleura against the bonyprominence of the spine. Regardless of the cause,the DPF remains open because of the pressuregradient between the positive pressure of the subarachnoidspace and the negative pressure in thepleural space.Although laminectomy is a common neurosurgicalprocedure to treat spinal injuries, there havebeen only two reports of DPF as a complication. 2,3Neurosurgery may be a more common cause thantrauma; however, these DPFs may be recognizedearly by neurosurgeons and repaired without generatingreports in the literature and requestingpulmonary consultation. DPF has been documentedas a complication of thoracotomy in sixcases and rupture of intrathoracic meningocele inone case.Clinical PresentationA high degree of suspicion is necessary toestablish the diagnosis of a DPF. A DPF should beconsidered as a cause of a persistent transudativeeffusion. Symptoms that suggest a CSF leak includepostural headaches, nausea, and vomiting. Oftensymptoms related to the effusion, such as chestpain, dyspnea, or fever, may be minimal or overshadowedby concomitant injuries, often delayingthe diagnosis. In addition to the almost universalfindings of severe spinal cord injury, other clues tothe diagnosis may include pneumocephalus ormeningitis.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 557ACC-PULM-09-0302-034.indd 5577/10/09 8:56:52 PM


Chest RadiographShortly after the development of a DPV, thechest radiograph may be normal. Effusions rangefrom small to massive depending on the size andduration of the fistula. Mediastinal widening maybe an early manifestation, and once aortic injuryhas been excluded, a diagnosis of subarachnoidmediastinalor subarachnoid-pleural fistula shouldbe considered.Pleural Fluid AnalysisPleural fluid in nontraumatic DPF is clear andlooks like water. The nucleated cell count is lowand the pleural fluid glucose value is approximately0.6 of the serum glucose. An importantfeature of the transudative effusion is that the totalprotein is consistently 1.0 g/dL (if preciselymeasured should be 0.05.); the LDH is alsoclearly in the transudative range. 1,4 Other causes oftransudates with total protein consistently 1.0g/dL include urinothorax, peritoneal dialysis, andextravascular migration of a central venous catheterwith saline or glucose infusion. Previousauthors have described both transudates and exudatesin traumatic DPF. Because concomitantpleural processes secondary to trauma can be present,the diagnosis of DPF by pleural fluid analysiscan be confounded by other pleural processes, suchas hemothorax. 1 β 2-Transferrin is produced byneuraminidase activity in the brain and is founduniquely in the CSF and inner ear perilymph thisprotein is accepted as a sensitive and specificmarker to identify CSF leaks into the pleural spacefollowing head or spine trauma or spinal surgery;testing for β 2-transferrin is available to practicingclinicians. 1,4,5DiagnosisAlthough the presence of β 2-transferrin in pleuralfluid diagnostic of DPF, definitive identificationof the fistula requires radiographic visualization.Conventional myelography and radionuclidemyelography are most commonly used. Severalauthors have concluded that isotope myelographyis more sensitive than conventional myelography. 6However, there are reports of false-negative resultswhen myelography is used alone, with slow orintermittent leaks. CT scanning performed concomitantlywith myelography will add to thedelineation of the anatomic defect. In addition, thepresence of contrast in the pleural space aftermyelography confirms the diagnosis in patientswith slow CSF leaks.TreatmentBecause of the limited number of cases, thereis no strong consensus regarding management ofDPF. However, spontaneous resolution is rare, andmost patients require either surgical ligation orclosed-tube drainage for management. Of the 19posttraumatic DPFs reported in one series, 13(68%) patients were treated definitively withlaminectomy or thoracotomy, while 3 (16%)patients responded to chest tube drainage. Two(11%) of 19 patients were treated with thoracentesis,although one was treated with “conservative”measures. In 9 (82%) of the 11 patients in whomlaminectomy was performed, closure of the fistulawas successful. The appropriate timing of surgicalintervention is unknown. Some surgeons advocatechest tube drainage for up to 2 weeks before surgicalintervention, whereas others recommend earlyintervention, noting the low rate of spontaneousclosure. 7Clinical Pearls• The most common reported cause of DPF isblunt or penetrating trauma; other causesinclude transthoracic diskectomy and thoracotomy.• Symptoms that suggest CSF leak include posturalheadaches, nausea, and vomiting.• DPF is a cause of a persistent, transudativepleural effusion.• In nontraumatic cases, the pleural fluid lookslike water, is paucicellular with a mono nuclearpredominance, has an alkaline pH, has a glucoselevel of approximately 0.6 of the serum glucose,and has a protein concentration 1 g/dL (typically 0.05).• The presence of β 2-transferrin in the pleuralfluid confirms the diagnosis.• Definitive identification of the fistula requiresradiographic visualization, either by radionuclideor conventional myelography.558 Pleural Pearls (Sahn)ACC-PULM-09-0302-034.indd 5587/10/09 8:56:52 PM


• Spontaneous resolution is rare, and most patientsrequire surgical closure for definitive treatment.References1. Lloyd C, Sahn SA. Subarachnoid pleural fistuladue to penetrating trauma. Chest 2002; 122:2252–22562. Monla-Hassan J, Eichenhorn M, Spickler E, et al.Duropleural fistula manifested as a large pleuraltransudate: an unusual complication of transthoracicdiskectomy. Chest 1998; 114:1786–17883. Assietti R, Kibble MB, Bakay R. Iatrogenic cerebrospinalfluid fistula to the pleural cavity: case reportand literature review. Neurosurgery 1993; 33:1104–11084. Huggins JT, Sahn SA. Duro-pleural fistula diagnosedby β2-transferrin. Respiration 2003; 70:423–4255. Skedros DG, Cass SP, Hirsch BE, et al. Beta-2 transferrinassay in clinical management of cerebralspinal fluid and perilymphatic fluid leaks. J Otolaryngol1993; 22:341–3446. Rosen PR, Chaudhuri TK. Radioisotope myelographyin the detection of pleural-dural communicationas a source of recurrent meningitis. Clin NuclMed 1983; 8:28–307. Pollack II, Pang D, Hall W. Subarachnoid-pleuraland subarachnoid mediastinal fistulae. Neurosurgery1990; 26:519–525Cholesterol Pleural EffusionA cholesterol pleural effusion, also referred toas a pseudochylous or chyliform effusion, is characterizedby a high lipid content that is not a consequenceof leakage from the thoracic duct. It isthought to be a rare condition but probably occursmore commonly than reported in the literature.Some refer to pseudochylothoraces as lipid effusionswith cholesterol crystals and to chyliform effusionsas those without cholesterol crystals. 1−3PathogenesisThe pathogenesis of these effusions has not beenclearly elucidated. However, these effusions are aconsequence of chronic pleural inflammation in thesetting of lung entrapment and severe parietalpleural fibrosis that evolve over several years. 2 Theorigin of the elevated pleural fluid cholesterol hasbeen attributed to local degeneration of pleuralnucleated cells and RBCs. The lipoprotein patternof a cholesterol effusion shows a shift of cholesterolbinding towards the high-density lipoproteins,most likely occurring as the cholesterol is sequesteredin the pleural space and undergoes a changein lipoprotein binding characteristics. Cholesterolin these effusions may be complexed with triglyceridesand proteins. 4These chronic, inflammatory effusions resultin “permanent” lung entrapment that causes persistenceof the effusion as a result of hydrostaticimbalance; a concomitant, low-grade inflammatoryprocess continues, as evidenced by a pleuralfluid total protein and LDH in the exudativerange with concomitant neutrophilia. 2 Over time,pleural fibrosis progresses and large cholesteroldeposits are often demonstrated on histologicexamination.The high triglyceride concentration noted insome cholesterol effusions cannot be explained bylocal cellular degeneration and supports thehypothesis that plasma lipoproteins move into thepleural space bound to triglycerides. 4CausesA literature review published in 1999 revealed175 cases of cholesterol pleural effusions, with78% in men (age range, 17 to 81 years). The majorityof the cases were attributable to tuberculosis;rheumatoid pleurisy was a distant second in frequency.Other causes include paragonimiasis,lung cancer, empyema, hemothorax, and trauma(Table 1). 5Table 1. Causes of Cholesterol Effusions*Chronic tuberculous pleural effusion (most common)Rheumatoid pleurisy (2nd most common)ParagonimiasisLung cancerEchinococcosisHemothoraxCollapse therapy for TBTraumaHemothorax*From Garcia-Zamalloa et al. 5<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 559ACC-PULM-09-0302-034.indd 5597/10/09 8:56:52 PM


DiagnosisThe cardinal feature of these effusions isa high cholesterol concentration, which is independentof the serum cholesterol. A cholesterolpleural effusion may be differentiated from achylothorax by its etiology, chronicity, lack ofsymptoms, a cholesterol/triglyceride ratio 1.0,and a total cholesterol level 200 mg/dL (Table 2).The effusion can have a milky appearance orappear turbid with a satin-like sheen. These effusionsalso have been described as resembling“soft, white cheese” and “motor oil.” Cholesterolcrystals may be seen on microscopy. Cholesteroleffusions typically have cholesterol concentrationsfrom 300 to 1,500 mg/dL; however, lowerlevels may occur from a dilutional effect afterrepeated thoracenteses. Chylothoraces usuallyhave cholesterol concentrations 150 mg/dL.When the cholesterol concentration is 200 mg/dL,the most likely diagnosis is a cholesterol effusion,regardless of the triglyceride levels. A chylouseffusion is most likely when the triglyceridevalue is 110 mg/dL and the cholesterol level is 200 mg/dL. 2ManagementThese effusions represent the consequence ofprolonged lung entrapment (hydrostatic imbalance)and chronic inflammation. The underlyingdisease should be treated if possible. If the effusionis small and the patient relatively asymptomatic,observation with reassurance is indicated. Decorticationshould be considered for the symptomaticpatient with restrictive physiology as long as theunderlying lung is relatively normal. However, thepresence of an active inflammatory process wouldappear to make the prospect of successful decorticationless likely. 6Clinical Pearls• A cholesterol effusion is distinct from a chylothoraxand is a consequence of chronic pleuralinflammation and lung entrapment.• Lung entrapment causes hydrostatic imbalance,whereas pleural inflammation leads to acapillary leak and with high total protein andLDH concentrations.• The most common causes reported are chronictuberculous empyema and rheumatoid pleurisy.• The diagnosis can be suspected by observingfluid with a satin-like sheen and diagnosed if thecholesterol concentration is 200 mg/dL andcholesterol crystals are seen micro scopically.• If the effusion does not resolve with treatmentof the underlying disease, decortication shouldbe considered for the symptomatic patient withrestrictive physiology.References1. Coe JE, Aikawa JK. Cholesterol pleural effusion.Arch Intern Med 1961; 108:163–174Table 2. Differentiating a Cholesterol Effusion From Chylothorax*Condition Chylothorax Cholesterol EffusionIncidence Uncommon RareCauses Lymphoma, trauma, surgery, lymphangioleiomyomatosis Lung entrapment; tuberculosis, rheumatoidarthritis, empyemaOnset Acute to subacute Insidious; chronic course (years)Symptoms Dyspnea None; dyspneaAppearance Serous, milky, turbid, bloody Milky; satin-like sheenPleural fluidanalysisProtein-discordant exudate with 80% lymphocytes;triglycerides 110 mg/dLNeutrophil-predominant exudate; cholesterol, 200 mg/dL; cholesterol/triglycerideratio 1.0Diagnosis Presence of chylomicrons; cholesterol 60 andCholesterol 200 mg/dL; cholesterol crystals200 mg/dLTreatment Manage underlying disease; pleurodesis Observe; decortication*From Agrawal V, Sahn SA. 2560 Pleural Pearls (Sahn)ACC-PULM-09-0302-034.indd 5607/10/09 8:56:53 PM


2. Agrawal V, Sahn SA. Lipid pleural effusions. Am JMed Sci 2008; 335:16–203. Hillerdal G. Chylothorax and pseudochylothorax.Eur Respir J 1997; 10:1157–11624. Hamm H, Pfalzer B, Fabel H. Lipoprotein analysis ina chyliform pleural effusion: implications for pathogenesisand diagnosis. Respiration 1991; 58:294–3005. Garcia-Zamalloa A, Ruiz-Irastorza G, Aguayo FJ,et al. Pseudochylothorax: report of two cases andreview of the literature. <strong>Medicine</strong> 1999; 78:1–136. Goldman A, Burford TH. Cholesterol pleural effusions:a report of three cases with cure by decortication.Dis Chest 1950; 18:586–594Spontaneous Bacterial Empyema(Spontaneous Bacterial Pleuritis)DefinitionSpontaneous bacterial empyema is the termused in the literature when bacteria are culturedfrom pleural fluid of a cirrhotic patient withhepatic hydrothorax who has a neutrophil count 250/μL, similar to spontaneous bacterial peritonitis(SBP). It can also be diagnosed if pleuralfluid culture is negative and the pleural fluidneutrophil count is 500 cells/μL, there isabsence of an intra-abdominal source of infection,and the patient has not received antibiotics in thepast 30 days. However, spontaneous bacterialempyema is a misnomer because, in the vastmajority of cases, the pleural fluid is not purulentand, therefore, is not an empyema. The appropriateterm should be spontaneous bacterial pleuritis(SBPL).PathogenesisConn reported the first case of SBP, 1 which hasa prevalence of 10 to 20% in hospitalized patientswith cirrhosis and clinical ascites. The organismsresponsible for this infection are enteric bacteria,which may reach the ascitic fluid by migrationacross the bowel wall or through hematogenousspread. In normal individuals, enteric bacteria inportal blood are removed by the reticuloendothelialsystem. In patients with cirrhosis who haveimpaired reticuloendothelial system function,intrahepatic and portosystemic shunting of bloodis likely to cause persistent bacteremia. Cirrhoticpatients have a number of defects in host defenses,including defects in serum bactericidal function,chemoattraction, opsonization, and impaired cellularfunction of neutrophils and monocytes. Inaddition, ascitic fluid provides an excellent culturemedium for bacteria.Approximately 6% of patients with cirrhosisand clinical ascites have a pleural effusion (hepatichydrothorax). 2 These effusions occur most commonlyon the right side (70%) and less often on theleft (15%) or bilaterally (15%). A hepatic hydrothoraxdevelops as ascitic fluid moves into the pleuralspace through congenital diaphragmatic defectsalong a pressure gradient. This occurs more commonlyon the right, as there are a greater numberof diaphragmatic defects compared with the lefthemidiaphragm.Clinical FeaturesThe exact incidence of SBPL is unknown. Inaddition to case reports, there have been only tworetrospective series totaling 15 episodes and aprospective study of 24 episodes. 3,4 In the prospectivetrial, 120 patients with hepatic hydrothoraxwere hospitalized during a period of 4 years. 4Ninety-five (79%) of the patients had detectableascites in addition to their pleural effusion. Sixteen(13%) of the 120 patients had 24 episodes of SBPL.All patients had advanced cirrhosis, and most hadbeen hospitalized on previous occasions withclinical signs of progressive liver dysfunction.Clinical ascites was not detectable during 6 (25%)of the 24 episodes of SBPL. In the 18 episodes withascites, the ascitic fluid was infected in 14 (78%)patients and sterile in 4 (22%) patients. In 10 (43%)of the 24 episodes, SBPL was not associated withSBP (6 without ascites and 4 with noninfectedascites). Blood culture results were positive in 11(46%) of 24 episodes, 4 of 10 patients without SBP,and 7 of 14 with SBP. Pleural fluid culture resultswere positive in 18 (75%) of the 24 episodes usinga more sensitive method of culture (inoculation ofpleural fluid into a tryptic soy broth blood culturebottle at the patient’s bedside). 4−6The microorganisms identified in pleural fluidwere Escherichia coli in eight patients, Streptococcussp in four patients, Enterococcus sp in threepatients, Klebsiella pneumoniae in two patients, andPseudomonas stutzeri in one patient. 4 Six patients<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 561ACC-PULM-09-0302-034.indd 5617/10/09 8:56:53 PM


had culture-negative SBPL. The etiologic diagnosiswas confirmed in four of the six culture-negativepatients, two by ascitic fluid culture, one by bloodculture, and one by both ascitic and bloodcultures.Only the pleural fluid LDH increased duringinfection; there was no significant change in totalprotein and glucose with SBPL. The absolute pleuralfluid neutrophil count in the 24 episodes ofSBPL was 1,000/μL in 21 (88%) episodes, witha range of 530 to 13,300/μL. In 8 (36%) of 22 episodes,the pleural fluid pH was 7.30 (range 7.04to 7.50). 4TreatmentAll patients were treated with antibiotics for 7to 10 days. A chest tube was not inserted in any ofthe patients with SBPL. 4OutcomeFive (20%) of the 24 patients died. The causesof death were septic shock, esophageal varicealhemorrhage, and hepatic insufficiency. In theremainder of the patients, the infection was curedafter 7 to 10 days of antibiotic treatment. Four of the10 patients with SBPL who were discharged diedbetween 1 and 19 months after their episode. In fivepatients, liver transplantation was performed 2 to24 months after the episode of SBPL. 4Clinical Pearls• SBPL is a frequent complication of hepatichydrothorax; the 13% incidence is similar tothat for patients with SBP with ascites.• SBPL is rarely diagnosed because thoracentesisis not performed routinely in patients withhepatic hydrothorax, and hepatic hydrothoraxin patients with clinical ascites is relativelyuncommon.• There are patients with culture-positive SBPLwho have culture-negative SBP and vice-versa.Therefore, a thoracentesis should be performed,in addition to paracentesis and bloodcultures, in cirrhotic patients with suspectedinfection.• Ascites is not a prerequisite for SBPL, suggestingthat enteric microorganisms reach the pleuralspace hematogenously.• The inoculation of 10 mL of pleural fluid andascitic fluid into a tryptic soy broth blood culturebottle at the patient’s bedside increases thesensitivity for the diagnosis of both SBPL andSBP compared with traditional delayed inoculationtechniques.• If patients with SBPL do not have purulentfluid, they should be treated with antibioticsalone without the insertion of a chest tube,which can result in significant morbidity.References1. Conn HO. Spontaneous peritonitis and bacteremiain Laennec’s cirrhosis caused by entericorganisms: a relatively common but rarely recognizedsyndrome. Ann Intern Med 1964; 60:568–5802. Lieberman FL, Hidemura R, Peters RL, et al. Pathogenesisand treatment of hydrothorax complicatingcirrhosis with ascites. Ann Intern Med 1964; 61:385–4013. Xiol X, Castellote J, Baliellis C, et al. Spontaneousbacterial empyema in cirrhotic patients: analysis of11 cases. Hepatology 1990; 11:365–3704. Xiol X, Castellvi JM, Guardiola J, et al. Spontaneousbacterial empyema in cirrhotic patients: a prospectivestudy. Hepatology 1996; 23:719–7235. Castellote J, Xiol X, Verdaguer R, et al. Comparisonof 2 ascitic fluid culture methods in cirrhotic patientswith spontaneous bacterial peritonitis. Am J Gastroenterol1990; 85:1605–16086. Runyon BA, Antillon MR, Akriviadis EA, et al.Bedside inoculation of blood culture bottles withascitic fluid is superior to delayed inoculation inthe detection of spontaneous bacterial peritonitis.J Clin Microbiol 1990; 28:2811–2812562 Pleural Pearls (Sahn)ACC-PULM-09-0302-034.indd 5627/10/09 8:56:53 PM


Lung TransplantationStephanie M. Levine, MD, FCCPObjectives:• Define the indications for lung transplantation• <strong>Review</strong> the guidelines for recipient selection for lungtransplantation• Describe the relative and absolute contraindications to lungtransplantation• Describe outcomes after transplantation, including survivaland physiologic results• <strong>Review</strong> the complications after lung transplantation• Give an overview of the immunosuppressive medicationsused in lung transplantationKey words: acute rejection; cytomegalovirus; immunosuppression;lung transplantation; obliterative bronchiolitisDuring the last 2.5 decades, lung transplantation(LT ) has become a successful therapeutic optionfor patients with end-stage pulmonary parenchymaland vascular diseases. Dr. James Hardy at theUniversity of Mississippi performed the first LTin 1963. However, the patient survived for only 18days. Subsequent attempts at LT were complicatedby bronchial anastomotic dehiscence and early graftfailure, resulting in limited survival. Advances indonor and recipient selection, improved surgicaltechniques, new immunosuppressive drugs,and better management of infections have allcontributed to improved survival. Despite theseadvancements, numerous complications still existafter LT.According to the 2008 report from the InternationalSociety of Heart Lung Transplantation(ISHLT), 24,904 LT procedures have been performedworldwide; in 2006 alone, 2,168 procedureswere performed. The 1-year, 3-year, 5-year,and 10-year survival rates are 78%, 63%, 51%, and28%, respectively, as reported by the Registry ofthe ISHLT. 1 The major rate-limiting factors to thelong-term survival of LT patients remain chronicrejection and infection.Most LT referrals come from pulmonary physiciansfrom outside of tertiary transplant centers.With the increase in the number of transplantrecipients, much of the subsequent follow-up careis now conducted by the referring pulmonologistfor logistical convenience and often is dictated byinsurance company mandates. This section willattempt to concisely review the topic of LT for thepulmonary physician.Indications for LTHeart-Lung TransplantationOne of the original successful LT procedures,performed primarily in the late 1980s and early1990s, was heart-lung transplantation (HLT). Currently,HLT is performed at only a few transplantcenters and should be reserved for patients whocannot be treated by LT alone. The most frequentindications for HLT are Eisenmenger syndromewith a surgically uncorrectable cardiac anomalyor severe end-stage lung disease with concurrentsevere heart disease.Bilateral LTBilateral LT (BLT) is performed in patients withsuppurative pulmonary lung disease (ie, cysticfibrosis [CF] and bronchiectasis). In fact, 28% of allBLTs are performed as treatment for CF. Initially,double-lung transplantation was the procedureof choice with the anastomosis performed at thelevel of the trachea; however, the rate of ischemicairway complications was prohibitive. Now, BLT(essentially, sequential single-lung transplantation[SLT]), in which the anastomoses are performedat the level of the mainstem bronchi, isthe preferred surgical technique. Many centersalso recommend BLT for younger patients withsevere COPD (secondary to tobacco use [24% ofBLTs] or 1-antitrypsin deficiency [9% of BLTs])because of a longer posttransplant life expectancyfrom the increased reserve provided during timesof graft complications. In addition, most centersprefer to perform a BLT procedure in patients with<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 563ACC-PULM-09-0302-035.indd 5637/10/09 8:57:19 PM


primary pulmonary hypertension (PPH; 6% ofBLTs). Although a single-lung allograft with normalpulmonary vasculature can accommodate theentire right ventricular output without elevationof pulmonary artery pressures, in times of graftcompromise, such as rejection or infection, severeventilation-perfusion abnormalities can develop.Overall, BLT procedures are now being performedmore than SLT procedures.SLTSLT is performed for the treatment of obstructivenonsuppurative lung disease such as emphysemasecondary to tobacco use or 1-antitrypsindeficiency (50% and 7%, respectively, of all SLTsare performed for these indications). SLT wasinitially thought to be a poor choice of procedurefor the treatment of patients with COPD as theresult of concerns of preferential ventilation tothe compliant native lung, but these concernshave proven to be largely unfounded. Otherindications for SLT include idiopathic pulmonaryfibrosis (28% of all SLTs), familial pulmonaryfibrosis, drug-induced or toxin-induced lungdisease, occupational lung disease, sarcoidosis,limited scleroderma, lymphangioleiomyomatosis,eosinophilic granuloma, and other disordersresulting in end-stage fibrotic lung disease. SLTis rarely performed for pulmonary vascular disease( 1% of SLTs ); the immediate postoperativeperiod can be difficult because of the largevolume of blood flow going to the transplantedlung. The theoretical advantages of SLT includereduced rate of surgical morbidity, shortenedhospitalization and, often, the avoidance ofcardiopulmonary bypass. This procedure alsoresults in the optimization of the use of donororgans, which are in critical shortage.Guidelines for Recipient SelectionThere has been a revision of the originalconsensus-based guidelines for the selection of LTcandidates. 2 The new guidelines are also consensusbased but have taken into consideration a greaterbody of evidence and transplant experience sincethe earlier version in 1998 and have several importantmodifications compared with the originalguidelines (Table 1).Table 1. Guidelines for Recipient Selection for LungTransplantation*CriteriaGeneral selectioncriteriaAgeRelativecontraindicationsAbsolutecontraindications*Modified from Orens et al.DescriptionUntreatable end-stage obstructiveor restrictive pulmonary parenchymalor pulmonary vasculardisease;No other significant medicaldiseases;Substantial limitation of dailyactivity;Limited life expectancy;Ambulatory with rehabilitationpotential; andSatisfactory psychosocial profile andemotional support system 65 yrCritical or unstable medicalcondition;Systemic or untreated extrapulmonicdisease;Colonization with resistant bacteria,fungus or atypical mycobacteria;Symptomatic osteoporosis;Severely limited functional statusMechanical ventilation;Ideal body weight 70% or 130%(body mass index 30 kg/m 2 );Extrapulmonic disease (ie, renal[creatinine clearance, 50 mg/mL/min]);HIV infection;Malignancy within prior 2 yr;Hepatitis B antigen positivity;Hepatitis C biopsy-proven liverdisease;Severe musculoskeletal disease;Substance addiction in prior 6 mo(including tobacco use);Absence of a reliable support system;Untreatable psychosocial problems;and medical noncomplianceAny patient with end-stage pulmonary orcardiopulmonary disease with the capacity forrehabilitation can be considered for transplantation.The patient should have untreatable end-stagepulmonary disease, no other significant medicalillness, and a limited life expectancy. The candidateshould be ambulatory with rehabilitation potential.The patient must be psychologically stable, committedto the idea of transplantation, and willingto comply with the rigorous medical protocols andregimens required for a successful LT outcome.564 Lung Transplantation (Levine)ACC-PULM-09-0302-035.indd 5647/10/09 8:57:19 PM


AgeThe 2006 international guidelines for the selectionof transplant candidates 2 now suggest an agelimit of 65 years as, generally, but not firmly, theupper limit regardless of procedure type. Althoughthis limit is somewhat arbitrary, numerous patientswith end-stage pulmonary disease are young tomiddle aged, and there is a relative lack of availabledonors. In addition, older patients have a somewhatpoorer outcome than younger patients.Relative ContraindicationsSystemic or Multisystem Disease: Transplantationis not contraindicated in patients with systemicdiseases that are limited to the lungs such as scleroderma,systemic lupus erythematosus, polymyositis,and rheumatoid arthritis. These cases shouldbe considered on an individual basis. Patients withdiabetes mellitus, hypertension, or peptic ulcerdisease should be evaluated carefully before beingconsidered as candidates for LT, and they shouldonly be accepted if their disease is well controlledand there is no resulting end-organ damage.Patients with active sites of infection are notconsidered to be good transplant candidates.Treated tuberculosis and fungal disease pose aparticular problem but are not contraindicationsfor LT. Many centers will not consider performinga transplant in a patient with chronic colonizationwith a resistant organism (eg, Burkholderia cepacia,methicillin-resistant Staphylococcus species,atypical mycobacterium, or Aspergillus species).Centers should try to eradicate these organisms inthe pretransplant period and consider each patienton an individual basis. However, if considered,these patients should be candidates only for BLTprocedures because the remaining colonized lungcould pose a serious threat to the new graft in thecase of an SLT.Osteoporosis: Osteoporosis has become a significantproblem in the posttransplant period, andpreexisting symptomatic osteoporosis has beenidentified as a relative contraindication to transplantation.Bone densitometry should be part ofthe pretransplant evaluation, and treatment shouldbe initiated in those patients with evidence of osteoporosisthat is symptomatic or in those who areasymptomatic.Mechanical Ventilation: A requirement for invasivemechanical ventilation is a strong relativecontraindication to transplantation, althoughLT has been performed successfully in small numbersof CF patients receiving mechanical ventilation.However, patients who are receiving noninvasiveventilatory support can be considered fortransplantation.Nutritional Status: To be considered for transplantation,patients should have a body weight of 70% or 130% of predicted ideal body weight.Those patients with poor nutritional status may betoo weak to withstand the surgical procedure; thosepatients who are obese make more difficult surgicalcandidates and may have greater mortality ratesthan nonobese patients. 3Corticosteroids: Initial data implicated corticosteroidsas a cause of tracheal bronchial dehiscence.At most centers, patients were required to havecompletely discontinued therapy with corticosteroids.This requirement certainly eliminated alarge number of patients with chronic obstructivelung disease and pulmonary fibrosis. Subsequently,pretransplant low-dose therapy with corticosteroidshas been proven to be acceptable for patientswho cannot discontinue corticosteroid therapycompletely. Currently, transplant programs willconsider patients who can be maintained in the longterm on a regimen of prednisone of 20 mg/d andmay consider patients who are receiving greaterdoses.Previous Thoracic Surgery: Previous thoracotomyor pleurodesis was once considered to be a relativecontraindication to patients undergoing transplantationbecause of increased technical difficultiesand increased bleeding. Despite these factors,transplantation can be successfully performed inthese patients.Absolute ContraindicationsThe 2006 international guidelines 2 identifiedseveral absolute contraindications to LT, includingmajor organ dysfunction (ie, renal creatinineclearance of 50 mg/mL/min), HIV infection,hepatitis B-antigen positivity, and hepatitis C withbiopsy-documented liver disease. Active malignancywithin the previous 2 years is also a contraindicationto transplantation. For patients with ahistory of breast cancer greater than stage 2, colon<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 565ACC-PULM-09-0302-035.indd 5657/10/09 8:57:20 PM


cancer greater than Duke A stage, renal carcinoma,or melanoma greater than or equal to level 2, thewaiting period should be at least 5 years. Restagingis suggested before transplant listing.Severe Musculoskeletal Disease: Severe nonosteoporoticskeletal disease, such as kyphoscoliosis, isoften an absolute contraindication to transplantation,primarily because of the technical difficultiesencountered during surgery.Substance Abuse: Drug abuse and alcoholism areconsidered to be contraindications to transplantationbecause patients with these conditions are athigh risk for noncompliance. Patients who continueto smoke despite having end-stage pulmonary diseaseare not candidates for LT. Transplant centersrequire patients to abstain from cigarette smoking,alcohol use, or narcotics use for 6 months to 2 yearsbefore being considered for LT evaluation.Psychological Criteria: The patient must be wellmotivated and emotionally stable to withstandthe extreme stress of the pretransplant and perioperativeperiod. A history of noncompliance orsignificant psychiatric illness is an absolute contraindication,although many patients will presentwith reactive depression or anxiety in the terminalphase of their pulmonary illness.Disease-Specific Guidelines for LTOne of the most difficult decisions whenreferring a patient for transplantation is definingthe appropriate time for transplantation (alsocalled the transplant window). The potential candidateshould be sick enough to have a limitedlife expectancy but not so disabled that the individualwill be unable to withstand the procedure.In 2006, revised guidelines were established todefine the transplant window for specific diseases(Table 2). 2COPDThe variability in the natural course of COPDmakes it especially difficult to predict whenpatients should be referred for LT. On the basisof the available data from several large studiesin which authors determined the predictors ofmortality in COPD, guidelines have been established.The 2006 revised guidelines recommendthe inclusion of the BODE index (an index derivedTable 2. Disease-Specific Criteria for Lung Transplantation*DiseaseCOPDIPFCFPPHCriteriaBODE index 7 or at least one of the following:FEV 1 20% predicted (nonreversible) anddiffusing capacity 20% or homogenousemphysema on HRCT;Hospitalization with Paco 2 50 mm Hg;Cor pulmonale or pulmonary hypertension; andO 2-dependent hypercapnic patients (refer early)BODE index 5 (refer to lung transplant center)Diagnosis of usual interstitial pneumonitis withany of the following:Diffusing capacity 39% predicted;A decrease in FVC of 10% in 6 mo of followup;A decrease in pulse oximetry below 88% on a6-min walk test;Honeycombing seen on HRCT;Histologic or radiographic evidence of UIPirrespective of vital capacity (refer to lungtransplant center);FEV 1of 30% predicted (refer);Clinical deterioration with FEV 1of 30%predicted (exacerbation, hemoptysis, andpneumothorax) [refer];Paco 2of 50 mm Hg;Pao 2on room air of 55 mm Hg;<strong>Pulmonary</strong> hypertension; andYoung, female patients (refer early)NYHA class III–IV despite vasodilator treatment;Cardiac index of 2 L/min/m 2 ;Right atrial pressure of 15 mm Hg; andLow ( 350 m) or declining 6-min walk testdistance*IPF idiopathic pulmonary fibrosis; UIP usual interstitialpneumonitis. Adapted from Orens et al.from measurements of body mass index, degree ofairway obstruction, dyspnea score, and exercisecapacity) when evaluating patients with COPD.The guidelines suggest that a BODE index scoreof 7 should be among the criteria for transplantationin patients with COPD, and a BODE indexof 5 as appropriate for transplant referral. Theguidelines also suggest that patients with COPDare in the transplant window if the FEV 1is 20% ofpredicted without reversibility with bronchodilatoradministration, the diffusing capacity is 20% orthere is homogenous distribution of emphysemaon a high-resolution CT (HRCT) scan, the Paco 2is 50 mm Hg, and/or cor pulmonale or pulmonaryhypertension are present. Those patients who arehypercapnic and hypoxemic and require oxygensupplementation should be given preference.566 Lung Transplantation (Levine)ACC-PULM-09-0302-035.indd 5667/10/09 8:57:20 PM


A 1998 analysis 4 of transplant survival datahas shown that patients with COPD may achievea significant improvement in quality of life (QOL),but a clear survival benefit with transplantationhas not been documented. Rehabilitation and longtermsupplemental oxygen therapy, if the criteriaare met, should be initiated while the patient isawaiting LT.Idiopathic <strong>Pulmonary</strong> FibrosisOn the basis of studies showing poor survivaltimes in patients with idiopathic pulmonary fibrosisand the fact that most patients have a progressivedownhill course despite receiving standardimmunosuppressive therapy, patients with pulmonaryfibrosis should be referred for transplantationearly. In fact, this is the group of patientswho have the greatest mortality rate while on thetransplant list. The 2006 revised guidelines havesuggested that these patients should be referredto a transplant center at the time of a diagnosis ofusual interstitial pneumonitis and should undergotransplantation when diffusing capacity is 39%of predicted, when there is a decrease in FVC of 10% in the 6 months of follow-up, when there isa decrease in pulse oximetric saturation to 88%on a 6-min walk test, or when honeycombing isseen on a HRCT scan.CFAlthough there has been an improvement inlife expectancy for patients with CF, most patientsstill die from respiratory failure and cor pulmonale.The international guidelines have suggested thatthe following criteria be used to define the transplantwindow for patients with CF: FEV 1 30%of predicted or an FEV 130% of predicted withprogressive deterioration, such as an increasingnumber of hospitalizations, a rapid deteriorationin FEV 1, cachexia, and/or massive hemoptysis;Pao 2 55 mm Hg on room air; Paco 250 mm Hg;or pulmonary hypertension. Female patients andpatients 18 years old have a more progressivecourse and should be considered for LT earlier. In2002, a predictive model 5 for selecting patients withCF was proposed to more accurately predict thesurvival effect of LT in these patients. The modelincludes the following: age, sex, percentage ofpredicted FEV 1, weight for age Z score, the presenceof pancreatic sufficiency, diabetes mellitus,infection with Staphylococcus aureus or B cepacia,and the number of acute exacerbations requiringtreatment. The use of the FEV 1criteria and of amore complicated multivariate logistic model todefine transplant referral for CF have been evaluatedin a large group of CF patients. 6 The studyfound that these referral criteria had high negativepredictive values (98% and 97%, respectively) butonly modest positive predictive values (33% and28%, respectively) and can result in prematuretransplant referral.Recently, a controversial article on the impactof LT on survival in the pediatric CF populationwas published. The authors used data from the USCystic Fibrosis Foundation registry and the OrganProcurement and Transplantation Network. Theauthors examined 514 children wait-listed for LTduring a 10-year time period. A total of 248 patientsunderwent LT. Risk factors for survival on thewaiting list and survival after LT were studied. Theauthors found that B cepacia was a risk factor fordeath both before and after LT, diabetes was a riskfactor for death before LT but not after LT, older ageaffected post LT survival but not wait-list survival,and S aureus infection increased wait-list survivalbut decreased post LT survival. The authors concludedthat of the 514 patients listed, 5 patientshad a survival benefit, 76 had insignificant benefit,118 patients had insignificant harm, and 315 had asignificant risk of harm with LT.This study included a large number of patientsbut was retrospective. In addition, QOL improvementswere not addressed. Significant controversyhas arisen over this article, including criticisms ofthe statistical methods used, resulting in subsequentpublished corrections by the authors, butwithout significant changes in the overall conclusionsreached. 7Patients with CF often have multiple, resistantorganisms that are defined as being resistant to allagents in two of the following classes: -lactams,aminoglycosides, and/or quinolones. Panresistantcolonization includes organisms, particularlyPseudomonas aeruginosa and B cepacia, that are resistantin vitro to all groups of antibiotics. As for theoutcomes of CF patients with panresistant P aeruginosaand B cepacia vs those patients with organismssusceptible to therapy, including the number<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 567ACC-PULM-09-0302-035.indd 5677/10/09 8:57:20 PM


of postoperative ventilator days, the length ofhospital stay, and the number of days receivingantibiotic therapy, the incidence of bronchitis andpneumonia, and 1-year survival rates appear to besimilar between groups.However, a subanalysis of those patients withB cepacia infection revealed that they had a lower1-year survival rate (50%) compared with thosewith resistant P aeruginosa infection (90%). Anotherstudy 8 found a 1-year survival rate of 67% for Bcepacia-positive CF patients, compared with 97%in B cepacia-negative CF patients. Some newer datahave suggested that those patients with colonizationwith B cepacia of the genomovar III type havea particularly poor prognosis. Thus, colonizationwith B cepacia of the genomovar III type shouldbe considered a strong relative contraindicationfor transplantation. A recent study examined theresults of LT for CF in patients with panresistantorganisms other than B cepacia (eg, P aeruginosa, Stenotrophomonasmaltophilia, Achromobacter xyloxoxidans).The authors 9 found that there was decreasedsurvival in patients with panresistant bacteriacompared with patient with pansensitive bacteria(58.3% vs 85.6% at 5 years, respectively), but thesesurvival rates were similar or better than thosepredicted by United Network for Organ Sharing.PPHOn the basis of the extrapolation of data fromthe National Heart, Blood, and Lung TransplantRegistry for PPH, selection guidelines have beenestablished for patients with PPH. These guidelinessuggest that patients who are in New York HeartAssociation (NYHA) functional class III or IV, despitereceiving optimal therapy including vasodilatorssuch as prostacyclin, and those with a depressedcardiac index ( 2 L/min/m 2 ), a right atrial pressureof 15 mm Hg, or low or declining 6-min walk testdistances should be considered for transplantation.Other patients should be observed closely and reassessedfor transplantation at 6-month intervals.<strong>Pulmonary</strong> Hypertension Secondary toCongenital Heart Disease (EisenmengerSyndrome)Patients with this type of secondary pulmonaryhypertension have been found to have asignificantly better prognosis than those patientswith primary disease. Therefore, because of theindividuality of the progression of this disease,precise criteria for transplantation have notbeen established. In general, patients are usuallyreferred to transplant centers when they are inNYHA functional class III or IV.Pretransplant EvaluationOnce a patient is deemed to be a potential candidatefor transplantation, several studies are usuallyperformed for further assessment. Typically, thesetests include pulmonary function tests, includinglung volumes, spirometry, and diffusing capacity,and a measure of exercise performance such as a6-min walk test. Cardiac evaluation includes an ECGand an echocardiogram, in addition to a functionalcardiac study, such as dobutamine echocardiographyand/or coronary angiography, in patients40 years of age or those with significant risk factorsfor coronary artery disease. An HRCT scan is oftenobtained to look for bronchiectasis, which couldindicate the necessity for a bilateral procedure, orto look for focal nodules that were not apparent onplain chest radiographs. Renal and liver functionsare assessed by 24-h creatinine clearance and liverfunction tests, respectively. Serologies for hepatitisand HIV should also be obtained. The majority ofthese investigations can be performed at the referringcenter. In those patients being evaluated forSLT, a ventilation/perfusion scan with quantitationmay suggest the preferred side to be transplanted ifthere is unequal distribution of perfusion.While awaiting transplantation, patientsshould have close follow-up by the transplant centeror their referring physician for both physiologicand emotional support. In those patients who arein the pretransplant phase for the treatment ofsuppurative lung disease, sputum cultures shouldbe obtained at 3-month intervals to assist in theantibiotic regimen in the immediate postoperativeperiod. Those patients who are able to should beinvolved in a rehabilitation program before undergoingtransplantation.Donor AllocationUntil the spring of 2005, as established by TheUnited Network for Organ Sharing, lungs were568 Lung Transplantation (Levine)ACC-PULM-09-0302-035.indd 5687/10/09 8:57:21 PM


allocated primarily by time on the waiting list andnot by necessity. In the spring of 2005, the systemfor donor allocation for lungs was revised, and theassigned priority for lung offers became based onthe lung allocation score (LAS), which is calculatedusing the following measures: (1) waitinglist urgency measure (ie, the expected number ofdays lived without a transplant during an additionalyear on the waiting list); (2) posttransplantsurvival measure (ie, the expected number of dayslived during the first year after transplant); and(3) the transplant benefit measure (ie, the posttransplantsurvival measure minus waiting listurgency measure).It is too early to determine the effects that thisnew allocation system will have on LT, but preliminarydata after the first year of the new allocationsystem suggest that patients with pulmonaryfibrosis have moved significantly higher (ie, shorterwait time) on the waiting list in comparison withtheir ranking when the previous time accrualallocation system was used. The distribution ofpatients undergoing LT changed after the institutionof the LAS, with more idiopathic pulmonaryfibrosis and CF patients undergoing transplantationand fewer COPD patients. The mean LASincreased significantly after the institution of theLAS. Despite this, survival rates were comparablebefore and after the implementation of the LAS.Recent data suggest that the early results of theLAS appear to indicate that the LAS is achievingmany of its goals. The long-term effects of the LASremain to be determined, and further refinementof the system will be necessary.Organs are first distributed locally, then regionally,and finally nationally. Currently, the averagetime spent on the waiting list is approximately 18to 24 months; therefore, close management of thelisted transplant patient is required. Despite thisclose attention, a significant percentage of patientsdie while awaiting transplantation. Non—heartbeatingdonors are now also being used in isolatedcases for LT and may prove to partially alleviatethe shortage of donor organs.Living Donor TransplantationSeveral institutions worldwide are now performingliving donor transplantations. The primaryindication for living donor transplantation isCF. Generally; two blood group-compatible livingdonors each provide a lower lobe to the recipient.Ideally, the donors should be larger than the recipientso that the donor lobes fill the hemithorax. Thisprocedure is not performed at the majority of LTcenters because of certain technical and ethicalissues involved.Donor CriteriaDonor organ shortage remains the greatestlimiting factor in the numbers of LTs performed.The majority of potential lung donors are braindead and, as stated previously, only a small numberof living donor-related LT procedures havebeen performed. The usual donor selection criteriainclude age 60 to 65 years, no history of significantlung disease, and a limited smoking history.In addition, potential donors should have clearlung fields on chest radiographs and adequate gasexchange as assessed by a Pao 2of 300 mm Hg ona fraction of inspired oxygen (Fio 2) equal to 1 anda positive end-expiratory pressure of 5 cm H 2Oor a Pao 2/Fio 2ratio of 250 to 300 mm Hg. A normalsputum Gram stain finding and/or endobronchialinspection are also part of the donor evaluationexamination. Marginal or extended donors (ie,those not meeting all of the aforementioned criteria)are being used more frequently to expand thedonor pool. 10,11 By instituting a protocol, includingeducational and donor management interventions,and changing donor classification and selection criteria,a single organ-procurement organization wasable to increase the percentage of lungs procuredper donor from 11.5 to 22.5% and the number ofprocedures performed without adverse recipientoutcomes. 10Donors are excluded from potential lung donationif they have evidence of active infection, HIVinfection, hepatitis infection, and/or malignancy.Donor and recipient compatibility are assessed bymatching by A, B, and O blood types and chest wallsize. Human leukocyte antigen (HLA) matching isnot routinely performed in LT.Surgical ProcedureFor an SLT, the recipient surgery is performedvia a posterolateral thoracotomy incision orsternotomy. The recipient left atrium and donor<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 569ACC-PULM-09-0302-035.indd 5697/10/09 8:57:21 PM


pulmonary vein undergo anastomosis first, followedby the bronchial anastomosis, and finallythe arterial anastomosis. BLT is usually performedthrough a transverse thoracosternotomy (clamshell incision or median sternotomy). Cardiopulmonarybypass may be required in cases of pulmonaryhypertension. Organ preservation remains amajor area of research in LT. Currently, the lung canonly be preserved for a period of approximately 4to 6 h without experiencing significant ischemia/reperfusion injury. Newer preservative agents mayincrease this time period and permit a larger areafor donor allocation.Management After the PostoperativePeriodPatients are typically discharged from the hospitalwithin 7 to 14 days after surgery. Follow-upis in the outpatient clinic on a weekly, biweekly,and finally monthly basis. After this time, patientsoften return home for follow-up with the referringpulmonologist. Weekly monitoring includes measuringthe levels of immunosuppressive medicationssuch as cyclosporine or tacrolimus; a CBCcount to monitor leukocyte concentrations, plateletcounts because of therapy with azathioprine ormycophenolate mofetil; blood chemistry measurementsto follow creatinine levels because of theuse of therapy with cyclosporine or tacrolimus; achest radiograph; routine spirometry and exerciseoximetry; or a 6-min walk test. In addition, patientsare given home spirometers and are instructed tobring in their home spirometry measurements ateach visit.Some institutions perform surveillance bronchoscopyon a routine schedule to detect asymptomaticrejection or infection, whereas otherinstitutions reserve this procedure for clinical deteriorationwithout compromised survival or ratesof bronchiolitis obliterans syndrome (BOS). 12,13 Thechest radiograph has not been shown to be specificfor the early detection of rejection or infection.Close monitoring of pulmonary function hasalso been studied as a way of detecting graft complications.<strong>Pulmonary</strong> function tests have beenshown to have very low specificity and good sensitivity(85%) for the detection of graft complicationsbut, again, are unable to distinguish rejectionfrom infection. Therefore, pulmonary function testsremain the most sensitive test in the late postoperativeperiod for detecting infection or rejection,although they cannot differentiate between theseand other possible complications.OutcomeSurvivalActuarial survival rates after transplantationare reported at 92%, 78%, 62%, and 50%,respectively, for 1 month, and 1, 3, and 5 years, asreported to the ISHLT Registry, 1 and are significantlylower than those reported for other types ofsolid-organ transplant recipients. Early mortality(ie, at 30 days) is most often to the result of primarygraft failure, mortality between 30 days and1 year is most often to the result of infection, andlate mortality (ie, at 1 year) is most often relatedto rejection. Previously, there was no apparentsurvival difference between those patients undergoingSLT and BLT. However, it now appears thatBLT recipients may have a survival advantagebeginning at 3 years after LT. Patients undergoingHLT have lower survival rates. Survival rates aregreater for patients undergoing LT for CF or COPDin comparison with those patients undergoingtransplantation for idiopathic pulmonary fibrosis;older patients have a poorer long-term survivalrate. Younger COPD patients (ie, 50 to 60 yearsof age) undergoing BLT appear to have a survivaladvantage over those undergoing SLT.Physiologic ResultsThe degree of improvement in lung functionpostoperatively is the product of many factors.In the uncomplicated transplant recipients, onecan expect a gradual improvement and ultimateplateau of lung function by 3 to 6 months afterthe procedure. In patients receiving a single lung,the FEV 1can be expected to improve to 50 to 70%of predicted. SLT for the treatment of nonsepticobstructive lung disease results in moderate residualobstructive pulmonary dysfunction seen inspirometry findings. SLT in patients with underlyingrestrictive lung disease can be expected to havemild residual restrictive physiology. These resultslikely reflect the physiology of the remainingobstructed or restricted lung. After SLT procedures,570 Lung Transplantation (Levine)ACC-PULM-09-0302-035.indd 5707/10/09 8:57:21 PM


the majority of the ventilation and perfusion goesto the transplanted lung. BLT procedures usuallyresult in normal spirometry findings with anequal division of ventilation and perfusion. Thosepatients who undergo transplantation for PPHhave no significant change in pulmonary function,marked improvement in gas exchange, andnear-normal hemodynamics after transplantation,including a nearly complete recovery of right ventricularfunction.The majority of those patients who undergoeither LT procedure without complications areable to performed activities of daily living withoutcompromise, although formal cardiopulmonaryexercise testing generally reveals a reduction inmaximum oxygen consumption to 40 to 60% ofpredicted in all transplant groups. The reasonsfor this remain unclear and do not appear to becardiac or pulmonary in origin. The leading physiologichypothesis is that the immunosuppressiveagents may have an effect on peripheral skeletalmuscles, resulting in impaired peripheral oxygenutilization.QOLQOL issues are a relatively recent area ofresearch in LT. Several small studies have shownimprovement in overall health-related QOL. Thelarge majority of patients have expressed satisfactionwith their transplant decision. Even if thesurvival benefit is in question, the improvement inQOL may be worth the sacrifice to many patients.Of interest, 40% of patients return to work on apart-time or full-time basis after transplantation.ComplicationsFigure 1 shows some of the common posttransplantcomplications and the approximate timeperiod in which they usually occur.Primary Graft DysfunctionThe most significant early postoperative complicationafter LT is the development of primarygraft dysfunction (PGD), pulmonary reimplantationresponse (PRR), or primary graft failure. It isestimated that up to 80% of patients will experiencesome degree of reimplantation injury, and in15% of patients it can be severe. A 2005 consensusconference 14 attempted to standardize the gradingof PGD based on gas exchange and the presenceof radiographic infiltrates. When using the acutelung injury definition of ARDS as a Pao 2/Fio 2ratio of 200, the authors found that PGD has areported incidence of 11 to 25%. PGD is characterizedclinically by the presence of new radiographicinfiltrates, a decrease in pulmonary compliance,an increase in pulmonary vascular resistance, anddisrupted gas exchange.Radiographic findings include patchy alveolarconsolidation and/or dense perihilar haze andlower lobe alveolar consolidation. Pathology frombiopsy specimens, autopsy specimens, or lungexplants removed during retransplantation revealsdiffuse alveolar damage. The typical course ofPGD includes progressive worsening or stabilizationover the subsequent hours up to 2 to 4 days,followed by resolution. However, PGD can persistfor days after the surgery. It is thought that thelikely mechanism for PGD is ischemia/reperfusioninjury. Management includes the differentiationof other perioperative complications such asvolume overload, rejection, infection, and venousanastomotic problems. The latter can be evaluatedwith a transesophageal echocardiogram. Treatmentincludes supportive care and therapy with diuretics.Reports 15,16 have documented the successful useof protective ventilatory strategies, nitric oxide,and/or extracorporeal membrane oxygenationin these patients. The prophylactic administrationof inhaled nitric oxide was not found to beof benefit for hemodynamics, reperfusion injury,oxygenation, time to extubation, ICU or hospitalstay, or 30-day mortality rate in one randomized,placebo-controlled trial 16 of 84 patients.A large single center study 17 examined patientswith primary graft failure. The investigators foundan incidence of 15% for this complication, and itwas associated with a prolonged hospital course,prolonged mechanical ventilation, a poor 1-yearsurvival rate (40% vs 69%, respectively), and compromisedfunction among survivors. The authorsfound no clear risk factors for primary graft failure,including age, sex, underlying disease, pulmonaryartery pressure, type of transplant procedureperformed, ischemic times, or use of cardiopulmonarybypass. The study did note that inductionimmunotherapy was used less frequently in those<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 571ACC-PULM-09-0302-035.indd 5717/10/09 8:57:21 PM


COMPLICATIONS FOLLOWING LUNG TRANSPLANTATIONBacterial(line sepsis, pneumonia,tracheobronchitis)INFECTIOUSBacterial pneumonia and/or tracheobronchitis (increasedincidence with treatment for chronic rejection)CytomegalovirusFungi (Aspergillus,Candida)Opportunistic Infections(Pneumocystis carinii)PrimaryGraftDysfunctionNONINFECTIOUSAirwaydehiscence/stenosis Disease recurrenceAcute RejectionChronic RejectionPosttransplant Lymphoproliferative Disorder (PTLD)Transplant1stwk2ndwk1stmo3rdmo6thmo9thmo12thmo18thmoTIME POSTTRANSPLANTATIONFigure 1. The common complications after LT and the typical time periods in which they develop. Reprinted with permission from Melo J, Levine SM. Lung transplantation for thereferring pulmonologist. Pulm Crit Care Update 1999; 13:lesson 16.ACC-PULM-09-0302-035.indd 5727/10/09 8:57:22 PM


patients in whom primary graft failure developed.A study 18 from a different large transplant centeridentified cardiopulmonary bypass as a risk factorfor PRR and also observed prolonged mechanicalventilation and ICU stays in those patients withPRR. It did not observe a difference in survivalrates between patients with and without PRR.Another study 19 found the incidence of severePGD to be 11.8% among 255 transplant recipients.Risk factors associated with the developmentof PGD when subjected to multivariate analysisincluded the following: a recipient diagnosis ofPPH, donor of female sex, donor of African-Americanrace, and donor age 21 and 45 years. 19 Thetechnique used for preservation may also prove tobe a risk factor for PGD. A French study 20 concludedthat the development of PGD, even in the mildestform, is associated with a prolonged duration ofmechanical ventilation in addition to increased ICUmorbidity and mortality. The long-term outcomesin survivors, such as pulmonary function andincidence of bronchiolitis obliterans, vary amongstudies, with some reporting compromise andothers reporting no long-term adverse effects. 21Hemorrhage and phrenic nerve paralysis have alsobeen reported in the early postoperative period.Airway ComplicationsAirway problems were significant causes ofmorbidity and mortality after early attempts at LTand developed in 20 to 50% of transplant recipients.Airway complications can be classified into theearly and late time periods. Early airway complicationsusually develop during the first 4 to 8 weeks.They present as a partial or complete anastomoticdehiscence and/or fungal anastomotic infection(usually Aspergillus or Candida species) or bacterialanastomotic infection (usually Staphylococcusor Pseudomonas species) anastomotic infectionand can subsequently result in anastomotic strictures.Bronchomalacia can also develop.The theoretical causes of airway complicationsinclude ischemia at the site of the anastomosis,infection, poor organ preservation, and/or rejection.Because revascularization of the bronchialcirculation is generally not performed after LT,anastomotic techniques such as the omental wrapor pericardial fat have been developed, in which anintact portion of the omentum or pericardial fat iswrapped around the bronchial anastomosis; or thetelescoping anastomotic technique, in which onebronchus overlaps one to two cartilaginous ringswithin the larger bronchus to improve blood flow.These techniques have resulted in a decrease in theincidence of anastomotic complications of 10 to20%, with low rates of mortality and morbidity.Clinically, patients with bronchial anastomoticcomplications, including stenosis and bronchomalacia,can present with cough, shortness of breath,wheezing, dyspnea on exertion, and worseningobstruction on pulmonary function testing. Thecharacteristic flow-volume loop demonstrates aconcave appearance in both the inspiratory andexpiratory loops. Bronchial strictures or stenosesmay also be visualized on chest radiographs, CTscans, and/or bronchoscopy. Partial or completebronchial dehiscence can also present with mediastinalemphysema seen on chest radiographs orwith air adjacent to the bronchial anastomosis seenon CT scans.The therapeutic options for anastomotic complicationsinclude balloon dilation of a stricture,stent placement, laser therapy and, rarely, surgery.If anastomotic infection with fungal or bacterialorganisms is suspected bronchoscopically or isdiagnosed by endobronchial washing, brushingor biopsy culture, or histology, then therapy withappropriate antibiotics should be instituted.RejectionGraft rejection is categorized clinically accordingto the time of onset after transplantationand the histopathologic pattern. The three typesof rejection are hyperacute, acute, or chronic.Hyperacute rejection is mediated by preexistingalloantibodies that immediately bind to the donorvascular epithelium, leading to vessel thrombosisvia complement activation. Fortunately, thiscomplication is a rare one in LT and will not bediscussed further.Acute Rejection: Acute rejection can developin up to 50% of patients in the first postoperativemonth, and as many as 90% of patients will have atleast one episode of acute rejection within the firstyear. The typical time period for acute rejection is10 to 90 days after LT. It is not uncommon (20% ofpatients) for a single patient to experience two tothree episodes within the first few months after<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 573ACC-PULM-09-0302-035.indd 5737/10/09 8:57:22 PM


transplantation that are either recurrent (ie, morethan two episodes) or persistent (failure to resolvewith standard therapy). Acute rejection is usuallynot seen as frequently after the first year aftertransplantation. Risk factors for acute rejection arepoorly defined, but HLA mismatches and gastroesophagealreflux may have a correlation.Clinically, patients with acute rejection presentwith cough, shortness of breath, malaise, and fever.Occasionally, the presentation is asymptomatic,and some centers advocate surveillance bronchoscopyto detect this, although outcome data arenot available. A physical examination may revealrales or wheezing. The utility of the chest radiographdepends on the time after transplantation.Typically, in the first month, the chest radiographfindings can be abnormal in up to 75% of rejectionepisodes; however, in those episodes of rejectionoccurring after 1 month after transplantation only25% of patients have abnormal radiograph findings.The most common radiographic patternsnoted with acute rejection include a perihilarflare, and alveolar, or interstitial, localized, or diffuse,infiltrates with or without associated pleuraleffusion.Physiologic findings during periods of acuterejection include hypoxemia and deterioration inpulmonary function. <strong>Pulmonary</strong> function abnormalitiesare characterized by a decrease in FEV 1of at least 10 to 15% from baseline as well as adecrease in the forced expiratory flow, midexpiratoryphase of an expired vital capacity of 10 to15%. Once again, these changes are nonspecificand can also be found in patients with infectiousetiologies.Because clinical criteria alone cannot differentiateamong acute rejection, infection and,less common, graft complications, transbronchialbiopsy with BAL fluid has emerged as the primaryprocedure for diagnosis. A diagnosis of acute rejectionby transbronchial biopsy has ranged from 61to 94%, and specificity has ranged from 90 to 100%.A histologic grading system for acute pulmonaryrejection was initially proposed in 1990 and revisedin 1996 and 2007. 22 Pathologically, acute rejectionis characterized by perivascular, mononuclear, andlymphocytic infiltrates with or without airwayinflammation and is graded histologically fromA0 to A4, based on the degree of perivascularinflammation. In addition, airway involvementwith lymphocytic bronchitis or bronchiolitis candevelop and is graded from B0, B1R (low grade),B2R (high grade), and BX (ungradable). As rejectionprogresses, the perivascular lymphocytic infiltratessurrounding the venules and arterioles becomedense and extend into the perivascular and peribronchiolaralveolar septa. In patients with severerejection, the alveolar space may be involved, andparenchymal necrosis, hyaline membranes, andnecrotizing vasculitis have been described. 22Once acute rejection has been diagnosed,treatment consists of the augmentation of immunosuppression.Methylprednisolone, 10 to 15mg/kg/d IV for 3 days, followed by an increasein the maintenance prednisone regimen to 0.5 to1 mg/kg/d with a taper over the next severalweeks is a standard treatment regimen. Maintenanceimmunosuppression therapy should alsobe augmented. Typically, the resolution of symptomsoccurs in days, and histologic follow-up in3 to 4 weeks should show resolution. Recurrentor persistent acute rejection may initiate conversionin the baseline immunosuppression regimen.Lympholytic therapy, methotrexate, photopheresis,total lymphoid irradiation, and/or aerosolizedcyclosporine therapy have been used with variablesuccess. Therapy with mycophenolate mofetil initiatedde novo has resulted in a decreased incidenceof acute rejection in several small studies.Obliterative Bronchiolitis: Chronic rejection hasbeen equated with the histologic finding of obliterativebronchiolitis (OB) and remains a major causeof morbidity and mortality after LT and the leadingsingle cause of death after the first posttransplantyear. The current incidence of OB ranges from 35to 50% among different centers. OB has been definedclinically by an obstructive functional defectand histologically by the obliteration of terminalbronchioles. The mean time to diagnose OB is 16to 20 months after the LT procedure but has beenreported as early as 3 months after transplantation.Some degree of OB will develop in 50% of transplantrecipients by 5 years after transplant.The etiology and risk factors for OB remainunclear. Several possible causes have been proposed,including uncontrolled acute rejection;lymphocytic bronchiolitis; cytomegalovirus (CMV)pneumonitis; CMV infection without pneumonitis;HLA-A mismatches; total HLA mismatches; theabsence of donor antigen-specific hyporeactivity;574 Lung Transplantation (Levine)ACC-PULM-09-0302-035.indd 5747/10/09 8:57:22 PM


non-CMV infection; older donor age; and bronchiolitisobliterans with organizing pneumonia.The most consistently identified risk factor isacute rejection, particularly in those patients whoexperience recurrent, high-grade episodes of acuterejection. It is probable that lymphocytic bronchiolitisand CMV pneumonitis are also importantrisk factors for OB. 23,24 Other risk factors are beingidentified. In the past several years, gastric aspirationand gastroesophageal reflux disease (GERD)have been recognized as an important cause and/or contributing risk factor to the development ofpulmonary dysfunction after LT, particularly BOS.Primary graft dysfunction may also be a risk factorfor BOS. Clinically, OB can manifest as an upperrespiratory tract infection and can be mistakenlytreated as such. Other patients present withoutclinical symptoms but with gradual obstructivedysfunction seen on pulmonary function testresults. FEV 1has been the standard spirometricparameter used, but midexpiratory flow rates maybe a more sensitive parameter for early detection.Typically, chest radiographs are not helpful inthe diagnosis of OB because most patients haveradiographic findings that are unchanged fromtheir baseline posttransplant radiographs. HRCTscans may reveal peripheral bronchiectasis, patchyconsolidation, decreased peripheral vascular markings,trapped air, and bronchial dilation, whichmay aid in the diagnosis of OB. Trapped air onend-expiratory HRCT scans has been shown tobe a sensitive (91%) and accurate (86%) radiologicindicator of OB but may not be able to provide anearly diagnosis of this disorder. 25 BAL fluid neutrophilia,exhaled nitric oxide levels, bronchial hyperresponsiveness,and increased nitrogen washoutmay also be surrogate markers of BOS.As with acute rejection, transbronchial biopsyis used to diagnose OB, although primarily toexclude other diagnoses. The classic pathologicfinding is constrictive bronchiolitis. Unfortunately,the sensitivity for the diagnosis of OB by transbronchialbiopsy is low (range, 15 to 87%), andthe diagnosis of OB is often made by exclusionand is graded physiologically based on the degreeof change in pulmonary function (ie, FEV 1) frombaseline. Because of the variability in obtainingbronchioles by transbronchial biopsy, the ISHLThas established a BOS staging system. 26 Thisstaging system is based on a reduction in FEV 1incomparison with a posttransplant baseline FEV 1with or without the pathologic documentation ofOB. This staging system was revised by the ISHLTin 2002, and the new system includes an earlierBOS category of potential BOS using changes inFEV 1and/or mid-flows compared with posttransplantbaseline values (FEV 1, 81 to 90% of baselinevalues; and/or forced expiratory flow, midexpiratoryphase, 75% of baseline values; Table 3). 24Once OB has been diagnosed histologically orclinically by the exclusion of alternate diagnoses,treatment is begun with high-dose methylprednisolonetherapy followed by a tapering course oforal corticosteroids. Lympholytic agents such asantilymphocyte globulin, monoclonal antibodyto the CD 3lymphocyte receptor (OKT 3), or antiinterleukin(IL)-2 agents such as daclizumab canbe considered for therapy if there is no clinicalresponse to steroid treatment. Therapy may stabilizepulmonary function but uncommonly resultsin significant improvement. Other common physiologicscenarios may be relentless deteriorationin pulmonary function despite therapy, gradualsteady deterioration in pulmonary function, ora series of decreases in function with plateaus. Arecent study found that the majority of patients hadthe steepest decrease in the percentage of predictedFEV 1within the first 6 months after BOS diagnosis.The authors noted a steeper decrease in the first 6months after the onset of BOS in patients with adiagnosis before LT of pulmonary fibrosis and inpatients who were female. Patients with a rapidonset of BOS also had a steeper decrease in FEV 1percentage of predicted. Recipients of an SLT hada worse percentage of predicted FEV 1after diagnosis.After 24 months, the course of the percentageTable 3. Clinical Staging System for OB*Stage † FEV 10 (no OB) 90% of baseline and FEF 25–75%of 75%of baseline0-p (potential OB) 81–90% of baseline and/or FEF 25–75%of 75% of baseline1 (mild OB) 66–80% of baseline2 (moderate OB) 51–65% of baseline3 (severe OB) 50% of baseline*Adapted from Estenne et al. FEF forced expiratory flow.†Each stage is subdivided into a and b, where a is withouthistologic documentation of OB and b is with histologic documentationof OB. Adapted from the ISHLT staging system.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 575ACC-PULM-09-0302-035.indd 5757/10/09 8:57:22 PM


of predicted FEV 1in BLT recipients was better thanthat in SLT recipients. 27Alternate immunosuppressive agents such asmycophenolate mofetil, tacrolimus, and sirolimushave also been associated with the stabilization ofpulmonary function when used as rescue treatmentfor BOS. Methotrexate, total lympholytic radiation,aerosolized cyclosporine, photopheresis, andnewer immunosuppressive agents have been usedfor treatment in refractory cases of OB. Inhaledcorticosteroids may be added to therapy in casesof lymphocytic bronchiolitis. Small studies 28 haveshown benefit from the use of low-dose azithromycin(via an antiinflammatory mechanism) for treatingpatients with BOS. Therapy directed towardGERD (both pharmacological and surgical) shouldalso be considered.Infection, including bronchiectasis, frequentlycomplicates intensive immunosuppression therapyfor OB and may result in death. Pseudomonas is acommon offender, and aerosolized aminoglycosideantibiotics or suppressive quinolone treatment maybe considered. Bronchomalacia can also develop.Survival rates after the diagnosis of OB have beenreported at 74%, 50%, and 43%, respectively, at 1,3, and 5 years in one small series. 29 Because mostpatients with OB can only be stabilized, strategiesdirected at prevention, early diagnosis, and treatmentare necessary for the preservation of lungfunction. Retransplantation has been performedwith variable results. Survival is somewhat lessthan that for patients receiving de novo transplants.In addition, the limited donor supply does notallow for the common practice of this procedure.Humoral Rejection: Recently, humoral rejectionhas been described in LT recipients as a componentof either acute or chronic rejection. Humoral rejectionis suggested when high levels of donor-specificHLA antibodies are present in the serum of therecipient, and complement deposition is presentin the tissue. Preliminary studies are examiningthe use of plasmapheresis, rituximab, and IV Ig inthis setting.Infectious ComplicationsInfections have been a major cause of earlyand late morbidity and mortality after transplantation,and they remain the leading single specificcause of death in the first posttransplant year.The incidence of infection is significantly greaterthan that reported in other solid-organ recipientsand may be related to continuous environmentalexposure of the allograft. Other predisposing factorsinclude diminished cough reflex secondary todenervation, poor lymphatic drainage, decreasedmucociliary clearance, recipient-harbored infectionand, occasionally, the transfer of infection from thedonor organ. Nosocomial infections at the site ofthe surgical wound, vascular access, and urinarytract or ventilator-associated pneumonias alsooccur in the early postoperative period. In mostcircumstances, the allograft or graft is the primarylocation of infection.Bacterial Infections: Bacterial pneumonia is themost common life-threatening infection to developin the early postoperative period and had oncebeen reported to have an incidence as high as 35%in the first 2 postoperative weeks. Common organismsinclude P aeruginosa and Staphylococcusspecies. The incidence of perioperative bacterialpneumonia has been decreased to as low as 10% bybroad-spectrum antibiotic prophylaxis, usually anantipseudomonal cephalosporin and clindamycin,and routine culture of the trachea of the donor andthe recipient at the time of the surgery. Prophylactictherapy with antibiotics is usually discontinuedat 3 days if the culture results are negative and aretailored to the cultured organisms if the results arepositive. Patients with bronchiectasis who receivetransplants usually receive postoperative bacterialprophylaxis for 7 to 10 days. The prevalence ofbacterial pneumonia remains high during the first6 months after transplantation and decreases subsequently,although a second late peak in incidenceoften occurs during the augmentation of immunosuppressionfor the treatment of chronic rejection.Bacterial infection caused by Staphylococcus or,less commonly, Pseudomonas species can developat or distal to the site of the anastomosis in the earlyposttransplant period. Of interest, the incidence oflower respiratory tract infections does not appearto be increased in patients who receive transplantsfor CF in comparison with patients who receivetransplants for other indications.It is often difficult to distinguish pneumoniafrom other early graft complications such asreperfusion injury, pulmonary edema, rejection,and other infectious etiologies. In addition, differentiationbetween colonization and invasion may576 Lung Transplantation (Levine)ACC-PULM-09-0302-035.indd 5767/10/09 8:57:22 PM


e difficult and often requires invasive proceduressuch as bronchoscopy with BAL fluid, quantitativesterile brush sample, and/or transbronchialbiopsy.Other Early Infections: Atypical pneumoniassuch as Legionella, Mycobacteria, and Nocardiaspecies are uncommon in patients undergoingLT but have been reported to occur in 2 to 9% ofrecipients. The numbers of viral infections, suchas those attributable to herpes simplex virus, havebeen significantly reduced with the common useof acyclovir or ganciclovir prophylaxis. Candidalinfections may be seen in the early postoperativeperiod but usually do not cause invasive disease.Trimethoprim sulfamethoxazole prophylaxis(usually on a three times per week schedule) hassignificantly reduced the incidence of Pneumocystispneumonia to 1% in LT recipients.Viral Infections (CMV): In the period 1 to 6months after transplantation, CMV accounts forthe majority of the viral infections in LT recipients.The typical time period for the development ofCMV infection is 30 to 150 days postoperatively,with an incidence of illness (ie, infection and disease)of approximately 50%. Risk factors for CMVdisease are dependent on the serology of the donorand recipient as well as the use of high-intensityimmunosuppressive therapy, including cytolytictherapy. Approximately 25 to 35% of the time CMVdisease develops in CMV-positive recipients fromeither positive or negative donors, whereas CMVnegativerecipients have an 85% chance of thedisease developing when implanted with a CMVpositivelung. There are data showing that CMVpneumonitis may contribute to the developmentof chronic rejection. 24CMV can cause a wide spectrum of diseasefrom asymptomatic infection (ie, shedding of thevirus in the urine or BAL fluid) to widespread dissemination.Various presentations of CMV in LTpatients can include pneumonitis and, most commonly,gastroenteritis, hepatitis, and colitis. CMVpneumonitis can often be confused with acuterejection. Clinical findings of CMV pneumonitisinclude fever, cough, hypoxemia, the presence ofan interstitial or alveolar infiltrate, and leukopenia.The diagnosis of invasive disease requires cytologicor histologic changes in cell preparation or tissue.Therefore, flexible bronchoscopy with transbronchialbiopsy and BAL fluid is often necessary andcan diagnose CMV pneumonia in 60 to 90% ofpatients.The diagnosis of CMV infection may occasionallybe made on the basis of positive cultures andthe compatible clinical setting after other causesof pulmonary disease have been excluded bronchoscopically.The pathologic hallmark of CMVinfection is a cytomegalic 250-nm cell containinga large central basophilic intranuclear inclusion.This inclusion is referred to as an owl’s eye becauseit is separated from the nuclear membrane bya halo. These inclusions may be well seen withhematoxylin-eosin or Papanicolaou staining. Theidentification of CMV cytologically is very specific(98%) but lacks sensitivity (21%) for the presenceof infection. Other pathologic findings in the lungparenchyma include a lymphocytic and mononuclearcell interstitial pneumonitis.Ganciclovir is currently the mainstay oftherapy for invasive CMV disease. Initial dosesof 5 mg/kg bid for 3 to 4 weeks have been shownto reduce mortality from 60 to 80% to 15 to 20%in patients with symptomatic CMV pneumonitis.Some centers also use CMV-specific hyper-Ig in thetreatment of CMV disease.Prophylaxis against CMV infections hasbecome a major strategy in most transplant centers.Initially, some centers attempted to match CMVnegativerecipients with CMV-negative donorswhen possible; however, the limited donor supplydid not allow the continuation of this practice. Theuse of CMV-negative blood products is advocated.Ganciclovir prophylaxis seems to be effective indelaying the onset of CMV infection. At some centers,prophylaxis is administered only to the CMVmismatchedpatients (ie, a CMV-negative recipientreceiving an organ from a CMV-positive donor).Other centers administer prophylaxis to all but theCMV-negative-to-CMV-negative patient group.Prophylactic treatment approaches usually include2 to 4 weeks of therapy with ganciclovir (5 mg/kgqd), although some centers continue prophylaxisfor up to 90 days, particularly in the donor-positiveand recipient-negative subset of patients. In thegroups with the greatest risk, CMV hyper-Ig maybe added to the regimen. The oral form of ganciclovirmay be of potential use in the LT population,but the data are inconclusive. Preemptive strategiessuch as the initiation of treatment when a highlevel of CMV blood antigenemia or polymerase<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 577ACC-PULM-09-0302-035.indd 5777/10/09 8:57:23 PM


chain reaction copies is detected may also delayand decrease the severity of CMV infection, andmay become the standard of care.Other viral infections that have been describedin LT patients include herpes simplex virus (earlyafter LT) respiratory syncytial virus, other paramyxoviruses(such as parainfluenza), influenza,metapneumovirus, and adenovirus. Acyclovirprophylaxis for herpes infection is initiated in mostprograms after the discontinuation of therapy withganciclovir.Fungal Infections: Fungal infections are morecommon in the LT recipient than in those who havereceived other solid-organ transplants. The overallincidence of invasive fungal infection in LT rangesfrom 10 to 22% and usually develops in the first fewmonths after transplantation. Fungal infections accountfor the most significant morbidity and mortalityof all infectious agents after transplantation, andthe mortality rate can range from 40 to 70%.Aspergillus species including Aspergillusfumigatus, Aspergillus flavus, Aspergillus terreus, andAspergillus niger can be colonizing organisms or canresult in infection presenting as indolent, progressivepneumonia or as an acute fulminant infectionthat rapidly disseminates. Aspergillus speciesexhibits the propensity to invade blood vessels andmay present as an infarct or with hemoptysis. Theradiographic findings of pulmonary aspergillosisinclude focal, lower-lobe infiltrates, patchy bronchopneumonicinfiltrates, single or multiple noduleswith or without cavitation, thin wall cavities,and opacification of the entire lung graft. A HRCTscan may reveal a halo sign thought to be pathognomonicfor invasive aspergillosis. Other forms ofpresentation of Aspergillus infection can includepseudomembranous tracheobronchitis, often at anddistal to the site of the anastomosis. The diagnosisof invasive aspergillosis requires the identificationof organisms within tissues. These organisms canappear as septated hyphae that branch at acuteangles, and can be detected with hematoxylin-eosinand methenamine silver staining.Survival with Aspergillus infection has beenimproved with the early initiation of advancedgeneration azoles such as itraconazole and voriconazoleand/or high-dose amphotericin therapyand a reduction in immunosuppressive therapy.Surgical resection may occasionally be required tomaximize cure rates in patients with aspergillosis.Therapy with a lipid formulation of amphotericinB should also be considered in the managementof invasive fungal infections in patients who areintolerant or in whom nephrotoxicity developswith conventional amphotericin B and in patientswith progressive fungal infection despite therapywith conventional amphotericin. Prophylaxis withamphotericin B, azoles (particularly itraconazolefor 3 to 6 months), or aerosolized amphotericinhas shown promise in decreasing the incidence ofAspergillus infection after LT.Candida species can cause a variety of syndromesin the patient undergoing LT, includingmucocutaneous disease, line sepsis, woundinfection and, rarely, pulmonary involvement.Fluconazole has emerged as an effective alternativefor infections caused by Candida albicans, butamphotericin B is still the agent of choice for awidespread disease. Fluconazole appears to beless active against other Candida species such asCandida glabrata and Candida krusei.Other rare causes of fungal infection in LTrecipients have included Cryptococcus neoformans,as well as the dimorphic fungi such as coccidioides,histoplasma, and blastomyces. Azoles andamphotericin B remains the antifungal agents ofchoice for these infections, although azole agentscan be used for maintenance therapy.Mycobacteria: Mycobacterial diseases both typicaland atypical have been successfully treated inthe LT patient.Vaccinations: Standard vaccination regimensshould be used in patients in the LT population.Posttransplant Lymphoproliferative DisordersPosttransplant lymphoproliferative disorders(PTLDs) are reported more frequently after LTthan after other types of solid-organ transplantation.Lymphomas comprise a majority (22%) ofPTLDs. The PTLDs are composed of a heterogeneousgroup of lymphoid proliferations of variableclonality. The B-cell non-Hodgkin lymphomas arethe most frequent form of PTLD and have beenassociated with Epstein-Barr virus (EBV) activityeither serologically or by the identification of viralDNA in tissue. There is no clear correlation amongepisodes of rejection, specific immunosuppressivedrugs, and the development of PTLD. The incidenceof PTLD after LT has been reported to be578 Lung Transplantation (Levine)ACC-PULM-09-0302-035.indd 5787/10/09 8:57:23 PM


1.8% to 9.4%. Those patients who have negativeEBV serology before transplantation and receivean organ from an EBV-positive donor resultingin seroconversion are at a significantly greaterrisk for the development of a PTLD. Childrenare at greater risk because of their frequent EBVnegativestatus.The clinical features of PTLD in LT recipientsinclude development in the first posttransplantyear, involvement of the allograft, and radiographicfindings of solitary or multiple pulmonary nodules.Disseminated disease has also been reported. Treatmentincludes the anti-CD20 monoclonal antibodyrituximab, a reduction in immunosuppression, aswell as consideration for adjunctive treatment withantiviral therapy, radiation, chemotherapy, surgery,and adoptive immunotherapies extrapolated fromthe bone marrow transplant population. Mortalitymay be significant with high-grade PTLD. Anincreased incidence of other nonlymphomatousmalignancies has been reported in the LT population,and careful screening is recommended.Miscellaneous ComplicationsThe recurrence of primary disease has beendescribed in patients who have received transplantsfor the treatment of sarcoidosis, lymphangiomyomatosis,giant-cell interstitial pneumonitis,desquamative interstitial pneumonia, idiopathicpulmonary hemosiderosis, bronchioalveolar carcinoma,eosinophilic granuloma, alveolar proteinosis,and diffuse panbronchiolitis. However, diseaserecurrence may be an incidental finding in some ofthese patients, particularly those with sarcoidosis.Pleural effusions may develop because of thefact that the lymphatics are not reanastomosedafter transplant and/or in the setting of rejection.Prolonged postoperative air leaks have been welldescribed. Gastroparesis and an increased incidenceof GI emergencies have also been describedin LT recipients. GERD, both symptomatic andasymptomatic, is now known to be a significantproblem after LT. This complication may be morecommon in those patient populations predisposedto the development of GERD, such as patients withCF. Osteoporosis remains a significant problemin the posttransplant period and is best managedwith bisphosphonate agents such as alendronateor pamidronate. 30,31 Deep venous thrombosisand pulmonary embolism were reported to havean incidence of 8.6% in one series of LT recipients.32 Hypogammaglobulinemia has also beendescribed.ImmunosuppressionBecause transplantation of a lung graft into arecipient invokes a specific immune response, theneed to suppress the recipient’s immune systemexists immediately. Rejection results from theactivation, differentiation, and proliferation of Tlymphocytes acting against donor cells that arerecognized as foreign, the major determinants ofwhich are cell-surface markers known as HLAs.Current immunosuppressive regimens and strategiesare based on the inhibition of this whole-cellresponse to foreign antigens.Currently, most transplant centers use maintenanceimmunosuppression regimens includingcyclosporine or tacrolimus, azathioprine or mycophenolatemofetil, and prednisone. In addition, inthe first 2 weeks after LT, immunosuppression maybe enhanced by the use of cytolytic agents such asOKT 3(a monoclonal antibody to the OKT 3), antithymocyteglobulin, or new human-derived IL-2receptor antibodies such as daclizumab.A typical maintenance immunosuppressiveregimen consists of cyclosporine, 5 mg/kg bid (withdose adjusted to serum levels of 250 to 350 ng/mL);azathioprine, 1 to 2 mg/kg/d (adjusted to maintaina leukocyte count of 4,000 to 4,500 cells/µL); andprednisone, approximately 0.5 mg/kg/d for thefirst 3 months tapered over the next 3 months to15 mg/d and to 5 mg/d by the 12th posttransplantmonth. Some programs have completely discontinuedthe use of prednisone after 1 year aftertransplantation. Tacrolimus may be used in lieu ofcyclosporine at a dose of approximately 0.1 mg/kgbid (adjusted to a serum level of 8 to 15 ng/mL);and mycophenolate mofetil, 1 to 3 g/d, may be usedin lieu of azathioprine.All immunosuppressive drugs are wroughtwith complications, which can pose a significantproblem after transplantation (Table 4). Cyclosporineand tacrolimus act by inhibiting IL-2 genetranscription, thus decreasing T-lymphocyte activationand subsequent proliferation. Chronic nephrotoxicityis a common complication of both agentsand can develop in 25 to 75% of patients receiving<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 579ACC-PULM-09-0302-035.indd 5797/10/09 8:57:23 PM


Table 4. Immunosuppressive Drugs Used in Lung TransplantationDrugCyclosporineTacrolimusAzathioprineMycophenolate mofetilPrednisoneOKT 3Antithymocyte globulinSirolimusCommon Adverse EffectsNephrotoxicity; hypertension; neurotoxicity (tremor, seizures, and headache); hyperlipidemia;hyperkalemia; hypomagnesemia; GI disturbance; hemolytic uremic syndrome; hirsutism; andgingival hyperplasiaNephrotoxicity; hypertension; neurotoxicity (tremor, seizures, headache); hyperlipidemia;hyperkalemia; hypomagnesemia; GI disturbance; hemolytic uremic syndrome; hyperglycemia;and diabetesLeukopenia, macrocytic anemia, thrombocytopenia, hepatotoxicity, and pancreatitisDiarrhea, abdominal pain, nausea, leukopenia, and anemiaHyperglycemia, hypertension, hyperlipidemia, osteoporosis, myopathy, insomnia, cataracts, andweight gainCytokine release syndrome, hypotension, acute respiratory distress syndrome, and aseptic meningitisSerum sickness, leukopenia, and thrombocytopeniaAnemia, thrombocytopenia, leukopenia, anastomotic dehiscence (early postoperative period),hyperlipidemia, arthralgias, interstitial pneumonitis, and lower extremity edemathese drugs and, to some degree, in nearly 100%of patients with long-term use. Acute renal toxicityis usually related to dose and typically is reversible.In addition, other potentially nephrotoxicagents, including amphotericin B, trimethoprimsulfamethoxazole,nonsteroidal antiinflammatoryagents, and aminoglycoside antibiotics, which maycompound the toxic effects of cyclosporine andtacrolimus, may be used in transplant patients.Both cyclosporine and tacrolimus are also associatedwith systemic hypertension, which candevelop in up to 25 to 50% of LT recipients.The occurrence of hypercholesterolemia hasalso been well described. The incidence of bothposttransplant hypertension and hyperlipidemiamay be lower with the use of tacrolimus than withcyclosporine. Other well-described side effects oftacrolimus and cyclosporine include neurologictoxicity such as tremors, paresthesias, headaches,confusion, depression, somnolence, seizures, whitematter changes, coma, and death. Peripheral neurologicfindings have also been described. Neurotoxicityappears to be significantly more commonwith tacrolimus therapy than with cyclosporinetherapy. Cyclosporine use is also associated withhirsutism, gingival hyperplasia, and gastroparesis.Tacrolimus use is associated with hyperglycemiaand diabetes. Both agents have been reported tocause a hemolytic uremic syndrome or thromboticthrombocytopenic purpura.It is important that physicians who are caringfor transplant patients are aware of the numerousdrug interactions with tacrolimus and cyclosporine.For example, use of the azoles results in a significantincrease in cyclosporine and tacrolimus levelsin the same dose. Likewise, the discontinuation oftherapy with azole agents without increasing thedose of cyclosporine or tacrolimus can result inan acute and life-threatening drop in therapeuticlevels of these drugs. Interactions with macrolideantibiotics, calcium-channel blockers, and gastricmotility drugs have also been reported. Levels ofboth agents are decreased with the use of rifampinor anticonvulsant agents.Azathioprine inhibits nucleic acid synthesisand suppresses the proliferation of lymphocytes.Toxicities of this drug include cytopenias such asleukopenia and thrombocytopenia. Macrocyticanemia can develop with long-term use. Pancreatitisand cholestatic hepatitis have been welldescribed with azathioprine use.Mycophenolate mofetil also inhibits purinesynthesis and blocks lymphocyte proliferation.Typical side effects include GI symptoms such asdiarrhea, nausea, and abdominal pain. In addition,leukopenia and anemia have been described.Corticosteroids play a main role in posttransplantimmunosuppression by binding to cytoplasmicglucocorticoid receptors, undergoingtranslocation into the nucleus, and then blockingcytokine transcription of genes and secretion fromphagocytes. The side effects of corticosteroid useare well known and include hyperglycemia, hypertension,hyperlipidemia, osteoporosis, proximalmyopathy, mood disturbance, cataract formation,and weight gain.580 Lung Transplantation (Levine)ACC-PULM-09-0302-035.indd 5807/10/09 8:57:23 PM


The lympholytic agents such as OKT 3andantithymocyte globulin act by depleting CD 3lymphocytesby lysis, opsonization, and phagocytosis.The administration of OKT 3may be associatedwith the cytokine release syndrome manifested byhypotension, noncardiogenic pulmonary edema,renal insufficiency, and aseptic meningitis. Antithymocyteglobulin is associated with leukopenia,thrombocytopenia, and serum sickness. It shouldnot be administered to those patients who havemanifested an allergic reaction to horse serum inthe past.Sirolimus is one of the newer available immunosuppressantagents. It is being used for casesof acute and chronic LT rejection and, in smallstudies, as de novo therapy after LT. Sirolimus is amacrocyclic triene antibiotic (structurally related totacrolimus) with immunosuppressive, antitumor,and antifungal properties. Sirolimus has been demonstratedto block the proliferative response andthe activation of T cells, B cells, and other cell linesby cytokines and growth factors, thus preventingcell-cycle progression and proliferation. In contrast,tacrolimus (FK-506) and cyclosporine inhibit theproduction of cytokines. The use of sirolimus inthe immediate posttransplant period is discouragedbecause of the association with bronchialanastomotic dehiscence when used in combinationwith tacrolimus and corticosteroids. The major sideeffects of sirolimus include cytopenias (particularlythrombocytopenia and leukopenia), hyperlipidemia,arthralgias, and interstitial pneumonitis,which can become a serious problem in the LTrecipient. Other side effects include those similarto the side effects described with cyclosporine andtacrolimus although, in general, the nephrotoxicitywith this agent appears to be less that that of theother two agents. The role of sirolimus in the LTarmamentarium remains to be established.References1. Christie JD, Edwards LB, Aurora P, et al. The Registryof the International Society for Heart and LungTransplantation: twenty-fifth official adult lung andheart-lung annual report; 2008. J Heart Lung Transplant2008; 27:957–9692. Orens JB, Estenne M, Arcasoy S, et al. Internationalguidelines for the selection of lung transplant candidates:2006 update; a consensus report from the<strong>Pulmonary</strong> Scientific Council of the InternationalSociety for Heart and Lung Transplantation. J HeartLung Transplant 2006; 25:745–7553. Kanasky WF Jr., Anton SD, Rodrigue JR, et al. Impactof body weight on long-term survival after lungtransplantation. Chest 2002; 121:401–4064. Hosenpud JD, Bennett LE, Keck BM, et al. Effect ofdiagnosis on survival benefit of lung transplantationfor end-stage disease. Lancet 1998; 351:24–275. Liou TG, Adler FR, Cahill BC, et al. Survival effectof lung transplantation among patients with cysticfibrosis. JAMA 2002; 286:2683–26896. Mayer-Hamblett N, Rosenfeld M, Emerson J, et al.Developing cystic fibrosis lung transplant referralcriteria using predictors of 2-year mortality. Am JRespir Crit Care Med 2002; 166:1550–15557. Liou TG, Adler FR, Cox DR, Cahill BC. Lung transplantationand survival in children with cysticfibrosis. N Engl J Med 2007; 357:2143–21528. Chaparro C, Maurer J, Gutierrez C, et al. Infectionwith Burkholderia cepacia in cystic fibrosis: outcomefollowing lung transplantation. Am J Respir CritCare Med 2001; 163:43–489. Hadjiliadis D, Steele MP, Chaparro C, et al. Survivalof lung transplant patients with cystic fibrosisharboring panresistant bacteria other than Burkholderiacepacia, compared with patients harboringsensitive bacteria. J Heart Lung Transplant 2007;26:834–83810. Angel LF, Levine DJ, Restrepo MI, et al. Impact of alung transplantation donor-management protocolon lung donation and recipient outcomes. Am JRespir Crit Care Med 2006; 174:710–71611. Bhorade SM, Vigneswaran W, McCabe MA, et al.Liberalization of donor criteria may expand thedonor pool without adverse consequence in lungtransplantation. J Heart Lung Transplant 2000;12:1199–120412. Valentine VG, Taylor DE, Dhillon GS, et al. Success oflung transplantation without surveillance bronchoscopy.J Heart Lung Transplant 2002; 21:319–32613. Levine SM. Transplant/Immunology Network ofthe American College of Chest Physicians. A surveyof clinical practice of lung transplantation in NorthAmerica. Chest 2004; 125:1224–123814. Christie JD, Carby M, Bag R, et al. Report of theISHLT working group on primary lung graftdysfunction: part II. Definition: a consensus statementof the ISHLT. J Heart Lung Transplant; 2005;24:1454–1459<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 581ACC-PULM-09-0302-035.indd 5817/10/09 8:57:24 PM


15. Shargall Y, Guenther G, Ahya VN, et al. Report ofthe ISHLT working group on primary lung graftdysfunction: part VI. Treatment: a consensus statementof the ISHLT. J Heart Lung Transplant; 2005;24:1489–150016. Meade MO, Granton JT, Matte-Martyn A, et al.A randomized trial of inhaled nitric oxide to preventischemia-reperfusion injury after lung transplantation.Am J Respir Crit Care Med 2003; 167:1483–148917. Christie JD, Bavaria JE, Pavlevsky HI, et al. Primarygraft failure following lung transplantation. Chest1998; 114:51–6018. Khan S, Salloum J, O’Donovan PB, et al. Acutepulmonary edema after lung transplantation: thepulmonary reimplantation response. Chest 1999;116:187–19419. Christie JD, Kotloff RM, Pochettino A, et al. Clinicalrisk factors for primary graft failure following lungtransplantation. Chest 2003; 124:1232–124120. Thabut G, Vinatier I, Stern JB, et al. Primary graftfailure following lung transplantation: predictivefactors of mortality. Chest 2002; 121:1876–188221. Arcasoy SM, Fisher A, Hachem RR, et al. Report ofthe ISHLT working group on primary lung graftdysfunction: part V. Predictors and outcomes: aconsensus statement of the ISHLT. J Heart LungTransplant; 2005; 24:1483–148822. Stewart S, Fishbein MC, Snell GI, et al. Revision ofthe 1996 working formulation for the standardizationof nomenclature in the diagnosis of lung rejection.J Heart Lung Transplant 2007; 26:1229–124223. Sharples LD, McNeil K, Stewart S, et al. Riskfactors for bronchiolitis obliterans: a systematicreview of recent publications. J Heart Lung Transplant2002; 21:271–28124. Estenne M, Maurer JR, Boehler A, et al. Bronchiolitisobliterans syndrome 2001: an update of thediagnostic criteria. J Heart Lung Transplant 2002;21:297–31025. Leung AN, Fisher K, Valentine V, et al. Bronchiolitisobliterans after lung transplantation. Chest 1998;113:365–37026. Cooper JD, Billingham M, Egan T, et al. A workingformulation for the standardization of nomenclatureand for clinical staging of chronic dysfunction inlung allografts: International Society for Heart andLung Transplantation. J Heart Lung Transplant 1993;12:713–71627. Lama VN, Murray S, Lonigro RJ, et al. Course ofFEV 1after onset of bronchiolitis obliterans syndromein lung transplant recipients. Am J Respir Crit CareMed 2007; 175:1192–119828. Yates B, Murphy DM, Forrest IA, et al. Azithromycinreverses airflow obstruction in established bronchiolitisobliterans syndrome. Am J Respir Crit CareMed 2005; 172:772–77529. Valentine VG, Robbins RC, Berry GJ. Actuarialsurvival of heart-lung and bilateral sequential lungtransplant recipients with obliterative bronchiolitis.J Heart Lung Transplant 1996; 15:371–38330. Spira A, Gutierrez C. Chaparro C, et al. osteoporosisand lung transplantation: a prospective study. Chest2000; 117:476–48131. Aris RM, Lester GE, Renner JB, et al. Efficacy ofpamidronate for osteoporosis in patients with cysticfibrosis following lung transplantation. Am J RespirCrit Care Med 2000; 162:941–94632. Izbicki G, Bairey O, Shitrit D, et al. Increased thromboembolicevents after lung transplantation. Chest2006; 129:412–416Annotated BibliographyArcasoy SM, Fisher A, Hachem RR, et al. Report of theISHLT working group on primary lung graft dysfunction:part V. Predictors and outcomes: a consensus statementof the ISHLT. J Heart Lung Transplant; 2005;24:1483–1488One article in a series of articles focusing on PGD.Arcasoy SM, Kotloff RM. Lung transplantation. N EnglJ Med 1999; 340:1081–1091This is a concise review of the field of LT.Belperio JA, Weigt SS, Fishbein MC, et al. Chronic lungallograft rejection: mechanisms and therapy. Proc AmThorac Soc 2009; 6:108–121This issue of this journal is devoted to LT.Blondeau K, Mertens V, Vanaudenaerde BA, et al. Gastrooesophagealreflux and gastric aspiration in lung transplantpatients with or without chronic rejection. EurRespir J 2008; 31:707–713This study further examines the question of GERD andchronic rejection.Boehler A, Estenne M. Post-transplant bronchiolitisobliterans. Eur Respir J 2003; 22:1007–1018A review article discussing pathogenesis, risk factors, clinicalpresentation, diagnosis, and treatment of bronchiolitis obliterans.Briffa N, Morris RE. New immunosuppressive regimensin lung transplantation. Eur Respir J 1997; 10:2630–2634582 Lung Transplantation (Levine)ACC-PULM-09-0302-035.indd 5827/10/09 8:57:24 PM


This article gives an excellent overview of the mechanisms ofaction of the immunosuppressive agents commonly used in LT.Chaparro C, Maurer J, Gutierrez C, et al. Infection withBurkholderia cepacia in cystic fibrosis: outcome followinglung transplantation. Am J Respir Crit Care Med 2001;163:43–48An article examining survival of LT patient with CF andresistant bacteria.Christie JD, Carby M, Bag R, et al. Report of the ISHLTworking group on primary lung graft dysfunction:part II. Definition: a consensus statement of the ISHLT.J Heart Lung Transplant 2005; 24:1454–1459One article in a series of articles focusing on PGD.Christie JD, Edwards LB, Aurora P, et al. The Registry ofthe International Society for Heart and Lung Transplantation:twenty-fifth official adult lung and heart-lungannual report; 2008. J Heart Lung Transplant 2008;27:957–969This is the annual report from the ISHLT. It presents data onindications, numbers, survival, rates of morbidity and mortality,and some complications after LT.Cooper JD, Billingham M, Egan T, et al. A working formulationfor the standardization of nomenclature andfor clinical staging of chronic dysfunction in lungallografts: international Society for Heart and LungTransplantation. J Heart Lung Transplant 1993; 12:713–716This is the original article defining staging for the BOS.Corris PA, Christie JD. Update in transplantation 2007.Am J Respir Crit Care Med 2008; 177:1062–1067A succinct update on this subject.de Antonio DG, Marcos R, Laporta R, et al. Results ofclinical lung transplant from uncontrolled non-heartbeatingdonors. J Heart Lung Transplant 2007; 26:529–534This article examines the use of non--heart-beating donors inLT.DeSoyza A, McDowell A, Archer L, et al. Burkholderiacepacia complex genomovars and pulmonary transplantationoutcomes in patients with cystic fibrosis. Lancet2001; 358:1780–1781This article examines transplant outcomes with differentgenomovars of B cepacia.Estenne M, Maurer JR, Boehler A, et al. Bronchiolitisobliterans syndrome 2001: an update of the diagnosticcriteria. J Heart Lung Transplant 2002; 21:297–310This article includes the proposed new staging system forBOS, lists BOS risk factors, reviews the pathology of BOS,reviews surrogate markers of BOS, and discusses the responseto treatment.Gries CJ, Mulligan MS, Edelman JD, et al. Lung allocationscore for lung transplantation: impact on disease severityand survival. Chest 2007; 132:1954–1961This article examines the early impact of the LAS on waitingtimes, death on the waiting list, diseases transplanted, and earlysurvival before and after the implementation of the LAS.Hadjiliadis D, Steele MP, Chaparro C, et al. Survival oflung transplant patients with cystic fibrosis harboringpanresistant bacteria other than Burkholderia cepacia,compared with patients harboring sensitive bacteria.J Heart Lung Transplant 2007; 26:834–838An article examining survival of LT patient with CF andresistant bacteria.Izbicki G, Bairey O, Shitrit D, et al. Increased thromboembolicevents after lung transplantation. Chest 2006;129:412–416Another article on this subject.Lama VN, Murray S, Lonigro RJ, et al. Course of FEV(1)after onset of bronchiolitis obliterans syndrome in lungtransplant recipients. Am J Respir Crit Care Med 2007;175:1192–1198This study examined the progression in the percentage ofpredicted FEV 1after LT and attempted to identify risk factorsfor rapid decline.Levine SM, Angel L, Anzueto A, et al. A low incidence ofpost transplant lymphoproliferative disorder in 109 lungtransplant recipients. Chest 1999; 116:1273–1277This article presents data from a single center on PTLD. Thearticle also reviews data from some of the other published serieson this subject.Levine SM, Peters JI. Fungal infection in the lungtransplant recipient: update. Pulm Crit Care 1998; 12:Lesson 17This is a review article describing the incidence, outcome,and varied clinical presentations of fungal infections in LTrecipients.Levine SM, Transplant/Immunology Network of theAmerican College of Chest Physicians. A surveyof clinical practice of lung transplantation in NorthAmerica. Chest 2004; 125:1224–1238This article presents the results of a large survey sent to all NorthAmerican LT centers regarding the clinical practice of LT.Levy RD, Ernst P, Levine SM, et al. Exercise performanceafter lung transplantation. J Heart Lung Transplant 1996;15:1045–1058This article examines exercise physiology in SLT and HLTgroups.Liou TG, Adler FR, Cahill BC, et al. Survival effect oflung transplantation among patients with cystic fibrosis.JAMA 2002; 286:2683–2689<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 583ACC-PULM-09-0302-035.indd 5837/10/09 8:57:24 PM


This article proposes a detailed predictive model for a survivaleffect when referring CF patients for BLT.Liou TG, Adler FR, Cox DR, et al. Lung transplantationand survival in children with cystic fibrosis. N Engl J Med2007; 357:2143–2152The purpose of this study was to determine the impact of LTon survival in children ( 18 years old) with CF.Martinu T, Chen DF, Palmer SM. Acute rejection andhumoral sensitization in lung transplant recipients. ProcAm Thorac Soc 2009; 6:54–65A review of acute rejection including a discussion on humoralrejection.Meade MO, Granton JT, Matte-Martyn A, et al. A randomizedtrial of inhaled nitric oxide to prevent ischemiareperfusioninjury after lung transplantation. Am J RespirCrit Care Med 2003; 167:1483–1489A negative study of the use of nitric oxide to prevent reperfusioninjury.Mehrad B, Paciocco G, Martinez FJ, et al. Spectrum of aspergillusinfection in lung transplant recipients: case seriesand review of the literature. Chest 2001; 119:169–175This article reviews the various presentations of Aspergillusin the LT population.Meyer DM, Bennett LE, Novick RJ, et al. Single vs bilateral,sequential lung transplantation for end-stageemphysema: influence of recipient age on survival andsecondary end-points. J Heart Lung Transplant 2001;20:935–941This study supports the practice at some centers of performingBLT over SLT for younger patients with COPD as the resultof a survival advantage.Orens JB, Estenne M, Arcasoy S, et al. Internationalguidelines for the selection of lung transplant candidates:2006 update; a consensus report from the <strong>Pulmonary</strong>Scientific Council of the International Society for Heartand Lung Transplantation. J Heart Lung Transplant 2006;25:745–755These are updated consensus-determined guidelines regardingselection of LT candidates.Pochettino A, Kotloff RM, Rosengard BR, et al. Bilateralversus single lung transplantation for chronic obstructivepulmonary disease: intermediate-term results. AnnThorac Surg 2000; 70:1813–1818This study also supports performing BLT in younger patientswith COPD and SLT in older COPD patients.Reams BD, McAdams HP, Howell DN, et al. Posttransplantlymphoproliferative disorder: incidence, presentation,and response to treatment in lung transplantrecipients. Chest 2003; 124:1242–1249Includes information on the treatment of PTLD withrituximab.Shargall Y, Guenther G, Ahya VN, et al. Report of theISHLT working group on primary lung graft dysfunction:part VI: treatment: a consensus statement of theISHLT. J Heart Lung Transplant 2005; 24:1489–1500One article in a series of articles focusing on PGD.Sharples LD, McNeil K, Stewart S, et al. Risk factors forbronchiolitis obliterans: a systematic review of recentpublications. J Heart Lung Transplant 2002; 21:271–281A review of published human studies on the risk factors forOB.Snyder LD, Palmer SM. Immune mechanisms of lungallograft rejection. Semin Respir Crit Care Med 2006;27:534–543A review of the immunology of graft rejection, including adiscussion on humoral rejection.Stewart S, Fishbein MC, Snell GI, et al. Revision of the1996 working formulation for the standardization ofnomenclature in the diagnosis of lung rejection. J HeartLung Transplant 2007; 26:1229–1242This is a revision of the pathology of acute and chronic rejectionin LT recipients.Trulock EP. Lung transplantation. Am J Respir Crit CareMed 1997; 155:789–818This is the most comprehensive review of the field of LT. It isdirected toward the pulmonary and critical care medicineaudience.United Network for Organ Sharing. Available at: http://www.unos.org. Accessed March 29, 2009An important Web site for statistics regarding lung and othersolid-organ transplants.Young LR, Hadjiliadis D, Davis RD, et al. Lung transplantationexacerbates gastroesophageal reflux disease.Chest 2003; 124:1689–1693This article describes the prevalence and possible etiologies ofthis complication.Proc Am Thorac Soc 2009; 6:1–100An entire issue devoted to many aspects of LT.584 Lung Transplantation (Levine)ACC-PULM-09-0302-035.indd 5847/10/09 8:57:24 PM


Rare Interstitial Lung Diseases: <strong>Pulmonary</strong> LangerhansCell Histiocytosis, Lymphangioleiomyomatosis, andCryptogenic Organizing PneumoniaJoseph P. Lynch III, MD, FCCPObjectives:• Describe the salient epidemiologic, clinical, physiologic,and radiographic features of Langerhans cell histiocytosis(LCH), lymphangioleiomyomatosis (LAM), and cryptogenicorganizing pneumonia (COP)• Compare and contrast the salient features seen on highresolutionCT scans in these disorders• <strong>Review</strong> the characteristic histopathologic features of eachof these disorders and the role of immunohistochemicaltechniques or markers (for pulmonary LCH and LAM)• <strong>Review</strong> the differing therapeutic strategies for these disordersKey words: bronchiolitis obliterans organizing pneumonia;cryptogenic organizing pneumonia; cystic lung diseases;Langerhans cell granulomatosis; Langerhans cell histiocytosis;lymphangioleiomyomatosis; pulmonary eosinophilicgranuloma;<strong>Pulmonary</strong> Langerhans cell histiocytosis (LCH;also termed Langerhans cell granulomatosis) andlymphangioleiomyomatosis (LAM) are rare andpoorly understood chronic lung disorders characterizedby extensive cyst formation within thelung parenchyma. These disorders share certainfeatures, but LAM affects exclusively women,whereas pulmonary LCH (PLCH) may affect eithersex. Cryptogenic organizing pneumonia (COP),formerly termed bronchiolitis obliterans organizingpneumonia (BOOP), is a rare disease of unknownetiology characterized by dense alveolar infiltrates,subacute course, and excellent responsiveness tocorticosteroid (CS) therapy. This chapter reviewsthe salient features of these disorders and presentsapproaches to diagnosis and therapy.<strong>Pulmonary</strong> LCH<strong>Pulmonary</strong> LCH is a rare disorder usuallyseen in smokers. It presents as chronic interstitiallung disease (ILD) or with pneumothoraces. 1,2Histologically, the lesions of PLCH are identicalto LCH (formerly termed eosinophilic granuloma orhistiocytosis X), involving bone or extrapulmonaryorgans. 3 However, 20% of adults with PLCHmanifest extrapulmonary involvement. 3-8 Moreimportantly, LCH predominantly affects children,whereas isolated PLCH is a rare occurrence inchildren. 9 The clinical and radiographic features ofPLCH overlap with those of myriad chronic ILDs.However, CT scans are highly characteristic andmay distinguish PLCH from idiopathic pulmonaryfibrosis and other chronic ILDs.EpidemiologyThe prevalence of PLCH is estimated at two tofive cases per million persons. 1,3 PLCH accountsfor 5% of cases of ILDs. 3 PCH typically occursin adults 20 to 50 years of age; children are rarelyaffected. 3,4,8 PLCH is almost exclusively seen inCaucasian patients, suggesting a genetic predisposition.3-5 However, no familial or inheritabletrait has been identified. Multisystemic LCH inchildren or adults is monoclonal, 10 whereas PLCHis polyclonal, suggesting a reactive process. 11 Thelow proliferative rate of Langerhans cell (LCs)in PLCH suggests a nonmalignant process. 6,12The authors of some studies suggest a distinctmale predominance, 5,13 but others cite a slightfemale predominance. 4,6,7 These variations inincidence by sex may reflect differences in smokinghabits of the populations studied. 1 More than90% of cases of PLCH occur in smokers, 3-5,7,14which suggests that constituents of cigarette smokehave a strong role in the pathogenesis.Clinical FeaturesThe clinical features and natural history ofPLCH are variable. Symptoms of cough anddyspnea are present in twothirds of patientsand often develop insidiously over the course of<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 585ACC-PULM-09-0302-036.indd 5857/10/09 10:24:04 PM


several months or years. 1 Pneumothoraces occursin 6 to 20% of patients and may be the presentingfeature. 1,4-6,15-17 Recurrent pneumothoracesmay require surgical pleurodesis for control. 5,17Rupture of subpleural cysts or blebs accountsfor the predilection for recurrent pneumothoraces.Hemoptysis or wheezing occur in 5%of patients. 3 Constitutional symptoms of fever,malaise, weight loss, or anorexia are present in15 to 30% of patients. 3-5 A total of 10 to 25% ofpatients are asymptomatic, with incidental findingson chest radiographs. 1,4-6<strong>Pulmonary</strong> arterial hypertension (PAH) iscommon in severe PLCH 18 and likely representsintrinsic pulmonary vascular disease. In a seriesof 21 patients with advanced PLCH who wereawaiting lung transplantation, all had PAH(mean pulmonary arterial pressure, 59 mm Hg)that was disproportionate to the degree of pulmonaryfunctional impairment or hypoxemia. 18Histopathology demonstrated proliferativevasculopathy involving muscular arteries andveins with prominent venular involvement. Afatal case of veno-occlusive disease complicatingPLCH was described. 19 Extrapulmonary involvementoccurs in 20% of patients with PLCH. 3-7,13Solitary punched-out lesions of bone and diabetesinsipidus (from involvement of the pituitary)are the most common sites of extrapulmonaryinvolvement. 1,3 Blood or serologic studies arenot helpful. In contrast to chronic eosinophilicpneumonia, peripheral blood eosinophilia is nota feature of PLCH. 1The prognosis of PLCH is generally good,but disparate results have been reported in theliterature. The disease stabilizes or improvesspontaneously in approximately two thirds ofpatients, usually within 18 months of onset ofsymptoms. 4-7,15,16 However, the disease progressesin 15 to 30%, with destruction of lung parenchymaand respiratory impairment; fatality rates rangefrom 2 to 27%. 4-7,13,16,20,21The risk of lung cancer appears to be increasedin patients with PLCH. 22-24 In one study of 93patients with PLCH, 5 developed bronchogeniccarcinoma, for an annual risk of 1,040/100,000. 22In contrast, none of 48 patients in a series fromthe National Institutes of Health had lung cancer. 7Because virtually all patients are smokers, cigarettesmoke likely contributes to the cancer risk. Anassociation between lymphoma and PLCH has alsobeen noted. 23,25Radiographic Features of the ChestConventional chest radiographs usually demonstratediffuse reticular, reticulonodular, or cysticlesions with a predilection for the mid- and upperlung zones (Fig 1). 3-5 The costophrenic angles areusually spared. 4,5,15 Both nodular and cystic componentsare usually present concomitantly. Thecystic radiolucencies correspond to walls of cystsor dilated bronchi or bronchioles. Reticular orreticulonodular lesions, typically ranging in sizefrom 1 to 5 mm, represent the walls of cysts (fromdestroyed lung parenchyma) or peribronchiolarcellular lesions. 1 With advanced disease, volumeloss and fibrosis may be extensive. Pneumothoracesoccur in 6 to 20% of patients and may be thepresenting feature (Fig 2). 4-6 Pleural thickening oreffusions are uncommon. 15 Cavitation of nodulesis occasionally observed on CT scans 1,15 but is notseen on plain radiographs. Hilar or mediastinallymphadenopathy is rare. 1,4,26Chest CT CharacteristicsCertain findings on thin-section high-resolutionCT (HRCT) scans are highly suggestive of PLCH,and HRCT is far superior to conventional chestFigure 1. LCH. Posteroanterior chest radiograph demonstratingfinely nodular densities throughout the lung parenchyma,with a predominance in the middle and upper lung fields, in a44-year-old man with pulmonary LCH.586 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 5867/10/09 10:24:05 PM


Figure 2. LCH. Posteroanterior chest radiograph demonstratingfar advanced cystic and nodular changes throughout thelung parenchyma. A persistent right pneumothorax is evident,despite the presence of a right thoracoscopy tube. Surgical clipsare from prior thoracotomies for attempted pleurodesis.Figure 4. LCH. HRCT scan from the patient in Fig 3 reveals numerouscysts of various sizes throughout the lung parenchyma.Multiple small nodules can also be seen dispersed throughoutthe peribronchial regions and alveolar interstitium.Figure 3. LCH. Posteroanterior chest radiograph demonstratingfinely nodular densities throughout the lung parenchyma andhyperinflation in a 38-year-old woman with pulmonary LCH.Figure 5. LCH. HRCT scan of the upper lobes from a 56-year-old man with progressive cough and dyspnea. Note themultiple well-defined cystic spaces and coalescence of blebsin the medial aspect of the right apex. A few tiny, scatteredperibronchiolar nodules are also present. Transbronchial lungbiopsy specimens demonstrated typical histologic features ofLCH. S100 stains were also positive.radiographs in discerning the cystic nature of thedisease (Figs 3–5). 3,27,28 The combination of cysticand nodular lesions, with a proclivity to involvethe upper lobes, strongly suggests PLCH. 1,29 Thereis no central or peripheral predominance. Multiplethin-walled cysts are observed in 85% ofpatients. 15,27,30 These cysts are usually round and 10 mm but may coalesce, occasionally exceeding3 cm in size. 15,27 Nodules ranging in size from1 to 15 mm are a prominent component of PLCH;they are found in approximately two thirds ofpatients. 15,27 The nodules surround small bronchiolesand represent peribronchiolar granulomas. Asthe disease progresses, the nodules are replaced by<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 587ACC-PULM-09-0302-036.indd 5877/10/09 10:24:06 PM


cysts, some of which become confluent. The extentof nodular opacities on HRCT correlates with thedensity of granulomatous inflammatory lesionsin lung tissue; cystic lesions on CT correlate withthe density of cavitary lesions (both inflammatoryand fibrotic) on biopsy. 8 Serial studies of individualpatients indicate that lesions evolve from nodulesto cavitary nodules and then to thick-walled andthin-walled cysts. 1,15,27 These cysts or nodules areassociated with areas of intervening normal lungtissue. 15,27 The cystic radiolucencies resemble LAM,but the nodular component is lacking in LAM. Inaddition, LAM affects all lung zones, includingthe basilar regions. The cysts of PLCH lack thesubpleural, peripheral, and basilar distribution ofthe honeycomb cysts noted in idiopathic pulmonaryfibrosis (Fig 6, 7). 29 Ground-glass opacities(GGOs) have been described in 3 to 20% of patientswith PLCH. 3,14,15 but are rarely dominant. WhenGGOs are prominent, concomitant respiratorybronchiolitis-ILD often is present. 14 Pleural effusionsand hilar adenopathy are rare. 1,15,29 However,mediastinal or paratracheal lymphadenopathymay be detected on CT scans in up to one third ofpatients. 3Figure 6. Idiopathic pulmonary fibrosis. HRCT scan from a55-year-old woman with usual interstitial pneumonia (UIP)(documented by thoracoscopic lung biopsy) demonstratesextensive honeycombing throughout both lungs (upper lobes).Note that the cysts are distributed preferentially in the peripheral(subpleural) regions of the lung, with relative sparing ofthe central regions. This pattern differs from the more uniformdistribution of cysts seen in patients with pulmonary LCH.<strong>Pulmonary</strong> Function Tests<strong>Pulmonary</strong> function tests (PFTs) are abnormalin 80% of patients with PLCH. 1,4-6,13,31 The diffusingcapacity of the lung for carbon monoxide(Dlco) is reduced in 75% of patients. 3-6,13,15,31Severe reductions in Dlco are associated witha worse prognosis and correlate with radiologiccyst formation. Reductions in vital capacity ortotal lung capacity (TLC) are present in 50 to80% of patients. 4-7,13,31 Pure obstructive or mixedobstructive-restrictive patterns are present inone third of patients. 4-6,31 Air trapping (increasedresidual volume) occurs in nearly half of patientswith PLCH, but hyperinflation (TLC 110% ofpredicted) is rare. Exercise tests typically demonstrateworsening gas exchange and increased deadspace. 31 Aberrations in PFTs do not consistentlycorrelate with radiographic findings. However, ina study of 26 cases of PLCH, correlations betweenPFTs and CT (semiquantitative) were noted. 32 Theextent of cystic lesions on CT correlated inverselywith FEV 1/FVC, Pao 2, Dlco, whereas nodularopacities did not. In most patients, PFTs improveFigure 7. Idiopathic pulmonary fibrosis. HRCT scan from thepatient in Fig 6 (lower lobes). Foci of honeycombing are notedthroughout both lungs. Note the peripheral (subpleural) distributionof honeycomb cysts, with relative sparing of the centralregions. In addition, prominent septal lines, bands, or fibrosisare apparent. This pattern differs from the more uniform distributionof cysts seen in patients with pulmonary LCH.or stabilize over time, but this finding is variable.Serial PFTs should be performed to monitor thecourse of the disease.HistopathologyHistologically, PLCH is characterized by inflammatory,cystic, nodular, and fibrotic lesions distributedin a bronchocentric fashion. 4,6,31,33 Vasculitis is nota feature. The diagnosis of PLCH usually requires588 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 5887/10/09 10:24:08 PM


thoracoscopic lung biopsy, but transbronchial biopsiesmay be adequate provided the salient featuresare present. 7,13 Early in the course of the disease, theproliferation of atypical histiocytes (histiocytosis Xcells or LCs) dominates. 4,6,34 Later, granulomatousinflammation ensues. Fibrosis is the end result of thechronic inflammatory cellular lesions. In individualpatients, several histologic features may be presentconcomitantly (eg, histiocytic proliferation, granulomatousinflammation, fibrosis, healing, and repair).The diagnosis of PLCH is suggested on lowpowermicroscopy by numerous peribronchiolarnodules, numerous cysts, and a distinctive stellateor star-shaped pattern of fibrosis. 4,6 The stellate patternof fibrosis was observed in 84 of 100 cases ofPLCH reported by Friedman and colleagues. 4 Thestellate border reflects extension of the cellular infiltrateinto adjacent alveolar interstitium. Numerousdiscrete nodular infiltrates distributed throughoutthe lung, interspersed with large areas of normallung parenchyma, can be seen under low-powermagnification (Fig 8). These nodules are peribronchiolarin 80% of cases and subpleural in 70%. 7Because most patients with LCH are smokers,respiratory bronchiolitis is a frequent concomitantfinding. 3,14 Under high-power light microscopy,atypical histiocytes (LCs or histiocytosis X cells)are prominent (encompassing up to 50% of cells)and are surrounded by variable numbers ofeosinophils, lymphocytes, plasma cells, fibroblasts,and foci of fibrosis. 3,4,6 Aggregates of these LCscomprise the key histologic feature of PLCH andmay be seen within the nodular lesions, airspaces,and alveolar interstitium. 4,6,34 LCs can be identifiedby conventional stains (eg, hematoxylin-eosin) onhigh-power light microscopy by an experiencedpathologist. 34 They are distinctive large, ovoidmononuclear phagocytes with moderate amountsof eosinophilic cytoplasm, a prominently grooved,folded nucleus, inconspicuous nucleoli, and finelydispersed chromatin (Fig 9). 3,6Electron microscopy, which is not necessary forclinical purposes, demonstrates distinctive 42-nmpentilaminar intracytoplasmic inclusions (termedX bodies or Birbeck granules) within LCs. 1,4,5 LCsalso express the CD1a antigen (the common thymocyteantigen) 3 and stain for S100 protein. 34 LCsare distributed in normal lung but rarely constitute 4% of cells. 3,34 In PLCH, the cellular lesions arecomposed of LCs, admixed with other inflammatorycells. Large numbers of macrophages may beprominent in the alveolar spaces and interstitium,simulating desquamative interstitial pneumonia(DIP). 3,5−7 Eosinophils are prominent in some cases,mimicking chronic eosinophilic pneumonia. 5−7However, in some patients, eosinophils are sparseor absent. 3 Early lesions form in terminal andrespiratory bronchioles and invade and destroythe bronchial wall in the process. 1 Cavitation mayrepresent the lumen of the preexisting bronchioleFigure 8. Photomicrograph of a subpleural granulomatousnodule in a patient with pulmonary LCH. The darker stainingareas within the nodule correspond to inflammatory cellularinfiltrates. The process extends into the alveolar interstitium.The pleural surface is at the lower left corner (hematoxylineosin,low-power magnification).Figure 9. Photomicrograph of pulmonary LCH. Scatteredatypical histiocytes (LCs) are present within a cellular nodule.LCs are moderately large cells with pale eosinophilic cytoplasmand an indented, clefted nuclei (hematoxylin-eosin, oil immersion).From Lynch JP III, Raghu G. Major disease syndromesof unknown etiology. In: Baum G, Crapo J, Karlinsky J, et al,eds. Textbook of pulmonary diseases. 7th ed. Philadelphia, PA:Lippincott Williams & Wilkins, 2004; 629–655.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 589ACC-PULM-09-0302-036.indd 5897/10/09 10:24:10 PM


destroyed by the granulomatous process. 1 As theprocess evolves, bronchioles and alveolar walls aredestroyed and replaced by fibrotic connective tissue.Dilated and distorted bronchioles and alveolarparenchymal cysts result. Dense fibrosis (consistingof mature collagen) and intraluminal fibrosis(consisting of young, organizing connective tissuewithin lumens of bronchioles and alveoli) maybe seen. 4,7 Intraluminal buds of connective tissueresemble COP. 7 Thus, the histopathologic featuresof PLCH overlap with those of other chronic ILDs.Identification of large numbers of LCs is the key tothe diagnosis. However, late- or end-stage PLCHmay be relatively acellular, with 5% LCs andextensive alveolar and honeycomb cysts. 1,34 In thiscontext, distinguishing PLCH from other endstagechronic fibrotic lung disorders is difficult,if not impossible. The retention of a nodular orstellate configuration is a clue to the diagnosis. 4,6Because the pattern and distribution of lesions isa key to the diagnosis of PLCH, surgical (open orthoracoscopic) lung biopsies often are required tosubstantiate the diagnosis. 3 Transbronchial lungbiopsies or BAL fluid are diagnostic in 10 to 40%of patients, 3,7 but significant potential for samplingerror exists. 1,35 When bronchoscopic specimens areassessed, ancillary techniques such as S100 or CD1astains are important to substantiate the diagnosis(discussed in the section “Immunologic Techniques” to follow).Immunologic TechniquesImmunohistochemical stains may be invaluablein establishing the diagnosis of PLCH whenlight microscopic features are not definitive. Positivestaining for S100 protein, CD1a antigen, andhuman leukocyte antigen DR are used to identifythe characteristic LCs. 3,7 S100 protein can beassayed in paraffin-embedded biopsy specimens;some monoclonal antibodies recognizing theCD1a antigen require fresh or frozen tissue, butanti-CD1a antibody O10 recognizes the antigenin fixed and paraffin-embedded tissues. 1 Theseimmunohistochemical methods are less expensiveand less time-consuming than electron microscopy,and their use avoids the sampling problems associatedwith electron microscopy.Large aggregates of S100-positive histiocytesin stellate nodules or granulomatous lesions arevirtually pathognomonic of PLCH. 3 Small numbersof S100-positive cells may be seen in normal lungtissue or BAL fluid but rarely exceed 4% of cells. 36Staining is most intense in active PLCH lesionsand diminishes in the fibrotic regions. 36 Similarly,LCs express CD1a antigen (recognized by themonoclonal antibody OKT6), but lymphocytes andmonocytes do not. Rare CD1a-positive histiocytesmay be found in BAL fluid from normal smokersor patients with chronic ILDs (mean, 0.20%). 36 Afinding of 4% of cells expressing CD1a in BALfluid is relatively specific for PLCH, 1,36 but sensitivityis low (ie, 25%). 1 In one study, positiveimmunohistochemical stains to a murine monoclonalantibody (Mab O10) were demonstrated in33 of 34 paraffin-embedded biopsies from patientswith LCH. 37 LCs also express CD4 antigens andboth CD1a and CD1c markers, a family of antigenpresentingmolecules. 38 In addition, LCs in PLCHexpress surface markers associated with activation,such as B7 molecules, that are not present in normalpulmonary LCs. 1PathogenesisThe pathogenesis of PLCH likely reflects anexaggerated immune response to cigarette smokethat is initiated or regulated by LCs. 30,34 LCs arespecialized dendritic cells that express the CD1areceptor 39 and function to regulate pulmonaryresponses to exogenous (inhaled) and endogenous(self) antigens. 40 The striking bronchocentricnature of the inflammatory process suggests areaction to inhaled antigens or irritants. 6,7 Accumulationof LCs is the earliest lesion to occur inPLCH. 1 LCs replicate in the alveolar structuresand play a role in maintaining the alveolitis inPLCH. 6,7 In addition to large numbers of LCswithin the cellular lesions, immune complexesand other inflammatory cells (eg, lymphocytes,plasma cells, eosinophils, mononuclear phagocytes,and foam cells) may be present. Interactionsbetween these cells, in addition to the release ofcytokines, may drive the immune and fibroticresponse molecules. 1 LCs serve as accessory cells,activating T lymphocytes and recruiting otherimmune effector cells to the lung. 1,34 Tobaccosmoke is strongly linked to PLCH. More than90% of patients are smokers, 3-7,13,20 and constituentsof cigarette smoke are known to serve as590 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 5907/10/09 10:24:12 PM


T-cell mitogens, stimulate macrophage cytokineproduction, 41 and induce epithelial cell productionof cytokines (eg, granulocyte-macrophagestimulating factor and transforming growthfactor-β) 1,3,42 that recruit LCs and up-regulate lymphostimulatoryactivity of LCs. 1,34The number of pulmonary LCs is increasedin asymptomatic smokers. 43 Cigarette smoking isassociated with hyperplasia of pulmonary neuroendocrinecells and increased levels of bombesinlikepeptides in the lower respiratory tract. 44 Largenumbers of bombesin-positive neuroendocrinecells have been noted in open lung biopsy specimensin patients with PLCH. 44 These neuroendocrinecells may contribute to the recruitment andactivation of mononuclear phagocytes and LCsto the lung and stimulate growth of fibroblasts. 3Furthermore, cigarette smoke impairs CD10/neutralendopeptidase, an enzyme that plays a role ininactivating bombesin-like peptides. 44 In one study,exposure to cigarette smoke in mice evoked interstitialgranulomatous inflammation, with featuresconsistent with PLCH in humans. 45 After cessationof exposure, the density of pulmonary LCs in micereturned to normal levels. These data support theassumption that tobacco products are the majorrisk factor for PLCH. 46 However, the vast majorityof cigarette smokers do not develop PLCH despitehaving an increased population of LCs, suggestingthat a “second-hit” is necessary for the disease tooccur. 34 T lymphocytes (predominantly CD4 ) areabundant in LCH lesions and associate closely withLC cells, 1 suggesting local production of antigen atthese sites. The source of antigen is unknown butcould be derived from cigarette smoke-damagedairway epithelium. 34Course and PrognosisThe natural history of PLCH is variable,but most patients improve or stabilize within 6to 12 months of the onset of symptoms. 1,2,4-7,15,16However, data are limited to retrospective studies,with variable treatment regimens. In 1978,French investigators reported 78 cases of PLCH. 5Among 67 cases with follow-up data, outcomeswere as follows: improved (13%), stable (40%),worse (21%), and death (25%). The followingfactors were associated with a worse prognosis:multisystemic generalized disease; extensivehoneycombing on chest radiographs; severereductions in Dlco, extremes of age; and multiplepneumothoraces. 4,5,13,20 In 1981, Friedmanet al 4 reviewed 100 cases of PLCH confirmed bysurgical lung biopsy. Follow-up was availablein only 60 patients (both treated and untreated).Complete resolution of symptoms was noted in 33patients (55%), 22 (37%) had persistent symptoms,5 worsened, and only 1 (2%) died. The authors ofa European study followed 45 patients with PLCHfor a median of 6 years; 12 patients (27%) died orrequired lung transplantation. 13 Median survivalfrom the time of diagnosis was 13 years. ReducedFEV 1/FVC ratio was an independent predictor ofmortality by multivariate analysis.Investigators from the Mayo Clinic identified87 patients with isolated PLCH diagnosed between1946 and 1996 (83 were adults; 84 were smokers).47 Of these 87 patients, 74 (85%) ultimatelyachieved “disease-free survival.” Treatment regimensincluded prednisone (58%), chemotherapy(10%), surgical excision (7%), or no treatment(25%). Ten patients (11%) died; an additional 3patients developed progressive pulmonary disease.A subsequent review of 102 cases of PLCHin adults from the Mayo Clinic from 1976 to 1998cited a median survival of 12.5 years; 5- and 10-year survival rates after the diagnosis were 74%and 64%, respectively. 21 Only 15 patients (15%)died from respiratory failure. Treatment regimensincluded smoking cessation (100%), administrationof prednisone alone (39%), prednisone pluschemotherapeutic agents (15%), and lung transplantation(1%). Variables associated with worsesurvival included older age, lower FEV 1, highresidual volume, low FEV 1/FVC ratio, and reducedDlco. 21 Another study from the Mayo Clinic notedthat the presence of PAH on echocardiogram markedlyincreased mortality (hazard ratio, 28.8). 48 Thispoint bears emphasis because PAH is common inadvanced PLCH. In one study of 21 patients withPLCH referred for lung transplantation, all patientshad PAH (mean pulmonary arterial pressure of59 mm Hg). 18TherapyBecause of the rarity of PLCH and its highlyvariable natural history, the role of therapy iscontroversial. 1 Smoking cessation is the primary<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 591ACC-PULM-09-0302-036.indd 5917/10/09 10:24:12 PM


and essential component of therapy; the role ofpharmacological agents has not been studiedin randomized trials. 34 For multisystemic LCHin children, the Histiocyte Society recommendstreatment with vinblastine and corticosteroids(CS). 49 A variety of chemotherapeutic regimenshave been used to treat disseminated LCH inchildren, including vinca alkaloids (eg, vincristineor vinblastine), etoposide, cladribine (2-chlorodeoxyadenosine),cyclophosphamide, methotrexate,cyclosporine A, and antithymocyte globulin. 47,49-53In one study of disseminated LCH, cyclosporineA was administered alone (10 patients)or in combination with vinblastine, etoposide,prednisolone, and/or antithymocyte globulin(16 patients). 54 A single patient had a completeresponse, 3 had partial responses, and 22 (85%)did not respond.In several case reports or small series of bone,skin, or disseminated LCH, anecdotal responseswere achieved with 2-chlorodeoxyadenosine(2-CdA), a purine nucleoside that is toxic to lymphocytesand monocytes. 51,55-58 Further, favorableresponses were noted in 3 cases of PLCH treatedwith 2-CdA. 56,58,59 In anecdotal cases, interleukin(IL)-2 60 and etanercept 61 were associated withfavorable responses in children with disseminatedLCH, but data are sparse. Moreover, PLCHin adults likely has a different pathogenesis thanchildhood LCH, and optimal therapy for PLCHis controversial. 30 No controlled or prospectivestudies have been performed. However, giventhe strong link between cigarette smoking andPLCH, patients must be urged in the strongestpossible terms to discontinue smoking. 3 In mostpatients, the disease will stabilize or even improveafter cessation of smoking. 3,62 Anecdotal responseshave been noted in PLCH treated with CS andimmunosuppressive and cytotoxic agents, 5,7,9,13,16,20but the data may be confounded by the impact ofsmoking cessation or the potential for spontaneousremission.One uncontrolled study cited radiographicimprovement in 12 of 14 patients treated withprednisone for progressive PLCH; the other twopatients were stable. 20 In another study, CS therapywas associated with worse survival, 13 but thisfinding may reflect a selection bias. In the cohortof PLCH reported from the Mayo Clinic, 54 of102 patients (53%) were treated with prednisone(alone or combined with immunosuppressiveagents). 21 Although data were not provided, theauthors stated “no specific therapeutic interventionshave been shown to improve survival.” Amulticenter therapeutic trial is needed to clarifythe impact (if any) of therapy but has not beendone. Radiation therapy may be effective for solitarylesions of bone but has no role in pulmonaryLCH. The role of immunosuppressive or cytotoxicagents in isolated PLCH in adults has not beenestablished.The International Histiocyte Society initiateda registry of LCH in adults. 63 That registry identified274 adults with LCH from 13 countries; 44(16%) had isolated pulmonary involvement, and188 (69%) had multisystemic disease. 63 Treatmentregimens varied. The most commonly used therapiesincluded vinblastine (with or without CS) in30% and etoposide in 10%. Only 13 patients (30%)with isolated PLCH were treated. Five-year survivalrates were 89% among patients with isolatedPLCH and 92% among patients with multisystemicinvolvement. The impact of therapy of PLCH couldnot be ascertained.Given the paucity of data and the potentialfor PLCH to remit spontaneously or after cessationof smoking, 1 this author reserve therapy forpatients with severe, progressive, or debilitatingdisease. In this context, an initial trial of CS (0.5 to1 mg/kg/d, with gradual taper) for 2 to 4 monthsis reasonable. Prolonged therapy should be continuedonly in patients manifesting unequivocal andobjective improvement. Immunosuppressive orcytotoxic agents may be considered for patientswith progressive PLCH recalcitrant to CS therapyand cessation of smoking. Pneumothoraces occurin 15% of patients with PLCH, and rate of recurrenceexceeds 50% when managed conservativelyby observation or chest tube without pleurodesis. 17Early surgical pleurodesis is justified in managingpneumothoraces in PLCH patients. The incidenceof PAH is relatively common in patients withPLCH 48 and may reflect a primary vasculopathy(including veno-occlusive disease). 19 Given thehigh incidence of PAH in PLCH, this author recommendstransthoracic echocardiography in newlydiagnosed cases to estimate right ventricular systolicpressures. The use of vasodilators for treatingpulmonary hypertension in patients with PLCHhas not been extensively studied. 18 However, two592 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 5927/10/09 10:24:13 PM


patients with PLCH treated with IV epoprostenoldeveloped acute pulmonary edema. 18Lung transplantation has been successfullyaccomplished in PLCH patients with severe respiratoryfailure refractory to medical therapy. 64,65Data from the International Society for Heartand Lung Transplantation (ISHLT) Registry fromJanuary 1995 to June 2004 identified only 39patients with PLCH among 13,007 lung transplantrecipients (0.3%). 66 French investigators reported39 patients with PLCH who underwent LT at7 centers in France; 65 rates of 1-, 2-, and 5-yearsurvival were 77%, 64%, and 54%, respectively.The disease recurred in 21% but did not influencesurvival. Recurrent PLCH appears to be morecommon among those who resumed cigarettesmoking 64,67,68 or patients with extrapulmonaryinvolvement. 65 Patients must be vigorously counselednot to resume smoking after LT. Criteriafor listing patients with PLCH for LT are lacking.This author believes that LT should be consideredwhen the following factors are present: (1) severerespiratory failure despite smoking cessation (eg,FVC 55% predicted; FEV 1 40% predicted;Dlco 40% predicted); (2) need for supplementaloxygen; and (3) poor and declining quality of life;severe PAH. 69LAMLAM is a rare disorder affecting exclusivelywomen that is characterized by hamartomatousproliferation of atypical smooth muscle along lymphaticsin the lung, thorax, abdomen, and pelvis. 70-73Mean age of onset is between 30 and 45 years ofage; 90% of cases have been in premenopausalwomen. 70,71,74-79 Clinical features include progressiveairflow obstruction, pneumothorax, hemoptysis, orchylothorax. 74 LAM is exceptionally rare, with anestimated prevalence of 1 to 3 per million in theUnited States, 80 United Kingdom, 75 France, 71 andSingapore. 80 The cardinal features of LAM weredescribed in 1974 and 1975 in two series comprising32 patients 81 and 28 patients, respectively. 79In 1990, Taylor et al 78 described 32 patients withLAM who were followed at Stanford and the MayoClinic. In 1995, Kitaichi and colleagues described46 patients with LAM from Japan, Korea, andTaiwan. 77 Subsequent series from the United Kingdom,82 France, 71 Korea, 83 and the United States 72,76further elucidated the clinical spectrum and naturalhistory of this disorder. In the United States,a national registry studied 230 patients enrolledinto a national registry established by the NationalHeart, Lung, and Blood Institute (NHLBI). 72 In aseparate registry (the LAM Foundation), 1,300patients had been registered in the United Statesas of July 7, 2007. 74 In addition to sporadic LAM(which affects women), LAM can complicate tuberoussclerosis complex (TSC), a genetic disorderdiscussed later. 74 LAM occurs virtually exclusivelyin women. 74 Only four cases of biopsy-confirmedLAM have been reported in men; three had definiteor probable TSC, 84-86 whereas one did not. 87Clinical FeaturesThe most common symptom of LAM is dyspnea,which usually begins in the third or fourthdecade of life and progresses inexorably overyears. 71,72,75-78 Pneumothorax occurs in 50 to 80% ofpatients. 71,75-78 The rate of recurrence is 70%, thehighest among all chronic lung diseases. 74 Recurrentpneumothoraces in women of childbearingage may be a clue to the diagnosis. Chylous effusions,resulting from rupture of involved lymphaticsof the pleura, mediastinal lymph nodes,or thoracic ducts, have been noted in 7 to 39%of women with LAM. 71,72,75-78 Chyloptysis, withexpectoration of white, sticky sputum, may alsooccur. 72,75,81 Hemoptysis or focal alveolar hemorrhage,which is caused by obstruction of venules,occurs in 28 to 40% of patients with LAM. 71,72,75-78The course of LAM is usually indolent but inexorable;most patients ultimately die of respiratoryfailure (typically 10 years after diagnosis). However,the prognosis of LAM is heterogeneous, andthe impact of therapy is controversial (discussedlater).Differential DiagnosisBecause of the rarity of LAM and the nonspecificityof clinical features, the diagnosis is oftendelayed for 3 to 5 years after the onset of symptoms.74,88 Symptoms may be erroneously ascribedto chronic bronchitis, asthma, emphysema,PLCH, or chronic ILD. 89 <strong>Pulmonary</strong> lesions identicalto LAM may occur in patients with TSC, anautosomal-dominant familial disorder associated<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 593ACC-PULM-09-0302-036.indd 5937/10/09 10:24:13 PM


with mental retardation and cutaneous manifestations.72,74,89,90 By contrast, LAM is not familial and isnot associated with either cutaneous or neurologicmanifestations. 75,89Chest Radiographic ChangesPlain chest radiographs in LAM are nonspecificbut may demonstrate pneumothoraces,cystic or reticulonodular shadows, pleural effusions,or hyperinflation. 71,75-78 Pneumothorax hasbeen noted in 39 to 53% of patients at the time ofpresentation and occurs in approximately 80%of patients during the course of the disease (Fig10). 71,75-78 Reticulonodular infiltrates are evidentin 47% to 85% of patients. 71,76-78 The reticulationactually represents summation of numerous cysticwalls. 91 Cysts or bullae are detected in 41 to 58%of patients. 71,76-78 Hyperinflation develops as thedegree of airways obstruction worsens, and maybe seen in up to two thirds of patients late in thecourse of the disease. Pleural effusions (oftenchylous due to lymphatic obstruction) occur in 7to 39% of patients. 71,72,76-78,92 Mediastinal and hilarlymphadenopathy are not features of LAM. 75,77Chest radiographs may be normal early in thecourse of disease. 71,75,76HRCT ScansThin-section HRCT scans in LAM revealnumerous thin-walled cysts, ranging in size froma few millimeters to 6 cm throughout both lungs;the intervening lung parenchyma is normal (Figs11–14). 29,71,75-78 Cysts usually are round but mayassume polygonal or bizarre shapes as multiplecysts coalesce (Fig 12). In LAM, the cysts are distributeddiffusely without predilection for specificFigure 11. LAM. Posteroanterior chest radiograph demonstratingmodest hyperinflation and some cystic and emphysematouschanges in a 42-year-old woman with LAM. Surgical clipsare present in the left lung from a previous thoracotomy for thetreatment of recurrent pneumothorax.Figure 10. LAM. HRCT scan from a 33-year-old woman witha history of recurrent pneumothoraces shows numerous welldefinedcysts scattered throughout the lung parenchyma.Figure 12. LAM. CT scan from the same patient demonstratingmultiple thin-walled cysts of varying sizes in the lower lobes,with several large confluent cysts. Note that cysts are scatteredrelatively evenly throughout the lung, with neither peripheralnor central predominance. A nodular component is lacking.594 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 5947/10/09 10:24:13 PM


Figure 13. LAM. Posteroanterior chest radiograph from a 41-year-old woman with a history of recurrent pneumothoracesdemonstrating marked hyperinflation and emphysematouschanges. A reticular pattern is noted in the lower lobes, whichlikely reflects the walls of cysts.Figure 15. LAM. HRCT scan in a 44-year-old woman withLAM demonstrates multiple, thin-walled cystic radiolucenciesbilaterally. Note the two large lesions, representing confluentcysts. From Lynch JP III, Raghu G. Major disease syndromesof unknown etiology. In: Baum G, Crapo J, Karlinsky J, et al,eds. Textbook of pulmonary diseases. 7th ed. Philadelphia, PA:Lippincott Williams & Wilkins, 2004; 629–655.Figure 14. LAM. CT scan from the patient in Fig 13 demonstratingmultiple thin-walled cystic radiolucencies throughout lungparenchyma consistent with LAM. Note that both the centraland peripheral regions of the lungs are involved.regions or lobes. 91 Cavitation is not a feature ofpulmonary LAM. Nodules, interstitial fibrosis, orirregular lung-pleural interfaces—features commonlyobserved in other chronic ILDs—are absentor a minor feature in LAM. 29 GGOs were not citedin early reports of HRCT in LAM 78,91 but were notedin 12% (8 of 66) 71 and 59% (22 of 37) 77 of patients intwo more recent series. GGOs may reflect foci ofFigure 16. LAM. CT scan (upper lobes) from a 36-year-oldwoman demonstrating small (1 to 3 mm) round cysts throughoutboth lung fields, which is consistent with early changes inLAM. A diffuse background haze (ground-glass opacification)is present, which is consistent with alveolar hemorrhage. Shehad a 12-month history of intermittent hemoptysis. Transbronchiallung biopsy specimens demonstrated hemosiderin-ladenmacrophages (which is consistent with previous hemorrhage)and clusters of smooth-muscle cells (positive for HMB45),which is consistent with LAM.alveolar hemorrhage, pulmonary hemosiderosis, ordiffuse proliferation of smooth-muscle cells (Figs 15and 16). 71,75,76 Mediastinal or intrathoracic lymphadenopathyis unusual. 29,77,93 However, retrocrural adenopathyis common, found in 26% of patients with<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 595ACC-PULM-09-0302-036.indd 5957/10/09 10:24:15 PM


LAM in one series. 76 Semiquantitative and quantitativeanalyses of extent of disease by thin-sectionCT scan correlate well with physiologic parameters(eg, FEV 1, Dlco, and gas exchange at rest and exercise)91,94 and exercise performance. 95 Abdominal CTmay reveal (1) cysts or angiomyolipomas (AMLs) inthe kidney, spleen, or pelvic organs or (2) retrocruralor para-aortic lymphadenopathy. 75,76,89,96,97Differential Diagnosis of HRCTCystic radiolucencies may be observed inpulmonary LCH, emphysema, and any chronicend-stage lung disease. However, salient differencesdistinguish these entities from LAM. Cystsin PLCH are preferentially distributed in the upperand mid-lung zones, and a nodular component isusually evident. 15,27 Cavitation may be observedin PLCH but is rare in LAM. Honeycomb cysts, aprominent feature of idiopathic or collagen vasculardisease–associated pulmonary fibrosis, aredistributed in the peripheral (subpleural) regionsof the lungs and are nearly invariably accompaniedby linear bands, distortion of lung parenchyma,or GGOs. In those disorders, the processis heterogeneous, with focal areas of involvedlung interspersed among large areas of normallung parenchyma. Severe, end-stage LAM maygive rise to large cysts that may resemble bullousemphysema; however, the emphysematous lesionslack discernible walls and have an upper lobe predominance.The cysts in LAM are more regular andhave well-formed walls. 76,78,91 Other rare causes ofcystic lung disease include lymphocytic interstitialpneumonia, 98 Sjögren syndrome, 99 hypersensitivitypneumonitis, 100 amyloidosis, 99 light chain depositiondisease, 101 and Birt-Hogg Dube’ syndrome. 102Ventilation-Perfusion Lung ScansAn unusual “speckling” pattern on ventilationlung scanning was cited in 28 of 39 of patients withLAM (72%) in one study. 103 On CT scans, all 39 hadpulmonary cysts. Univariate analysis showed thatthe extent of disease on chest radiographs and CTscans, cyst size, ventilation-perfusion abnormalities,and the degree of speckling were inverselycorrelated with FEV 1, Dlco, and ratio of FEV 1/FVCbut not with FVC or TLC. 103<strong>Pulmonary</strong> Function TestsEarly in the course of LAM, spirometry maybe normal. 71,72,74,77 The Dlco is reduced in 65%of patients. 71,72,76-78,104 An obstructive ventilatorydefect, often with air trapping, has been cited in29 to 78% of patients; lung volumes are normal orincreased. 71,76-78,105 Restrictive defects in LAM mayreflect the presence of chylous pleural effusions oreffects of previous pneumothoraces or pleurodesis.94,104 Impairments in exercise performanceand gas exchange are typical 76 and correlate withthe extent of cystic disease on quantitative CTscanning. 95,106Cardiopulmonary exercise testing typicallydemonstrates impairments in maximum oxygenuptake (V • o 2max). 107 Reductions in V • o 2max correlatewith increasing severity of disease by CTand histologic criteria. 107 By multivariate analysis,FEV 1and Dlco correlate with V • o 2max. 107 ReducedDlco is the best predictor of exercise-inducedhypoxemia. 107 In one study, exercise-inducedhypoxemia occurred in 10 of 24 patients (41%)with Dlco between 60% and 70% predicted, 7 of29 patients (21%) with Dlco between 70% and 80%predicted, and 7 of 69 patients (10%) with Dlco 90% predicted. 107Proliferation of smooth-muscle cells withinthe pulmonary interstitium may cause airwaysobstruction, dilation of proximal airspaces, anddestruction of pulmonary capillaries. Airspacecystic lesions appear to be more important thanmuscular proliferation in small airways as acause of airflow obstruction. 95 Loss of alveolarsupport and parenchymal interdependence maybe an important contributory mechanism ofairflow obstruction. In one study of 143 patientswith LAM, a predominantly solid (as opposedto cystic) histologic pattern of LAM lesions inlung biopsy specimens was associated with amore rapid decrease in FEV 1. 104 Further, a positiveresponse to bronchodilators was associated withmore severe airflow obstruction and a greaterrate of decline in expiratory flow. 104 Interestingly,airway inflammation did not correlatewith severity of airflow obstruction or responseto bronchodilators. 104Reductions in the FEV 1/FVC ratio and in creasedTLC are associated with a worse prognosis andpoor survival. 77,105 By contrast, FVC, FEV 1, arterial596 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 5967/10/09 10:24:18 PM


lood gases, and alveolar-arterial gradient do notpredict survival. 77,105 The rate of progression of airflowobstruction is highly variable, ranging from afew years to more than two decades. 75,78,104 A reviewof 47 patients with LAM in the United Kingdomcited a mean decrease in FEV 1of 118 mL/year, butthere was marked variability between patients. 82In a retrospective review of 31 cases of LAM,French investigators cited a mean decrease in FEV 1of 106 mL/year. 71 The largest study (comprising275 patients with LAM in the United States followedfor 5 years) cited annual decreases in FEV 1ranging from 52 to 119 mL. 108 Overall annual ratesof decrease were 1.7% predicted for FEV 1and 2.4%predicted for Dlco. The most significant predictorsof functional decrease were initial lung functionand age. The rate of decrease in FEV 1was lowerin older patients. Conversely, patients with greaterinitial FEV 1and Dlco (less severe disease) hadmore rapid decreases in Dlco. 108 In another studyof LAM patients with mild disease, the presenceof pneumothorax was associated with a lowerFEV 1and more rapid decrease in FEV 1than thosewithout pneumothorax. 109 These correlations werenot observed in patients with more profuse cysticchanges.Although airflow obstruction is the mostimportant factor responsible for exercise limitation,pulmonary vascular involvement plays acontributory role. 95 PAH at rest in LAM is uncommon( 10% of cases) 110 but may be elicited byexercise. 111 Taveira-DaSilva et al 111 performedechocardiography at rest and during exercisein 95 patients with LAM. By echocardiographiccriteria, PAH was present in 8 patients at rest butwas elicited by exercise (systolic pulmonary arterypressure 40 mm Hg) in 56 (59%). Exercise arterialoxygen saturation was the best predictor ofexercise-induced PAH. 111The smooth muscle cells are heterogeneous andmay exhibit features of large spindle cells, smallercells with little cytoplasm, or epithelioid cells. 75,89These smooth muscle cells are often arranged innodules or in linear fascicles. 75 The LAM cells growin a haphazard arrangement, unlike the orderlypatterns of organization of normal smooth musclecells. 112In pulmonary LAM, open lung biopsies demonstrateboth cystic lesions and muscular lesions(ie, proliferation of immature smooth muscle cells)(Fig 17). 75,77 Nodular proliferations of immaturesmooth muscle surround airways, blood vessels,and lymphatics and extend into bronchioles andalveolar interstitium. Destruction of alveolarparenchyma forms cysts, which are surroundedby immature smooth muscle. The walls of thecysts are typically 2 mm thick. 91 Pneumothoracesresult from rupture of subpleural cysts. 71,75,77Proliferating smooth muscle may obstruct pulmonaryvenules (causing focal edema and pulmonaryhemorrhage) or lymphatics (leading tochylothorax). 71,75,78,79 Hemosiderin-laden macrophageswithin alveolar spaces or interstitium maybe a clue to previous episodes of hemorrhage. 71,75,77The extent of cystic or smooth muscle lesions onopen lung biopsy has prognostic significance.Predominantly cystic LAM lesions suggest aworse prognosis and survival, whereas gradesof smooth muscle proliferation or hemosiderosisdo not correlate with survival. 77,105 LAM histologyscores, based on the extent of replacement ofPathologyGrossly, the lungs in LAM may first appearemphysematous. However, as the disease progresses,the entire lung is replaced by cysts.Histologically, LAM is characterized by proliferationof atypical interstitial smooth muscle andthin-walled cysts within lung, renal parenchyma,uterus, or affected organs. 75,77,79,81,89 Unlike otherforms of ILD, there is very little fibrosis in LAM.Figure 17. Photomicrograph of LAM. Open-lung biopsy specimendemonstrating the proliferation of atypical smooth-musclecells within the alveolar interstitium, which is consistent withLAM (hematoxylin-eosin, high-power magnification).<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 597ACC-PULM-09-0302-036.indd 5977/10/09 10:24:18 PM


lung tissue by cystic lesions and infiltration ofLAM cells, are a predictor of death and time totransplantation. 93Immunohistochemical Stains and AncillaryDiagnostic TechniquesSmooth muscle cells in pulmonary LAMare similar histologically and by immunohistochemicalanalysis to AMLs, which are benignmesenchymal tumors consisting of blood vessels,abnormal smooth muscle, and fat. 71,75,113Immunohistochemical stains are positive formuscle-specific actin, desmin, and human melanomablack-45 (HMB-45). 75,114 HMB-45 producesa granular pattern of staining within the cytoplasm,often with a perinuclear accentuation. 114HMB-45 immunoreactivity in smooth muscle ishighly specific for LAM, AMLs, and clear celltumor of the lung but is never found in normalsmooth muscle. 75,114 The specificity of HMB-45may allow a diagnosis of LAM by transbronchiallung biopsy, provided the immunohistochemicalmarker is positive. 71,76 The immunohistochemicalreactivity of LAM cells is localized more frequentlyin the large epithelioid cells. 93 In addition,LAM cells contain increased amounts of matrixmetalloproteinases (MMPs), especially MMP-2but also MMP-1 and MMP-9. 115 These MMPslikely play a critical role in the development ofpulmonary cystic lesions in LAM. 115 Receptorsfor progesterone and estrogen have been notedin the nuclei of proliferating smooth-musclecells in some patients with LAM, 71,78 but thesefindings are not uniform. 116 Electron microscopyof smooth cells in LAM demonstrates indentedround to ovoid nuclei with prominent nucleoli,prominent microfilament bundles with densebodies, a well-developed rough endoplasmicreticulum, and numerous clusters of intracytoplasmic,electron-dense, membrane-bound,crystalloid granules. 117 Similar cells are found inrenal or hepatic AMLs. 117BAL FluidDiagnosisIn two studies, hemosiderin-laden macrophageswere found in BAL fluid in 60 to 81%of LAM patients; BAL differential counts were71, 76normal.of LAMHistorically, the diagnosis of LAM requiredopen lung biopsy. 78,79,81 It was often diagnosedinadvertently at the time of thoracotomy forrecurrent pneumothoraces. 78 The presence of innumerablesmall cysts throughout the lung surfacewithout nodularity or interstitial fibrosis is virtuallypathognomonic for LAM. 78 Transbronchiallung biopsy may be adequate to substantiate thediagnosis but requires a skillful pathologist toperform. 71,77,78 Unless the pathologist is familiarwith the salient histologic features, the abnormalsmooth muscle cells are often misinterpreted asfibrocytes. In equivocal cases, positive immunohistochemicalstains (eg, HMB-45) may substantiatethe diagnosis of LAM, provided that clinicaland radiographic features (especially HRCT)are consistent. 71,75,76 The diagnosis of LAM canbe established without lung biopsy providedHRCT features are characteristic, clinical criteriaare consistent, and any of the following featuresare present: typical renal AMLs, chronic chylousascites with abdominal lymphadenopathy, orcharacteristic histopathologic findings on lymphnode biopsy. 71,74-76 Blood tests have no diagnosticvalue in LAM. However, serum vascular endothelialgrowth factor-D (VEGF-D) shows promiseas a screen for pulmonary cyst development inwomen with TSC. 118Extrapulmonary InvolvementAMLs or cysts may involve abdominal orretroperitoneal lymph nodes, spleen, kidney,periadrenal blood vessels, liver, uterus, andovaries. 71,75,76,97,112,113,119 Although the source ofLAM cells (atypical proliferating smooth musclecells) is not known, there is direct evidence forblood-borne and lymphatic dissemination ofLAM. 74,119-121 Induction of lymphangiogenesismay be important. Lymphatic endothelial markersare expressed abundantly in LAM, includingpodoplanin, VEGF receptor-3, and VEGF-C. 121VEGF-D is increased more than threefold in theserum of patients with LAM 122 and may prove tobe a useful biomarker for LAM. Possible sourcesof LAM cells include AMLs, lymphatic or bonemarrow cells, uterine or ovarian cells, or cells ofneural crest origin. 74598 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 5987/10/09 10:24:20 PM


Renal or abdominal AMLs (HMB-45 positive)have been noted in 15 to 54% of patients withpulmonary LAM and often antedate the diagnosisof pulmonary LAM. 71,76,97,113 In one study, uterineleiomyomas were cited in 41% of patients withLAM. 71 AMLs often are incidental findings on CT,but the growth of these smooth muscle tumors maycause local pain, bleeding, chylous ascites, or compressionof contiguous structures. 75,76,113 A study of12 AMLs from women with LAM demonstratedpositive estrogen receptor immunoreactivity in10 (83%); all expressed progesterone receptors(PRs). 123 A previous study noted positive PRs in55% of AMLs associated with TSC but in only 7%without LAM or TSC. 124AMLs are hyperechogenic on ultrasonographyand reveal fat attenuation on CT. 75,76 CT scanningis the best screening test to assess the presence,size, and extent of renal or intra-abdominal cysticor angiomyolipomatous lesions in patients withpulmonary LAM. 97,113 In one prospective study,80 patients with LAM underwent chest andabdominopelvic CT and abdominopelvic ultrasonography.97 Intra-abdominal lesions were detectedin 61 patients (76%), including renal AMLs in 43(54%), enlarged abdominal lymph nodes in 31(39%), lymphangiomyoma in 13 (16%), ascites in8 (10%), dilatation of the thoracic duct in 7 (9%),and hepatic AMLs in 3 (4%). 97A significant correlation (p 0.02) was foundbetween enlarged abdominal lymph nodes and theseverity of lung disease. 97 Ultrasonography andabdominal CT scans may have complementaryroles in the diagnosis and follow-up of abdominaland pelvic manifestations of LAM. 97 Appropriatemanagement of renal AMLs or cysts depends onsize, growth rate, and the presence or absence ofsymptoms. AMLs often are incidental findingsnoted on CT but may cause pain, local bleeding,compression of renal parenchyma, and impairedrenal function. 75,76,113 The risk of local complicationsis increased for lesions 4 cm in size. 75,113Asymptomatic patients with lesions 4 cmin greatest dimension can be followed with yearlyultrasonography or CT; lesions 4 cm in size orcausing symptoms should be followed no lessoften than at 6-month intervals. 113 Expanding orsymptomatic lesions 4 cm in size should betreated with surgery or embolization. 113 Partialor total nephrectomy may be required to managecomplications (eg, hemorrhage and refractorypain). 76,113 An increased risk of renal cell carcinomahas been noted in patients with TSC andAML. 110 The prevalence of meningiomas, whichpossess PRs, appears to be increased in patientswith LAM and TSC. 125 In one prospective study,MRI or CT scans of the brain were performed in250 women with LAM; lesions consistent withmeningiomas on MRI were found in 8 (5 hadTSC). 125TSCBoth LAM and TSC are caused by mutationsin the tuberous sclerosis genes, TSC1 and TSC2,that control cell growth, survival, and motilitythrough the Akt/mammalian target of rapamycinsignaling pathway. 74 Sporadic LAM (S-LAM) andTSC-associated LAM (TSC-LAM) share many commonclinical and radiographic features. However,pulmonary manifestations are typically milderin patients with TSC-LAM. 89 Further, comparedwith S-LAM, patients with TSC-LAM have a lowerincidence of abdominal lymphangioleiomyomas(9% vs 12%), thoracic duct dilation (0% vs 4%),pleural effusion (6% vs 12%), and ascites (6% vs10%). 126 Additionally, patients with TSC-LAM havea greater frequency of hepatic (33% vs 2%) andrenal (93% vs 32%) AMLs, previous nephrectomy(25% vs 7%), and noncalcified pulmonary nodules(12% vs 1%). 126TSC, an autosomal-dominant disorder withhigh penetrance and variable expressivity, sharessome features with LAM (eg, extrapulmonaryAMLs; hamartomatous proliferation of smoothmuscle; pulmonary LAM). 74,127 TSC affects 1 in5,800 individuals 128 and is far more common thansporadic LAM. 74 TSC affects the migration anddifferentiation of neural crest cells and affectsmultiple tissues, including skin, CNS, lymphatics,bone, and lung. 113, 127 Up to 60% of cases of TSC aresporadic, with no family history. 80 Both male andfemale patients can be affected with TSC. Clinicalfeatures commonly observed with TSC includemental retardation, seizures, digital fibromas,sebaceous adenomas, depigmented skin lesions,sclerotic bone lesions, and cranial calcifications. 71,129These features are not present in sporadic LAM. 75,128The phenotypic expression of tuberous sclerosisis variable, but 75% of patients die by the age of<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 599ACC-PULM-09-0302-036.indd 5997/10/09 10:24:20 PM


20 years. 113 Early prevalence studies cited pulmonarylesions indistinguishable radiographicallyand histopathologically from LAM in 1 to 4% ofpatients with TSC. 71,75,90,127 However, subsequentstudies suggest that the prevalence of LAM amongwomen with TSC is 26 to 39%. 93,128,130In one prospective study, screening HRCTscans were performed in 48 patients with TSC whohad no pulmonary symptoms, no previous historyof LAM, and normal PFTs. 93 <strong>Pulmonary</strong> cysts consistentwith LAM were detected in 13 of 38 (34%)female patients but in none of 10 male patientswith TSC. Similarly, in a retrospective study of 78women with TSC, 20 (26%) had evidence of LAM(by CT in 13; by surgical lung biopsy or autopsyin 7). 130 AMLs are found in 80% of patients withTSC. 71,75,91,127 AMLs in both LAM and TSC expressthe antimelanoma monoclonal antibody HMB-45,suggesting a common origin from a progenitorsmooth muscle cell. 71,75,113,128 Chromosomal abnormalities(13q14 and 14q24) are noted in TSC, LAM,and uterine leiomyomas. 74,128 Further, the markerfor LAM cells (HMB45) is located on chromosome12q13-15, a region frequently rearranged in uterineleiomyomas. 80 Both leiomyomas and LAM involveproliferation of smooth muscle in women of childbearingage. 74,128In both TSC and LAM, mutations in one oftwo genes (TSC1 or TSC2) are present (mutationsin TSC2 are more common in both disorders). 74,128Tumor suppressor genes (TSC1 and TSC2) encodeTSC1 and TSC2 proteins, also known as hamartinand tuberin, respectively. 74,128 In TSC, a germlinemutation (ie, first hit) is present in all cells in thebody. Neoplasms and dysplasias occur whensomatic “second-hit” mutations result in lossof heterozygosity for the normal allele. 74,128 Thisresults in loss of function of either TSC1 or TSC2proteins. 128 Because TSC genes suppress tumorgrowth, this additional mutation allows cellularproliferation and the formation of hamartomas. 128TSC1 mutations in general are associated withmilder manifestations of disease severity comparedwith TSC2 mutations. 131 Germline mutationsare present in all cells of the body in TSC orTSC-LAM, whereas in sporadic LAM, both thefirst-hit and second-hit mutations are confined tolesions in the lung, kidney, and lymph nodes. 74Loss of heterozygosity for TSC2 (but not TSC1)was found in renal AMLs from 7 of 14 patientswith LAM without clinical features of TSC. 132A similar lesion was found in AMLs in patientswithout LAM or TSC. 75 Because a germ-linemutation is not found in LAM, sequential somaticmutations of the TSC2 gene likely occur in LAM.The relationships between LAM, TSC, and othersmooth muscle proliferative disorders (eg, AMLs,leiomyomas) are likely complex, and additionalstudies are required to elucidate the genetics andpathogenetic mechanisms involved in these disorders.74,128PathogenesisSignificant advances in the understandingof the pathogenesis of the smooth-muscle proliferationin LAM have been made within thepast decade. 74,128,133,134 Because LAM only affectswomen (with rare exceptions), estrogens likelyplay a key role in the pathogenesis. 70,74 Steroidreceptors for estrogen were detected in LAM butnot normal lungs, 134 suggesting that steroids mayalso play a role in LAM pathology. LAM occursnearly invariably in premenopausal women, andexogenous or endogenous estrogens acceleratethe course of the disease. 70,74 One postulate isthat estrogen signals through Akt and releasesthe tuberin-deficient or hamartin-deficient cellfrom feedback inhibition, resulting in abnormalproliferation. 74LAM cells have high mitotic and low apoptoticactivities that are important in their proliferativebehavior. 80,135 LAM cells consist of two subpopulations:myofibroblast-like spindle-shaped cellsand epithelioid-like polygonal cells. 70,119 Spindleshapedcells express smooth muscle-specific proteinssmooth muscle α-actin, desmin, and vimentinand display high rates of proliferation. 70 By contrast,the epithelioid-like cells exhibit lower ratesof proliferation. 70 Both cell types express melanoma-associatedproteins HMB-45 and CD63. 128Until recently, LAM was considered benign. However,LAM cells may be found in multiple sites(eg, renal AMLs, lung, chylous effusions, lymphnodes, blood, urine, etc), suggesting metastaticbehavior. 119,128,135 Recent evidence suggests thatLAM is a cancer resulting from biallelic mutationsat a single genetic locus, resulting in unregulated600 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 6007/10/09 10:24:20 PM


growth, lymphatic or vascular spread, and tissuedestruction. 74,136Course and Prognosis of LAMThe course of LAM is indolent, but most patientsultimately die of respiratory failure. However, thepace of the disease is extremely variable. Earlyretrospective reviews 79,81 cited 5- and 10-yearsurvival rates of only 60% and 20%, respectively.Subsequent studies noted improved survival,which likely reflects earlier diagnosis (particularlywith the advent of CT scans). In 1990, Taylor et al 78reported that 25 of 32 patients with LAM (78%) werealive a mean of 8.5 years after onset of the disease.Data from Asia cited survival rates of 70% at 5years after the onset of LAM but only 38% (10 of 26patients) at 8.5 years. 77 A study of 69 LAM patientsfrom France cited survival rates (by Kaplan-Meieranalysis) of 91% at 5 years, 79% at 10 years, and71% after 15 years. 71 In a recent study of 57 LAMpatients from the United Kingdom, 10-year survivalfrom the onset of symptoms was 91%. 137 Similarly,a recent report of 227 LAM patients cited 10-yearsurvival rate of 92%. 109TreatmentBecause of the rarity of LAM, optimal therapyhas not been determined. 72,74,108,138 No randomizedor controlled studies have been performed. Corticosteroids,immunosuppressive agents, cytotoxicdrugs, and radiation therapy have no role in LAM.Because LAM is exclusively a disease of womenand appears to be exacerbated by estrogens, physiciansshould advise against pregnancy or theuse of exogenous estrogens. 71,75,82,138 Historically,diverse estrogen-ablative therapies have been tried(eg, oophorectomy, administration of antiestrogenregimens [eg, progesterone, tamoxifen, androgens,lynestrenol]; luteinizing hormone–releasing hormoneantagonists [eg, goserelin]; gonadotropinreleasinghormone agonists [eg, leuprolide acetate];and somatostatin). 71,75−78,82,83 However, these therapiesare of unproven benefit. Data are discouragingbut are limited by retrospective analyses and smallnumbers of patients.In a retrospective review from Stanford and theMayo Clinic, oophorectomy alone was ineffectivein all 16 patients. 78 None of 9 patients treatedwith tamoxifen improved; 2 of 19 treated withIM medroxyprogesterone acetate (MPA) alonestabilized or improved. 78 In 1995, Kitaichi et al 77retrospectively evaluated 40 treatment coursesin patients with LAM. Only 2 patients improved;9 stabilized and 29 deteriorated. Oophorectomy orprogesterone alone was never effective. The combinationof oophorectomy and progesterone wasassociated with deterioration in 9, stabilization in1, and improvement in 1. One patient improvedwith combined treatment with progesterone,tamoxifen, and oophorectomy. Gonadotropinreleasinghormone agonists were ineffective in all 6patients. Tamoxifen (alone or in combination withother agents), was given to 13; only 1 improved.A review from Korea identified 21 cases of LAMseen from 1984 to 1997. 83 Follow-up for ≥ 12 monthswas available in only 10 patients, all of whom weretreated with MPA and/or tamoxifen. No patientunderwent oophorectomy. The disease progressedin 8 of 10; no patient improved. 83Tamoxifen has partial estrogen-agonist activity,which raises concerns about its potential toexacerbate the proliferative process. 75 Chu andcolleagues 76 reported 35 women with LAM evaluatedat the National Institutes of Health from 1995to 1997. Fifteen patients (43%) underwent bilateraloophorectomy before their entry into the study; 28(80%) had received medical antiestrogen therapyand 7 (20%) were untreated. Only one patientwith chylothorax improved with MPA; the valueof therapy in the remaining patients was not discussed.The administration of a synthetic analog ofluteinizing hormone–releasing hormone wasassociated with improvement in one patient withLAM described by Rossi and coworkers 139 butwas ineffective in two other patients reported byRadermecker et al. 140 Triptorelin, a gonadotropinreleasinghormone analog, was ineffective in a prospectivestudy of 11 premenopausal women withLAM. 141 Anecdotal case reports cited responses tointerferon 2b 142 and somatostatin, 143 but the value ofthese modalities is unproven. French investigatorsidentified 69 women with LAM from 1973 to 1996. 71Diverse hormonal therapies were adminis -tered in 57 patients (84%). Treatments includedprogesterone (n 46), tamoxifen (n 22),<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 601ACC-PULM-09-0302-036.indd 6017/10/09 10:24:20 PM


gonadotropin-releasing hormone agonists (n 14),somatostatin (n 6), and oophorectomy (n 5). 71Among 34 patients with serial PFTs, FEV 1improvedby 15% in only 4 patients; 19 patients stabilizedand 11 deteriorated. Treatment regimens usedamong the patients who improved includedtamoxifen and progesterone in 2, progesterone in1, and oophorectomy in 1. A retrospective analysisof 43 patients with LAM in the United Kingdomsuggested that progesterone may slow the rateof decrease of pulmonary function tests (FEV 1,TLC), 82 but differences did not achieve statisticalsignificance.A subsequent retrospective study by theseinvestigators analyzed factors influencing diseaseprogression in a cohort of 57 LAM patients, 36of whom had been treated with progesterone. 137There was more rapid progression of dyspneaamong those who became pregnant after the onsetof symptoms (hazards ratio [HR] of 2.7), patientstreated with progesterone (HR, 2.2) and cigarettesmokers (HR, 2.0). These associations were ofborderline statistical significance. These data donot support a beneficial effect of progesterone onlung function.A larger study from the United States followed275 patients with LAM for approximately4 years. 108 Overall, 139 patients (50%) were treatedwith progesterone (67 PO, 72 IM). No benefit inrates of decrease in FEV 1or Dlco were notedamong patients receiving progesterone comparedwith untreated patients. Absolute annualdecreases in FEV 1were 59 mL among patients nottreated with progesterone vs 119 mL/year and 76mL/year decreases among patients treated withPO or IM progesterone, respectively. Absoluteannual decreases in Dlco were 0.57 mL/min/mmHg among patients not treated with progesteronevs 0.95 and 0.70 mL/min/mm Hg amongpatients treated with PO or IM progesterone,respectively. 108Italian investigators followed 36 women withLAM for a mean of 9.8 years; all received sometype of estrogen ablative therapy. 144 Five- and 10-year survival rates were excellent (97% and 90%,respectively), but the impact of therapy couldnot be assessed. Data from these various clinicalseries. 71,77,78,82,83,108,137 suggest that current therapiesare of limited or no value. Unfortunately, there areno clear guidelines for optimal therapy. Estrogenreceptors and PRs are found in lung LAM lesionsin some patients, 71,74 but the presence or absenceof hormonal receptors does not correlate withresponse to hormonal therapy. 76,78 I do not believethat antihormonal therapies have a role to treatLAM. Sirolimus, a macrolide with immunosuppressiveproperties, inhibits the activity of themammalian target of rapamycin, suppressessmooth muscle proliferation, and suppresses DNAsynthesis of LAM cells in vitro. 128,145 Further, sirolimusinduced remission of angiolipomas in twopatients with TSC. 145,146 In an open-label trial, 25patients with AMLs were treated with sirolimusfor 12 months; 5 had TSC only and 18 womenhad LAM (sporadic or associated with TSC). 147After 12 months of therapy, the mean volume ofAML lesions decreased by nearly 50%. Further,among 11 female patients with LAM, mean FEV 1increased above baseline by 118 mL and FVCincreased by 390 mL at 12 months. 147 These beneficialeffects tended to reverse after the drug waswithdrawn.In one woman with sporadic LAM, sirolimusled to complete disappearance of abdominaland pelvic masses and right chylous effusionand improved lung function. 148 Additional casereports cited improvement with sirolimus post-LT chylothorax, 149,150 retroperitoneal AML, 151 andpulmonary LAM. 150,152 These data are encouraging,but given the potential for serious adverseeffects with long-term use of sirolimus, additionalstudies are required to ascertain the role for thisagent in LAM. A 3-year randomized, placebocontrolledclinical trial evaluating sirolimus inpulmonary LAM is in progress (Multicenter InternationalLAM Efficacy of Sirolimus Trial). 74 Recentstudies found that simvastatin inhibits migrationof human LAM cell cultures and inhibits LAMcell proliferation; furthermore, combined treatmentwith simvastatin and sirolimus abrogatedcell proliferation to a greater degree than eitheragent alone. 128 Future studies exploring the role ofstatins as therapy for LAM would be of interest.Another clinical trial in progress at the NHLBI isevaluating octreotide, 74 which has shown promisefor treating chylous complications in otherdiseases. 153Bronchodilators and supplemental oxygenhave adjunctive roles in selected patients withpulmonary LAM. 71,75,76 Sildenafil or vasodilators602 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 6027/10/09 10:24:21 PM


have a theoretical role among patients with PAHcomplicating LAM, 111 but data are lacking.Management of Chylous Effusions andPneumothoraces in LAMPneumothoraces complicate LAM in 50 to 80%of patients; chylothoraces, in 10 to 39%. 71,72,75,76,88Both complications may occur repetitively in individualpatients, mandating an aggressive surgicalapproach to therapy. Thoracostomy tubes may beadequate for pneumothorax or chylothorax in somepatients, but recurrences are common. 71,74,76,92 Videoassistedthoracoscopic surgery with pleurodesisis warranted for recurrent pneumothoraces orchylothoraces complicating LAM. 92 For recurrentpneumothoraces, partial pleurectomy is preferred;chemical pleurodesis should be avoided. 71 Managementof chylous pleural effusions or ascites in LAMpatients is difficult. Dietary fat restriction, peritoneal-jugularshunts, and sclerosing agents havebeen tried but are usually ineffectual. 71,75,76,92 Giventhe rarity of LAM, optimal management of chylouseffusions is controversial. The approach should beindividualized, taking into account the severity orpersistence of chylothorax, comorbidities, and localexpertise. 92 For refractory chylothorax, this authorfavors video-assisted thoracoscopic surgery withtalc pleurodesis and thoracic duct ligation. In onestudy, placement of a pleurovenous shunt led tosuccessful resolution of chylous pleural effusionsin one LAM patient with persistent chylothoracesrefractory to pleurodesis and thoracic ductligation. 154LT for LAMSingle or double LT has been successful inpatients with LAM and end-stage pulmonary insufficiency.155-158 Data from the ISHLT Registry fromJanuary 1995 to June 2006 identified 175 patientswith LAM among 17,616 LT recipients worldwide(1%). 159 Survival rates after LT for LAM are generallysimilar to other indications. 155-158 A review of61 LAM patients who had LT in Europe from 1997to 2007 cited 1- and 3-year survival rates of 79%and 73%, respectively. 157 Survival rates among 44LAM patients who had LT in France between 1988and 2006 were similar (1- and 3-year survival ratesof 80% and 74%, respectively. 158 Experience in theUnited States (79 patients between from 1987 to2002) was similar: 1- and 3-year survival rates of85% and 76%, respectively. 155When to list LAM patients for LT is difficultbecause the prognosis is varies widely amongpatients. Guidelines for LT in LAM include thefollowing: disease progression despite medicaltherapy, FEV 1/FVC 50%, TLC 130%, FEV 130%, and severe cystic disease on HRCT. 80LAM-associated complications of LT includemassive operative hemorrhage caused byextensive pleural adhesions, pneumothoraxin the native lung (for single lung transplantrecipients), and postoperative chylothorax. 88,157,158Previous pleurectomy or talc pleurodesis can createdifficulties with tissue plane dissection andbleeding during removal of the native lung(s). 88In one study of 80 LAM patients who had LT,pleural-related postoperative bleeding occurredin 13 of 45 patients with previous pleurodesiscompared with 1 of 35 without prior pleurodesis.88 Management strategies for chylothoraxinclude drainage, diet containing mediumchaintriglycerides, progesterone, thoracic ductligation, pleurodesis, and pleurectomy. 74,82 Therecurrence of LAM in the donor lung allografthas been noted in 3 to 8% of LT recipients. 74,157,158Molecular techniques confirmed that foci ofLAM in lung allografts were derived from therecipient, 160,161 indicating that LAM cells canmigrate or metastasize despite histologicallybenign features. 119COPCOP is a rare immune disorder characterizedby granulation tissue that obstructs small bronchiolesand extends into the distal alveolar ducts andalveoli. 162-165 The term BOOP is synonymous, 166 butthe use of this term should be discouraged. 164,167Clinical features of COP include focal alveolarinfiltrates, cough, fever, dyspnea, and a subacutecourse, mimicking community-acquired pneumonia.163,164 Obliterative bronchiolitis (OB), which ischaracterized by obstruction of bronchioles andsmall airways but lacks the organizing pneumonia(OP) component, is a distinct disorder withdiffering clinical and radiographic features andprognosis. 164,167 OP (with or without OB) maycomplicate collagen vascular disease, 168,169 Sjögren<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 603ACC-PULM-09-0302-036.indd 6037/10/09 10:24:21 PM


syndrome, 170 inflammatory bowel disease, 171Wegener granulomatosis, 172 lower respiratorytract infections (eg, legionellae, viruses, and Mycoplasmasp), 165,173 toxic fume inhalation; radiationtherapy, 174 chemotherapy, 175,176 malignancy, 177,178lung cancer, 179 or bone marrow, 180 stem cell, 181,182or lung or heart-lung 183 transplants. It can alsooccur as a reaction to pharmacologic or exogenousagents. 167,184-186 Only idiopathic (cryptogenic)organizing pneumonia, a disorder in which nospecific etiology can be established, will be discussedhere.EpidemiologyThe incidence of COP is difficult to ascertain,as historically a variety of terms were used interchangeablyto refer to this disorder (eg, OP, COP,BOOP, bronchiolitis obliterans, and resolvingpneumonia). Nonetheless, COP (or idiopathicBOOP) is rare, and most referral centers see nomore than two to five cases per year. In 1983,Davison and colleagues 187 described eight patientswith “cryptogenic organizing pneumonia.” In1985, Epler and colleagues 166 described 67 patientswith BOOP (including 50 with the idiopathic form)gleaned from 2,500 open lung biopsies performedfrom 1950 to 1980. A retrospective reviewof pathologic files at the University of Alabamaidentified 24 cases from 1972 to 1984; five hadconnective tissue disease. 188 French investigatorsat a leading referral center identified 16 patientswith idiopathic BOOP/COP in 6 years. 189 A surveyin Japan detected 29 cases of BOOP/COP from200 hospitals from 1986 to 1988; 5 patients hadconnective tissue disease. 190 In a Spanish communityhospital, 33 cases of COP were observed in19 years. 179A survey at a leading referral hospital inVancouver, British Columbia, from 1985 to 1992detected 25 cases of COP. 162 This represented aprevalence of 1.2 per 10,000 admissions. Theirreview of 63 published papers identified 296patients with COP (24 were single case reports).Of the published cases, 218 were idiopathic and29 had connective tissue disease. Gudmundssonet al 191 retrospectively evaluated all biopsy-provencases of OP diagnosed in Iceland between 1984and 2003. 191 The mean annual incidence of OP per100,000 population was 1.10 for COP and 0.87 forsecondary OP. Most patients with idiopathic COPpresent in the fourth through sixth decades of life,but all ages may be affected. 162-164,166 There is nopredominance by sex. 191 COP is more common innonsmokers (2:1 ratio). 163Clinical FeaturesPredominant symptoms are cough (typicallynonproductive) and dyspnea. 162,163,166 Extrapulmonaryinvolvement does not occur, but fever,weight loss, and malaise may be prominent. 162-164,166The course is subacute, with symptoms developing 2 weeks to 6 months. 162,163,166 An antecedentupper or lower respiratory tract infection precedesthe onset of symptoms in at least onethird of patients. 162,166 The constellation of thesefeatures simulates an infectious etiology. Thediagnosis is often not suspected until repetitivecourses of antibiotics fail. Physical examinationreveals mid-inspiratory squeaks or rhonchi in40% of patients; wheezing does not occur. 162,166Rales may be present, but frank consolidationchanges on chest auscultation are lacking. 163,164Clubbing is not a feature of COP. 163 No consistentlaboratory aberrations are present. The erythrocytesedimentation rate is usually increased;peripheral blood leukocytosis is noted inonethird. 162,163,166Chest RadiographsFocal lobar or segmental alveolar infiltrates,often with air bronchograms, are the cardinalfeatures, observed in 60 to 80% of patients withCOP (Fig 18, 19). 162,163,164,166,187 These infiltratesare often in the periphery of the lung (resemblingchronic eosinophilic pneumonia) 164,192 but lack thestriking upper-lobe predilection characteristic ofthat disorder. The infiltrates may wax and wane,even before initiation of therapy. A more diffusepattern, associated with reticulonodular infiltrates,occurs in 20 to 30% of patients. 162,164,166 Inthis context, focal opacities are lacking, and thepattern resembles other chronic ILDs. Chest radiographsare normal or reveal only hyperinflation in4 to 10% of patients. 162,163,166 Focal airspace diseasewith consolidation is associated with an improvedprognosis and greater rate of responsiveness to604 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 6047/10/09 10:24:21 PM


Figure 19. BOOP. Posteroanterior chest radiograph demonstratingextensive patchy alveolar infiltrates in the left lung anda patchy infiltrate in the right lower lobe in a 58-year-old man.The open-lung biopsy specimen was consistent with BOOP.After 4 weeks of therapy with prednisone, 60 mg/d, he wasasymptomatic, and the infiltrates seen on chest radiographshad almost completely resolved. From Neagos GR, Lynch JPIII. Making sense out of bronchiolitis obliterans. J Respir Dis1991; 12:789-814.Figure 18. BOOP. Top, A: posteroanterior chest radiographfrom a 75-year-old woman with a 6-month history of recurrent“pneumonias” demonstrating focal right lower lobe consolidation.A patchy left lower lobe infiltrate is also present. Bottom,B: posteroanterior chest radiograph from the same patient asin top, A, demonstrates right upper lobe and left lower lobealveolar infiltrates; a patchy right lower lobe infiltrate is alsopresent. A transbronchial lung biopsy specimen demonstratedthe classic features of BOOP. Following the initiation of therapywith prednisone, 40 mg/d, complete radiographic and symptomaticresolution was evident within 3 weeks.corticosteroids compared with other radiographicpatterns. 163 Pleural effusions, cavitation, and intrathoraciclymphadenopathy are not features ofCOP. However, multiple nodular mass lesions, 193miliary lesions, 194 or cavitary nodules 195,196 havebeen noted in anecdotal cases.HRCT scans depict the salient features ofCOP more clearly than conventional chest radiographs163,197 but are not essential for diagnosis ormanagement. Typically, multiple segmental or lobardense alveolar opacities with air bronchograms areseen in the peripheral regions of the lungs. 163,192 Insome cases, a feeding vessel or bronchus leadinginto the area of consolidation or nodular opacitymay be seen. Reticulonodular infiltrates are notedin one quarter of patients. Other findings includeGGOs and peribronchiolar nodules extendinginto the lung parenchyma in a centrifugal patternfrom the involved airways. 163,192,194 Honeycombingis rare.One study of 38 patients with COP noted thefollowing features on CT: consolidation (87%),GGOs (58%), bronchial dilation (58%), nonseptallinear or reticular opacities (45%), nodules ormass lesions (32%), and peribronchial distributionof consolidation (29%). 192 Mediastinal lymphadenopathymay be present but is usually minor(lymph nodes 15 mm in size). The presence ofconsolidation on CT predicts a favorable prognosis,with complete or partial resolution with CStherapy; by contrast, reticular patterns suggest an<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 605ACC-PULM-09-0302-036.indd 6057/10/09 10:24:22 PM


incomplete response or progression to fibrosis. 198Occasionally, COP presents as a solitary focalmass lesion, mimicking neoplasm. 163,199,200 In thissetting, the diagnosis is usually established aftersurgical excision. 199,200 Medical therapy is usuallynot required for isolated OP noted incidentallyat surgery. 200 Increased uptake of fluorodeoxyglucoseon PET scan is characteristic in patientswith OP; 199,201 however, this finding has no clinicalvalue.PFTsPFTs in COP demonstrate reductions in lungvolumes (eg, vital capacity and TLC) and Dlco anda widened alveolar-arterial oxygen gradi ent. 162-164,166Despite the involvement of small airways, anobstructive component is rare in nonsmokers. 163,166The lack of an obstructive defect in COP has beenascribed to the patchy nature of involvement. Someregions are completely obstructed (resulting inloss of entire units and proportionate reductionin lung volumes), whereas other units are spared.Bronchodilators generally are ineffective. Thesedeficits usually reverse promptly with corticosteroidtherapy. 162-164,166Figure 20. Photomicrograph of BOOP. Open-lung biopsyspecimen demonstrating a plug of organizing granulationtissue within a respiratory bronchiole. A peribronchiolarmononuclear inflammatory cell infiltrate is also evident(hematoxylin-eosin, high-power magnification). From NeagosGR, Lynch JP III. Making sense out of bronchiolitis obliterans.J Respir Dis 1991; 12:789-814.HistologyThe cardinal histologic feature of COP is an exuberantinflammatory and fibrotic process involvingterminal and respiratory bronchioles. 163,165 Tuftsof granulation tissue and aggregates of neutrophils,edema, debris, fibrin, connective tissue,myofibroblasts, and fibroblasts plug the terminalbronchioles (Fig 20, 21). Inflammatory cells arepresent within bronchiolar lumens and extend intothe peribronchiolar regions, alveolar ducts, andalveolar spaces. Mononuclear cells predominate,but scattered neutrophils and eosinophils are seen.Multinucleated giant cells are present in up to 20%of cases. Foamy alveolar macrophages within thealveolar spaces may simulate DIP. 202 The disease ispatchy, but the peribronchiolar distribution (whichcan be identified on lower-power magnification) isa clue to the diagnosis.Extension of the process to contiguous alveolarducts and spaces results in the “organizingpneumonia” component (Fig 22). 163,165 Despitethe presence of extensive granulation tissue andFigure 21. Photomicrograph of BOOP. Open-lung biopsyspecimen demonstrating a plug of granulation tissue withinthe bronchioles. A peribronchiolar inflammatory cellularinfiltrate is present extending into the alveolar interstitium(hematoxylin-eosin, high-power magnification).edema, the alveolar architecture is preservedand fibrosis is absent. The lesions in COP appearto be of uniform age, suggesting a stereotypicresponse to prior injury. By contrast, constrictivebronchiolitis, also termed OB, lacks the organizingpneumonia component of COP and is associatedwith a distinctly worse prognosis. 164 In OB, thesmall airways and bronchioles are obliteratedby transmural fibrosis and inflammation, but aprominent inflammatory component is lacking. 164The airway lumens are narrowed by a concentricdeposition of collagen and fibrous connective tissue.The histologic lesions of OB may be seen as a606 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 6067/10/09 10:24:23 PM


Figure 22. Photomicrograph of BOOP. Open-lung biopsyspecimen demonstrating a plug of granulation tissue withinthe bronchioles. A peribronchiolar inflammatory cellularinfiltrate is present extending into the alveolar interstitium(hematoxylin-eosin, low-power magnification). From Clinicalpulmonary medicine. Boston, MA: Little Brown, 1992;193-263. 197complication of rheumatoid arthritis, toxic fumeinhalation, bone marrow or lung transplantation,and diverse exposures. 165,167,184,185 In contrast to COP,OB is associated with a poor prognosis and low rateof responsiveness to corticosteroids or cytotoxicagents. The course is characterized by progressiveairflow obstruction, eventually resulting in respiratoryfailure and death. This entity is beyond thescope of this chapter.The histologic features of COP overlap withthose of several other entities, including DIP,chronic eosinophilic pneumonia, hypersensitivitypneumonitis, 163 cystic fibrosis, 203 infectiousetiologies, 163 and more recently described entitiessuch as airway centered interstitial fibrosis, 204idiopathic bronchiolocentric interstitial pneumonia,205 and bronchiolitis interstitial pneumonitis. 206The differentiating features of these entities isbeyond the scope of this chapter and is addressedelsewhere. 204-206 Given the patchy nature of COP,fiberoptic bronchoscopy with transbronchialbiopsy is usually not adequate to establish thediagnosis with certainty. In such cases, videoassistedthoracoscopic surgical lung biopsy isrequired. However, the diagnosis can sometimesbe confirmed with transbronchial biopsyif clinical features are consistent and infectiousetiologies are reliably excluded. 163,164,207 Grossmacroscopic findings at the time of bronchoscopyand analysis of BAL fluid are helpful in excludinginfectious etiologies. The absence of bronchitis orpurulent secretions within airways and negativesmears and cultures of BAL fluid makes infectionless likely. BAL in COP often reveals increasesin lymphocytes (20 to 40%), polymorphonuclearleukocytes (5 to 15%), or eosinophils (5%), sometimeswith plasma cells or mast cells. 162-164 TheCD4/CD8 ratio is decreased. 163,164 These aberrationsare nonspecific, and do not distinguishCOP from other infectious or immune-mediateddisorders. Percutaneous CT-guided needle biopsymay establish the diagnosis in some patients withCOP, 208 but caution should be applied owing tosmall sample size.PathogenesisThe pathogenesis of COP has not been elucidatedbut likely represents an exaggerated hostresponse to a variety of inflammatory or injuriousstimuli. 163,164 The uniformity of the histologic lesionand the peribronchiolar distribution suggest thatinhaled stimuli (eg, possibly inhaled antigens ornoxious agents) may initiate lung injury. This isfollowed by activation and recruitment of inflammatorycells and an attempt to repair. The frequentassociation of antecedent viral or respiratory tractinfection in COP suggests that inhaled viral or bacterialantigens may elicit an intense local immuneresponse.BAL fluid in patients with COP often demonstrateslymphocytic alveolitis with expansionof CD8 cells and increased levels of Th1-relatedcytokines including γ-interferon, IL-12, andIL-18. 209 Newly formed intraluminal fibromyxoidtissue in COP demonstrates increased capillarizationcompared with fibroblastic foci of usualinterstitial pneumonia (UIP). 210 Angiogenesismay be mediated by growth factors, VEGF, andbasic fibroblast growth factors. 210 Regulation ofangiogenesis may influence reversibility of thefibrotic lesions.Imbalance between matrix MMPs and tissueinhibitors of metalloproteinases (TIMPs) may playa role in the airway remodeling observed in COP. 211Concentrations of MMP-9 and TIMP-1 wereincreased in BAL fluid from patients with COPand UIP compared with control patients, but thehighest concentrations were observed in patientswith COP. Lymphocytes are important cellular<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 607ACC-PULM-09-0302-036.indd 6077/10/09 10:24:25 PM


sources of MMP-9. It is possible that MMP-2 preferentiallydegrades collagens and that TIMPs playa role in remodeling after parenchymal damage. 211Suga et al 212 also found increased expression ofMMP-9 in UIP, whereas COP cases showed predominantMMP-2 expression in BAL fluids andtissues. MMP-2 activity in COP correlated withincreases in BAL lymphocytes. 212 Japanese investigatorsnoted increased serum levels of KL-6 (amucin-like glycoprotein) in 17 of 30 patients withCOP). 213 KL-6 may be produced by damaged orregenerating type 2 epithelial cells and may beelevated in diverse idiopathic interstitial pneumonias(including UIP and nonspecific interstitialpneumonitis). 214The pivotal factors that initiate, direct, and regulatethe inflammatory and reparative process inCOP have yet to be elucidated. However, therapiesthat ablate the inflammatory response are usuallyhighly efficacious in COP.TherapyCorticosteroids are the cornerstone of therapyfor COP. 165,179,215 The optimal dosage and durationof CS therapy for COP have not been studied inrandomized trials. In early studies, initial treatmentwith high dose prednisone (1 mg/kg/d)was advocated, 166 but lower doses (eg, 0.5 to 0.75mg/kg/d) are often adequate in patients withmild-to-moderate disease. 165,179,215 Responses to CSare often dramatic (symptomatic improvementwithin a few days). 165 Radiographic clearing is usuallyevident within 1 to 4 weeks after institutionof CS therapy. 162,165,166 Complete remissions occurin 80% of patients. 165,179,214,215 In the remainingpatients, minor impairments persist, but serioussequela are uncommon.<strong>Pulmonary</strong> fibrosis occurs in 20% of patientsand death in 5%. 162,164,166,179,215 The dose and rateof taper of CS should be individualized accordingto clinical symptoms, chest radiographs, andPFTs. Among responders, the dose of prednisonecan be tapered to 20 mg within 2 to 3 months. 215Relapses may occur as the steroid is tapered ordiscontinued. 163,179,215 In one retrospective study,one or more relapses (mean, 2.4) occurred in58% of 48 patients with COP. 215 Among patientsexperiencing multiple (≥ 3) relapses, the meandelay between first symptoms and treatmentonset was significantly longer compared withpatients who never relapsed (22 vs 11 weeks,respectively; p 0.02). 215 Most relapses occurredwithin 1 year of the initial episode of COP; only2 patients experienced relapses 15 monthsafter the initial episode. 215 Another study notedresponses to steroids in 32 of 33 (97%) patients;relapses occurred in 18 (56%). 179 In both series,treatment of relapses was highly efficacious.Relapses did not affect survival or result in longtermfunctional morbidity. 179,215 In a recent study,relapse rates were greater in COP patients withincreased serum KL-6 levels (37.5%) comparedwith normal KL-6 levels. 213Although the optimal duration of therapy hasnot been studied, a minimum of 6 to 12 months oftreatment is advised. 165 This author prefers to continuelow-dose CS (10 to 15 mg prednisone everyother day) for a minimum of 12 months unlessadverse effects require earlier discontinuation.Patients exhibiting a propensity to relapse eachtime CSs are tapered require chronic low-dose CStherapy. 164 Factors associated with a worse prognosisinclude a predominantly interstitial patternon chest radiographs or HRCT 163 and lack of lymphocytosison BAL fluid. 216 In one study, relapseswere more frequent in patients with more severehypoxemia at initial presentation. 217 In addition,relapses occurred in five of eight patients withassociated disease compared with only 2 of 8 withidiopathic COP. 217Immunosuppressive or cytotoxic agents (eg,azathioprine, cyclophosphamide, cyclosporine A)are reserved for patients who do not respond orwho experience adverse effects from CSs. 163,218-220Data evaluating these agents are limited to anecdotalcases and small series. Macrolide antibiotics(particularly azithromycin) have been used totreat constrictive bronchiolitis (OB) complicatinglung 183,221 or bone marrow transplantation, 180 withfavorable responses noted in small series. Theefficacy of macrolides reflects immunomodulatoryproperties 222,223 rather than antimicrobial activity.Although data are limited for COP, Stover et al 174cited favorable responses to azithromycin in sixpatients with OP (three attributable to radiationtherapy, three cryptogenic). 174 Others noted favorableresponses with macrolide for COP, 224,225 butdata are limited to a few cases. Given the favorableside effect profile of macrolide antibiotics,608 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 6087/10/09 10:24:26 PM


oral macrolides are logical therapeutic agents forpatients with OP (irrespective of etiology) whoare refractory to or experiencing adverse effectsfrom CS.Rapidly Progressive COPA subset of patients with COP exhibit a fulminantcourse, with bilateral alveolar infiltrates,crackles, and severe hypoxemic respiratory failuredeveloping within a few days or weeks. 226,227 Thissyndrome is rare; only 10 such patients were identifiedin a retrospective analysis at three medicalcenters from 1979 to 1992. 226 In that sentinel report,nine patients required mechanical ventilatorysupport. Despite CS therapy in all cases, sevenpatients died of progressive respiratory failure.Five patients received additional therapy withcyclophosphamide or azathioprine. Possible riskfactors included underlying connective tissuedisease (n 3) and exposure to drugs or environmentalagents (n 4). Nine were current or formersmokers. Histologic features of COP were presentin all cases. In addition, necropsies in six patientsrevealed prominent alveolar septal inflammation,diffuse alveolar damage, interstitial fibrosis, andhoneycombing. Korean investigators reported sixpatents with rapidly progressive, ultimately fatal,OP from a cohort of 45 patients with OP. 228 Becauseof the high mortality rate associated with rapidlyprogressive COP, this author recommends IV pulsemethylprednisolone (1,000 mg daily for 3 days),following by high-dose prednisone (1 mg/kg/dor equivalent, with a gradual taper). The durationof therapy may be relatively brief once a clinicalresponse has been achieved. For CS-recalcitrantpatients, cyclophosphamide or azathioprineshould be added.References1. Tazi A, Soler P, Hance AJ. Adult pulmonary Langerhans’cell histiocytosis. Thorax 2000; 55:405–4162. Tazi A. Adult pulmonary Langerhans’ cell histiocytosis.Eur Respir J 2006; 27:1272–12853. Vassallo R, Ryu JH, Colby TV, et al. <strong>Pulmonary</strong>Langerhans’-cell histiocytosis. N Engl J Med 2000;342:1969–19784. Friedman PJ, Liebow AA, Sokoloff J. Eosinophilicgranuloma of lung: clinical aspects of primaryhistiocytosis in the adult. <strong>Medicine</strong> (Baltimore) 1981;60:385–3965. Basset F, Corrin B, Spencer H, et al. <strong>Pulmonary</strong> histiocytosisX. Am Rev Respir Dis 1978; 118:811–8206. Colby TV, Lombard C. Histiocytosis X in the lung.Hum Pathol 1983; 14:847–8567. Travis WD, Borok Z, Roum JH, et al. <strong>Pulmonary</strong>Langerhans cell granulomatosis (histiocytosis X). Aclinicopathologic study of 48 cases. Am J Surg Pathol1993; 17:971–9868. Soler P, Bergeron A, Kambouchner M, et al. Is highresolutioncomputed tomography a reliable tool topredict the histopathological activity of pulmonaryLangerhans cell histiocytosis? Am J Respir Crit CareMed 2000; 162:264–2709. Ladisch S, Gadner H. Treatment of Langerhans cellhistiocytosis—evolution and current approaches. BrJ Cancer Suppl 1994; 23:S41–S4610. Willman CL, Busque L, Griffith BB, et al. Langerhans’-cellhistiocytosis (histiocytosis X)–a clonal proliferativedisease. N Engl J Med 1994; 331:154–16011. Yousem SA, Colby TV, Chen YY, et al. <strong>Pulmonary</strong>Langerhans’ cell histiocytosis: molecular analysisof clonality. Am J Surg Pathol 2001; 25:630–63612. Brabencova E, Tazi A, Lorenzato M, et al. Langerhanscells in Langerhans cell granulomatosis arenot actively proliferating cells. Am J Pathol 1998;152:1143–114913. Delobbe A, Durieu J, Duhamel A, et al. Determinantsof survival in pulmonary Langerhans’ cell granulomatosis(histiocytosis X). Groupe d’Etude en PathologieInterstitielle de la Societe de Pathologie Thoraciquedu Nord. Eur Respir J 1996; 9:2002–200614. Vassallo R, Jensen EA, Colby TV, et al. The overlapbetween respiratory bronchiolitis and desquamativeinterstitial pneumonia in pulmonary Langerhanscell histiocytosis: high-resolution CT, histologic,and functional correlations. Chest 2003; 124:1199–120515. Moore AD, Godwin JD, Muller NL, et al. <strong>Pulmonary</strong>histiocytosis X: comparison of radiographic and CTfindings. Radiology 1989; 172:249–25416. Selman M, Carillo G, Gaxiola M, et al. <strong>Pulmonary</strong>histiocytosis X (Eosinophilic granuloma): clinicalbehavior, pathogenesis, and therapeutic strategiesof an unusual interstitial lung disease. Clin PulmMed 1996; 3:191–19817. Mendez JL, Nadrous HF, Vassallo R, et al. Pneumothoraxin pulmonary Langerhans cell histiocytosis.Chest 2004; 125:1028–1032<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 609ACC-PULM-09-0302-036.indd 6097/10/09 10:24:27 PM


18. Fartoukh M, Humbert M, Capron F, et al. Severepulmonary hypertension in histiocytosis X. Am JRespir Crit Care Med 2000; 161:216–22319. Hamada K, Teramoto S, Narita N, et al. <strong>Pulmonary</strong>veno-occlusive disease in pulmonary Langerhans’cell granulomatosis. Eur Respir J 2000; 15:421–42320. Schonfeld N, Frank W, Wenig S, et al. Clinical andradiologic features, lung function and therapeuticresults in pulmonary histiocytosis X. Respiration1993; 60:38–4421. Vassallo R, Ryu JH, Schroeder DR, et al. Clinical outcomesof pulmonary Langerhans’-cell histiocytosisin adults. N Engl J Med 2002; 346:484–49022. Sadoun D, Vaylet F, Valeyre D, et al. Bronchogeniccarcinoma in patients with pulmonary histiocytosisX. Chest 1992; 101:1610–161323. Egeler RM, Neglia JP, Puccetti DM, et al. Associationof Langerhans cell histiocytosis with malignantneoplasms. Cancer 1993; 71:865–87324. Tomashefski JF, Khiyami A, Kleinerman J. Neoplasmsassociated with pulmonary eosinophilicgranuloma. Arch Pathol Lab Med 1991; 115:499–50625. Neumann MP, Frizzera G. The coexistence of Langerhans’cell granulomatosis and malignant lymphomamay take different forms: report of seven caseswith a review of the literature. Hum Pathol 1986;17:1060–106526. Brambilla E, Fontaine E, Pison CM, et al. <strong>Pulmonary</strong>histiocytosis X with mediastinal lymph node involvement.Am Rev Respir Dis 1990; 142:1216–121827. Brauner MW, Grenier P, Tijani K, et al. <strong>Pulmonary</strong>Langerhans cell histiocytosis: evolution of lesionson CT scans. Radiology 1997; 204:497–50228. Hidalgo A, Franquet T, Gimenez A, et al. Smokingrelatedinterstitial lung diseases: radiologic-pathologiccorrelation. Eur Radiol 2006; 16:2463–247029. Koyama M, Johkoh T, Honda O, et al. Chronic cysticlung disease: diagnostic accuracy of high-resolutionCT in 92 patients. AJR Am J Roentgenol 2003;180:827–83530. Sundar KM, Gosselin MV, Chung HL, et al. <strong>Pulmonary</strong>Langerhans cell histiocytosis: emergingconcepts in pathobiology, radiology, and clinicalevolution of disease. Chest 2003; 123:1673–168331. Crausman RS, Jennings CA, Tuder RM, et al. <strong>Pulmonary</strong>histiocytosis X: pulmonary function andexercise pathophysiology. Am J Respir Crit Care Med1996; 153:426–43532. Canuet M, Kessler R, Jeung MY, et al. Correlationbetween high-resolution computed tomographyfindings and lung function in pulmonary Langerhanscell histiocytosis. Respiration 2007; 74:640–64633. Wang CW, Colby TV. Histiocytic lesions and proliferationsin the lung. Semin Diagn Pathol 2007;24:162–18234. Vassallo R, Patel R, Aubry MC. Langerhans cell histiocytosis.In: Lynch JP III, ed. Interstitial pulmonaryand bronchiolar disorders (Vol. 227). New York:InformaUSA; 2008; 733–74535. Leslie KO, Gruden JF, Parish JM, et al. Transbronchialbiopsy interpretation in the patient with diffuse parenchymallung disease. Arch Pathol Lab Med 2007;131:407–42336. Chollet S, Soler P, Dournovo P, et al. Diagnosis ofpulmonary histiocytosis X by immunodetection ofLangerhans cells in bronchoalveolar lavage fluid.Am J Pathol 1984; 115:225–23237. Emile JF, Wechsler J, Brousse N, et al. Langerhans’cell histiocytosis: definitive diagnosis with the useof monoclonal antibody O10 on routinely paraffin-embeddedsamples. Am J Surg Pathol 1995;19:636–64138. Tazi A, Bonay M, Grandsaigne M, et al. Surfacephenotype of Langerhans cells and lymphocytesin granulomatous lesions from patients with pulmonaryhistiocytosis X. Am Rev Respir Dis 1993;147:1531–153639. Jaksits S, Kriehuber E, Charbonnier AS, et al.CD34 cell-derived CD14 precursor cells developinto Langerhans cells in a TGF-beta 1-dependentmanner. J Immunol 1999; 163:4869–487740. Banchereau J, Steinman RM. Dendritic cells and thecontrol of immunity. Nature 1998; 392:245–25241. Youkeles LH, Grizzanti JN, Liao Z, et al. Decreasedtobacco-glycoprotein-induced lymphocyte proliferationin vitro in pulmonary eosinophilicgranuloma. Am J Respir Crit Care Med 1995;151:145–15042. Churg A, Tai H, Coulthard T, et al. Cigarette smokedrives small airway remodeling by induction ofgrowth factors in the airway wall. Am J Respir CritCare Med 2006; 174:1327–133443. Casolaro MA, Bernaudin JF, Saltini C, et al. Accumulationof Langerhans’ cells on the epithelial surfaceof the lower respiratory tract in normal subjects inassociation with cigarette smoking. Am Rev RespirDis 1988; 137:406–41144. Aguayo SM, King TE Jr., Waldron JA Jr., et al.Increased pulmonary neuroendocrine cells withbombesin-like immunoreactivity in adult patients610 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 6107/10/09 10:24:27 PM


with eosinophilic granuloma. J Clin Invest 1990;86:838–84445. Zeid NA, Muller HK. Tobacco smoke induced lunggranulomas and tumors: association with pulmonaryLangerhans cells. Pathology 1995; 27:247–25446. Wells AU, Nicholson AG, Hansell DM. Challengesin pulmonary fibrosis. 4: smoking-induced diffuseinterstitial lung diseases. Thorax 2007; 62:904–91047. Howarth DM, Gilchrist GS, Mullan BP, et al. Langerhanscell histiocytosis: diagnosis, natural history,management, and outcome. Cancer 1999; 85:2278–229048. Chaowalit N, Pellikka PA, Decker PA, et al. Echocardiographicand clinical characteristics of pulmonaryhypertension complicating pulmonaryLangerhans cell histiocytosis. Mayo Clin Proc 2004;79:1269–127549. Gadner H, Grois N, Arico M, et al. A randomizedtrial of treatment for multisystem Langerhans’ cellhistiocytosis. J Pediatr 2001; 138:728–73450. Giona F, Caruso R, Testi AM, et al. Langerhans’ cellhistiocytosis in adults: a clinical and therapeuticanalysis of 11 patients from a single institution.Cancer 1997; 80:1786–179151. Saven A, Burian C. Cladribine activity in adultLangerhans-cell histiocytosis. Blood 1999; 93:4125–413052. Zeller B, Storm-Mathisen I, Smevik B, et al.Multi system Langerhans-cell histiocytosis withlife-threatening pulmonary involvement—goodresponse to cyclosporine A. Med Pediatr Oncol 2000;35:438–44253. Braier J, Chantada G, Rosso D, et al. Langerhans cellhistiocytosis: retrospective evaluation of 123 patientsat a single institution. Pediatr Hematol Oncol 1999;16:377–38554. Minkov M, Grois N, Broadbent V, et al. CyclosporineA therapy for multisystem Langerhans cell histiocytosis.Med Pediatr Oncol 1999; 33:482–48555. Choi SW, Bangaru BS, Wu CD, et al. Gastrointestinalinvolvement in disseminated Langerhans cell histiocytosis(LCH) with durable complete response to2-chlorodeoxyadenosine and high-dose cytarabine.J Pediatr Hematol Oncol 2003; 25:503–50656. Goh NS, McDonald CE, MacGregor DP, et al. Successfultreatment of Langerhans cell histiocytosiswith 2-chlorodeoxyadenosine. Respirology 2003;8:91–9457. Saven A, Foon KA, Piro LD. 2-Chlorodeoxyadenosine-inducedcomplete remissions inLangerhans-cell histiocytosis. Ann Intern Med 1994;121:430–43258. Pardanani A, Phyliky RL, Li CY, et al. 2-Chlorodeoxyadenosinetherapy for disseminated Langerhanscell histiocytosis. Mayo Clin Proc 2003; 78:301–30659. Aerni MR, Christine Aubry M, et al. Completeremission of nodular pulmonary Langerhans cellhistiocytosis lesions induced by 2-chlorodeoxyadenosinein a non-smoker. Respir Med 2008;102:316–31960. Hirose M, Saito S, Yoshimoto T, et al. Interleukin-2therapy of Langerhans cell histiocytosis. Acta Paediatr1995; 84:1204–120661. Henter JI, Karlen J, Calming U, et al. Successfultreatment of Langerhans’-cell histiocytosis withetanercept. N Engl J Med 2001; 345:1577–157862. Mogulkoc N, Veral A, Bishop PW, et al. <strong>Pulmonary</strong>Langerhans’ cell histiocytosis: radiologic resolutionfollowing smoking cessation. Chest 1999; 115:1452–145563. Arico M, Girschikofsky M, Genereau T, et al. Langerhanscell histiocytosis in adults: report from theInternational Registry of the Histiocyte Society. EurJ Cancer 2003; 39:2341–234864. Etienne B, Bertocchi M, Gamondes JP, et al. Relapsingpulmonary Langerhans cell histiocytosis after lungtransplantation. Am J Respir Crit Care Med 1998;157:288–29165. Dauriat G, Mal H, Thabut G, et al. Lung transplantationfor pulmonary Langerhans’ cell histiocytosis: a multicenteranalysis. Transplantation 2006; 81:746–75066. Hertz MI, Boucek MM, Deng MC, et al. Scientific Registryof the International Society for Heart and LungTransplantation: introduction to the 2005 annualreports. J Heart Lung Transplant 2005; 24:939–94467. Habbib SB, Congelton J, Carr D, et al. Recurrenceof Langerhan’s cell histiocytosis following bilaterallung transplantation. Thorax 1998; 53:323–32568. Gabbay E, Dark JH, Ashcroft T, et al. Recurrence ofLangerhans’ cell granulomatosis following lungtransplantation. Thorax 1998; 53:326–32769. Lynch JP 3rd, Saggar R, Weigt SS, et al. Overviewof lung transplantation and criteria for selectionof candidates. Semin Respir Crit Care Med 2006;27:441–46970. Taveira-DaSilva AM, Steagall WK, Moss J.Lymphangioleiomyomatosis Cancer Control 2006;13:276–28571. Urban T, Lazor R, Lacronique J, et al. <strong>Pulmonary</strong>lymphangioleiomyomatosis: a study of 69 patients.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 611ACC-PULM-09-0302-036.indd 6117/10/09 10:24:27 PM


Groupe d’Etudes et de Recherche sur les Maladies“Orphelines” Pulmonaires (GERM“O”P). <strong>Medicine</strong>(Baltimore) 1999; 78:321–33772. Ryu JH, Moss J, Beck GJ, et al. The NHLBI LymphangioleiomyomatosisRegistry: characteristics of 230patients at enrollment. Am J Respir Crit Care Med2006; 173:105–11173. Johnson SR. Lymphangioleiomyomatosis. EurRespir J 2006; 27:1056–106574. McCormack FX. Lymphangioleiomyomatosis: aclinical update. Chest 2008; 133:507–51675. Johnson S. Rare diseases. 1: lymphangioleiomyomatosis:clinical features, management and basicmechanisms. Thorax 1999; 54:254–26476. Chu SC, Horiba K, Usuki J, et al. Comprehensiveevaluation of 35 patients with lymphangioleiomyomatosis.Chest 1999; 115:1041–105277. Kitaichi M, Nishimura K, Itoh H, et al. <strong>Pulmonary</strong>lymphangioleiomyomatosis: a report of 46 patientsincluding a clinicopathologic study of prognostic factors.Am J Respir Crit Care Med 1995; 151:527–53378. Taylor JR, Ryu J, Colby TV, et al. Lymphangioleiomyomatosis:clinical course in 32 patients. N Engl JMed 1990; 323:1254–126079. Corrin B, Liebow AA, Friedman PJ. <strong>Pulmonary</strong>lymphangiomyomatosis: a review. Am J Pathol 1975;79:348–38280. National Heart, Lung, and Blood Institute. WorkshopSummary. Report of workshop on lymphangioleiomyomatosis.Am J Respir Crit Care Med 1999;159:679–68381. Silverstein EF, Ellis K, Wolff M, et al. <strong>Pulmonary</strong>lymphangiomyomatosis. AJR Am J RoentgenolRadium Ther Nucl Med 1974; 120:832–85082. Johnson SR, Tattersfield AE. Decline in lung functionin lymphangioleiomyomatosis: relation to menopauseand progesterone treatment. Am J Respir CritCare Med 1999; 160:628–63383. Oh YM, Mo EK, Jang SH, et al. <strong>Pulmonary</strong> lymphangioleiomyomatosisin Korea. Thorax 1999; 54:618–62184. Kim NR, Chung MP, Park CK, et al. <strong>Pulmonary</strong>lymphangioleiomyomatosis and multiple hepatic angiomyolipomasin a man. Pathol Int 2003; 53:231–23585. Miyake M, Tateishi U, Maeda T, et al. <strong>Pulmonary</strong>lymphangioleiomyomatosis in a male patientwith tuberous sclerosis complex. Radiat Med 2005;23:525–52786. Aubry MC, Myers JL, Ryu JH, et al. <strong>Pulmonary</strong>lymphangioleiomyomatosis in a man. Am J RespirCrit Care Med 2000; 162:749–75287. Schiavina M, Di Scioscio V, Contini P, et al. <strong>Pulmonary</strong>lymphangioleiomyomatosis in a karyotypicallynormal man without tuberous sclerosis complex. AmJ Respir Crit Care Med 2007; 176:96–9888. Almoosa KF, Ryu JH, Mendez J, et al. Managementof pneumothorax in lymphangioleiomyomatosis:effects on recurrence and lung transplantation complications.Chest 2006; 129:1274–128189. McCormack FX. Lymphangioleiomyomatosis. In:Lynch JP III, ed. Interstitial pulmonary and bronchiolardisorders (Vol. 227). New York: InformaUSA,2008; 747–76890. Castro M, Shepherd CW, Gomez MR, et al. <strong>Pulmonary</strong>tuberous sclerosis. Chest 1995; 107:189–19591. Aberle DR, Hansell DM, Brown K, et al. Lymphangiomyomatosis:CT, chest radiographic, andfunctional correlations. Radiology 1990; 176:381–38792. Ryu JH, Doerr CH, Fisher SD, et al. Chylothorax inlymphangioleiomyomatosis. Chest 2003; 123:623–62793. Matsui K, Beasley MB, Nelson WK, et al. Prognosticsignificance of pulmonary lymphangioleiomyomatosishistologic score. Am J Surg Pathol 2001;25:479–48494. Avila NA, Kelly JA, Dwyer AJ, et al. Lymphangioleiomyomatosis:correlation of qualitative and quantitativethin-section CT with pulmonary functiontests and assessment of dependence on pleurodesis.Radiology 2002; 223:189–19795. Crausman RS, Lynch DA, Mortenson RL, et al.Quantitative CT predicts the severity of physiologicdysfunction in patients with lymphangioleiomyomatosis.Chest 1996; 109:131–13796. Woodring JH, Howard RS 2nd, Johnson MV. Massivelow-attenuation mediastinal, retroperitoneal,and pelvic lymphadenopathy on CT from lymphangioleiomyomatosis:case report. Clin Imaging 1994;18:7–1197. Avila NA, Kelly JA, Chu SC, et al. Lymphangioleiomyomatosis:abdominopelvic CT and US findings.Radiology 2000; 216:147–15398. Johkoh T, Muller NL, Pickford HA, et al. Lymphocyticinterstitial pneumonia: thin-section CT findingsin 22 patients. Radiology 1999; 212:567–57299. Jeong YJ, Lee KS, Chung MP, et al. Amyloidosis andlymphoproliferative disease in Sjögren syndrome:thin-section computed tomography findings andhistopathologic comparisons. J Comput AssistTomogr 2004; 28:776–781612 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 6127/10/09 10:24:27 PM


100. Silva CI, Churg A, Muller NL. Hypersensitivitypneumonitis: spectrum of high-resolution CT andpathologic findings. AJR Am J Roentgenol 2007;188:334–344101. Colombat M, Stern M, Groussard O, et al. <strong>Pulmonary</strong>cystic disorder related to light chain depositiondisease. Am J Respir Crit Care Med 2006;173:777–780102. Ayo DS, Aughenbaugh GL, Yi ES, et al. Cystic lungdisease in Birt-Hogg-Dube syndrome. Chest 2007;132:679–684103. Avila NA, Chen CC, Chu SC, et al. <strong>Pulmonary</strong>lymphangioleiomyomatosis: correlation of ventilation-perfusionscintigraphy, chest radiography, andCT with pulmonary function tests. Radiology 2000;214:441–446104. Taveira-DaSilva AM, Hedin C, Stylianou MP, et al.Reversible airflow obstruction, proliferation ofabnormal smooth muscle cells, and impairment ofgas exchange as predictors of outcome in lymphangioleiomyomatosis.Am J Respir Crit Care Med2001; 164:1072–1076105. Crausman RS, Jennings CA, Mortenson RL,et al. Lymphangioleiomyomatosis: the pathophysiologyof diminished exercise capacity. Am J RespirCrit Care Med 1996; 153:1368–1376106. Muller NL, Chiles C, Kullnig P. <strong>Pulmonary</strong> lymphangiomyomatosis:correlation of CT with radiographicand functional findings. Radiology 1990;175:335–9107. Taveira-DaSilva AM, Stylianou MP, Hedin CJ,et al. Maximal oxygen uptake and severity of diseasein lymphangioleiomyomatosis. Am J RespirCrit Care Med 2003; 168:1427–31108. Taveira-Dasilva AM, Stylianou MP, Hedin CJ,et al. Decline in lung function in patients withlymphangioleiomyomatosis treated with or withoutprogesterone. Chest 2004; 126:1867–18874109. Steagall WK, Glasgow CG, Hathaway OM, et al.Genetic and morphologic determinants of pneumothoraxin lymphangioleiomyomatosis. Am J PhysiolLung Cell Mol Physiol 2007; 293:L800–L808110. Boehler A. Lung transplantation for cystic lungdiseases: lymphangioleiomyomatosis, histiocytosisX, and sarcoidosis. Semin Respir Crit Care Med2001; 22:509–516111. Taveira-DaSilva AM, Hathaway OM, Sachdev V,et al. <strong>Pulmonary</strong> artery pressure in lymphangioleiomyomatosis:an echocardiographic study. Chest2007; 132:1573–1578112. Matsui K, Tatsuguchi A, Valencia J, et al. Extrapulmonarylymphangioleiomyomatosis (LAM):clinicopathologic features in 22 cases. Hum Pathol2000; 31:1242–1248113. Bernstein SM, Newell JD Jr., Adamczyk D, et al.How common are renal angiomyolipomas inpatients with pulmonary lymphangiomyomatosis?Am J Respir Crit Care Med 1995; 152:2138–2143114. Chan JK, Tsang WY, Pau MY, et al. Lymphangiomyomatosisand angiomyolipoma: closely relatedentities characterized by hamartomatous proliferationof HMB-45-positive smooth muscle. Histopathology1993; 22:445–455115. Matsui K, Takeda K, Yu ZX, et al. Role for activationof matrix metalloproteinases in the pathogenesisof pulmonary lymphangioleiomyomatosis. ArchPathol Lab Med 2000; 124:267–275116. Matsui K, Takeda K, Yu ZX, et al. Downregulationof estrogen and progesterone receptors in theabnormal smooth muscle cells in pulmonarylymphangioleiomyomatosis following therapy: animmunohistochemical study. Am J Respir Crit CareMed 2000; 161:1002–1009117. Kalassian KG, Doyle R, Kao P, et al. Lymphangioleiomyomatosis:new insights. Am J Respir CritCare Med 1997; 155:1183–1186118. Young LR, Inoue Y, McCormack FX. Diagnosticpotential of serum VEGF-D for lymphangioleiomyomatosis.N Engl J Med 2008; 358:199–200119. Glasgow CG, Taveira-Dasilva AM, Darling TN, etal. Lymphatic involvement in lymphangioleiomyomatosis.Ann N Y Acad Sci 2008; 1131:206–214120. Kumasaka T, Seyama K, Mitani K, et al. Lymphangiogenesis-mediatedshedding of LAM cell clustersas a mechanism for dissemination in lymphangioleiomyomatosis.Am J Surg Pathol 2005; 29:1356–1366121. Kumasaka T, Seyama K, Mitani K, et al. Lymphangiogenesisin lymphangioleiomyomatosis: itsimplication in the progression of lymphangioleiomyomatosis.Am J Surg Pathol 2004; 28:1007–1016122. Seyama K, Kumasaka T, Souma S, et al. Vascularendothelial growth factor-D is increased in serumof patients with lymphangioleiomyomatosis. LymphatRes Biol 2006; 4:143–152123. Logginidou H, Ao X, Russo I, Henske EP. Frequentestrogen and progesterone receptor immunoreactivityin renal angiomyolipomas from women withpulmonary lymphangioleiomyomatosis. Chest2000; 117:25–30<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 613ACC-PULM-09-0302-036.indd 6137/10/09 10:24:28 PM


124. Henske EP, Ao X, Short MP, et al. Frequent progesteronereceptor immunoreactivity in tuberoussclerosis-associated renal angiomyolipomas. ModPathol 1998; 11:665–668125. Moss J, DeCastro R, Patronas NJ, et al A. Meningiomasin lymphangioleiomyomatosis. JAMA 2001;286:1879–1881126. Avila NA, Dwyer AJ, Rabel A, et al. Sporadiclymphangioleiomyomatosis and tuberous sclerosiscomplex with lymphangioleiomyomatosis: comparisonof CT features. Radiology 2007; 242:277–285127. Torres VE, Bjornsson J, King BF, et al. Extrapulmonarylymphangioleiomyomatosis and lymphangiomatouscysts in tuberous sclerosis complex. MayoClin Proc 1995; 70:641–618128. Goncharova EA, Krymskaya VP. <strong>Pulmonary</strong>lymphangioleiomyomatosis (LAM): progress andcurrent challenges. J Cell Biochem 2008; 103:369–382129. Roach ES, Smith M, Huttenlocher P, et al. Diagnosticcriteria: tuberous sclerosis complex: report of theDiagnostic Criteria Committee of the NationalTuberous Sclerosis Association. J Child Neurol1992; 7:221–224130. Costello LC, Hartman TE, Ryu JH. High frequencyof pulmonary lymphangioleiomyomatosis inwomen with tuberous sclerosis complex. Mayo ClinProc 2000; 75:591–594131. Sancak O, Nellist M, Goedbloed M, et al. Mutationalanalysis of the TSC1 and TSC2 genes in a diagnosticsetting: genotype–phenotype correlations andcomparison of diagnostic DNA techniques inTuberous Sclerosis Complex. Eur J Hum Genet2005; 13:731–741132. Smolarek TA, Wessner LL, McCormack FX, et al.Evidence that lymphangiomyomatosis is causedby TSC2 mutations: chromosome 16p13 loss ofheterozygosity in angiomyolipomas and lymphnodes from women with lymphangiomyomatosis.Am J Hum Genet 1998; 62:810–815133. Krymskaya VP. Smooth muscle-like cells in pulmonarylymphangioleiomyomatosis. Proc Am ThoracSoc 2008; 5:119–126134. Juvet SC, Hwang D, Downey GP. Rare lung diseasesI–Lymphangioleiomyomatosis. Can Respir J 2006;13:375–380135. Crooks DM, Pacheco-Rodriguez G, DeCastro RM,et al. Molecular and genetic analysis of disseminatedneoplastic cells in lymphangioleiomyomatosis.Proc Natl Acad Sci U S A 2004; 101:17462–27467136. Henske EP. Metastasis of benign tumor cells intuberous sclerosis complex. Genes ChromosomesCancer 2003; 38:376–381137. Johnson SR, Whale CI, Hubbard RB, et al. Survivaland disease progression in UK patients with lymphangioleiomyomatosis.Thorax 2004; 59:800–803138. Johnson SR, Tattersfield AE. Clinical experience oflymphangioleiomyomatosis in the UK. Thorax2000; 55:1052–1057139. Rossi GA, Balbi B, Oddera S, et al. Responseto treatment with an analog of the luteinizinghormone-releasing hormone in a patient withpulmonary lymphangioleiomyomatosis. Am RevRespir Dis 1991; 143:174–176140. Radermecker M, Broux R, Corhay JL, et al. Failureof buserelin-induced medical castration to controlpulmonary lymphangiomyomatosis in twopatients. Chest 1992; 101:1724–1726141. Harari S, Cassandro R, Chiodini J, et al. Effect of agonadotrophin-releasing hormone analogue onlung function in lymphangioleiomyomatosis. Chest2008; 133:448–454142. Klein M, Krieger O, Ruckser R, et al. Treatment oflymphangioleiomyomatosis by ovariectomy, interferonalpha 2b and tamoxifen–a case report. ArchGynecol Obstet 1992; 252:99–102143. DeBove P, Murris-Espin M, Buscail L, et al. Somatostatinreceptors in pulmonary lymphangioleiomyomatosis:therapeutic relevance. Eur Respir J1997; 10(suppl 25):297S.144. Schiavina M, Contini P, Fabiani A, et al. Efficacy ofhormonal manipulation in lymphangioleiomyomatosis:a 20-year-experience in 36 patients. SarcoidosisVasc Diffuse Lung Dis 2007; 24:39–50145. Krymskaya VP. Tumour suppressors hamartin andtuberin: intracellular signalling. Cell Signal 2003;15:729–739146. Wienecke R, Fackler I, Linsenmaier U, et al. Antitumoralactivity of rapamycin in renal angiomyolipomaassociated with tuberous sclerosis complex.Am J Kidney Dis 2006; 48:e27–e9147. Bissler JJ, McCormack FX, Young LR, et al. Sirolimusfor angiomyolipoma in tuberous sclerosis complexor lymphangioleiomyomatosis. N Engl J Med 2008;358:140–151148. Taille C, Debray MP, Crestani B. Sirolimus treatmentfor pulmonary lymphangioleiomyomatosis. AnnIntern Med 2007; 146:687–688614 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 6147/10/09 10:24:28 PM


149. Ohara T, Oto T, Miyoshi K, et al. Sirolimus amelioratedpost lung transplant chylothorax in lymphangioleiomyomatosis.Ann Thorac Surg 2008;86:e7–e8150. Sugimoto R, Nakao A, Yamane M, et al. Sirolimusamelioration of clinical symptoms of recurrentlymphangioleiomyomatosis after living-donorlobar lung transplantation. J Heart Lung Transplant2008; 27:921–924151. Morton JM, McLean C, Booth SS, et al. Regressionof pulmonary lymphangioleiomyomatosis (PLAM)-associated retroperitoneal angiomyolipoma postlungtransplantation with rapamycin treatment.J Heart Lung Transplant 2008; 27:462–465152. Chen F, Omasa M, Kondo N, et al. Sirolimus treatmentfor recurrent lymphangioleiomyomatosisafter lung transplantation. Ann Thorac Surg 2009;87:e6–e7153. Kalomenidis I. Octreotide and chylothorax. CurrOpin Pulm Med 2006; 12:264–267154. Fremont RD, Milstone AP, Light RW, et al. Chylothoracesafter lung transplantation for lymphangioleiomyomatosis:review of the literature and utilizationof a pleurovenous shunt. J Heart LungTransplant 2007; 26:953–955155. Kpodonu J, Massad MG, Chaer RA, et al. The USexperience with lung transplantation for pulmonarylymphangioleiomyomatosis. J Heart LungTransplant 2005; 24:1247–1253156. Maurer JR, Ryu J, Beck G, et al. Lung transplantationin the management of patients with lymphangioleiomyomatosis:baseline data from the NHLBILAM Registry. J Heart Lung Transplant 2007;26:1293–1299157. Benden C, Rea F, Behr J, et al. Lung transplantationfor lymphangioleiomyomatosis: the Europeanexperience. J Heart Lung Transplant 2009; 28:1–7158. Reynaud-Gaubert M, Mornex JF, Mal H, et al.Lung transplantation for lymphangioleiomyomatosis:the French experience. Transplantation 2008;86:515–520159. Trulock EP, Christie JD, Edwards LB, et al. Registryof the International Society for Heart and LungTransplantation: twenty-fourth official adult lungand heart-lung transplantation report-2007. J HeartLung Transplant 2007; 26:782–795160. Bittmann I, Rolf B, Amann G, Lohrs U. Recurrenceof lymphangioleiomyomatosis after single lungtransplantation: new insights into pathogenesis.Hum Pathol 2003; 34:95–98161. Karbowniczek M, Astrinidis A, Balsara BR, et al.Recurrent lymphangiomyomatosis after transplantation:genetic analyses reveal a metastatic mechanism.Am J Respir Crit Care Med 2003; 167:976–982162. Alasaly K, Muller N, Ostrow DN, et al. Cryptogenicorganizing pneumonia: a report of 25 cases and areview of the literature. <strong>Medicine</strong> (Baltimore) 1995;74:201–211163. Cordier JF. Organising pneumonia. Thorax 2000;55:318–328164. Ryu JH, Myers JL, Swensen SJ. Bronchiolar disorders.Am J Respir Crit Care Med 2003; 168:1277–1292165. Cordier JF. Cryptogenic organising pneumonia.Eur Respir J 2006; 28:422–446166. Epler GR, Colby TV, McLoud TC, et al. Bronchiolitisobliterans organizing pneumonia. N Engl J Med1985; 312:152–158167. Ryu JH. Classification and approach to bronchiolardiseases. Curr Opin Pulm Med 2006; 12:145–151168. White E, Tazelaar H, Lynch JPI. Bronchiolar complicationsof connective tissue diseases. SeminRespir Crit Care Med 2003; 24:543–564169. Cavallasca JA, Caubet M, Helling CA, et al. Cryptogenicorganizing pneumonia (COP), as presentationof rheumatoid arthritis. Rheumatol Int 2008;29:99–101170. Ioannou S, Toya SP, Tomos P, et al. Cryptogenicorganizing pneumonia associated with primarySjögren’s syndrome. Rheumatol Int 2008; 28:1053–1055171. Baron FA, Hermanne JP, Dowlati A, et al. Bronchiolitisobliterans organizing pneumonia andulcerative colitis after allogeneic bone marrowtransplantation. Bone Marrow Transplant 1998; 21:951–954172. Uner AH, Rozum-Slota B, Katzenstein AL. Bronchiolitisobliterans-organizing pneumonia (BOOP)-like variant of Wegener’s granulomatosis: aclinicopathologic study of 16 cases. Am J SurgPathol 1996; 20:794–801173. Wachowski O, Demirakca S, Muller KM, et al.Mycoplasma pneumoniae associated organisingpneumonia in a 10 year old boy. Arch Dis Child2003; 88:270–272174. Stover DE, Mangino D. Macrolides: a treatmentalternative for bronchiolitis obliterans organizingpneumonia? Chest 2005; 128:3611–3617175. Radzikowska E, Szczepulska E, Chabowski M, et al.Organising pneumonia caused by transtuzumab<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 615ACC-PULM-09-0302-036.indd 6157/10/09 10:24:28 PM


(Herceptin) therapy for breast cancer. Eur Respir J2003; 21:552–555176. Garrido M, O’Brien A, Gonzalez S, et al. Cryptogenicorganizing pneumonitis during oxiliplatinchemotherapy for colorectal cancer: case report.Chest 2007; 132:1997–1999177. Mokhtari M, Bach PB, Tietjen PA, et al Bronchiolitisobliterans organizing pneumonia in cancer: a caseseries. Respir Med 2002; 96:280–286178. Martinez-Gallo M, Puy C, Ruiz-Hernandez R, et al.Severe and recurrent episodes of bronchiolitisobliterans organising pneumonia associated withindolent CD4 CD8 T-cell leukaemia. Eur RespirJ 2008; 31:1368–1372179. Barroso E, Hernandez L, Gil J, et al. Idiopathicorganizing pneumonia: a relapsing disease. 19years of experience in a hospital setting. Respiration2007; 74:624–631180. Khalid M, Al Saghir A, Saleemi S, et al. Azithromycinin bronchiolitis obliterans complicating bonemarrow transplantation: a preliminary study. EurRespir J 2005; 25:490–493181. Freudenberger TD, Madtes DK, Curtis JR, et al.Association between acute and chronic graft-versus-hostdisease and bronchiolitis obliterans organizingpneumonia in recipients of hematopoieticstem cell transplants. Blood 2003; 102:3822–3828182. Ditschkowski M, Elmaagacli AH, Trenschel R, et al.T-cell depletion prevents from bronchiolitis obliteransand bronchiolitis obliterans with organizingpneumonia after allogeneic hematopoietic stem celltransplantation with related donors. Haematologica2007; 92:558–561183. Belperio JA, Weigt SS, Fishbein MC, et al. Chroniclung allograft rejection: mechanisms and therapy.Proc Am Thorac Soc 2009; 6:108–121184. Kreiss K, Gomaa A, Kullman G, et al. Clinical bronchiolitisobliterans in workers at a microwavepopcornplant. N Engl J Med 2002; 347:330–338185. Mann JM, Sha KK, Kline G, et al. World Trade Centerdyspnea: bronchiolitis obliterans with functionalimprovement: a case report. Am J Ind Med2005; 48:225–229186. Scott AI, Sharples LD, Stewart S. Bronchiolitisobliterans syndrome: risk factors and therapeuticstrategies. Drugs 2005; 65:761–771187. Davison AG, Heard BE, McAllister WA, et al. Cryptogenicorganizing pneumonitis. Q J Med 1983;52:382–394188. Katzenstein AL, Myers JL, Prophet WD, et al. Bronchiolitisobliterans and usual interstitial pneumonia:a comparative clinicopathologic study. Am JSurg Pathol 1986; 10:373–381189. Cordier JF, Loire R, Brune J. Idiopathic bronchiolitisobliterans organizing pneumonia: definition ofcharacteristic clinical profiles in a series of 16patients. Chest 1989; 96:999–1004190. Yamamoto M, Ina Y, Kitaichi M, et al. Clinical featuresof BOOP in Japan. Chest 1992; 102(1Suppl):21S–25S.191. Gudmundsson G, Sveinsson O, Isaksson HJ, et al.Epidemiology of organising pneumonia in Iceland.Thorax 2006; 61:805–808192. Arakawa H, Kurihara Y, Niimi H, et al. Bronchiolitisobliterans with organizing pneumonia versuschronic eosinophilic pneumonia: high-resolutionCT findings in 81 patients. AJR Am J Roentgenol2001; 176:1053–1058193. Akira M, Yamamoto S, Sakatani M. Bronchiolitisobliterans organizing pneumonia manifesting asmultiple large nodules or masses. AJR Am J Roentgenol1998; 170:291–295194. Fruchter O, Solomonov A, Guralnik L, et al. Anunusual radiographic manifestation of bronchiolitisobliterans organizing pneumonia. J ThoracImaging 2007; 22:263–264195. Haro M, Vizcaya M, Texido A, et al. Idiopathicbronchiolitis obliterans organizing pneumoniawith multiple cavitary lung nodules. Eur Respir J1995; 8:1975–1977196. Froudarakis M, Bouros D, Loire R, et al. BOOPpresenting with haemoptysis and multiple cavitarynodules. Eur Respir J 1995; 8:1972–1974197. Pipavath SJ, Lynch DA, Cool C, et al. Radiologicand pathologic features of bronchiolitis. AJR Am JRoentgenol 2005; 185(2):354–63198. Lee JS, Lynch DA, Sharma S, et al. Organizingpneumonia: prognostic implication of high-resolutioncomputed tomography features. J ComputAssist Tomogr 2003; 27:260–265199. Maldonado F, Daniels CE, Hoffman EA, et al. Focalorganizing pneumonia on surgical lung biopsy:causes, clinicoradiologic features, and outcomes.Chest 2007; 132:1579–1583200. Melloni G, Cremona G, Bandiera A, et al. Localizedorganizing pneumonia: report of 21 cases. AnnThorac Surg 2007; 83:1946–1951201. Tateishi U, Hasegawa T, Seki K, et al. Disease activityand 18F-FDG uptake in organising pneumonia:616 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 6167/10/09 10:24:28 PM


semi-quantitative evaluation using computedtomography and positron emission tomography.Eur J Nucl Med Mol Imaging 2006; 33:906–912202. Ryu JH, Myers JL, Capizzi SA, et al. Desquamativeinterstitial pneumonia and respiratory bronchiolitis-associatedinterstitial lung disease. Chest 2005;127:178–184203. Hausler M, Meilicke R, Biesterfeld S, et al. Bronchiolitisobliterans organizing pneumonia: a distinctpulmonary complication in cystic fibrosis. Respiration2000; 67:316–319204. Churg A, Myers J, Suarez T, et al. Airway-centeredinterstitial fibrosis: a distinct form of aggressivediffuse lung disease. Am J Surg Pathol 2004; 28:62–68205. Yousem SA, Dacic S. Idiopathic bronchiolocentricinterstitial pneumonia. Mod Pathol 2002; 15:1148–1153206. Mark EJ, Ruangchira-urai R. Bronchiolitis interstitialpneumonitis: a pathologic study of 31 lungbiopsies with features intermediate between bronchiolitisobliterans organizing pneumonia andusual interstitial pneumonitis, with clinical correlation.Ann Diagn Pathol 2008; 12:171–180207. Crestani B, Valeyre D, Roden S, et al. Bronchiolitisobliterans organizing pneumonia syndromeprimed by radiation therapy to the breast: theGroupe d’Etudes et de Recherche sur les MaladiesOrphelines Pulmonaires (GERM“O”P). Am J RespirCrit Care Med 1998; 158:1929–1935208. Poulou LS, Tsangaridou I, Filippoussis P, et al.Feasibility of CT-guided percutaneous needlebiopsy in early diagnosis of BOOP. CardiovascIntervent Radiol 2008; 31:1003–1007209. Forlani S, Ratta L, Bulgheroni A, et al. Cytokineprofile of broncho-alveolar lavage in BOOP andUIP. Sarcoidosis Vasc Diffuse Lung Dis 2002;19:47–53210. Lappi-Blanco E, Soini Y, Kinnula V, et al. VEGF andbFGF are highly expressed in intraluminal fibromyxoidlesions in bronchiolitis obliterans organizingpneumonia. J Pathol 2002; 196:220–227211. Choi KH, Lee HB, Jeong MY, et al. The role of matrixmetalloproteinase-9 and tissue inhibitor of metalloproteinase-1in cryptogenic organizing pneumonia.Chest 2002; 121:1478–1485212. Suga M, Iyonaga K, Okamoto T, et al. Characteristicelevation of matrix metalloproteinase activity inidiopathic interstitial pneumonias. Am J Respir CritCare Med 2000; 162:1949–1956213. Okada F, Ando Y, Honda K, et al. Comparison ofpulmonary CT findings and serum KL-6 levels inpatients with cryptogenic organizing pneumonia.Br J Radiol 2009; 82:212–218214. Ishii H, Mukae H, Kadota J, et al. High serum concentrationsof surfactant protein A in usual interstitialpneumonia compared with non-specificinterstitial pneumonia. Thorax 2003; 58:52–57215. Lazor R, Vandevenne A, Pelletier A, et al. Cryptogenicorganizing pneumonia: characteristics ofrelapses in a series of 48 patients. The Grouped’Etudes et de Recherche sur les Maladles “Orphelines”Pulmonaires (GERM“O”P). Am J Respir CritCare Med 2000; 162:571–577216. Costabel U, Guzman J, Teschler H. Bronchiolitisobliterans with organizing pneumonia: outcome.Thorax 1995; 50(suppl 1):559–564217. Watanabe K, Senju S, Wen FQ, et al. Factors relatedto the relapse of bronchiolitis obliterans organizingpneumonia. Chest 1998; 114:1599–1606218. Schlesinger C, Koss MN. The organizing pneumonias:an update and review. Curr Opin Pulm Med2005; 11:422–430219. Kobayashi I, Yamada M, Takahashi Y, et al. Interstitiallung disease associated with juvenile dermatomyositis:clinical features and efficacy of cyclosporinA. Rheumatology (Oxford) 2003; 42:371–374220. Laszlo A, Espolio Y, Auckenthaler A, et al. Azathioprineand low-dose corticosteroids for the treatmentof cryptogenic organizing pneumonia in anolder patient. J Am Geriatr Soc 2003; 51:433–434221. Verleden GM, Dupont LJ. Azithromycin therapyfor patients with bronchiolitis obliterans syndromeafter lung transplantation. Transplantation 2004;77:1465–1467222. Rubin BK, Henke MO. Immunomodulatory activityand effectiveness of macrolides in chronic airwaydisease. Chest 2004; 125(2 Suppl):70S–78S.223. Labro MT, Abdelghaffar H. Immunomodulation bymacrolide antibiotics. J Chemother 2001; 13:3–8224. Kastelik JA, Greenstone M, McGivern DV, et al.Cryptogenic organising pneumonia. Eur Respir J2006; 28:1291225. Ichikawa Y, Ninomiya H, Katsuki M, et al. Lowdose/long-termerythromycin for treatment ofbronchiolitis obliterans organizing pneumonia(BOOP). Kurume Med J 1993; 40:65–67226. Cohen AJ, King TE Jr., Downey GP. Rapidly progressivebronchiolitis obliterans with organizing pneumonia.Am J Respir Crit Care Med 1994; 149:1670–1675<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 617ACC-PULM-09-0302-036.indd 6177/10/09 10:24:29 PM


227. Perez de Llano LA, Soilan JL, Garcia Pais MJ, et al.Idiopathic bronchiolitis obliterans with organizingpneumonia presenting with adult respiratory distresssyndrome. Respir Med 1998; 92:884–886228. Chang J, Han J, Kim DW, et al. Bronchiolitis obliteransorganizing pneumonia: clinicopathologicreview of a series of 45 Korean patients includingrapidly progressive form. J Korean Med Sci 2002;17:179–186618 Rare Interstitial Lung Diseases (Lynch)ACC-PULM-09-0302-036.indd 6187/10/09 10:24:29 PM


Tuberculosis and Other Mycobacterial DiseasesDavid Ashkin, MDObjectives:• Become familiar with the epidemiology of tuberculosis• <strong>Review</strong> the pathogenesis and clinical presentation oftuberculosis• Address issues concerning the prevention of tuberculosis,including the diagnosis and treatment of tuberculosisinfection• <strong>Review</strong> issues concerning the diagnosis and treatment oftuberculosis disease• Outline pertinent topics concerning mycobacteria otherthan tuberculosis (eg, Mycobacterium avium complex andMycobacterium kansasii)Key words: multidrug-resistant tuberculosis; nontuberculousmycobacteria; tuberculosisTuberculosisEpidemiologyAt the turn of the 20th century, tuberculosis(TB) was one of the leading causes of death in theUnited States. After the discovery of effectivechemotherapy, the rate of TB significantlydecreased at an average rate of 8% per yearbetween 1953 and 1983. From 1985 to 1992, however,TB unexpectedly increased by approximately20%. Factors associated with the resurgenceincluded the HIV epidemic, increased immigrationof foreign-born persons from areas of high TBincidence, an increased number of medicallyunderserved persons (eg, homeless persons anddrug abusers), and probably most importantly, thedeterioration of the public health infrastructurefor the control of TB. Associated with this increasedincidence was an increase in the number of casesof drug-resistant TB.With improvements in TB control, the overallnumber of cases in the United States has decreasedduring the last 16 years, but due to the increasingnumber of TB cases worldwide and the large numberof international individuals arriving in theUnited States every year, TB remains a significantnational public health problem. In 2008, there werea total of 12,898 cases of TB in the nation, with arate of 4.2 per 100,000 persons, which is the lowestrate and total number of TB cases ever documentedsince national reporting began in the United States.The highest numbers of cases have been reportedin California, Florida, New York, and Texas, whichcollectively account for 49.2% of TB cases nationally.It is estimated that approximately 15 millionpeople are infected with TB in the United States,making for a large potential reservoir of disease inthe population. The infection rate among foreignbornindividuals in the United States is approximatelyseven times greater than that for theUS-born population. Although the number of casesin the United States has been decreasing, theproportion of cases in the foreign-born populationhas increased from 27% in 1992 to 59% in 2008. Inaddition, 82% of the reported multidrug-resistantcases in the United States occurred in foreign-bornindividuals.Worldwide, the rate of TB has not decreased asit has in the United States. According to the mostrecent statistics from the World Health Organization,the number of new cases of TB in 2007 wasapproximately 9.27 million worldwide, with a totalprevalence of 13.7 million cases of TB disease.Ninety-five percent of TB cases occur in developingcountries (especially in Africa and Asia) where HIVinfection is common, and resources are scarce orunavailable to ensure proper TB treatment. It isestimated that 1.8 million people died of the disease,with a fatality rate as high as 50% in somecountries with high rates of HIV coinfection. Inaddition, approximately one third of people in theworld, or 1.86 billion people, have Mycobacteriumtuberculosis (MTB) infection. To make mattersworse, the incidence of drug-resistant TB worldwideis increasing, with new strains harboringmore extensive forms of resistance (eg, extensivelyresistant TB [XDR-TB]) being more commonlyrecognized. During the 1990s, an estimated 30 millionpeople died as a result of TB, making a strongargument for its being the most important pathogenin the world today.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 619ACC-PULM-09-0302-037.indd 6197/10/09 8:57:51 PM


Causative Agent of TBTB is caused by a group of closely relatedmycobacteria (MTB, Mycobacterium bovis, Mbovis bacillus Calmette-Guérin (BCG), Mycobacteriumafricanum, Mycobacterium caprae, Mycobacteriummicroti, Mycobacterium pinnipedii,and Mycobacterium canetti), which form the MTBcomplex. The MTB organism is responsible forthe vast majority of cases of TB in the UnitedStates.Transmission of MTBThe risk factors for acquiring TB infection, aprerequisite for the development of disease, arerelated to having contact with a source case. In theUnited States, important risk factors for infectionare as follows: close contact with a person with TB;immigration from an endemic area (eg, Africa, Asia,or Latin America); exposure to untreated TB casesin congregate living facilities (eg, homeless shelters,correctional facilities, nursing homes, or otherhealth-care facilities); age; and residence in highincidencelocations (eg, inner cities and travel toforeign endemic areas).TB is usually spread human to human throughthe air by droplet nuclei, which are particles 1 to5 μm in diameter containing MTB complex organisms.MTB enters the air when patients with activepulmonary TB cough, speak, sneeze, or sing,although coughing remains the most effectivemethod of aerosolization. Droplet nuclei may alsobe produced by aerosol treatments, sputum induction,aerosolization during bronchoscopy, manipulationof lesions, or processing of tissue or secretionsin the hospital or laboratory. These droplet nucleiare so small that air currents normally present inan indoor environment can keep them airborne forlong periods. Individuals who are in prolonged,close contact with patients with active TB, especiallyin environments with poor ventilation, aremost likely to inhale the organism and acquireinfection (especially in congregate settings such asjails, prisons, shelters, or hospitals); however, onlyapproximately one third of these exposed individualsacquire infection. In countries in which TBis endemic, the chances of being exposed to a personwith active TB is more likely; thus, the chancesfor transmission are increased.When a person inhales a droplet nucleus,which may contain 1 to 400 organisms, it usuallyis trapped in the upper respiratory tract andcleared. Organisms deposited on intact mucosa orskin do not invade the tissue. The smallest droplets,those measuring 5 μm, make it to the alveoli.The chances of the transmission of MTB areinfluenced by the following four factors: (1) thenumber of organisms entering the air; (2) the concentrationof organisms in the air, determined bythe size of the space and the adequacy of ventilation;(3) the length of time an exposed personbreathes the contaminated air; and (4) possibly theimmune status of the exposed individual. Somebelieve that persons with HIV and others withimpaired cell-mediated immunity may be morelikely to acquire MTB infection after exposure thanpersons with normal immunity; however, diseaseis more likely to develop in those persons withimpaired cell-mediated immunity if they haveinfection (discussed in the section “Pathogenesis,”below).PathogenesisAfter inhalation, the droplet nucleus is carrieddown the bronchial tree to a respiratory bronchioleor alveolus, where it is phagocytized by residentalveolar macrophages. The nuclei are able to surviveat this primary site of infection; bacilli multiplyinitially within the macrophage and within 2weeks are transported through the lymphatics toestablish secondary sites. The development of animmune response, heralded by the developmentof delayed-type hypersensitivity during the next4 weeks, leads to granuloma formation with asubsequent decrease in bacillary numbers. Thisstage of disease is called latent TB infection (LTBI)and is most commonly detected by a reaction tothe tuberculin skin test (TST). Patients with LTBIare asymptomatic, but the condition has the potentialto progress to active disease later. For mostindividuals with normal immune function, theproliferation of MTB is arrested once cell-mediatedimmunity develops, although small numbers ofviable bacilli may remain within the granuloma. Aprimary complex can sometimes be seen on chestradiograph, but most pulmonary TB infections areclinically and radiographically inapparent. Mostcommonly, a positive TST result is the only620 TB and Other Mycobacterial Diseases (Ashkin)ACC-PULM-09-0302-037.indd 6207/10/09 8:57:51 PM


indication that infection has taken place. Individualswith LTBI but not active disease are not infectiousand thus cannot transmit the organism.After TB infection, active disease develops in3 to 5% of “immunocompetent” individuals within2 years (defined as progressive primary TB, which isseen more commonly in patients with a largeinoculation or immunosuppression), whereasactive TB disease develops in an additional 3 to 5%of infected persons during the remainder of theirlifetime. Most infected individuals are able tomount an effective immune response that encapsulatesthese organisms, usually for the rest of thehost’s life, thus preventing the progression frominfection to disease. Years after the initial infectionand in a small proportion of patients (approximately5%), the immune system may not be ableto contain these latent organisms, so reactivationof TB disease develops. Most cases of TB werethought to be caused by reactivation from remoteinfection (about 90% of cases), but studies in whichthe authors used DNA fingerprinting indicate thatrecent transmission (especially among HIV- positiveindividuals) probably accounts for as much as 40%of cases of TB.Some individuals appear to be more susceptiblethan others to progression to TB disease. Thisis particularly apparent among individuals withcertain medical conditions associated with varyingdegrees of immunosuppression (ie, HIV, diabetesmellitus, certain cancers, chemotherapy or immunosuppressivetherapy, including tumor necrosisfactor [TNF]-α blockers); silicosis; gastrectomy; oldage; malnutrition; IV drug use; and renal insufficiency)and in children 4 years old. Most of theseindividuals have conditions that are thought toimpair cellular immunity. In fact, patients with HIVinfection, with their severe defect in cellular immunity,are significantly more likely to have TB progressfrom infection to disease (relative riskincreased 80-fold to 170-fold compared withpatients without HIV infection. Unlike immunocompetentindividuals, who have a 5 to 10% chanceof TB progressing from infection to disease duringtheir lifetimes, individuals patients with HIV andTB coinfection have a rate that may be as great as8% per year. The risk of TB among anergic withHIV infection may be increased.The partnership between HIV and TB has augmentedthe deadly potential of each disease. Bydestroying the CD4 cells of the host’s immunesystem, HIV allows dormant TB to activate andrapidly cause disease. In response to the reactivationof TB, CD4 cells become stimulated and beginto replicate. This activation further renders the CD4cells that are vulnerable to invasion by the HIV andallows the virus to further replicate within thesecells, leading to a vicious cycle of increasing viralload and causing a further deterioration of thehost’s immune system. Studies have shown thatthe 1-year mortality rate for treated, HIV-relatedTB ranges from 20 to 35% and demonstrates littlevariation between cohorts from industrialized anddeveloping countries.In a person with intact cell-mediated immunitywho has previously been treated for MTB, someprotection against reoccurrence usually is presentif MTB exposure recurs. In an otherwise-healthy,previously treated person, any organisms that aredeposited in the alveoli after reexposure are likelyto be killed by the cell-mediated immune response.Exceptions may occur, but in immunocompetentindividuals, clinical and laboratory evidence indicatesthat disease produced by the inhalation of asecond infecting strain is uncommon. However,reinfection has been documented both in personswithout recognized immune compromise and inthose with advanced HIV infection.Classification of TBTable 1 outlines a classification system that isused mainly as an operational framework for publichealth programs, which may be helpful for practicingclinicians in categorizing their patients. Thisclassification is based on the broad host- parasiterelationships as described by exposure history, infection,and disease. Table 2 reviews the US Citizenshipand Immigration Services TB classification forTable 1. Classification of TBClassificationDescription0 No TB exposure, not infected1 TB exposure, no evidence of infection2 TB infection, no disease3 TB, clinically active4 TB, not clinically active5 TB suspect (diagnosis pending)<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 621ACC-PULM-09-0302-037.indd 6217/10/09 8:57:51 PM


Table 2. US Immigrant/Refugee TB Classification*Class Chest Radiograph AFB Smear RestrictionsA: TB, infectious Active TB PositiveNo entry to United States until treated andsmears are negativeReport to local health department for furtherB1: TB, clinically active, notActive TBNegativemedical evaluation within 30 d of arrivalinfectiousin United StatesB2: TB, not clinically active Inactive TBNot required unlesssymptomaticSame as aboveNo class (normal) Normal Not required None*Adapted from CDC Immigration Requirements: Technical Instructions for Tuberculosis Screening and Tuberculosis Treatment2007. (http: /www.cdc.gov/ncidod/dq/panel_2007.htm)immigrants and refugees developed by the Centersfor Disease Control and Prevention (CDC). Thepurpose of this classification is to identify new arrivalsthat are at high risk for TB and evaluate them totry to decrease the risk of active TB. It may be importantfor practicing physicians to be familiar withthese categories because they may be asked to assistin the evaluation of such individuals.Diagnosis and Treatment of LTBIThe TST is the most common method for identifyingMTB infection in persons who do not haveTB disease. Although the available TST antigensare substantially 100% sensitive and specific forthe detection of infection with MTB, no betterdiagnostic methods have been proven to be moreeffective. Blood tests approved by the US Food andDrug Administration (QuantiFERON-Gold; CellestisLimited; Carnegie, VIC, Australia; and T-Spot.TB; Oxford Immunotec; Abingdon, UK) based onthe quantification of interferon-γ released fromsensitized lymphocytes in whole blood incubatedovernight with antigens that are specific for MTBand control antigens hold the promise for the morespecific diagnosis of LTBI.These tests have been reported to be more specificthan the TST and may be able to differentiatethose individuals who may have traditionally hadfalse-positive reactions on a TST as the result of previousexposure to nontuberculous mycobacterium(NTM) and/or BCG from those persons who trulyhave MTB. In fact, guidelines from the CDC nowrecommend that the use of such interferon-γ releaseassays for MTB (IGRA) may be used in all circumstancesin which the TST is currently used, includingcontact investigations, the evaluation of recent immigrants,and sequential testing surveillance programsfor infection control (eg, those for health-care workers).In addition, the requirement of one visit to drawthe specimen (as opposed to two visits needed forconducting the TST and then subsequently to readthe reaction) and a more reproducible result are alsoproposed advantages of the IGRA. However, theneed for better definition of its performance in certainpopulations (eg, young children and immunosuppressedpersons) and the need to transfer thespecimen to a laboratory within 8 to 12 h are significantcurrent drawbacks of the IGRA.The TST is based on the fact that infection withMTB produces a delayed-type hypersensitivityreaction to certain antigens that are derived inextracts of a culture filtrate called tuberculin; thepreparation used for testing is called tuberculinpurified protein derivative (PPD). The diagnosis ofTB infection relies on determining the size of thedelayed-type hypersensitivity reaction to an intradermalinjection of 0.1 mL of 5 tuberculin units ofPPD (ie, the Mantoux method). Tests should beread 48 to 72 h after injection when the indurationis at its maximum, although induration may lastfor up to 7 days (Erythema should not be read.)The diameter of the induration (measured transverselyto the long axis of the arm) should bemeasured by the use of a ball-point pen. Resultsshould be recorded as millimeters of induration.Multiple puncture tests (eg, the tine test) are notrecommended because of their lower reliability.Because of the lower specificity of the TST inpopulations with a low risk of TB infection (especiallyin areas with an increased prevalence ofNTMs), specific cut points have been developed to622 TB and Other Mycobacterial Diseases (Ashkin)ACC-PULM-09-0302-037.indd 6227/10/09 8:57:52 PM


improve the specificity according to the individual’srisk of true TB infection and the risk of the activedisease acquisition if infected.An induration measurement of 5 mm is consideredto be indicative of TB infection in patientswith a high probability of infection or a high riskof disease when infection is present (eg, recent closecontacts with TB patients, individuals with radiographicevidence of old TB, individuals with HIVinfection, patients with organ transplants, patientsto whom TNF-α blockers have been prescribed,and other immunosuppressed patients administeredthe equivalent of 15 mg of prednisone eachday for 1 month). Many reports suggesting anincreased risk of progression of TB infection toactive TB disease have been described in patientsadministered TNF blockers for conditions such asrheumatoid arthritis or inflammatory bowel disease.As such, it is generally recommended thatpatients who are prescribed these agents first betested for LTBI and treated if found to have a reactioninduration of 5 mm.Indurations measuring 10 mm are consideredto be a positive result for individuals with amoderate-to-high probability of TB infection (eg,recent arri vals [ 5 years] from endemic areas, residentsand employees of high-risk congregate settings,or mycobacterial laboratory personnel) ormedical conditions that increase the likelihood ofdisease progression (Table 3). Patients who do notfit into these categories should be judged to have apositive reaction with indurations of 15 mm. Ingeneral, these individuals should not undergo testingfor LTBI unless otherwise indicated. Personswith HIV infection may have a compromised abilityTable 3. Medical Conditions (Other Than HIV) That IncreaseRisk of Progression of TB InfectionInjection drug useSilicosisDiabetes mellitusChronic renal failureLymphomas, leukemiasCancers of the head, neck and lungMalnutrition (weight loss 10% below ideal body weight)Gastrectomy or jejunoileal bypassChildren younger than 4 yr who are exposed to persons athigh risk for TBPatients who receive immunosuppressive agents (includingprednisone 15 mg/d for 1 mo and TNF blockingagents)to react to TSTs because of cutaneous anergyassociated with progressive HIV immunosuppression.However, the usefulness of anergy testing todetermine any lack of delayed-type hypersensitivityresponse in persons with HIV infection has not beenshown to be correlated and is not recommended.Confusion often arises when trying to apply theTST guidelines for LTBI in patients who have a historyof vaccination with BCG. It has been shown thatwhen a positive reaction occurs secondary to BCGvaccination, this reaction remains for at most 7 to 10years after inoculation. Because most countries thatuse BCG vaccinate infants, by the time most practitionerssee adults who have been vaccinated, thereaction should have waned. In those patients whohave been administered a BCG vaccination, the CDCrecommends ignoring the history of BCG vaccinationand applying the same principles and guidelinesconcerning LTBI that would be used forindividuals who have never been vaccinated.Persons with negative TST reactions who areto undergo repeat annual or semiannual skin testing(eg, health-care workers or residents of longtermcare facilities) should receive an initialtwo-step test. This test will help distinguishwhether the development of a subsequent positiveTST reaction was caused by an unrecognizedbooster reaction as opposed to a reaction causedby a true conversion from a recent TB exposure andsubsequent infection. An example would be a personwho had an initial false-negative reactionattributable to a past TB infection for which theperson’s immune system did not immediately reactto the tuberculin. Those patients with an initialtwo-step test negative reaction who are found onsubsequent TSTs to have an increase in the size ofthe reaction induration of 10 mm should be consideredto have TST conversion indicative of recentinfection with MTB. Two-step testing is not necessarywhen IGRA is used.The identification of persons with LTBI haspreviously been accomplished by screening individualsor groups at variable risk for TB (ie, widespreadprograms of TST). In many situations, thisscreening was performed with limited considerationof the risk of TB in the population beingtested. The emphasis is now on testing only thoseindividuals who are at high risk and would benefitfrom treatment. To focus on groups with thehighest risk of TB, the term targeted tuberculin<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 623ACC-PULM-09-0302-037.indd 6237/10/09 8:57:52 PM


testing is now being used. The dictum now is asfollows: “the decision to tuberculin test is the decisionto treat (and complete).” TSTs, for the mostpart, should not be performed in individuals forwhom treatment is not contemplated. A risk of TBthat is substantially greater than that in the generalUS population is found in persons with recent MTBinfection and in those with clinical conditionsassociated with an increased risk of LTBI progressionto active TB (Table 3). Targeted tuberculintesting should be offered only to those groups atrisk and should be discouraged in those individualsat low risk. Persons infected with TB who areconsidered to be at high risk for the developmentof active TB should be offered treatment for LTBIirrespective of age.The use of the terms preventive therapy andchemoprophylaxis has been confusing at times. Todescribe the intended intervention more accurately,the current guidelines recommend substituting theterm treatment of LTBI. It is hoped that the changein nomenclature will promote a greater understandingof the concept, resulting in more widespreadimplementation of this important TBcontrol strategy.The current guideline recommends isoniazidtherapy for 9 months (either 5 mg/kg/d to amaximum of 300 g, or 15 g/kg biweekly to amaximum of 900 g) for all individuals, instead oftherapy for 6 months for adults and 12 months forthose infected with HIV as was previously recommended.This change is based on studies thatrevealed 9 months of therapy to be superior to 6months (but as effective as 12 months of therapy)in preventing the development of active disease inthose persons with TB. If 9 months of isoniazidtherapy cannot be accomplished, 6 months of isoniazidtherapy is a less-efficacious alternative. Ninemonths of isoniazid therapy is recommended forchildren.In response to concerns about decreased ratesof completed preventive therapy, studies haveshown that rifampin (10 mg/kg to a maximum of600 mg) for 4 months or rifampin (or rifabutin) withpyrazinamide daily for 2 months in HIV-positiveindividuals to be effective alternatives to isoniazid.(Biweekly administration of rifampin/pyrazinamideby directly observed preventive therapy isan option if other regimens can not be used.) However,reports of increased hepatotoxicity in patientstreated with rifampin/pyrazinamide have promptedrecommendations that urge a cautious and extremelylimited role for this regimen, and it should only beprescribed in situations in which other recommendedregimens can absolutely not be used andwith close clinical and laboratory monitoring.Before beginning therapy in patients withLTBI, active disease should be ruled out. Baselinelaboratory testing is not routinely indicated for allpatients at the start of or during INH therapy. It isrecommended that all patients being treated forLTBI with isoniazid should receive an initialclinical evaluation and, at the least, monthly follow-upevaluations. Persons with HIV infection,pregnant women, and those in the immediatepostpartum period (within 3 months of delivery),persons with a history of chronic liver disease,persons who consume alcohol regularly, and personswho are at risk for liver disease (eg, individualsadministered other medications that affectliver function) should undergo baseline laboratorytesting. Those persons with abnormal baselineliver study findings may require continued monitoring.All patients should be educated about thesymptoms of hepatitis and be instructed to immediatelynotify the health-care provider if symptomsoccur, at which time therapy should bestopped immediately and specimens for transaminasestudies should be drawn. If the resultsare increased fivefold above the upper limit ofnormal or threefold above the upper limit of normalwith symptoms present, treatment for LTBIshould be withheld.Clinical Manifestations, Diagnosis, andTreatment of Active TB DiseaseActive TB remains primarily a disease of thepulmonary system; however, in individuals withHIV infection up to 60% of patients with TB willhave extrapulmonary involvement either alone(approximately 30%) or in addition to pulmonarydisease (about 33%), as opposed to 15% in individualswho are not without HIV infection. TBcan occur in almost any organ, but the mostcommon sites of extrapulmonary involvementinclude the pleura, lymph nodes (particularlycervical and hilar), CNS (as meningitis or tuberculoma),genitourinary system, blood, and bonemarrow.624 TB and Other Mycobacterial Diseases (Ashkin)ACC-PULM-09-0302-037.indd 6247/10/09 8:57:52 PM


The key element in the diagnosis of TB is tohave a high degree of suspicion, especially in thosegroups of persons who are at high risk. Early recognitionof the disease is essential to stop furthertransmission; however, no single clinical, radiographic,or laboratory tool is sensitive or specificenough to be diagnostic, and astute clinical assessmentis required. A careful history to elicit the mostcommon symptoms of pulmonary TB (ie, fever,productive cough, weight loss, wasting, nightsweats, shortness of breath, and occasionallyhemoptysis) should be obtained. The systemicnature of many of these symptoms is caused by thecytokine release associated with the inflammatoryresponse by the host to the organism. These symptomsare usually present for a prolonged period,characteristically weeks to months. A history ofpossible TB exposure risk (eg, previous exposure,history of homelessness, or prolonged stay in acorrectional or other congregate setting, history ofdrug abuse, migration from an endemic area, priorPPD test) should also be elicited. Physical examinationgenerally is not helpful in establishing thediagnosis.Patients with prolonged symptoms, especiallythose who are at high risk for TB, should have achest radiograph immediately (Table 4). <strong>Pulmonary</strong>TB in immunocompetent hosts nearly alwayscauses abnormalities on the chest radiograph,although an endobronchial lesion may not beassociated with a radiographic finding. In patientswith primary TB occurring as a result of recentinfection, the process is generally seen as a middle-lungor lower-lung zone infiltrate, often associatedwith ipsilateral hilar adenopathy. In patientswith primary pulmonary pleuritis, a unilateralpleural effusion may be present. The TST resultinitially may be negative in these patients. Pleuralfluid test findings are usually predominated byTable 4. General Indications for a Chest Radiograph To Detect TB*• Unexplained cough (for 3 wk)• Unexplained cough with fever ( 3 d)• Unexplained pleuritic chest pain, hemoptysis, and/ordyspnea• Unexplained fever, night sweats, and weight loss*From Pitchenik AE, Brooks R. The most common clinicalmistakes in prevention, diagnosis and therapy of tuberculosis.In: Tuberculosis in Florida: the clinician’s desktop reference.Gainesville, FL: Florida TB Control Coalition, 1999.lymphocytes, are negative for acid-fast bacilli(AFB), and are positive on culture in only 50 to60% of cases. The addition of pleural biopsy forhistologic studies and culture increases the yieldto approximately 80%.<strong>Pulmonary</strong> TB that develops as a result of theendogenous reactivation of LTBI in immunocompetenthosts usually causes abnormalities in theupper lobes of one or both lungs. Cavitation iscommon in this form of TB. The most frequent sitesare the apical and posterior segments of the lung,with the right lung affected slightly more oftenthan the left. Healing of the tuberculous lesionsusually results in a scar, with the loss of lung parenchymalvolume and, often, calcification. In theimmunocompetent adult with TB, intrathoracicadenopathy is uncommon but may occur, especiallywith primary infection. As TB progresses,infected material may be spread via the airwaysinto other parts of the lungs, causing a patchybronchopneumonia. The erosion of a parenchymalfocus of TB into a blood or lymph vessel may leadto dissemination of the organism and a miliarypattern (evenly distributed small nodules) on thechest radiograph.Nodules and fibrotic scars are seen most commonlyin old, healed TB and may contain slowlymultiplying tubercle bacilli with the potential forfuture progression to active TB. The risk of progressionis significant, and persons who have nodularor fibrotic lesions consistent with findings of previousdisease found on a chest radiograph and whohave a positive TST reaction should be consideredhigh-priority candidates for the treatment of LTBI,regardless of age.In patients with HIV infection, the nature ofthe radiographic findings depends to a certainextent on the degree of immunosuppression. TBthat occurs in patients with HIV infection and withpreserved or restored immune function tends tohave the typical radiographic findings that weredescribed in the proceeding two paragraphs. Inmore advanced HIV disease with associated immunosuppression,the radiographic findings becomemore atypical; cavitation is uncommon, and lowerlungzone or diffuse infiltrates and intrathoracicadenopathy are frequent. Clear lung fields may bepresent in up to 35% of patients with active TB andAIDS. Patients with symptoms or chest radiographfindings that are suspicious for TB disease should<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 625ACC-PULM-09-0302-037.indd 6257/10/09 8:57:52 PM


e immediately isolated if they are admitted to ahealth-care facility or congregate setting.The foundation of a rapid, accurate microbiologicaldiagnosis of TB is proper specimen collectionand rapid transport to the laboratory. At leastthree sputum specimens should be obtained forAFB smear and culture; recent recommendationsinclude at least one specimen to be sent for nucleicacid amplification to attempt to establish the diagnosismicrobiologically. A quality sputum specimenshould contain a volume of 5 to 10 mL. If thepatient is unable to produce an adequate specimen,sputum induction and/or bronchoscopy shouldbe considered (with proper infection control precautionsused).Sputum smears, the time-honored test for thediagnosis of TB, only yield a positive result for TBin approximately 50% of active cases. The advantageof the smear is that it is both rapid and cheap.The reason for its low sensitivity is the need for10,000 to 100,000 organisms in 1 mL of specimen.In addition, the smear is not specific because othermycobacteria also may have a similar appearance.Patients with smear-positive sputa are more likelyto transmit infection to contacts compared withthose with smear-negative studies, although thislatter group has clearly been shown to transmitinfection.Culture findings are positive in approximately80% of cases, but unfortunately organisms take aprolonged time to grow, up to 8 weeks for solidmedia and 1 to 3 weeks for liquid media. At least500 organisms per milliliter need to be present inthe specimen to obtain a positive culture result.Susceptibility testing (which should be performedon all initial isolates) may take another 1 to 3weeks, although tests that use molecular technologyare now becoming available that could shortenthe turn around time for susceptibility studies to1 to 2 days.Fifteen to twenty percent of diagnosed casescannot be confirmed microbiologically and areconsidered to be culture-negative or clinical TB. Inthese patients, the diagnosis of TB is based on thepresence of symptoms, positive TST results, aradiographic appearance compatible with TB, andan improvement in clinical status after treatmentwith antituberculous therapy, despite negativemicrobiological studies and no other etiologyaccounting for the illness. This is found more oftenin children and individuals from whom qualityspecimens are difficult to obtain. Bronchoscopicstudies that use BAL and transbronchial biopsy,which are performed with appropriate infectioncontrol precautions, may help to establish thediagnosis and rule out other etiologies when thediagnosis is in question. When TB is strongly suspected,the initiation of presumptive antituberculoustherapy is appropriate while awaitingmicrobiological confirmation.Nucleic acid amplification techniques hold thepromise of being able to detect a few strands ofnucleic acid in a sample, amplify it, and identifythe presence of TB within a matter of hours. Thetest is 90% sensitive and 99% specific when usedin smear-positive cases. Unfortunately, in smearnegativecases, the sensitivity may only be 60 to80%. The test has been approved by the US Foodand Drug Administration for use on respiratoryspecimens (not approved for nonrespiratoryspecimens) from untreated cases. The most recentrecommendations from the CDC recommendobtaining at least one (and many authorities recommendthe use of multiple) specimens to improvethe test sensitivity on patients suspected of havingactive TB disease; however, the clinician must beaware that a negative test result does not rule outthe possibility of TB and that a positive test resultmight not guarantee a diagnosis of TB.Since the advent and utilization of effectivechemotherapy against TB in the 1950s, 95% of allindividuals with pansusceptible TB who completetherapy are now cured. In those individuals withmultidrug-resistant strains (resistant to at leastisoniazid and rifampin) and XDR-TB (resistance toisoniazid, rifampin, a quinolone, and either capreomycin,kanamycin or amikacin), traditionalchemotherapy may be ineffective in 40% ofpatients.The American Thoracic Society (ATS) and theCDC recommended that four-drug therapy withisoniazid, rifampin, pyrazinamide, and either ethambutolor streptomycin be started initially inpatients from areas in which isoniazid resistanceexceeds 4% until susceptibility test results are available.In areas with 4% isoniazid resistance, thefourth drug may be omitted from the initial regimen.Once the results reveal susceptibility toisoniazid, rifampin, and pyrazinamide, therapywith ethambutol or streptomycin should be626 TB and Other Mycobacterial Diseases (Ashkin)ACC-PULM-09-0302-037.indd 6267/10/09 8:57:53 PM


discontinued. After 2 months of therapy, pyrazinamideis stopped, and isoniazid and rifampin arecontinued for an additional 4 months for a total of6 months of treatment (completion is also definedby the number of doses administered; Table 5).It is important to obtain sputum cultures at thetime of the completion of 2 months of therapy(called the initial phase of treatment) to identifypatients who are at increased risk of relapse. Ifculture findings are not negative after 2 months oftherapy and cavities are present on chest radiographs,then treatment should be extended for atleast 4 months after negative results are achieved.This regimen can be administered effectively eitherdaily or intermittently by the use of directlyobserved therapy (DOT; Table 5). It may also beTable 5. TB Treatment Options for Adults and Children With Culture-Positive <strong>Pulmonary</strong> TB Caused by Drug-Susceptible Organisms*Initial PhaseContinuation PhaseRating †(Evidence) ‡RegimenDrugsInterval and Doses §(Minimal Duration) Regimen DrugsInterval and Doses § ||(Minimal Duration)Range of TotalDoses (MinimalDuration) HIV HIV 1 INHRIFPZAEMB2 INHRIFPZAEMB3 INHRIFPZAEMB4 INHRIFEMB182–130 (26 wk) A (I) A (II)7 d/wk for 56 doses 1a INH/RIF 7 d/wk for 126 doses(8 wk) for 40 doses90 doses (18 wk) (8 wk) or 5 d/wk(18 wk) or 5 d/wk for1b INH/RIF Twice weekly for 92–76 (26 wk) A (I) A (II) #36 doses (18 wk)1c ** INH/RPT Once weekly for 18 74–58 (26 wk) B (I) E (I)7 d/wk for 14 doses(2 wk), then twiceweekly for 12 doses(6 wk) or 5 d/wkfor 10 doses(2 wk), then twiceweekly for 12 doses(6 wk)Three times per weekfor 24 doses (8 wk)doses (18 wk)2a INH/RIF Twice weekly for36 doses (18 wk)2b ** INH/RPT Once weekly for 18doses (18 wk)3a INH/RIF Three times weekly for54 doses (18 wk)62–58 (26 wk) A (II) B (II) §44–40 (26 wk) B (I) E (I)78 (26 wk) B (I) B (II)7 d/wk for 56 doses 4a INH/RIF 7 d/wk for 217 doses 273–195 (39 wk) C (I) C (II)for 40 doses (8 wk) # 155 doses (31 wk) #(8 wk) or 5 d/wk(31 wk) or 5 d/wk for4b INH/RIF Twice weekly for 62doses (31 wk)118–102 (39 wk) C (I) C (II)*Reprinted with permission from the Centers for Disease Control and Prevention from MMWR Morb Mortal Wkly Rep 2003:52;1-77. EMB ethambutol; RIF rifampin; RPT rifapentine; negative; positive.†Ratings: A preferred; B acceptable alternative; C offer when A and B cannot be given; E should never be given.‡Evidence: I randomized clinical trial; II data from clinical trials that were not randomized or were conducted in otherpopulations; III expert opinion.§When DOT is used, drugs may be administered 5 days per week and the necessary number of doses adjusted accordingly.Although there are no studies that have compared five daily doses with seven daily doses, extensive experience indicates thiswould be an effective practice.||Patients with cavitation seen on an initial chest radiograph and positive culture findings at the completion of 2 mo of therapyshould receive therapy with a 7-month continuation phase (31-week; either 217 doses [daily] or 62 doses [twice weekly]).Five-day-a-week administration is always given by DOT. Rating for 5 days per week regimens is AIII.#Not recommended for HIV-infected patients with CD4 cell counts of 100 cells/L.**Options 1c and 2b should be used only in HIV-negative patients who have negative sputum smears at the time of completionof 2 mo of therapy and who do not have cavitation on initial chest radiograph (see text). For patients started on this regimenand found to have a positive culture from the 2-month specimen, treatment should be extended an extra 3 mo.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 627ACC-PULM-09-0302-037.indd 6277/10/09 8:57:53 PM


used in individuals infected with HIV as well aspatients with extrapulmonary disease. However,for individuals with HIV infection with a CD4count of 100 cells/μL, daily or intermittenttherapy three times a week is recommended insteadof biweekly therapy as the result of reports of thedevelopment of relapse with rifampin-monoresistantdisease. In addition, children with miliary TB,bone/joint TB, or tuberculous meningitis shouldreceive a minimum of 9 to 12 months of therapy.If pyrazinamide cannot be administered for thefirst 2 months, a reasonable alternative is INH andrifampin administered for 9 months. A 4-monthregimen of isoniazid and rifampin is acceptabletherapy for adults who have active TB and negativeresults of smears and cultures (ie, culture-negativeor clinical TB).It is essential to treat pregnant women whohave active TB. Isoniazid and rifampin have beenshown to be safe for use in pregnant women andshould be administered. Pyrazinamide, althoughrecommended by many authorities, has not beenthoroughly studied in pregnant women and shouldbe used at the discretion of the treating clinician.Streptomycin has been shown to be harmful to thedeveloping fetus and should not be used.Corticosteroids have been shown to be of benefitin preventing cardiac constriction from TBpericarditis and in decreasing neurologic sequelaeresulting from TB meningitis. They may have a rolein preventing bronchial stenosis in cases of diffuseendobronchial TB. The need to treat with multipledrugs for a prolonged period leads to the majorobstacle facing the control of TB, ie, adherence totherapy. If patients do not take their medicationsas prescribed for the entire period, treatment failureand resistance have been shown to develop.Once medications are begun, adherence needs tobe ensured.The decrease of TB in the United States hasbeen attributed in large part to the implementationand utilization of DOT, which has beenshown to significantly improve the completionrates of TB therapy, as well as to impede the developmentof resistant strains. The responsibility forsuccessful treatment has clearly been assigned bycurrent guidelines to the public health programor private provider, not to the patient. Furthermore,it is strongly recommended that the initialtreatment strategy use patient-centered casemanagement with an adherence plan that emphasizesDOT.Through DOT, representatives of health-carefacilities (usually from the public health system)go into the community to observe and ensure thatpatients take their medications. Numerous studieshave demonstrated that there is no reliable way topredict which patient will be adherent to therapy;therefore, all patients with active TB disease shouldbe considered for DOT. Those who are not startedon DOT should be given combination pills (ie, pillscontaining isoniazid, rifampin, and pyrazinamidetogether) to avoid monotherapy and the subsequentdevelopment of resistance.Patients must be monitored for the effectivenessof treatment and toxicity. Soon after beginningtherapy, patients should have an improvement insymptoms (eg, cough, fevers, night sweats, andweight gain). Those who do not have a symptomaticimprovement or whose culture results fail toconvert to negative within 2 months of treatmentinitiation should be evaluated for treatment failure.More than 80% of patients receiving appropriatetherapy are culture negative after 2 months oftreatment.The potential reasons for treatment failureinclude the following: (1) resistance to prescribedmedications or the use of an inadequate combinationof drugs, resulting in increased drug resistance;(2) inadequate medication levels caused bymalabsorption, food or drug interactions, or concurrentmedical conditions; and (3) noncompliancewith prescribed treatment regimens. Of these, thelatter is by far the most common cause of treatmentfailure. It is important to seek the advice of a TBexpert when patients have resistant disease, complicatingconcurrent medical conditions, or a lackof expected response to therapy.Adverse reactions are not uncommon in thetreatment of TB. Isoniazid, which is best known forits hepatotoxicity, also causes peripheral neuropathy,for which vitamin B 6(pyridoxine) is prescribedas prophylaxis. Rifampin may cause hepatotoxicityin addition to muscle and joint stiffness and pain.Baseline liver function tests should be performedfor all patients beginning treatment, and monthlymonitoring is recommended for anyone with baselineabnormalities or concomitant liver disease.Clinical monitoring through regular inquiry of thepatient as to the development systems of toxicity628 TB and Other Mycobacterial Diseases (Ashkin)ACC-PULM-09-0302-037.indd 6287/10/09 8:57:53 PM


is recommended for all other patients. If transaminaselevels become elevated and symptoms ofhepatitis occur, therapy with all medicationsshould be discontinued. Once signs and symptomsresolve or improve, medications should be reintroducedsequentially, and the patient should bemonitored for the recurrence of hepatitis. Pyrazinamidecan also cause hepatotoxicity as well as highuric acid levels, resulting in gout-like symptoms.Ethambutol is associated with optic neuritis, especiallyin doses of 20 mg/kg/d and in patientswith renal insufficiency. The monthly assessmentof visual acuity and color discrimination is recommendedto identify patients who may have thistoxicity.Drug resistance has been increasingly recognizedin the United States as well as worldwide.The most common reasons for the development ofdrug resistance are patient nonadherence to therapyand/or physician mistakes (eg, adding a singledrug to an ineffective regimen). Multidrug-resistantTB is defined as strains that are resistant to atleast isoniazid and rifampin. XDR-TB is defined asresistance to at least isoniazid, rifampin, a quinolone,and either capreomycin, kanamycin, or amikacin.Outbreaks of such strains have been welldocumented, resulting in significant rates of morbidityand mortality, especially among populationswith HIV infection. The control and prevention ofsuch outbreaks have required the expenditure ofsignificant efforts and resources by hospital andTB control programs.The treatment of multidrug-resistant TB isfrequently unsuccessful, requiring the use of moretoxic, expensive drugs, in addition to possiblesurgery; thus, emphasis is on the prevention. Inpatients with confirmed multidrug-resistant disease,therapy with at least two drugs to which theisolate is susceptible is recommended for at least12 months after documented conversion of thecultures to negative results. Most experts recommend18 to 24 months of total therapy. Patientswith isoniazid-resistant TB can be treated withrifampin, pyrazinamide, and ethambutol for 6 to9 months. Patients with rifampin-resistant TBshould be treated with isoniazid, pyrazinamide,and ethambutol for at least 12 months after cultureresults convert to negative. Consultation with a TBexpert is recommended for the management ofdrug-resistant TB.Patients with HIV have a response to therapysimilar to that of individuals without HIV infection;however, the treatment of HIV has altered TBtherapy. Protease inhibitors (PIs), nonnucleosidereverse transcriptase inhibitors, and chemokine(CC motif) 5 receptor antagonists, three of the mostpotent agents available to control HIV, interferewith rifampin, which is the most important drugavailable in TB treatment. Rifampin is a potentinducer of the hepatic p450 system. For example,this induction enhances the metabolism of manyof the PIs, causing reduced serum levels and leadingto ineffective viral suppression, as well as thedevelopment of resistance. Conversely, many ofthe PIs may interfere with the metabolism of therifamycins, causing high serum levels and potentialtoxicity.On the basis of more current experience,rifabutin, which is a rifamycin derivative with lesseffect on the hepatic p450 system but equivalentefficacy against TB, may be used safely and effectivelywith the PIs. Consultation with a resourcefamiliar with the potential pharmacokinetic interactionsof these medications should be consultedto determine the best combinations and dosagesfor the particular patient (recent recommendationscan be found at http://www.cdc.gov/tb/TB_HIV_Drugs/default.htm).With the use of antiretroviral therapy and thesubsequent improvement in the host’s immuneresponse, patients with TB disease may actuallyexhibit a “paradoxical” worsening of their clinicalcondition, which is thought to be attributable toan inflammatory response from the immune reactivation(called the immune reconstitution inflammatorysyndrome). These patients may experiencethe development of new ascites, lymphadenopathy,fever, pleural effusions, or cerebral lesions.These may be life-threatening conditions, dependingon the site and size of the lesion. Clinicianstreating TB patients who have HIV must be awareof this phenomenon and rule out other etiologiesthat may be responsible for development of thelesions and/or clinical worsening. If no otheretiology can be found, continued treatment forHIV and TB usually results in an improvementwithout further intervention. In selected cases, theuse of immunomodulators (eg, steroids) maybe indicated to slow the progression of thisresponse.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 629ACC-PULM-09-0302-037.indd 6297/10/09 8:57:53 PM


Infection ControlGiven the increased number of immunosuppressedindividuals in health-care facilities, infectioncontrol efforts to stem nosocomial outbreaksof TB among patients and staff are essential. Allpatients who are suspected of having active TBshould be immediately isolated and remain so untilthey are no longer infectious or until TB has beenruled out.Airborne infection isolation rooms shouldhave negative air pressure to prevent infected airfrom entering hallways and should use at leastsix air exchanges per hour. Health-care workerswho are caring for these individuals should wearN95 masks to avoid inhaling infectious particleswhile working in the enclosed room. The observanceof these guidelines is especially importantin rooms in whim cough-inducing procedures areperformed. Patients in health-care facilities canbe removed from isolation once they have receivedmedications for 10 to 14 days, they are respondingclinically, and their AFB smear results are negative.Patients may be discharged from the hospitalbefore they are removed from respiratory isolationif the following conditions are met: (1) theyare returning to their previous residence (in noncongregatesettings), and the health departmenthas assessed that no vulnerable individuals (ie,immunocompromised persons and children 2years of age) are present and (2) others who havebeen exposed have been evaluated for LTBItherapy.VaccinationBCG, which is an attenuated strain of M bovis,was first used as a vaccine in humans in the early1920s. It has since become the most widely usedvaccine preparation in the world, despite questionsregarding its 0 to 80% efficacy in preventing pulmonaryTB in adults. Although the data are lessthan ideal, numerous studies have consistentlyshown the effectiveness of the BCG vaccine inreducing the incidence of TB in infants and youngchildren, particularly against potentially fatal disseminatedTB. For this reason, BCG remains animportant part of vaccination programs in most ofthe world because of its effectiveness in reducingpediatric mortality rates for disseminated TB. Itsinferior efficacy in preventing TB, compared withINH treatment for LTBI, and the interference withTST has limited its use in the United States.Diagnosis and Treatment of<strong>Pulmonary</strong> Disease Caused by NTMsEpidemiology, Transmission, and PathogenesisAlthough not as important worldwide as TB,diseases caused by other mycobacteria appear tobe more common than TB in the United States. Therecognition of these organisms as pathogens isrelatively recent. These organisms, characterizedas the NTM, share a number of features. UnlikeMTB, these NTMs are normal inhabitants of theenvironment (usually soil and water). Also, unlikeTB, pulmonary disease and other infections causedby the NTMs are not contagious, and patients withthese diseases are generally not isolated. Manyclinicians believe that these organisms are opportunisticrather than virulent pathogens becausesome local or systemic immune impairment isrequired for them to cause disease.Much remains to be understood about thepathogenesis of NTM infection and disease. It isassumed that most persons have exposure to theseorganisms through the environment. The aerosolizationof environmental NTMs may play a role inrespiratory disease, whereas ingestion may be thesource of infection for children with NTM lymphadenitisand for most patients with AIDS, whosedisseminated M avium or Mycobacterium genavensebegins as a GI colonization. Bacteremic spread ofthe organism in patients with AIDS then involvesmultiple organ systems, including bone marrow,lymph nodes, liver, and spleen. Direct inoculationwith NTM organisms from water or other materialis the likely source of infection in patients withsoft-tissue infections. It is not known whether NTMdisease develops soon after infection or, like TB,after a period of latency.Clinical ManifestationsEarly reports concerning patients with pulmonarydisease caused by NTMs involved older whitemen with a history of underlying lung disease. Thediseases appeared to regional, with most patientsfrom the rural southeastern United States having630 TB and Other Mycobacterial Diseases (Ashkin)ACC-PULM-09-0302-037.indd 6307/10/09 8:57:54 PM


disease caused by M avium, whereas most from thecentral states had disease caused by M kansasii.NTM disease is not reportable in the United States,and an accurate incidence of the disease is limited.Severely immunosuppressed (CD4 counts of 100cells/μL) individuals with HIV infection are particularlysusceptible to disseminated diseasecaused by M avium. Before the widespread use ofhighly active antiretroviral therapy, disseminatedM avium disease developed in individuals withHIV infection who had CD4 counts of 100 cells/μL at a rate of 20% annually. Localized pulmonarydisease caused by M avium occurs in 5% of casesin hosts with HIV infection. With the effective useof highly active antiretroviral therapy, however,the incidence of M avium disease seems to bedecreasing.Chronic pulmonary disease is the most commonlocalized manifestation of NTMs; it usuallyappears in older individuals, but not all patientshave a history of underlying lung disease. Mavium complex (MAC, consisting of M avium andthe less-common but difficult-to-distinguishMycobacterium intracellulare) and M kansasii arethe most frequent NTM pathogens causing lungdisease in the United States. Other less-commonNTM pathogens include Mycobacterium abscessus,Mycobacterium fortuitum, Mycobacteriumchelonae, Mycobacterium szulgai, Mycobacteriumsimiae, Mycobacterium xenopi, Mycobacterium malmoense,Mycobacterium celatum, and Mycobacteriumasiaticum.<strong>Pulmonary</strong> disease caused by MAC generallyis classified into the following three syndromes:1. Upper-lobe disease that mimics reactivated TBclinically and radiographically. This syndromeis the best-known one. Patients are generallyolder men and have a history of COPD.2. MAC that develops as a complication of previousbronchiectasis. This condition is usuallyseen in patients with a history of TB or M kansasiiinfection who present with a recurrence ofsymptoms and a new infiltrate in an area of olddisease. These patients also tend to be older,but there is no gender predilection and norelationship to smoking-related disease. Thissyndrome also may occur in patients with bronchiectasisas the result of previous virus-relatedor bacteria-related damage or cystic fibrosis.3. MAC with no underlying disease. Thesepatients are predominantly women, are nonsmokers,and have interstitial rather than cavitaryradiographic changes, which usually areconfined to the lingula and right middle lobes(referred to as Lady Windermere syndrome). Someinvestigators have noted a progression to respiratoryfailure in some individuals. The findingof MAC in these patients was thought to becolonization, but recent studies that have usedhigh-resolution CT scanning have suggestedthat the development of bronchiectasis in thesepatients is attributable to pathologic diseasecaused by MAC.Signs and symptoms of NTM pulmonary diseaseare variable and nonspecific. They includechronic cough, sputum production, and fatigue.Less commonly, malaise, dyspnea, fever hemoptysis,and weight loss can occur, usually withadvanced NTM disease. Evaluation is often complicatedby the symptoms caused by coexistinglung diseases, including COPD associated withsmoking, bronchiectasis, previous mycobacterialdiseases, cystic fibrosis, and pneumoconiosis.There are some differences in the radiographicfeatures of NTM lung disease compared with thoseproduced by MTB. NTMs tend to cause thin-walledcavities with less surrounding parenchymal infiltrate,have less bronchogenic but more contiguousspread of disease, and produce more markedinvolvement of the pleura over the involved areasof the lungs. Occasionally, they may produce densepneumonic disease or a solitary pulmonary nodulewithout cavitation. Basal pleural disease is notoften found, and pleural effusions are rare. HighresolutionCT scan studies have shown that up to90% of patients with mid- and lower-lung fieldnoncavitary disease with MAC have associatedmultifocal bronchiectasis, with many patients havingclusters of small ( 5 mm) nodules in associatedareas of the lung (ie, the “tree-in-bud”pattern).DiagnosisIn the absence of specific diagnostic features inthe history, physical examination, and chest radiographfindings, culture isolation of the NTM isessential; however, because these organisms are<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 631ACC-PULM-09-0302-037.indd 6317/10/09 8:57:54 PM


commonly found in nature, the contamination ofculture material or transient infection does occur.Thus, a single positive sputum culture finding,especially with a small numbers of organisms, isnot always sufficient to diagnose NTM disease. Theprevious notion that individuals had true colonizationwith NTMs (eg, no tissue invasion) is nowthought to probably be rare. Given the changingunderstanding of the disease caused by NTMs, theATS published an official statement concerning thediagnostic criteria of NTM lung disease in HIVseropositiveand HIV-seronegative hosts. Thesecriteria fit best with MAC, M abscessus, and Mkansasii because too little is known about otherNTMs to be certain how applicable these criteriawould be.The following criteria should be used to establisha diagnosis of pulmonary disease caused byNTMs. The criteria apply to symptomatic patientswith infiltrate, nodular, or cavitary disease, or highresolutionCT scans that show multifocal bronchiectasisor multiple small nodules. Other etiologiesfor the clinical and radiologic findings should beexcluded. At least three respiratory samples fromeach patient should be evaluated. Those havingone of the following should be considered to havedisease caused by NTMs:1. Positive culture results from at least two separateexpectorated sputum samples (if the resultsfrom the initial sputum samples are nondiagnostic,consider repeat sputum AFB smearsand cultures); or2. Positive culture results from at least one bronchialwash or BAL; or3. Transbronchial or other lung biopsy specimenswith mycobacterial histopathologic features(eg, granulomatous inflammation or AFB) andpositive culture findings for NTMs or biopsyspecimens showing mycobacterial histopathologicfeatures (eg, granulomatous inflammationor AFB) and one or more sputum or bronchialwashings that are culture positive forNTMs; or4. Patients who are suspected of having NTMlung disease but who do not meet the diagnosticcriteria should be followed up until thediagnosis is firmly established or excluded.Expert consultation should be obtained whenNTMs are recovered that are either infrequentlyencountered or usually represent environmentalcontamination. Routine susceptibility testing of allNTMs is discouraged, but such testing is warrantedto obtain baseline data for unresponsivedisease or when relapses occur.TreatmentNote that making the diagnosis of NTM lungdisease does not, per se, necessitate the institutionof therapy, which is a decision based on the potentialrisks and benefits of therapy for individualpatients.M kansasii: Treatment for pulmonary diseasecaused by this organism was disappointingbefore the advent of rifampin therapy. With theintroduction of rifamycins, the rate of treatmentfailures has decreased. The recommended treatmentfor adults consists of a regimen of dailyisoniazid (300 mg); rifampin (600 mg); and ethambutol(25 mg/kg for 2 months, then 15 mg/kgfor 18 months), with a minimum of 12 monthsof treatment after culture results are negative.Clarithromycin or rifabutin should be substitutedfor rifampin in HIV-positive patients whoreceive PIs.M avium: The medical treatment of pulmonarydisease caused by MAC in HIV-negative patientshistorically has been frustrating and disappointing.However, significant advances in the drugsavailable (eg, newer macrolides) have been made,and there is now greater expectation that pulmonaryMAC disease can be effectively treated(defined as high rates of sputum conversion withlong-term negative results on culture) with medicationsalone. Treatment consists of a regimen ofdaily clarithromycin (500 mg bid) or azithromycin(250 mg), rifampin (600 mg) or rifabutin (300 mg),and ethambutol (25 mg/kg for 2 months, then15 mg/kg) in adults without HIV infection. Therapywith streptomycin two to three times weeklyshould be considered for the first 8 weeks as tolerated.Patients should be treated until culture findingsare negative for 1 year.Disseminated MAC Disease in Patients WithHIV Infection: Therapy in adults should includeclarithromycin, 500 mg bid, or azithromycin,250 to 500 mg, plus ethambutol, 15 mg/kg/d.Consideration should be given to the additionof a third drug (preferably rifabutin at a dose of632 TB and Other Mycobacterial Diseases (Ashkin)ACC-PULM-09-0302-037.indd 6327/10/09 8:57:54 PM


300 mg/d). Therapy should be continued for lifeuntil more data become available. Prophylaxisshould be administered to adults with AIDS withCD4 counts of 50 cells/ μL, especially in thepresence of a history of previous opportunisticinfections, consisting of rifabutin, 300 mg/d, clarithromycin,500 mg bid, azithromycin, 1,200 mgonce weekly and azithromycin, 1,200 mg onceweekly, plus rifabutin, 300 mg/d.Annotated BibliographyAmerican Thoracic Society, Centers for Disease Controland Prevention. Targeted tuberculin testing andtreatment of latent tuberculosis infection. Am J RespirCrit Care Med 2000; 161:S221−S247American Thoracic Society, Centers for Disease Controland Prevention. Diagnostic standards and classificationof tuberculosis in adults and children. Am JRespir Crit Care Med 2000; 161:1376−1395American Thoracic Society, Centers for Disease Controland Prevention, Infectious Diseases Society ofAmerica. Treatment of TB. Am J Respir Crit Care Med2003; 167:603−662American Thoracic Society, Centers for Disease Controland Prevention and the Infectious Disease Society ofAmerica. Controlling tuberculosis in the United States.MMWR Recomm Rep 2005; 54:1−81The four official statements from the CDC and ATS thatset the standards for diagnosis and treatment of TB diseaseand infection. They also make recommendations for publichealth department programs for control of the illness. Eachis comprehensive and review the “science” behind the recommendations.American Thoracic Society, Infectious Diseases Societyof America. Diagnosis, treatment and prevention ofnontuberculous mycobacterial diseases. Am J RespirCrit Care Med 2007; 175:367−416An official statement from the ATS/Infectious Diseases Societyof America. A very good and thorough review on NTMdiagnosis and treatment.Behr MA, Warren SA, Salamon H, et al. Transmissionof Mycobacterium tuberculosis from patients smear negativefor acid-fast bacilli. Lancet 1999; 353:444−449A study using DNA fingerprinting that revealed that asmany as 17% of TB cases may be the result of transmissionfrom smear-negative, culture-positive patients.Burman WJ. Issues in the management of HIV-relatedtuberculosis. Clin Chest Med 2005; 26:283−294Good review of the treatment of individuals infected withHIV with TB, including necessary regimen adjustmentswhen using HIV and TB medications concomitantly.Centers for Disease Control and Prevention, NationalInstitutes of Health, and the HIV <strong>Medicine</strong> Associationof the Infectious Diseases Society of America. Guidelinesfor prevention and treatment of opportunisticinfections in HIV-infected adults and adolescents.MMWR Recomm Rep 2009; 58:1−207Most current recommendations from the CDC for the preventionand treatment of TB MAC among individualsinfected with HIV.Centers for Disease Control and Prevention. Updatedguidelines on managing drug interactions in the treatmentof HIV-related tuberculosis. Available at: http://www.cdc.gov/tb/TB_HIV_Drugs/default.htm.AccessedMay 4, 2009Recent recommendations concerning the concomitantadministration of HIV antiretroviral agents and rifamycins.Centers for Disease Control and Prevention. Guidelinesfor using the QuantiFERON-TB Gold test fordetecting Mycobacterium tuberculosis infection, UnitedStates. MMWR Recomm Rep 2005; 54:49−55Recent recommendations concerning the use of this test forthe diagnosis of LTBI.Colditz GA, Brewer TF, Berkey CS, et al. Efficacy of BCGvaccine in the prevention of tuberculosis: metaanalysisof the published literature. JAMA 1994; 271:698−702A meta-analysis showing that BCG therapy is about 50%effective at preventing TB.Mahmoudi A, Iseman MD. Pitfalls in the care ofpatients with tuberculosis: common errors and theiracquisitions of drug resistance. JAMA 1992; 270:65−68This classic article revealed that physicians caring for TBpatients made on average 3.93 errors that subsequently contributedto the acquisition of drug resistance. The most commonerrors were as follows: (1) adding a single drug to a failingregimen, (2) failure to identify preexisting or acquireddrug resistance, (3) initiation of an inadequate primaryregimen, (4) failure to address nonadherence to therapy, and(5) inappropriate isoniazid prophylaxis. The average cost oftreatment per patient was $180,000.Narita M, Ashkin D, Hollender ES, et al. Paradoxicalworsening of tuberculosis following antiretroviraltherapy in patients with AIDS. Am J Respir Crit CareMed 1998; 158:157−161Case series description of “paradoxical response” or“immune reactivation with inflammatory response” inpatients with TP and HIV coinfection who received antiretroviraltherapy.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 633ACC-PULM-09-0302-037.indd 6337/10/09 8:57:54 PM


Prince DS, Peterson DD, Steinger RM, et al. Infectionwith Mycobacterium avium complex in patientswithout predisposing conditions. N Engl J Med 1989;321:863−868Description of a group of patients who did not meet thepreviously recognized demographics for pulmonary diseasecaused by MAC.Wallace RJ. Nontuberculous mycobacterial infectionsin the HIV-negative host. In: Rom WN, Garay S, eds.Tuberculosis. Boston, MA: Little, Brown and Company;2004, 651−662Well-written chapter by one of the nation’s leading expertsreviewing NTM in HIV-negative hosts.634 TB and Other Mycobacterial Diseases (Ashkin)ACC-PULM-09-0302-037.indd 6347/10/09 8:57:55 PM


Idiopathic <strong>Pulmonary</strong> Fibrosis, Nonspecific InterstitialPneumonia/Fibrosis, and SarcoidosisJoseph P. Lynch III, MD, FCCPObjectives:• Describe the salient epidemiologic, clinical, physiologic,and radiographic features of idiopathic pulmonary fibrosis,nonspecific interstitial pneumonia/fibrosis, and sarcoidosis• Discuss the salient features seen on high-resolution CTscans and their impact on prognosis• <strong>Review</strong> the characteristic histopathologic features of eachof these disorders and the role of transbronchial or surgical(open or thoracoscopic) lung biopsies• Discuss the clinical role (if any) of ancillary studies such asradionuclide scanning or BAL to stage or follow up thesedisorders• <strong>Review</strong> therapeutic strategiesKey words: cryptogenic fibrosing alveolitis; idiopathic pulmonaryfibrosis; nonspecific interstitial pneumonia; sarcoidosis;usual interstitial pneumoniaand the histologic pattern of UIP. 3,4,8 Other histologicpatterns (eg, desquamative interstitial pneumonia[DIP] 9,10 ; respiratory bronchiolitis interstitiallung disease [ILD] [RBILD] 10 ; nonspecific interstitialpneumonia [NSIP] 11−14 ; acute interstitial pneumonia[AIP] 15,16 ; lymphoid interstitial pneumonia[LIP] 17,18 ; cryptogenic organizing pneumonia[COP] 19 ; and chronic hypersensitivity pneumonia20,21 ) are distinct entities with differing clinicalexpression and prognoses. 3,4 A definitive diagnosisof UIP requires surgical lung biopsy (SLB), 6 but thediagnosis of UIP can be affirmed with confidenceby thin-section high-resolution CT (HRCT) scansin some patients. 22−25 IPF is one of the most frustratingdisorders to manage because treatment islargely ineffective. 2,26−28EpidemiologyIdiopathic <strong>Pulmonary</strong> FibrosisIdiopathic pulmonary fibrosis (IPF), alsoknown as cryptogenic fibrosing alveolitis, is a specificform of chronic interstitial lung pneumonia associatedwith the histologic pattern of usual interstitialpneumonia (UIP). 1–5 Although UIP is a distincthistologic lesion, 6 this histologic pattern is notspecific for IPF and can also be found in otherdiseases (eg, connective tissue diseases [CTDs]),asbestosis, and diverse occupational, environmental,or drug exposures. 2–4,7 A diagnosis of IPF canbe established only when these and other alternativeetiologies have been excluded. 3,4The cardinal features of IPF include progressivecough, dyspnea, bilateral interstitial infiltrates onchest radiographs, a restrictive ventilatory defecton pulmonary function tests (PFTs), and progressivefibrosis and destruction of the lung parenchyma.2,3,8 The features are nonspecific and may beobserved with myriad other interstitial lung diseases(ILDs). 3,4,8 The term IPF should be restrictedto patients with the appropriate clinical featuresIPF is rare, but precise data regarding incidenceand prevalence are lacking. 29−31 Most populationbasedepidemiologic surveys 29,32−35 were based onclinical diagnosis, death certificates, or diagnosticcoding and lacked histologic confirmation; suchstudies included a mixture of ILDs, including disordersother than UIP. Studies from Europe 31−33,35and Japan 34 cited prevalence rates of IPF rangingfrom 3 to 8 cases per 100,000 population. In theUnited States, prevalence rates range from 13 to 42cases per 100,000. 29,30 The incidence of IPF appearsto be increasing. The incidence of IPF increasedprogressively in the United Kingdom between 1991and 2003. 31 Similarly, in the United States, deathsattributed to pulmonary fibrosis (PF) increasedsignificantly from 1992 to 2003 ( 28% increase). 36The incidence of IPF is much greater in menand in the elderly (peak onset after the sixth decadeof life). 25,28−31,34,36,37 In the United States, the rate ofmortality from IPF is increasing more rapidly inwomen than men, 36 possibly reflecting the impactof cigarette smoking. In one study in the UnitedStates, the prevalence of IPF in adults between<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 635ACC-PULM-09-0302-038.indd 6357/10/09 10:24:51 PM


35 to 44 years of age was 2.7 per 100,000. 29 By contrast,the prevalence exceeded 175 per 100,000 inadults > 75 years old. 29 Rates of mortality from IPFare markedly increased in the elderly. In the UnitedStates, projected deaths (per million) in 2008 wereas follows: 18 (ages 45 to 54 years); 71 (ages 55 to64 years); 306 (age 65 to 74 years); 827 (age 75 to 84years); and 1,380 (age 85 years). 36 The incidenceof idiopathic UIP is a rare occurrence in children.3,38,39 Despite its rarity, IPF accounts for 16,000 deaths annually in the United States 36 ; thismortality rate is greater than a number of malignancies,including bladder cancer, multiplemyeloma, and acute myelogenous leukemia. 36Risk Factors for IPFThe cause of idiopathic UIP is unknown, butenvironmental and occupational exposures likelyplay etiologic roles. 40−42 IPF is more common incurrent or former smokers. 25,26,42−44 The degree ofexposure (quantified as pack-years of smoking) isdirectly correlated with the occurrence of IPF. 43,45Importantly, ever-smoking remains a risk factor forthe development of IPF, even after smoking cessation.43 In familial IPF, smoking was discovered tobe the strongest associated risk factor and had anodds ratio of 3.6. 46 Risk factors for IPF cited in somestudies included the following: exposure to metaland wood dusts, 34,40,44–47 organic solvents, 34 andresidence in agricultural or polluted urbanareas. 34,37,44 However, a study from the British Isles 33found no evidence for an increased risk of IPFamong individuals exposed to wood or metaldusts. In that study, an increased rate of mortalitycaused by IPF was noted in electricians, electricalengineers, firemen, and cleaners, occupations associatedwith exposure to potentially toxic fumes orchemicals. 33A meta-analysis 42 of six case-control studiesfound the following six exposures associated withIPF: ever-smoking, agriculture farming, livestock,wood dust, metal dust, and stone/sand. ILD is anoccupational disease in coal miners; sandblasters;and workers exposed to asbestos, tungsten carbide,beryllium, and other metals, 41 suggesting that atleast some cases of idiopathic UIP represent pneumoconioses.The considerable variability that existsin the development of PF among workers exposedto similar concentrations of fibrogenic/organicdusts implies that genetic factors are likely importantin modulating the lung injury. 42Chronic aspiration secondary to gastroesophagealreflux (GER) is a possible cause (or contributoryfactor) in the pathogenesis of IPF. 48−50 PF is acommon complication of systemic sclerosis (scleroderma),a disorder with an extremely high prevalenceof esophageal dysmotility. 51 Aspiration ofstomach contents (acid or not), leads to lung injuryand fibrosis. 41,52 Esophageal reflux has been notedin more than two thirds of patients with IPF awaitinglung transplant (LT). 48−50,53 Importantly, 30 to50% of IPF patients with GER have no symptomsof reflux. 48,49,53 Further, the severity of IPF does notcorrelate with the severity of GER. 48,49 Among LTrecipients (with or without IPF), GER can causeallograft injury 52,54 and appears to be a risk factorfor bronchiolitis obliterans syndrome. 54,55In light of these findings, aggressive treatmentof GER is recommended to reduce the chance forrepetitive lung injury, which may elicit fibrosis. Ina small series of patients with early IPF, 48 aggressivetreatment of GER was associated with stabilizationor improvement of lung function. Further,retrospective studies 56,57 of LT recipients found thatearly fundoplication (within 3 months of LT) wasassociated with greater freedom from bronchiolitisobliterans syndrome and improved lung function 58in patients with GER. Additional studies arerequired to assess the role of GER or aspiration inthe pathogenesis or progression of IPF and therapeuticstrategies to prevent or reduce GER.Several lines of evidence suggest that geneticfactors are important in the pathogenesis of IPF.Clusters of IPs/fibrosis in families (familial interstitialpneumonia [FIP]) have been noted in 0.5 to3.7% of patients with IPF. 38,46,59−62 The mode oftransmission of FIP is not known but is believed tobe autosomal dominant with variable/reducedpenetrance. 46,59,61,62 Familial IP is indistinguishablefrom nonfamilial (sporadic) IPF, but patients tendto be younger with the familial variant. 60,61,63 Interestingly,multiple types of IPs have been noted inFIP (eg, UIP, NSIP, DIP, COP; 46,64). In one study 46of 78 patients with histologically confirmed FIP,histologic patterns included the following: UIP(86%), NSIP (10%), COP (2.5%), and unclassified(1.3%). Forty-five percent of pedigrees exhibitedmore than one histologic pattern among affectedfamily members. 46 In familial IPF, asymptomatic636 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6367/10/09 10:24:51 PM


ILD may progress to symptomatic IPF during aperiod of decades. 64Interestingly, gene expression profiles of lungtissues from patients with familial IP (either UIPor NSIP) exhibit striking similarities but differ fromgene expression profiles in sporadic IPF. 65−67 Mutationsin genes encoding proteins involved inchronic inflammation (ie, chemokines, complement,and Igs); smooth-muscle markers (ie,smooth-muscle cells and myofibroblasts); and/ormatrix mobilization and resolution 65−67 may beimportant in the pathogenesis of IPF. Genetic polymorphismsfor interleukin (IL)-1 receptor antagonistor tumor necrosis factor (TNF)-, 68 complementreceptor, 1,69 or transforming growth factor (TGF)- 70may influence risk of IPF or disease progression.Interestingly, polymorphism of the angiotensinogengene was associated with IPF progression butnot with disease predisposition. 71 Mutations in thegene for surfactant protein C may cause FIP(including UIP, NSIP, and DIP. 72−76Germ-line mutations in the genes hTERT andhTR, which encode telomerase reverse transcriptaseand telomerase RNA, were implicated indyskeratosis congenita, a rare hereditary disorderassociated with PF and aplastic anemia. 77 Thesemutations result in telomere shortening, whichlimits the replicative capacity of tissues and hasbeen implicated in age-related disease. 78 Interestingly,older age and smoking also cause telomereshortening. 77 The authors of two studies 79,80 foundthat short telomeres were more common in FIP andsporadic IPF compared with control patients, evenwhen mutations in hTERT and hTR were lacking.PF may also complicate diverse genetic disorderssuch as Hermansky-Pudlak syndrome. 81 familialhypocalciuric hypercalcemia, 82 and neurofibromatosis.83 IPF occurs in white and in nonwhite populations;the prevalence among different ethnicgroups has not been studied. 3 However, in theUnited States, mortality rates caused by IPF weregreater in non-Hispanic white compared withHispanic or black populations. 36 A retrospectivestudy 84 of IPF in New Zealand cited a lower incidencein those of Maori or Polynesian descent thanin those of European descent.Differences in susceptibility to fibrogenicagents may reflect genetic polymorphisms. 59,60Animal models involving different inbred strainsof rodents demonstrate dramatic variability in thelung inflammatory/fibrotic response to injuriousagents. 85,86 The aforementioned studies suggest thatIPF is a heterogeneous disorder caused by a numberof environmental/occupational exposures incombination with genetic predispositions.Clinical Features of IPFInitial symptoms of IPF are cough (typicallynonproductive) and dyspnea. 2 Over time, the coughmay become paroxysmal and debilitating, anddyspnea and exercise limitation worsen. Physicalexamination reveals end-inspiratory rales (oftenwith a “Velcro” quality) in 85% of patients withIPF. 2,28 Clubbing is noted in 25% of patients. 2,3,28Extrapulmonary involvement does not occur 2 andshould suggest other disorders (particularly CTDassociatedpulmonary fibrosis [PF]). 87 However,certain diseases, such as ischemia cardiacdisease, 88−90 deep venous thrombosis, 88 diabetesmellitus, 91,92 and GER, 49,91 are more common inpatients with IPF. The onset of the disease is usuallyindolent, but IPF progresses inexorably overmonths to years. 2,3,25,27,28The course is highly variable, with somepatients maintaining stability for years, 2,93 whereasin others the course is rapid, with fatal respiratoryfailure evolving over a few months. 94 Additionally,some patients have gradual progression over years,followed by acute exacerbations associated withabrupt and often fatal hypoxemia respiratory failure.93,95 Spontaneous remissions do not occur. 2,3Mean survival from the onset of symptoms approximates3 to 5 years. 25,26,28,96 Fewer than 15% ofpatients survive 10 years from the onset of symptoms.25−28 Older age 27,28,96 and more severe impairmentsin pulmonary function or HRCT scans 27,28,96,97are associated with a worse prognosis. Some studies27,28,45,94,98 cited greater rates of mortality in men,but others 28 did not.The major cause of death is respiratory failure99,100 ; surveys of deaths among IPF patients inthe United Kingdom and United States noted thatprogression of lung disease accounted for 72% 100and 60% 36 of deaths, respectively. Other causesinclude pulmonary embolism, 99,101 lung cancer, 99,101,102cardiac failure, cerebrovascular accidents (primarilyin the elderly). 101 Acute respiratory failure requiringmechanical ventilation may complicate IPF,either from infection or an acute exacerbation of<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 637ACC-PULM-09-0302-038.indd 6377/10/09 10:24:52 PM


the underlying disease process. 95,103−106 In this context,mortality is high ( 90%). For this reason,mechanical ventilation is usually ill-advised inpatients with severe IPF. Unfortunately, medicaltherapy has not been shown to alter the course ofthe disease.Lung Cancer Complicating UIPLung cancer occurs in 4 to 13% of patients withIPF. 102,107−109 The risk is greater in smokers, 107,109 butthe heightened risk of lung cancer is not solely aresult of the effects of cigarette smoking. 102,107 Surgicalresection may be curative in IPF patients withnon-small cell lung cancer, but postoperative morbidityand mortality is increased. 110,111Acute Exacerbations of UIPA subset of patients with IPF have an acceleratedcourse of the disease, often as a terminal event,characterized by severe hypoxemia, dyspnea, andpulmonary infiltrates on chest radiographs or CTscans, 95,104,112−114 This syndrome has been termedacute exacerbation of IPF. 95 Clinical and radiographicfeatures resemble the ARDS. 95,104,112,113 The cardinalhistologic feature on lung biopsy or necropsy isdiffuse alveolar damage and/or organizing pneumoniasuperimposed on a background of UIP. 104,113Idiopathic AIP 15,16 exhibits similar clinical and histologicfeatures as acute exacerbation of IPF, but itlacks the requisite features of UIP. The factorsresponsible for this accelerated phase of IPF areunknown, but viral infections, high concentrationsof oxygen, or drug reactions are plausible etiologicfactors. 16 This syndrome is usually fatal. Favorableresponses have been noted with high-dose IV pulsemethylprednisolone, but data are limited to anecdotalcases and small series. 15,95,104,112−114we obtain serologies for CTD (eg, antinuclear antibodyand antibodies to SSA, SSB, Scl-70 [scleroderma],Sm, RNP, Jo-1, double-stranded DNA) andhypersensitivity pneumonitis (HP) to rule outthose disorders. Elevations of the glycoproteinKL-6 115,116 and lung surfactant proteins A and D 116−118have been noted in serum and BAL fluid in patientswith IPF and may have prognostic value. Theseassays are available in only a few research laboratories,and additional studies are required to assesstheir specificity and clinical role.Chest RadiographsChest radiographs in IPF typically reveal bilateralinterstitial (reticular) infiltrates with a predilectionfor the basilar and peripheral (subpleural)regions of the lung (Fig 1). 2,3 With progression ofthe disease, all lung fields are affected and lungvolumes decrease. Similar radiographic featuresmay be observed in asbestosis and CTD-associatedPF. 119,120 <strong>Pulmonary</strong> hypertension and cor pulmonalemay be seen in far-advanced cases. 121 Intrathoraciclymphadenopathy and pleural thickeningare not features seen on plain chest radiographsbut may be noted on CT. 122 Chest radiographs havelimited prognostic value, but serial radiographsLaboratory TestsLaboratory test results in IPF are nonspecific.The erythrocyte sedimentation rate is increased in60 to 94% of patients; circulating antinuclear antibodiesor rheumatoid factor have been found in 10to 26% of patients with IPF. 2,3 These serologic studies2,3 do not correlate with the extent or activity ofthe disease and do not predict therapeutic responsiveness.However, for new cases of suspected IPF,Figure 1. IPF/UIP. Posteroanterior chest radiograph froma 61-year-old man with IPF/UIP demonstrates extensivereticulonodular infiltrates with a bibasilar predominance.A mixed alveolar and interstitial pattern is evident in the lowerlobes.638 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6387/10/09 10:24:52 PM


(including old films) may be used to gauge the paceand evolution of the disease. 2Thin-Section HRCT ScansHRCT Features and Value in the Differential Diagnosis:HRCT scans, using thin (1- to 2-mm) sectionswithout the use of contrast, are far more accuratethan conventional chest radiographs in assessingthe extent and nature of the disease. 123,124 HRCT hasdiagnostic 24,125,126 and prognostic 98,123,124,127 value andshould be part of the initial evaluation of suspectedILD. The HRCT features of IPF/UIP are stereotypicand predictable and include a patchy, heterogeneousdistribution with a predilection for theperipheral (subpleural) and basilar regions of thelungs; small cystic radiolucencies (ie, honeycombchange [HC]); coarse reticular or linear opacities(intralobular and interlobular septal lines); raggedpleural surfaces; irregular or thickened bronchialwalls or pulmonary vessels; and bronchiectasis andbronchiolectasis (Figs 2-6). 23,121Air bronchograms (1 to 2 mm in diameter)reflect dilated peripheral airways surrounded byfibrotic lung tissue. Bronchovascular thickening, acardinal feature of sarcoidosis, is minimal or absentin UIP. 121 Focal ground-glass opacities (GGOs)—hazy zones of increased alveolar attenuation—maybe present in IPF/UIP 127 but are not a dominantfeature. 23 HC is highly characteristic of IPF/UIPbut may be absent in mild cases. 22,127 In severe cases,severe volume loss, anatomic distortion, anddilated pulmonary arteries are observed. Zones ofFigure 3. HRCT from a 53-year-old man with UIP confirmedon open-lung biopsy specimen. Note extensive areas of GGOswith air-bronchograms in the lower lobes. Despite extensiveinvolvement of the posterior regions of both lower lobes, theanterior portion of the left lower lobe is relatively normal,emphasizing the patchy nature of the disease.Figure 4. HRCT from a 63-year-old woman with UIPconfirmed on open-lung biopsy specimen. HC is evident,particularly in the peripheral (subpleural) regions. Dilatedbronchi, indicating traction bronchiectasis, are also present.There are no GGOs.Figure 2. UIP. HRCT demonstrates foci of GGOs in the subpleuralregions of the lower lobes (arrows). A few small honeycombcysts are also present. The central portions of thelung are relatively spared.emphysema (particularly in the upper lobes) maybe present concomitantly in smokers. 128–131 Mediastinallymph node enlargement (LNE) has beennoted in 53 to 93% of patients with IPF 24,132–135 butis nonspecific. 134,135 LNE in IPF usually involvesonly one or two nodal stations, and the nodes usuallymeasure 15 mm. 133,135 The presence of LNEcorrelated with the extent of disease in some studies133,134 but not all studies. 122,132 One study 133 noted<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 639ACC-PULM-09-0302-038.indd 6397/10/09 10:24:53 PM


Figure 5. HRCT from a 44-year-old man with severe UIPon open-lung biopsy specimen. Note the marked cysticradiolucencies (some cysts are 3 cm in diameter) and honeycombinginvolving both lungs. Although most of the lungparenchyma has been destroyed, note the accentuation of thecystic disease process in the peripheral (subpleural) regions.Figure 6. HRCT from a 48-year-old woman with UIP, withextensive cystic radiolucencies (honeycomb cysts) scatteredthroughout the lung parenchyma. Despite treatment withcorticosteroids, AZA, and CP, she died of progressive respiratoryfailure.that increase in LNE over time was associated withprogression of fibrosis in UIP or NSIP.Differential Diagnosis by HRCT: Focal GGOsmay be observed in UIP, but the presence of predominantor extensive GGOs suggests an alternativediagnosis such as DIP, 136 HP, 4,137,138 cellularNSIP, 4,11,13,139 COP, 19 chronic eosinophilic pneumonia,140 or pulmonary alveolar proteinosis. 141 HC isa cardinal feature of UIP 121 but is absent in AIP 15and DIP. 4,136 Honeycombing is not a prominentfeature in NSIP, but focal HC may be seen in fibroticvariants of NSIP. 22,139 Cystic radiolucencies may beobserved in other disorders (eg, Langerhans-cellhistiocytosis, 142 sarcoidosis, 143 lymphangioleiomyomatosis,144 chronic HP, 145 LIP, 18 pneumoconiosis),but the distribution of lesions and presence ofconcomitant abnormalities differentiates thesedisorders from UIP. 146.Can HRCT Obviate the Need for SLB in IPF?:HRCT may obviate the need for SLB, provided theCT features are classical for UIP (including presenceof honeycombing. 22,24,98,125 The accuracy of aconfident diagnosis of UIP by a trained observeris 95%. 22,24,98, 125 However, classical features of UIP(confident diagnosis by CT) are present in only 34to 67% of patients with histologically confirmedUIP. 2,24,125,127 Interobserver and intraobserver variabilitymay be problematic 127 ; consensus is usuallyreached when extensive HC and subpleural predominanceare deemed to be present. 22 CT patternsoverlap between UIP and fibrotic NSIP. 139 Subpleuralpredominance, reticulation, and HC maybe found in fibrotic NSIP, and focal GGOs may beobserved in UIP. 127,139 Predominant GGOs make thediagnosis of UIP unlikely. 139,147When CT features are typical for UIP, specificityapproaches 100%. 22,24 In one study, 22 two radiologistsindependently assessed CT scans from acohort of patients with either idiopathic UIP(n 73) or NSIP (n 23). CT scans were categorizedas definite UIP, probable UIP, indeterminate,probable NSIP, or definite NSIP. CT features weretypical (definite or probable) of UIP in 27 of patients(37%); with histologic UIP on SLB. Importantly, all27 patients with a CT pattern of probable or definiteUIP had histologic UIP on SLB. In contrast, only18 of 44 patients (41%) with probable or definiteNSIP on CT had histologic NSIP on SLB (59% hadUIP). 22 These data suggest that CT features of UIPare associated with advanced, late-stage disease.Among patients with earlier phases of UIP, CTfeatures may be atypical 148 or indeterminate. 22When HRCT features are classical for UIP (includingHC), SLB is not necessary. However, SLBshould be performed when CT scans are nondiagnosticor atypical unless specific contraindicationsexist. 98Prognostic Value of HRCT in IPF: The prognosisof IPF can be inferred from the extent and nature(ie, predominant pattern) of HRCT abnormalities.123,124,127,148,149 The extent of disease on CT640 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6407/10/09 10:24:54 PM


correlates roughly with severity of functionalimpairment (ie, FVC and diffusing capacity of thelung for carbon monoxide [Dlco]) 128,150 ; the patternof CT does not. 150 When serial CT scans were performed,changes in CT were concordant withphysiologic parameters (eg, FVC and Dlco). 128,151PFTs improved only when GGOs regressed onHRCT 128,151 ; extensive fibrosis or HC on HRCTpredicted the progression of disease. 123,124,151Early reports 149,151–154 noted that a predominantGGO pattern in patients with IPF correlated withgreater rates of response to corticosteroid therapy(improvement in 33 to 44% of cases) and greatersurvival compared with reticular or HC patterns.However, it is likely that patients with extensiveGGO who were responsive to corticosteroidshad histologic patterns distinct from UIP, such asDIP, 136 NSIP, 139,155 or HP. 4,137 Combinations of patterns(eg, GGO, reticular, HC) may be present inindividual patients. 2,121 In some patients, GGOsevolve to HC or reticular lines. 149,151,154 Reticularpatterns or HC do not regress and may worsenover time. 23,128,130,149,156Severe HC on CT is a strong predictor of mortality( 80% mortality within 2 years). 98,123,124 Britishinvestigators 124 evaluated prognostic variablesin a cohort of 115 patients with UIP awaiting LT.By multivariate stepwise regression analysis, onlyDlco percentage of predicted and CT-fibrotic (CTfib)scores independently predicted 2-year survival.Receiver operating curve analysis gave the best fit(predictive value) with a combination of Dlco andCT-fib scores. The optimal points on the receiveroperating curves discriminating survivors fromnonsurvivors corresponded to a Dlco of 39% ofpredicted and CT-fib scores 2.25. 124 The curveresulting from a model combining these twoparameters yielded a specificity and sensitivity of84% and 82%, respectively, for discriminating survivorsfrom nonsurvivors.These data are remarkably similar to our owndata. 123 We extended our experience in a cohort of106 patients with UIP, 98 which confirmed that CTfibscore 2.0 was a significant predictor of mortality.A mean CT-fib score 2.0 in any lobe wasassociated with increased mortality (risk ratio, 3.35)compared with patients with a CT-fib score 2 inall lobes. A similar finding was noted when patientswere segregated by mean CT-fib 2 for all lobes(risk ratio, 2.5; p 0.002). However, when thehistologic diagnosis was incorporated into themodel, the diagnosis of UIP was the most powerfulpredictor of mortality and overshadowed the predictiveability of the CT-fib score. 98In a multicenter prospective study 23 of 315patients with IPF, greater CT-fib scores and lowerDlco correlated with increased mortality risk (bymultivariate analysis). Other investigators 148,155noted that CT patterns typical of IPF/UIP wereassociated with a greater rate of mortality comparedwith atypical CT scans, suggesting that CTfeatures typical of IPF reflect advanced disease.However, in a recent study, 127 the authors retrospectivelyreviewed 98 patients with a histologicdiagnosis of UIP who underwent CT scans. Patternsof CT scans were categorized as follows:(1) definite UIP, (2) probable UIP, and (3) suggestiveof alternative diagnosis. Mean survival rates were45.7, 57.9, and 76.9 months, respectively. Mediansurvival rates were 34.8, 43.4, and 112 months,respectively. Although these differences betweengroups did not achieve statistical significance,these data suggest that patients whose CT scansare interpreted as definite UIP have a worse prognosis.By multivariate analysis, the extent of tractionbronchiectasis and fibrosis scores influencedprognosis. More complex HRCT scoring systemsthat use stereologic methods correlate well withopen-lung biopsy 157 but are logistically difficult andof limited clinical utility. In summary, extensiveHC or reticulation on CT and no or minimal GGOsuggests a low likelihood of response to medicaltherapy and high mortality (98,123,124). Unlesscontraindications exist, these patients should belisted for LT. 158Several studies 128,130,155,156 assessed evolutionof CT over time. <strong>Pulmonary</strong> functional parametersimproved only when GGO regressed onHRCT. 128,130,156 Serial PFTs are more useful thanserial CT scans to assess prognosis. In one study 155of 60 patients with IPF, we found that increasingextent of GGO on CT ( 10% increase from baseline)at 6-month follow-up was associated withgreater rates of mortality, whereas change in CT-fibscore did not predict long-term survival. By multivariateanalysis, decreases in FVC at 6 monthsand UIP diagnosis were associated with worsesurvival compared with patients who had stableor improved FVC or a histologic diagnosis ofNSIP. 155 Changes in CT-alveolar or CT-fib scores<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 641ACC-PULM-09-0302-038.indd 6417/10/09 10:24:55 PM


did not contribute further information once thesefactors were included in the multivariate model.Given the potential for fibrosis to evolve over time,the value of CT in predicting long-term prognosisis modest. 22PhysiologyCharacteristic physiologic aberrations in IPFinclude reduced lung volumes (FVC and total lungcapacity [TLC]) consistent with a restrictive defect;normal or increased expiratory flow rates; impairedoxygenation, and impaired gas exchange. 3,159 TheDlco is reduced, often disproportionately to otheraberrations. 129,150,159 The Dlco is an indirect reflectionof the pulmonary vasculature; reductionsreflect destruction or loss of alveolar walls or capillaries.In nonsmokers with IPF, expiratory flowrates are preserved, and the ratio of FEV 1to FVCis increased. When emphysema also is present,lung volumes (ie, FVC and TLC) are preserved, andDlco is disproportionately reduced. 128,129,131,150The measurement of the Dlco is complex andvaries even among individual patients. Normalizingthe Dlco by the lung volume, yielding theDlco/lung volume ratio, does not improve thepredictive value of Dlco and is not necessary. 159Reduced compliance and alterations in pressurevolumecurves are characteristic of IPF, 28 but thesesophisticated studies require an esophageal balloonand lack clinical value. Impaired oxygenation,which is accentuated with exercise, is a cardinalfeature of IPF. 28,159,160 Early in the course of the disease,Pao 2may be preserved at rest but invariablyworsens with exercise. As the disease advances,hypoxemia at rest is a nearly universal feature.Cardiopulmonary exercise tests (CPETs) revealhypoxemia or widened alveolar-arterial oxygenpressure difference [P(A-a)O 2]; reduced oxygenconsumption; increased dead space; increasedminute ventilation for the level of oxygen consumption;high-frequency, low-tidal volumebreathing pattern; and low oxygen puls. 28,159,161,162Impact of Physiologic Tests on SurvivalNot surprisingly, severe derangements in PFTsor oxygenation predict a worse prognosis andlower rates of survival. 2,159 Survival is worse inpatients with severe impairments in vital capacity(VC) or Dlco. The 3-year rate of mortality is > 50%when VC decreases to 60% of predicted or Dlcodecreases to 45% of predicted. 2,159 Changes inTLC are less predictive of survival. 2,159 However,the prognostic value of physiologic parameters ata single time point is limited. 93 Further, disparateresults have been reported from different centers.For example, in one retrospective study 163 of 53patients with UIP, decreased percentage of predictedDlco was independently linked to the rateof mortality (p 0.01), whereas three other studies28,164,165 found no such relationship. In those studies,the following parameters correlated with ratesof mortality: percentage of predicted FVC (p 0.05)and widened P(a-a)O 2(p 0.005) 165 ; multistagePao 2on CPET (p 0.006) 164 ; and reduced lung volumesand abnormal oxygenation during maximalexercise. 28British investigators 124 examined 2-year survivalamong a cohort of 115 IPF patients awaiting LT. Thebest predictors of survival (assessed at 2 years) wereas follows: Dlco > 39% of predicted and less fibrosison HRCT scan. 124 In a separate study 166 by theseinvestigators, 116 nonsmokers with IPF wereprospectively assessed to evaluate predictors ofprognosis. The following variables predicted survival(by multivariate analysis): percentage ofpredicted TLC, percentage of predicted Dlco, age,and clearance of inhaled 99m Tc-diethylenetriaminepenta-acetic acid (DTPA) from the lungs.Several investigators 28,160,167−174 have found thathypoxemia (at rest or exercise) is an independentpredictor of mortality in patients with IPF orfibrotic NSIP. The 6-min walk test (6MWT) withoximetry is a noninvasive and relatively inexpensiveway to ascertain prognosis and need forsupplemental oxygen therapy and follow thecourse of IPF. We 160 performed the 6MWT in 83patients with UIP and 22 patients with NSIP;desaturation to 88% strongly correlated withmortality. Among patients with UIP, desaturationwas associated with increased risk of death (hazardratio, 4.2; p 0.01) after adjusting for age, sex,smoking, baseline Dlco, FVC, and resting oxygensaturation. 160 Further, 6-min walk distance (6MWD)correlates with percentage of predicted Dlco 124,168,170and has prognostic value.In one study 169 of IPF patients awaiting LT,survival time was shorter among patients with6MWD 350 m. A subsequent study 168 by these642 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6427/10/09 10:24:55 PM


investigators retrospectively analyzed the utilityof 6MWD as a predictor of mortality in a cohort of454 IPF patients awaiting LT. Lower 6MWD wasassociated with increased mortality (assessed at 6months) and was superior to FVC percentage ofpredicted as a predictor of mortality. Patients with6MWD 207 m had a more than fourfold-greatermortality rate than those with 6MWD 207 m,even after adjustment for demographics, FVCpercentage of predicted, pulmonary hypertension,and medical comorbidities. 168 Flaherty et al 175assessed the prognostic value of 6MWD and extentof desaturation on 6MWT in a cohort of 197patients with IPF. By multivariate analysis, 6MWDwas not a reliable predictor of mortality, but thedegree of desaturation during 6MWT had greaterprognostic value.Patients with oxygen saturation 88% duringtheir initial 6MWT had a median survival of 3.2years compared with 6.8 years for those with baselinearterial oxygen saturation 88% (p 0.006).The prognostic value of serial changes in FVC,Dlco, 6MWD, and desaturation area (DA) variedaccording to the extent of desaturation on 6MWT. 175Among patients with an oxygen saturation 88%during 6MWT, a decrease in Dlco emerged as theonly predictor of mortality; changes in FVC, DA,or walk distance were not predictive in this group.For patients with an oxygen saturation 88% during6MWT, decreases in FVC and increases in DApredicted subsequent mortality, whereas decreasesin 6MWD and Dlco were less predictive.These data emphasize the importance ofstratifying by baseline level of desaturation whenassessing prognosis. Recently, a 6-min step dosewas advocated as another way of assessing exercisecapacity and prognosis in patients with IPF orother ILDs. 176 Formal CPET provides additionaldata, including measurement of maximal oxygenuptake (V . o 2), an integrated measure of respiratory,cardiovascular, and neuromuscular function. 162One recent study 161 evaluated V . o 2as a predictor ofsurvival in a cohort of 117 patients with IPF.Patients with baseline V . o 2 8.3 mL/kg/min hadan increased risk of death after adjusting for age,smoking status, FVC, and Dlco. Further, V . o 2wasa stronger predictor than desaturation 88% on6MWT. However, V . o 2did not predict survivalwhen examined as a continuous variable. However,CPET with arterial cannulation is invasive,logistically difficult, difficult to perform for somepatients, and lacks practical value.Clinical-Radiographic-Physiologic ScoringSystemsWatters et al 177 developed the clinical-radiographic-physiologic(CRP) scoring system, a compositescore incorporating clinical (dyspnea),radiographic (chest radiographs), and physiologicparameters to more objectively monitor the courseof IPF. A modification of the CRP score (arbitrarytotal of 100 points) was developed; it predictedsurvival better than physiologic or radiographicfeatures alone. 28 In one prospective study 28 of 238patients with UIP, a comprehensive CRP score(obtained at the time of initial visit) was used topredict survival. Parameters included age, fingerclubbing, smoking history, the extent of profusionof interstitial opacities and pulmonary hypertensionon chest radiographs, percentage of predictedTLC, and Pao 2during maximal exercise. In addition,an abbreviated CRP score, excluding Pao 2during maximal exercise, was applied.The comprehensive CRP score was superior tothe abbreviated CRP score, but both scoring systemshad excellent prognostic value. The modifiedCRP score predicted 5-year survival with remarkableaccuracy. 28 Five-year survival rates at CRPscores of 20, 40, 60, and 80 points were 89%, 53%,4%, and 1%, respectively. The abbreviated CRPwas less accurate but more adaptable to clinicalpractice. Although HRCT data were not includedin that study, it is likely that incorporating HRCTdata in place of chest radiographs would enhancethe predictive value of such systems. Further, it canbe argued that other parameters (eg, FVC or Dlco)could be substituted in place of TLC.British investigators 165 developed a compositephysiologic index (CPI) incorporating CT andphysiologic parameters. The CPI score evaluateddisease extent observed by CT and selected functionalvariables (eg, percentage of predicted FVC,Dlco, and FEV 1). Exercise components were notincluded in this index. The CPI accounts for coexistingemphysema, which may confound pulmonaryfunctional indexes. Importantly, CPI predictedthe rate of mortality better than PFTs or CT parametersin isolation. Further, the CPI was comparedwith the original 177 or modified 28 CRP scoring<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 643ACC-PULM-09-0302-038.indd 6437/10/09 10:24:55 PM


systems in 30 patients with UIP who underwentexercise testing. The CPI was a superior predictorof outcome compared with the physiologic componentof the original CRP score (p 0.02) and thephysiologic component of the modified CRP score(p 0.009). The CPI score is promising but may notbe easily adapted into clinical practice settings.Additional studies using these or similar CRP scoringsystems would be of interest.What Is the Role of Sequential PhysiologicStudies To Follow UIP?Sequential physiologic studies are critical toassess the evolution of the disease. I obtain serialmeasurements of FVC, Dlco, and oxygen saturation(by pulse oximetry) at 3- to 4-month intervalsto follow the course of the disease. 2 The FVC islogistically easy to perform, relatively inexpensive,and less variable than TLC or Dlco. Thus, FVC isan ideal parameter to follow in individual patients.Because of inherent variability, serial changes inDlco are less reliable. Formal CPETs with arterialcannulation are inconvenient, expensive, and lackpractical value. 2 Serial 6MWTs with oximetry aremore convenient and cost-effective for monitoringthe course of the disease 160 and assessing the needfor supplemental oxygen. Although optimalparameters to assess response to therapy have notbeen validated, the American Thoracic Society(ATS) defined improvement as 10% increase inFVC or TLC, 15% increase in Dlco, or 4%increase in oxygen saturation or 4 mm increasein Pao 2during exercise. 3 Improvement or stabilityin VC or Dlco with corticosteroid therapy is associatedwith improved prognosis. 123,178,179 Conversely,deterioration in VC or Dlco at 3 months, 179 1 year, 178or later time points 180 predicts a greater rate ofmortality.In a retrospective study, 155 we assessed thevalue of serial PFTs in 80 patients with UIP and 29patients with NSIP. For patients with UIP, a changein FVC was the best physiologic predictor of mortality.By multivariate analysis, 10% decrease inFVC at 6 months was an independent risk factorfor mortality (hazard ratio, 2.47; p 0.006). 155 Inaddition, 10% increase in GGO on HRCT (ie, CTalveolar)at 6 months predicted a greater risk ofsubsequent death (relative risk, 2.88; p 0.001)compared with 10% change. 155 A change in Dlcoat 6 months did not add independent prognosticvalue. Changes in CT-fib or TLC did not predictlong-term survival when adjusted for correspondingbaseline values. Further, baseline alveolararterialgradient [P(A-a)O 2] was not a predictor ofsurvival.Collard et al 181 evaluated the prognostic valueof serial clinical (dyspnea score) and physiologicparameters in 81 patients with IPF. (All had histologicallyconfirmed UIP.) Survival correlated withdyspnea scores and pulmonary functional parameters[FVC percentage of predicted, P(a-a)O 2] atbaseline, as well as changes in these parameters at6- or 12-month time points. Not surprisingly, survivalwas worse among patients with deterioratingdyspnea scores or PFTs at 6 or 12 months. 181 Britishinvestigators 172 retrospectively reviewed the prognosticsignificance of histopathologic diagnoses,baseline PFTs, and serial trends in pulmonaryfunctional indexes (eg, FVC, FEV 1, Dlco) at 6 and12 months in 104 patients (UIP, n 63; fibroticNSIP, n 37). Survival was better in fibrotic NSIPcompared with UIP (p 0.001) but not in patientswith severe functional impairment. Mortality duringthe first 2 years was linked solely to the severityof functional impairment at presentation (ie,lower Dlco and FVC levels). The CPI score 165 wasthe strongest determinant of outcome (p 0.001). 172At 6 months, serial PFTs and histopathologic diagnosiswere prognostically equivalent. 172 However,at 12 months, serial PFT trends (Dlco, FVC, FEV 1,CPI) predicted mortality better than any othercovariates including histologic pattern (allp 0.0005). In this context, change in Dlco providedthe best prognostic information (2-yearsurvival); histologic pattern provided no additionalprognostic value. However, for late deaths (beyond3 years), the risk of mortality was greater with UIPthan with fibrotic NSIP.<strong>Pulmonary</strong> Arterial HypertensionComplicating IPF<strong>Pulmonary</strong> arterial hypertension (PAH) andright ventricular (RV) dysfunction are common inpatients with IPF. 182−184 The pathogenesis of PAH inIPF is complex and does not correlate with lungvolumes or extent of pulmonary dysfunction. 183,185−188<strong>Pulmonary</strong> artery remodeling and proangiogeniccytokines are likely central to developing PAH in644 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6447/10/09 10:24:56 PM


IPF, 183,185,189 but ablation of pulmonary vessels andvasoconstriction may play contributory roles. 183 Inpatients with advanced IPF, PAH was noted in 20to 84% of patients. 131,182,187,190−193 However, one prospectivestudy 186 of patients with all stages of IPFnoted PAH (mean pulmonary artery pressure[mPAP] > 25 mm Hg by right-heart catheterization[RHC]) in only 8.1%. One recent study 184 of 44patients with IPF awaiting LT found that mPAPincreased progressively over time (rate of change3.8 mm Hg per month in the entire cohort). Baselinelung function, oxygen requirements, or mPAP didnot predict the rate of change in mPAP nor whichcases progressed to New York Heart Associationfunctional class IV.The presence of PAH is associated with markedlyworse survival. 131,186,187,190,194 In one study, 190 88patients with IPF had estimates of systolic pulmonaryartery pressure (sPAP) by transthoracic echocardiography(TTE). Increased sPAPs were notedin 84% of patients; sPAP exceeded 50 mm Hg in 27patients (31%). Median survival was significantlyworse among patients with sPAP 50 mm Hg (0.7years) compared with patients with sPAP 35 mmHg (4.8 years) or sPAP of 36 to 50 mm Hg (4.1years). Others parameters associated with worsesurvival by univariate analysis included malegender; lower Dlco, use of oxygen, history ofcoronary artery disease, and worse New YorkHeart Association functional class. 190 In anothercohort of 79 IPF patients, 1-year mortality rateswere 28% in those with PAH (defined as mPAP 25 mm Hg by RHC) compared with 5.5% in thosewithout PAH. 187 Another prospective study 186evaluated 61 patients with IPF who had RHC. Fiveyearsurvival was 83% among patients with normalpulmonary artery pressure (PAP) and preservedDlco ( 40% of predicted) compared with only16% among patients with high PAP, low Dlco, orboth (p 0.001). 186 Several investigators 183,187,188,191,193found that the Dlco correlates with PAH, whereasFVC or lung volumes do not.Screening for PAH among patients with IPF isreasonable because it has prognostic value and mayinfluence therapy. TTE is invaluable to assess RVsize and function and estimate sPAP and is superiorto chest CT as a predictor of PAH. 193 Severe RVhypertrophy, reduced RV function, and abnormalbowing of the interventricular septum onTTE reflect RV dysfunction. TTE may allow theestimation of sPAP, which has prognostic value.However, TTE has only modest predictive valuefor PAH. 193,195 In one study 195 of 106 patients withILD, estimates of sPAP by TTE and RHC were discordantby 10 mm Hg in 63% of patients. Accuracyimproved when sPAP was 45 mg Hg. 195We 193 performed TTE and RHC in 61 IPFpatients; estimation of RV systolic pressure waspossible in 33 patients (54%). By using a RV systolicpressure 40 mm by RHC as a cutoff, we foundthat sensitivity and specificities for PAH were 76%and 38%, respectively. Exercise-induced desaturation,reduced Dlco ( 40% predicted), or need forsupplemental oxygen are surrogate markers forPAH in IPF. 190,193 The FVC/Dlco ratio and oxygensaturation may be markers for PAH. 190,193 In acohort of IPF patients. 193 a model incorporatingFVC/Dlco ratio and oxygen saturation was superiorto TTE (improved specificity and negativepredictive value) in assessing PAH. Dlco did notcontribute to mPAP prediction above and beyondoxygen saturation, whereas FVC/Dlco ratio did.However, RHC is invasive, and its role in patientswith IPF is controversial. RHC is not necessary inpatients with IPF when TTE results are normal butshould be considered when TTE results suggestRV dysfunction or PAH.Reduced 6MWT distance and increased plasmabrain natriuretic peptide (BNP) are surrogate markersof PAH in patients with PF. 188,194 In one study, 18839 patients with PF and severe restrictive lungdisease (FVC 55% predicted, including 28 withUIP) underwent RHC, measurement of BNP levels,and 6MWT. Among 28 patients with IPF, increasedBNP concentrations correlated with increased PAPand pulmonary vascular resistance and correlatedinversely with 6MWT distance. Importantly, PFTresults did not differentiate between patients withnormal and elevated PAPs. Korean investigators 194retrospectively analyzed 131 IPF patients with bothTTE and plasma BNP measurements; sPAP 40cm H 2O was used as the threshold for PAH. BothPAH and elevated BNP were independent predictorsof mortality. One-year mortality and meansurvival rates were worse among patients withPAH (61% and 10.8 months) compared with 20%and 23.7 months among patients without PAH.Further, prognosis was much worse amongpatients with elevated BNP levels compared withthose with normal BNP levels (1-year mortality<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 645ACC-PULM-09-0302-038.indd 6457/10/09 10:24:56 PM


70.5% vs 23.7%; mean survival 11.0 months vs 22.5months, respectively). The combination of bothPAH and BNP was a more robust marker of mortalitythan either value alone. These data 188,194 suggestthat plasma BNP concentration may be a useful anoninvasive marker to screen for PAH.The impact of treating PAH in patients withIPF has not been elucidated. Favorable responsesto sildenafil (a phosphodiesterase inhibitor), 192,196prostacyclin, 182 iloprost, 197 or endothelin-1 (ET-1)receptor antagonists 198 have been noted in IPFpatients with PAH, but impact on long-term outcomehas not been established. A multicenterplacebo-controlled trial sponsored by the Idiopathic<strong>Pulmonary</strong> Fibrosis Clinical Research Networkevaluating sildenafil for severe IPF is inprogress. The use of ET-1 receptor blockers hastheoretical value in IPF-associated PAH, 182,183 butdata are limited. In a double-blind randomizedcontrolled trial (DBRCT), 198 158 patients with IPFwere randomized to bosentan (an endothelin-1receptor antagonist) or placebo. Patients withsevere disease (FVC 50% of predicted or Dlco 30% of predicted or Pao 2 55 mm Hg) or PAHwere excluded. At 12 months, bosentan was notsuperior to placebo with regard to the primary endpoint (ie, 6MWD); further, physiologic parameters(FVC, Dlco, arterial oxygen saturation) did notdiffer between groups. However, a trend in favorof bosentan was noted in secondary end points(including time to death or disease progression[hazard ratio, 0.61, p 0.12] and quality of life anddyspnea scores). 198 Another randomized controlledtrial (RCT) 199 evaluating bosentan among IPFpatients with severe disease or PAH is in progress(Bosentan Use in Interstitial Lung Disease [BUILD-3]). Mediators such as TGF-, TNF-, plateletderivedgrowth factor, and other profibroticcytokines may be involved in the pathogenesis ofIPF-associated PAH. 183,200 Therapies targeting specificcytokines are worthy of study, 201 but data arelacking.Histologic FeaturesThe histologic pattern of UIP observed inpatients with IPF 3,5,6 can be found in other etiologies(eg, CTD, asbestosis, and diverse occupational,environmental, or drug exposures). 3,6 Historically,UIP was considered a subset of patients with IPF, 202but current recommendations restrict the term IPFto patients with idiopathic UIP. 3,6 In 2002, the ATSand European Respiratory Society published a classificationschema recognizing seven idio pathicinterstitial pneumonias (IIPs): UIP; NSIP, organizingpneumonia, AIP, RBILD, DIP, and LIP(Table 1). 4 UIP is the most common of the IIPs,comprising 47 to 71% of cases; NSIP accounts for13 to 48% of cases. 3,96,98,147,148,203−205 Clinical, physiologic,and radiographic features of UIP and NSIPoverlap; further, fibrotic NSIP and UIP may bedifficult to distinguish even by experiencedpathologists. 155,172,204,206,207 The remaining IIPs areless common and exhibit clinical features that arein sharp contrast to UIP. Salient features of theseother IIPs are reviewed elsewhere (eg, DIP, 10RBILD, 9,208 NSIP, 11,209,210 AIP, 4,16,112 COP, 19 and LIP 17,18 ).The literature is confusing, however, because thesevariants were often included in series of IPF, eventhough clinical, radiographic, and prognosticfeatures are disparate.UIP: The histologic diagnosis of UIP requiresan SLB, preferably with wedge biopsies from twoor three sites. 7,211 Care should be taken to avoid theworst areas (ie, areas with advanced HC). 7,211 Cardinalfeatures of UIP that distinguish this entityfrom other IIPs are temporal heterogeneity, profusionof fibroblastic foci (FF), and HC. 3,5,6 UIP exhibitsboth geographic and temporal heterogeneity.The lesions are bilateral but patchy and exhibit astriking predilection for the basilar and peripheral(subpleural) regions of the lung. This nonuniformdistribution can be appreciated at low-power magnification(Fig 7, top).In addition to geographic heterogeneity, thelesions also exhibit temporal nonuniformity. Areasof active injury, inflammation, and fibroblasticTable 1. Classification of IIPs*Histologic PatternUIPNSIPOPDiffuse alveolar damageRepiratory bronchiolitisDIPLIPCRP DiagnosisIPF/cryptogenic fibrosingalveolitisNSIP (provisional)COPAIPRB-ILDDIPLIP*Adapted from Amer J Respir Crit Care Med. 4646 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6467/10/09 10:24:56 PM


Figure 8. Photomicrograph of IPF. Open-lung biopsy specimenreveals end-stage honeycomb lung. The large cystic airspacesrepresent destroyed and coalescent alveolar septae.Minimal inflammatory cells are present at this late phase[reproduced with permission from Lynch JP III, Strieter RM.In: Internal medicine for the specialist. Montvale, NJ: MedicalEconomics, 1988; 61–84].Figure 7. Top: Photomicrograph of UIP in open-lung biopsyspecimen. Patchy subpleural fibrosis with dense scarringcause remodeling of lung architecture. Note the marked heterogeneity(hematoxylin-eosin) [reproduced with permissionfrom Lynch et al, J Respir Dis 2000; 21;197–214]. Bottom:Photo micrograph of UIP in an open-lung biopsy specimen.Fibrosis shows heterogeneity with dense eosinophilic collagenand a loose FF. The adjacent lung is relatively unaffected.These findings are consistent with UIP (hematoxylin-eosin)[reproduced with permission from Lynch et al, J Respir Dis2000; 21;197–214].proliferation coexist with areas of dense (old)fibrosis. Even within the same lobe, alternatingzones of interstitial inflammation, fibrosis, HC,and normal lung can be observed. Alveolar wallsare thickened by excessive collagen, extracellularmatrix, patchy mononuclear cell infiltrates(eg, lymphocytes, plasma cells), and fibroblasts. 3,6Intra-alveolar macrophages may be observedbut are not conspicuous. Scattered neutrophilsor eosinophils may be present. These inflammatorychanges are not prominent and are usuallyconfined to areas of collagen deposition orHC. 6The variability of UIP is also evident whenthe nature of the fibrosis is examined. Zonesof “old,” relatively acellular collagen bundlesare interspersed with aggregates of activelyprolifera ting myofibroblasts and fibroblasts.These aggregates, termed fibroblast foci, developin areas of previous lung injury and are associatedwith active collagen synthesis. FF are not pathognomonicbut are necessary for the diagnosis ofUIP (Fig 7, bottom). 6 Small cystic radiolucencies(termed honeycomb cysts) within the subpleuralregions reflect irreversible scarring and architecturalremodeling. 5,6 HC is an essential and oftenprominent feature of UIP but is not specificbecause it may be a sequela of severe lung injurycaused by diverse causes. 4 In late phases of UIP,the lung architecture is destroyed and replacedby large cystic air spaces (remnants of previousalveoli) and dense scar. Such cases are termedend-stage fibrosis or honeycomb lung (Fig 8). 5 Secondaryfeatures include pulmonary hypertensivechanges, smooth-muscle hypertrophy and hyperplasia,type II pneumocyte proliferation andhyperplasia, bronchiolectasis, and traction bronchiectasis.Subpleural fat and metaplastic bonereflect severe disease .203 Focal emphysematousblebs may be present in smokers.DIP: DIP is a histologic syndrome characterizedby dense collections of alveolar macrophageswithin air spaces (Fig 9), a homogeneous pattern,preserved alveolar architecture, and minimal orabsent fibrosis or honeycombing. 4,10,212 A significantinterstitial component is lacking and is overshadowedby the intra-alveolar component. The striking<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 647ACC-PULM-09-0302-038.indd 6477/10/09 10:24:56 PM


inhaled stimulus rather than persistent alveolarinjury characteristic of UIP. More than 90% ofpatients with DIP are smokers, suggesting thatconstituents of tobacco play a pathogenic role. 10,212Chest radiographs typically demonstrate reducedlung volumes and nonspecific linear or reticulonodularinterstitial infiltrates; in one fourth ofpatients, bibasilar, hazy GGOs are present. 9HRCT scans demonstrate diffuse GGO, oftenwith bibasilar or subpleural predominance; HC isabsent. 10 The average age of onset of symptoms forpatients with DIP is approximately 40 years. 10,212PFTs demonstrate a restrictive defect, with reducedDlco. 10,212 Prognosis is generally good, with longtermsurvival exceeding 90% at 5 years. 10,212 Althoughmost patients are treated with corticosteroids, spontaneousimprovement may occur. 10,212 Cessationof cigarette smoking is mandatory. Most patientsimprove or stabilize with the administration ofcorticosteroids, 9,10,212 but progressive, even fatal,respiratory failure can ensue. The role of cytotoxicand other immunosuppressive agents has not beenstudied.RBILDFigure 9. Top: Photomicrograph of DIP. Open-lung biopsyspecimen demonstrates extensive and diffuse filling of the alveolarspaces with alveolar macrophages. A single lymphoidaggregate is present. The alveolar architecture is preserved[reproduced with permission from Lynch et al, J Respir Dis2000; 21;197–214]. Center: Photomicrograph (high power) ofDIP showing dense aggregates of alveolar macrophages fillingthe alveolar spaces. The alveolar walls are thickened, butthe alveolar architecture is preserved. Fibrosis or honeycombcysts are absent [reproduced with permission from Lynchet al, J Respir Dis 2000; 21;197–214]. Bottom: DIP. Dense aggregatesof alveolar macrophages fill alveolar spaces (hematoxylin-eosin)[reproduced with permission from Lynch et al,J Respir Dis 2000; 21;197–214].heterogeneity and peripheral distribution characteristicof UIP are lacking in DIP. The uniformityof the lung lesion in DIP suggests a reaction to anRBILD, grouped with the IIPs, is characterizedby dense collections of pigmented alveolar macrophageswithin respiratory bronchioles but sparingthe distal lung parenchyma. 9,212,213 Honeycombingand severe fibrosis are not found. The pathologiclesion respiratory bronchiolitis was originallydescribed in an autopsy series of young cigarettesmokers who died of nonpulmonary causes. Thelesions were subsequently termed small airwaysdisease or smoker bronchiolitis. Histologic featuresoverlap with DIP, but DIP is more uniform andextensive than RBILD and exhibits a striking intraalveolarcomponent. 9,212 DIP and RBILD sharecommon histologic features and occur almostexclusively in smokers, 9,213,214 suggesting a commonpathogenesis. Patients with RBILD are generallyyoung (age 45 years) and present with mildsymptoms of cough, dyspnea, or sputum production.9,213 Chest radiographs demonstrate smallirregular opacities (“dirty lungs”) or reticular orreticulonodular infiltrates but are normal in up to28% of patients. 9,212 HRCT scans reveal numerous2- to 3-mm irregular peribronchiolar nodules;GGOs or emphysema may also be present. 9,213 PFTs648 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6487/10/09 10:24:58 PM


eveal reduced Dlco and mild airflow obstruction;lung volumes are variable. 9The course of RBILD is indolent or stable, andprognosis is generally excellent. However, data arelimited. Cessation of cigarette smoking is essential.9,212 Improvement and even complete resolutionoften occurs after the cessation of cigarette smoking.9,212 Corticosteroids are required in a minorityof patients. Late fibrosis or fatalities are rare. However,British investigators 215 identified 10 patientswith RBILD from 1980 to 1998, 7 of whom weretreated with prednisolone, which was combinedwith azathioprine or cyclophosphamide in 6patients. Despite cessation of smoking and aggressivetherapy, the condition of three patients deteriorated.Thus, the spectrum of RBILD is broaderthan the original descriptions. 216,217AIP: AIP, originally described by Hamman andRich in 1944, is a syndrome characterized by rapidlyprogressive respiratory failure (often within a fewdays) and histologic features of diffuse alveolar damageon lung biopsy. 15,16 Although this entity is distinctfrom IPF, some patients with IPF undergo an acceleratedphase and histologic features consistent withAIP. 95,115,218 Clinical features of AIP include bilateralalveolar infiltrates, an acute and rapid course,hypoxemic respiratory failure requiring mechanicalventilatory support, extensive GGOs on HRCT, ahigh mortality rate ( 50%), and potential responsivenessto high-dose corticosteroids. 15,16,115,218An acute viral-like prodrome often precedes theonset of AIP; fever is present in up to 50% ofpatients. 15,16 Histologic features include hyalinemembranes, fibrinous exudate, epithelial-cellnecrosis, and interstitial and intra-alveolar edema. 6,16As the process organizes and undergoes repair, typeII cells proliferate along alveolar walls, hyalinemembranes, and air space exudates resorb; and FBsproliferate within the alveolar interstitium andspaces. 6 The histologic features of AIP are nonspecificand are found with myriad disorders, includingARDS, inhalation or drug-induced injury,collagen vascular diseases, or infections. 6 Mortalityrates with idiopathic AIP range from 50 to 88%. 15,16Although data are limited, sustained and completeremissions have been achieved with the administrationof high-dose corticosteroids. 6,15,16,115,218 Patientssurviving the initial episode may heal with nosequelae or with variable degrees of fibrosis. 15,16Recurrent, even fatal, episodes of AIP may occurmonths or years after the initial episode. 15 Althoughdata are limited, aggressive treatment with highdoseIV pulse methylprednisolone is warranted forpatients with AIP.NSIP/Fibrosis: In 1994, Katzenstein proposedthe term NSIP/fibrosis to describe lung biopsyresults from immunocompetent patients withclinical syndromes resembling IPF 219 but with histologicfeatures distinct from UIP, DIP, RBILD, AIP,or LIP. 6 An inciting cause was not identified in mostpatients, but some had underlying CTD, drug reactions,or HP. 219 In the sentinel report, 219 the prognosisof NSIP was distinctly better than IPF. Aninternational ATS/European Respiratory Societyconsensus statement 4 on the classification of IIPspublished in 2002 recognized NSIP as a distincthistopathological pattern found in response tooccupational exposures, CTD, or as an idiopathicform. NSIP likely represents a stereotypic responseto diverse injuries or toxins. The relationshipbetween NSIP and UIP remains uncertain. 5,14,220Histopathology of NSIPSalient histologic features of idiopathic NSIPincluded varying degrees of inflammation andfibrosis that are temporally uniform, ie, occurringover a single time span (Fig 10). 219 The temporaluniformity in NSIP suggests a response to a singleinsult. Scattered foci of organizing pneumonia arenoted in nearly 50% of patients; prominent accumulationof intra-alveolar macrophages mimicking DIPis observed in one third of cases. 219 The key histologicfeature distinguishing NSIP from UIP is its temporaluniformity. 6,209 In NSIP, the lesions appear to be ofsimilar age, whereas in UIP, new and old lesions arepresent concomitantly. 6 Varying degrees of fibrosisand inflammation are observed in NSIP and UIP, buthoneycombing, a prominent feature of UIP, is rarein patients with NSIP. 5,148,203,210 UIP is characterizedby heterogeneity, greater destruction of the alveolararchitecture, extensive fibrosis, and minimal intraalveolarinflammation. 5,6Clinical Manifestations of NSIPClinical manifestations of NSIP are similar toIPF, with bibasilar crackles, cough, dyspnea,interstitial infiltrates, reduced Dlco, and a restrictivedefect on PFT results. 13,14,210,220 Because clinical<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 649ACC-PULM-09-0302-038.indd 6497/10/09 10:24:59 PM


Figure 10. Top: Photomicrograph of NSIP. Open-lung biopsyspecimen shows patchy interstitial fibrosis that lacks thesubpleural distribution and temporal heterogeneity of UIP[reproduced with permission from Lynch et al, J Respir Dis2000; 21;197–214]. Center: Photomicrograph of NSIP. Openlungbiopsy specimen. The alveolar septae are thickened byfibrosis and interstitial chronic inflammation (hematoxylin-eosin)[reproduced with permission from Lynch et al,J Respir Dis 2000; 21;197–214]. Bottom: Photomicrograph ofNSIP. Open-lung biopsy specimen. Prominent interstitial inflammationwithout dense scarring is evident. Some pneumocytesare hyperplastic. These inflammatory features areconsistent with the cellular form of NSIP [reproduced withpermission from Lynch et al, J Respir Dis 2000; 21;197–214].and radiographic features overlap with IPF, it ishighly likely that examples of NSIP were includedin historical series of IPF. However, NSIP and IPFdiffer in several important respects. The onset ofNSIP may be subacute, evolving over a few weeks,whereas IPF evolves over years. 13 Fever, noted inone third of patients with NSIP, is not found in IPF.Clubbing is found in 10 to 40% of patients withNSIP but in 66 to 93% of patients with IPF. 12,147,148Compared with patients with IPF, patients withNSIP are younger, and there is a slight female predominancein NSIP. 12,14,147,148 BAL lymphocytosis iscommon in NSIP 147,221 but rare with IPF. 147,222 However,significant overlap exists between NSIP andIPF. 223 HRCT features of NSIP and IPF/UIP differ.11,147,148,222 Bilateral GGOs are characteristic inNSIP but rare in IPF/UIP. 147 HC, a cardinal featureof UIP, is absent or sparse in NSIP. 20,147,148,222 Mostimportantly, the prognosis of NSIP is much betterthan IPF/UIP. 14 Some patients with NSIP improveeither spontaneously or in response to corticosteroidor immunosuppressive therapy. In strikingcontrast, improvement is rare in IPF ( 10%), evenwith aggressive therapy. 2,13,27,163,204In several early studies, 96,98,147,148,152,203 5-yearsurvival with NSIP exceeded 70%, compared with 30% with IPF. It is highly likely that manypatients in earlier publications labeled as IPF butmanifesting GGO on HRCT, BAL lymphocytosis,or “steroid responsiveness” in fact representedcases of NSIP. 151,153,224,225 Subsequent studies 13,147,203suggest that the designation NSIP is excessivelybroad. Segregation by histologic features into cellularnonspecific interstitial pneumonia (cellularNSIP) or fibrotic forms of NSIP identifies patientsubsets with differing prognoses. 147,203 Patients withthe cellular form have more GGOs and less honeycombingon HRCT, a greater rate of corticosteroidresponsiveness, and lower rates of mortality. 147,203Although NSIP has a better prognosis than IPF,mortality with fibrotic NSIP is high. British investigators204 monitored 28 patients with idiopathicNSIP up to 10 years after SLB. At a median followupof 42 months, 17 patients (61%) had died;median survival was only 52 months. These datasuggest that earlier studies 98,147,203 citing survivalrates 90% with NSIP were overly optimistic.Similarly, a retrospective review 172 of 104 patientswith IPF reclassified 61 patients as having UIP and43 (41%) as having fibrotic NSIP. At a medianfollow-up of 32 months, 25 of 43 NSIP patients(48%) had died vs 48 deaths among 61 patientswith UIP (79%).650 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6507/10/09 10:24:59 PM


Furthermore, the prognosis of fibrotic NSIPpatients with severe functional impairment is poor.Among NSIP patients with Dlco 35% of predicted,3-year survival was 50%. 172 Korean investigators221 cited 5-year survival rates of 56% among48 patients with fibrotic NSIP and 34% among 131patients with UIP. Further, early studies 147,203,226,227cited favorable responses to corticosteroids (aloneor combined with immunosuppressive agents) in60 to 83% of patients, whereas subsequent studies98,148,204 were less sanguine (response rates of 25to 40%). Short-term (ie, 2-year) survival ratesunderestimate long-term mortality associated withNSIP. 172,204,221 Additional studies are required tofurther elucidate the clinical features, outcomesand rate of progression associated with NSIP andhistologic subsets (eg, cellular or fibrotic forms).SLBThe optimal staging and role of SLBs in patientswith IPF are controversial. Video-assisted thoracoscopicsurgery (VATS) lung biopsies are requiredto establish the diagnosis of UIP unequivocally,7,228,229 but many clinicians are reluctant tosubject patients to the morbidity of surgery, particularlywhen therapeutic options for IPF are oflimited value. 124 Although VATS is relatively safe,acute exacerbation in respiratory status was notedin 2.1% of 236 patients undergoing SLB for ILD. 230SLB was associated with rapid deterioration inrespiratory status (consistent with acute exacerbations)in 3 to 4% of patients with idiopathic NSIPor CTD-associated ILD. 231Despite these caveats, SLB can be justified inotherwise-stable patients given the potential forserious toxicities associated with aggressive immunosuppressivetherapy and the prognostic valueof SLB. Clinical surveys have found that SLB isperformed in less than onethird of patients withIPF. 124,232,233 Within the past decade, HRCT hasincreasingly been used in lieu of SLB to diagnoseUIP/IPF. 27,124,148 However, SLB remains the “goldstandard” for the diagnosis of UIP and assessmentof prognosis. 6 The role of SLB in clinical practice isthreefold: (1) to rule out other etiologies, (2) to betterdefine the pattern of IIP (primarily UIP vsNSIP), and (3) to better define prognosis.VATS is less invasive than open-lung biopsiesand provides adequate histologic material. 7,97,164VATS can often be performed in the outpatientsetting with minimal morbidity. 234 Tissue shouldbe obtained from at least two sites (eg, upper andlower lobes of the same lung); representative areasshould be sampled (ie, avoid normal or veryfibrotic areas). 97,98,164,206 Specific histologic featuresor subtypes of IIPs have prognostic value. 6,98,164,204Establishing the histologic diagnosis of UIP is astrong predictor of increased mortality and poorresponse to therapy when compared with otherIIPs. 6,27,96−98 In one prospective study, 164 long-termsurvival was worse in patients with greater degreesof FF, whereas the degree of alveolar cellularity oralveolar wall fibrosis did not affect survival. In acohort of 53 patients with UIP, British investigators163 confirmed that increasing FF correlated withrate of mortality (p 0.006) and decreases in FVCand Dlco at 6 months and 12 months. Neither theextent of established fibrosis nor intra-alveolarmacrophage accumulation correlated with outcome.163 In contrast, in a cohort of 99 patients withidiopathic UIP, we 235 found that the extent ofFF, established fibrosis, or intra-alveolar macrophageaccumulation did not predict survival.Another study 236 of 43 patients with IPF found norelationship between profusion of FF on SLB andsurvival.Variation in histologic features in individualpatients limits the prognostic accuracy of evenSLB. 97,205,206 Although SLB is touted as a “gold standard,”a minute fraction of the lung parenchymais sampled. Evaluation of SLB is subject to interobserverand intralobar variation, even by expertpulmonary pathologists. 97,204,206 Further, discriminatingUIP from fibrotic NSIP is difficult. 203,204 Moreimportantly, foci of UIP and NSIP may be foundwithin the same lobe in individual patients. 97,205 Aconfident diagnosis of UIP or NSIP may be difficultto make, particularly by pathologists lacking expertisein pulmonary histology. Importantly, HRCTmay corroborate the diagnosis of UIP, providedthat radiographic features are classic. 22,125 Further,the extent and pattern of changes on HRCT haveprognostic value. 123,124,127,146,150Although the decision to perform VATS lungbiopsy should be individualized, we favor VATSbiopsy (if no contraindications exist) in patientswith recent onset of ILD of unknown cause, whentransbronchial lung biopsies (TBBs) are nondiagnostic.However, VATS biopsy is not warranted in<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 651ACC-PULM-09-0302-038.indd 6517/10/09 10:25:00 PM


patients who have a prolonged course ( 2 years)and clinical, physiologic, and HRCT features thatare characteristic of IPF/UIP. 3,125 Moreover, the riskof VATS lung biopsy is excessive in elderly ( 70years old) or debilitated patients, particularly whenclinical and HRCT features strongly suggest UIP.In such patients, HRCT is sufficient to determinean approach to therapy. Bronchoscopy with TBBsis most useful to diagnose alternative causes of ILD(eg, sarcoidosis, Langerhans-cell granulomatosis[LCG], pulmonary alveolar proteinosis, lymphangiticcarcinomatosis, COP). Because of the smallsample size (2 to 5 mm), TBBs cannot substantiatethe diagnosis of UIP, DIP, or NSIP and cannotbe used to assess the extent of inflammation orfibrosis.BALBAL is useful to exclude alternative etiologiesthat may mimic IPF (eg, specific infections such asPneumocystis carinii, Mycobacteria, and fungi, PAP,and LCG) 237 but has no role to diagnose or stageIPF. 3,4,223,238 Increases in polymorphonuclear leukocytes,mast cells, alveolar macrophages, andmyriad cytokines are noted in BAL fluid frompatients with IPF; lymphocyte numbers are usuallynormal. 238 However, BAL cell profiles in IPF do notpredict prognosis or therapeutic responsiveness. 3,238BAL has yielded significant insights into the pathogenicmechanisms responsible for IPF, but BAL isof doubtful clinical value in assessing the extent oractivity of IPF. 166,238Radionuclide Scans67Ga citrate scanning has been used as a surrogatemarker of alveolar inflammation (alveolitis) inIPF and other inflammatory pulmonary disorders.Increased intrapulmonary uptake of 67 Ga was associatedwith a more cellular biopsy in some earlystudies. However, even with careful quantification,67 Ga scans failed to predict responsiveness totherapy. 2 67 Ga scans are expensive and inconvenientbecause scanning is performed 48 h after injection.Despite initial enthusiasm for its application, 67 Gascanning is of no practical value. The clearance of99Tc-DTPA aerosol is accelerated in IPF and isa marker of increased lung permeability. 166,239,240British investigators 239 found that normal DTPAclearance in patients with IPF was associated withstable lung function; survival was not analyzed. Incontrast, accelerated clearance of the fast component(ie, reduced t 0.5F) was an independent predictorof mortality in patients with idiopathic UIP. 166In that study, 166 age and Dlco were the most importantpredictors of survival, followed by percentageof predicted TLC and t 0.5F.Thus, the incremental practical value of t 0.5Fappears to be small. Further, increased clearanceof DTPA occurs in smokers and patients with otherinflammatory lung disorders; its prognostic valueis debatable. 240 Clearance of inhaled pertechnegas(an aerosol of 99 Tc-labeled carbon particles) isincreased (compared with normal control subjects)in IPF, HP, and CTD-associated PF but not sarcoidosis,241 consistent with increased alveolar capillarypermeability. Increased metabolic activity may benoted by PET scans in patients with IPF. 156 At present,radionuclide imaging studies have no role inthe management of IPF.PathogenesisThe pathogenesis of IPF is complex and likelyinvolves myriad components, including repetitivelung injury (particularly to alveolar epithelial cells[AECs]), destruction of subepithelial basementmembranes, inflammation, cytokines and chemokines,exaggerated deposition of collagen andextracellular matrix, recruitment and proliferationof fibroblasts, inappropriate wound-healingresponse, and excessive angiogenesis .242 Thefibrotic/inflammatory process in UIP has beenlikened to abnormal wound healing, 243,244 with avigorous fibroblastic response to AEC injury occurring.Injury of AECs and destruction of the subepithelialbasement membranes lead to localrecruitment, differentiation, and proliferationof fibroblasts. 242 Excessive deposition of collagenand extracellular matrix contributes to the fibroticprocess. 242,245A complex interplay between inflammatoryand mesenchymal cell populations amplifies andperpetuates the fibrotic response. 242,245 Early theoriesof pathogenesis suggested that a heightenedinflammatory response led to injury and fibrosis.Current theories suggest a primary role for FBdysregulation; inflammatory cells may play a contributory,albeit minor, role. 242,245 Soluble mediators652 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6527/10/09 10:25:00 PM


(cytokines) produced by fibroblasts and AECsamplify and perpetuate the fibrotic response.TGF- likely plays a critical role in IPF. 242,245 Othercytokines that appear to be involved in the pathogenesisinclude TNF-, platelet-derived growthfactor, connective tissue growth factor, IL-8, andCXC chemokines. 242,245 Other possible mediatorsof the fibrotic process in IPF include integrinmediatedintercellular adhesion molecules, surfactantproteins, proteases, leukotrienes, oxygenradicals, and myriad soluble mediators. 2,242,243The relative deficiency of -interferon (-IFN),a cytokine that inhibits collagen synthesis andfibrosis, may contribute to the fibroproliferativeprocess. 2 The signals responsible for initiating andperpetuating this process are not known. However,the so-called FF in IPF/UIP may be the leadingedge of the inflammatory/fibrotic process. 6 Further,AEC death is most prominent immediatelyadjacent to FF. 246 Mesenchymal cells that form FFsin IPF/UIP are activated and display a contractilephenotype, commonly referred to as myofibroblasts.247 Myofibroblast differentiation and fate arecontrolled by soluble growth factors such asTGF- 1and matrix-derived signals. 245 Myofibroblastsin the lung in IPF may be derived fromextrapulmonary fibrocytes and bone marrowderivedprogenitor cells. 248Under the influence of TGF- 1, myofibroblastsdisplay increased production of collagen, vimentin,-actin, and tissue inhibitors of metalloproteinases,2 leading to a bias toward excessive matrixproduction and wound contraction in UIP. Additionally,a proangiogenic environment may favorfibrosis in IPF. 249 IL-8 and -IFN--inducible protein-10 (IP-10), members of the CXC chemokine family,affect fibrosis via angiogenic mechanisms. 245 IL-8stimulates neovascularization, whereas IP-10inhibits angiogenesis. In humans with IPF, IL-8 ismarkedly elevated in BAL fluid and serum, 250whereas IP-10 levels in IPF lung biopsies arereduced compared with control patients. 251 Geneticfactors may be important in the pathogenesis ofthe disease and its clinical expression. 59 Differencesin gene expression may influence the severity andcourse of the disease. 59,94Selman et al 94 examined gene expression profilesin 88 patients with slowly progressing IPF(symptoms 24 months) and 26 patients with therapid variant (symptoms 6 months). Two genes,adenosine 2B receptor and prominin-1/CD133,were expressed more strongly in the rapid group.It is evident that the pathophysiology of IPF iscomplex, with diverse cell types/mediators participatingin the initiation, orchestration, perpetuation,and evolution of the inflammatory/fibroticprocess. Unfortunately, as will be discussed ingreater detail below, therapeutic strategiesdesign ed to ablate inflammatory responses havehad little or no impact on the outcome of IPF/UIP.In light of this, novel therapies targeted againstFB proliferation or other profibrotic mediatorsare being investigated (discussed later in thischapter).TherapyThe management of patients with IPF isfrustrating because the disease progresses inexorablyand treatment has not been shown to alterthe natural history of the disease. 2,27,252 Initialtreatment strategies, based on the concept thatinflammation was a key factor in the pathogenesisof IPF, involved the administration high-dosecorticosteroids and/or immunosuppressiveagents. 27,164,224,225,253,254 However, RCTs were not performed.Several early studies 224,225,253,255 citedresponse rates of 10 to 30% with corticosteroids,but these older studies included a mix of histologicentities (eg, UIP, DIP, RBILD, NSIP, HP, etc) andcannot be extrapolated to UIP.Other studies 96,147,148,204 256 that included patientswith UIP cited response rates of 0 to 17%. Importantly,a survival benefit with corticosteroidtherapy has not been demonstrated. 13,26,27,57,164Unfortunately, the administration of corticosteroidsis associated with a plethora of adverseeffects, particularly in elderly or debilitatedpatients. 179,258 Given the significant toxicities, theadministration of corticosteroids in high dosesshould not be used to treat IPF. 179 Similarly, immunosuppressiveor cytotoxic agents have not beenshown to influence mortality or clinical outcomes.1,2,27,199,259−262 Anecdotal responses to azathioprine202,262 or cyclophosphamide 261,263 were noted inearly studies, but the preponderance of data suggestthat these agents are ineffective as therapy forIPF. 2,26,27,259 260 Several large retrospective studies26,27,164,264 failed to document survival benefit withany form of therapy.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 653ACC-PULM-09-0302-038.indd 6537/10/09 10:25:01 PM


Expert Consensus Statements and Recommendationsfor Therapy: International consensus statementsconclude that existing therapies are ofunproven benefit but acknowledge that physiciansmay wish to treat patients with deterioratingdisease. 3,232 Both statements advocated an individualizedapproach to therapy. For patientsrequiring treatment, the British Thoracic SocietyCommittee advocated combining prednisolone(0.5 mg/kg/d, with taper) plus azathioprine (2 to3 mg/kg/d) as initial therapy. 232 The ATS statementrecommended combining prednisone (0.5mg/kg/d for 4 weeks, with subsequent taper)with either oral azathioprine (2 to 3 mg/kg/d) ororal cyclophosphamide (2 mg/kg/d). 3 Forpatients unable to take corticosteroids, azathioprineor cyclophosphamide alone are used. Theserecommendations represented a substantialdeparture from earlier regimens, which advocatedmuch greater doses of corticosteroids. 45,225,262Therapy should be continued for 6 months unlessadverse effects develop and require modificationor discontinuation of therapy. Prolonging treatmentbeyond 6 months should be limited topatients who improve or remain stable. It shouldbe emphasized that these recommendations 3,232reflect expert opinion but have not been validatedin clinical trials.My Recommendations for Therapy for IPF: Giventhe lack of survival benefit, and potentially serioustoxicities of corticosteroids or immunosuppressivetherapy, I rarely recommend therapy for IPF. However,treatment may be offered to patients withsevere symptoms and a declining course, providedthat the pros and cons of therapy are discussedhonestly. We emphasize to patients that no therapyhas been proven to influence survival or clinicallyimportant outcomes. For patients exhibiting asubacute or deteriorating course and/or surrogatemarkers of alveolitis (GGO on CT scan), we offeroral azathioprine, 2 to 3 mg/kg/d, alone or combinedwith low-dose prednisone, 20 mg every otherday or equivalent.Azathioprine does not cause bladder injury orbladder carcinomas and has less oncogenic potentialthan cyclophosphamide. 265 Because of its suppressiveeffects on bone marrow, CBC and plateletcounts should be obtained every 2 weeks for thefirst 6 weeks and then every 4 to 8 weeks thereafter.Significant anemia, leukopenia ( 3,500/µL)or thrombocytopenia ( 120,000/µL) warrantsdose reduction. Azathioprine should be used withcaution in patients receiving allopurinol, and thedose should be reduced by at least 50%. 265 After6 months, efficacy should be examined. Unlessthere is unequivocal and objective improvement,we discontinue therapy after a 6-month trial. Further,treatment is discontinued among patientsexperiencing side effects or disease progression.Cyclophosphamide has potentially serious toxicities,265 and we do not use cyclophosphamide forIPF. Mycophenolate mofetil and other immunosuppressiveagents have been used by some clinicians,but these agents have not been evaluatedin clinical trials.In contrast to the dismal responses noted withimmunosuppressive or cytotoxic therapy in IPF/UIP, favorable responses to corticosteroids (aloneor combined with azathioprine) have been notedwith NSIP, particularly cellular variants. 210,220 ForNSIP, either cellular or fibrotic, I recommend acourse of therapy with corticosteroids and azathioprinefor a minimum of 6 months. Patients whoare intolerant of azathioprine or in whom azathioprinefails may be switched to mycophenolatemofetil or cyclophosphamide, but published datain this regard are lacking.Novel Therapeutic Options: TNF- antagonists(eg, etanercept and infliximab) have been tried insome cases of PF (idiopathic and CTD-associated),but data are limited. In a recent prospective, multicentertrial, 266 84 patients with IPF were randomizedto etanercept (twice weekly) or placebo. At 48weeks, there were no significant differences inprimary end points (physiologic parameters)between groups, but trends favoring etanerceptwere observed in some secondary end points. 266Anecdotal responses to infliximab were noted incase reports 267,268 of CTD-associated pulmonaryfibrosis. However, TNF- antagonists are expensiveand have serious potential toxicities. 268,269Additional studies are required before endorsingthese agents for IPF.Novel and Future Therapies: Major advancesawait the development of novel therapies thatprevent fibroproliferation and enhance alveolarreepithelialization. 201,245 Agents that have beentested in pilot studies include interferons, N-acetylcysteine (NAC), and pirfenidone (5-methyl-1-phenyl-2-[1H]-pyridine).654 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6547/10/09 10:25:01 PM


-IFN-1b: -IFN-1b is an endogenous cytokinethat down-regulates the expression of TGF- andmay have antifibrotic effects. 201 Initial studies 270,271employing recombinant -IFN-1b were encouraging,but the INSPIRE trial, which randomized 800patients with mild-to-moderate IPF to recombinant-IFN-1b or placebo in a 2:1 ratio, showed no benefit.The study was ended because of “futility” inMarch 2007 (InterMune, Brisbane, CA).NAC: NAC is an antioxidant that stimulatesglutathione synthesis, attenuates fibrosis in animalmodels, 272 and augments glutathione levelsin BAL in patients with IPF. 273 A phase III multicenter,double-blind, placebo-controlled trial(European Idiopathic <strong>Pulmonary</strong> Fibrosis InternationalGroup Exploring N-acetylcysteine [NAC]I Annual) 274 in Europe evaluated the efficacyof oral NAC in IPF. In this study, all patientsreceived conventional therapy with prednisone,0.5 mg/kg/d, with taper, plus azathioprine,2 mg/kg/d. Patients were then randomized tooral NAC, 1,800 mg/d, or placebo. At the end of1 year, PFT results had deteriorated in bothcohorts. However, the rates of decrease in FVCand Dlco were less in patients receiving NAC(p 0.05). 274 However, these changes in PFTs weresmall ( absolute difference in FVC of 4.8% and inDlco of 5.1%) and of doubtful clinical significance.The benefit if any of NAC as therapy forIPF remains controversial. Nonetheless, NAC isinexpensive and has few side effects, making thisan attractive option for IPF. 275 A multicenter RCTsponsored by the Idiopathic <strong>Pulmonary</strong> FibrosisClinical Research Network is planned to addressthe impact of NAC in IPF.Pirfenidone: Pirfenidone (5-methyl-1-phenyl-2-[1H]-pyridine) attenuates pulmonary fibrosis inanimal models, reduces synthesis of collagen I andIII and TNF-, inhibits TGF--stimulated collagensynthesis, decreases extracellular matrix, andblocks the mitogenic effect of profibrotic cytokines.201,276 . In a phase II open-label trial, 277 54 consecutivepatients with IPF were treated withpirfenidone. Conventional therapy failed in 46; 8were untreated. With pirfenidone, 1- and 2-yearsurvival rates were 78% and 63%, respectively.After 6 months of therapy, PFT results appeared tostabilize, but results are difficult to interpretbecause follow-up PFTs were not performed in allpatients.In another nonrandomized trial, 278 13 patientswith PF (idiopathic, n 11; associated with scleroderma,n 2) were treated with oral pirfenidonefor 1 year. At 2-year follow-up, nine patients haddied and two patients remained stable; the diseasehad progressed in the remaining two patients.Further, in a cohort 279 of 11 patients with PF complicatingHermansky-Pudlak syndrome, pirfenidonewas associated with a slower rate of decreasein PFTs (ie, FVC, FEV 1, Dlco) compared with 10patients receiving placebo.A phase II RCT 280 compared pirfenidone toplacebo (2:1 ratio) in a cohort of 107 patients withIPF. The study was discontinued prematurelybecause acute exacerbations were noted in fivepatients receiving pirfenidone (14%) comparedwith no cases in the placebo group. The primaryend point (change in lowest oxygen saturation on6MWT 6 or 9 months) was not met. There wereno significant differences between groups in rateof mortality, TLC, Dlco, or resting Pao 2. The rateof decrease in FVC at 9 months was lower in thepirfenidone group (p 0.037), but differencesbetween groups were small and of doubtful clinicalsignificance. These studies 277−280 are insufficient toassess the efficacy of pirfenidone but support theneed for additional therapeutic trials. Pirfenidonehas been approved for use in Japan based on arandomized trial, but published data are not available.Currently, pirfenidone is not commerciallyavailable elsewhere. Two placebo-controlledDBRCTs evaluating pirfenidone as therapy for IPFwere recently completed in the United States (Inter-Mune, Brisbane, CA) but have not yet beenpublished.ET-1 Receptor Antagonists: ET-1, a potent pulmonaryvasoconstrictor, displays profibrotic effectsand may play a role in the pathogenesis of IPF. 281Bosentan, an ET-1 receptor antagonist, reducedcollagen deposition in animal models (bleomycininducedpulmonary fibrosis), 282 but its efficacy inhumans with IPF has not yet been shown. Resultsof a multicenter DBRCT 198 evaluating BUILD-1were recently published. In that study, 158 patientswith IPF were randomized to bosentan or placebo.Patients with severe pulmonary dysfunction (FVC 50% of predicted or Dlco 35% of predicted)or concomitant PAH were excluded. At 12 months,the 6MWT worsened in both groups (no significantdifferences between groups). Mean changes from<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 655ACC-PULM-09-0302-038.indd 6557/10/09 10:25:01 PM


aseline in FVC at 12 months were 6.4% and7.7% in the bosentan and placebo groups, respectively.Mean changes from baseline in Dlco at 12months were 4.3% and 5.8% in the bosentan andplacebo groups, respectively. A trend in favor ofbosentan was noted in the secondary end point oftime to death or disease progression (hazard ratio,0.63; p 0.12). A larger DBRCT (BUILD-3) assessingthe impact of bosentan vs placebo in IPF is inprogress. Further, a DBRCT assessing ambrisentan(another ET-1 receptor antagonist) is planned.Phosphodiesterase-4 Inhibitors: Sildenafil, a phosphodiesterase-4inhibitor, has been shown to beeffective to treat idiopathic PAH. 283 Data regardingsildenafil for PAH complicating IPF are limited toanecdotal cases and small series. In an open-labelRCT, 16 patients with PAH secondary to PF underwentright-heart catheterization to assess the acutehemodynamic effects of vasodilators. 192 Eightpatients received sildenafil (50-mg oral dose), andeight received the maximal tolerated dose of IVepoprostenol. <strong>Pulmonary</strong> vascular resistancedecreased equally in both groups: epoprostenol( 36.9%), sildenafil ( 32.5%). Importantly, epoprostenolincreased intrapulmonary ventilation/perfusionmismatching, whereas sildenafil maintainedventilation/perfusion matching and improvedPao 2. However, this study only assessed parameters60 min after ingestion of sildenafil. Long-termeffects of sildenafil in patients with ILD (with orwithout PAH) are not known. Anecdotal responseswere noted in small series 196,284 of patients with IPFand PAH, but these data are inadequate to assessthe role of sildenafil for patients with ILD. A randomizedtrial evaluating the impact of sildenafilin patients with IPF is in progress (ClinicalTrials.gov identifier: NCT00352482).Anticoagulants: Inflammation and vascularinjury in IPF may lead to a prothrombotic state thatcould exacerbate lung injury. 285 Japanese investigators286 randomized 56 IPF patients to anticoagulants(warfarin) or placebo. Three-year survivaland freedom from acute exacerbations wereimproved in the anticoagulated group. However,the dropout rate was high, and it is possible thatselection bias may have influenced the study.Given the risk associated with anticoagulationtherapy, additional studies involving greater numbersof patients are required before endorsing thisform of therapy.Novel (Future) Agents: Unfortunately, currenttherapies for IPF are of unproven efficacy. Novelagents with potential antifibrotic properties havetheoretical value but have not yet been evaluatedin IPF. 199,201,244,272 Agents that are currently beingstudied in IPF include imatinib mesylate, sirolimus,captopril, inhaled iloprost, and other inhibitorsof fibrotic growth factors. 199,285,287LT: LT is the best option for patients with severeIPF. 158,252,288,289 Early listing for transplantation isurged in patients with progressive IPF because thewaiting time for procuring donor organs mayexceed 2 years. 158 Patients with severe functionalimpairment (eg, FVC 60% of predicted,Dlco 40% of predicted), oxygen dependency,and/or a deteriorating course should be referredpromptly for transplantation. The decision asto when to list should be left to the transplant center,according to local waiting times. Unfortunately,many patients with severe IPF die while awaitingorgans. 158Data from the International Society of Heartand Lung Transplantation (ISHLT) Registry 290 cited1-, 3-, and 5-year survival rates of 68%, 52%, and50%, respectively, among IPF patients after LT.Initial studies 291,292 from single centers noted similarmortality rates with single lung transplants (SLTs)and bilateral sequential lung transplants (BSLTs).However, a review 289 of 821 patients who receivedLTs for IPF (636 SLT, 185 BSLT) in the United Statesbetween 1994 and 2000 reported significantly betterearly (1-month) and late (3-year) survival rateswith SLT compared with BSLT. When posttransplantsurvival was reanalyzed contingent on survivalto 1 month, survival by procedure type (SLTvs BSLT) was similar.The reason for the increased mortality withBSLT likely reflects an increase in surgical problemsin the early transplant period with BSLT. Data fromthe ISHLT Registry 293 for recipients with IPF notedimproved survival with BSLT vs SLT (p 0.03);survival rates were similar up to 3 years butdiverged thereafter. The most recent data from theISHLT 294 cited lower survival rates at 3 months afterLT among patients with IPF (84%) or idiopathicPAH (74%) compared with those with cystic fibrosis(90%) and COPD (91%). Among patients survivingto 1 year, IPF and COPD had the worstlong-term survival, most likely reflecting older ageand comorbidities. 294656 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6567/10/09 10:25:02 PM


Secondary PAH is not a contraindication to LT,but its presence may influence the operative andperioperative management. Whelan et al 295reviewed 830 patients with IPF who underwenttransplantation between 1995 and 2002 in theIHSLT registry; 77% had SLT and 23% had BSLT.By multivariate analysis, mean PAP and BSLT wereindependent risk factors for mortality. Among SLTrecipients, there was a linear relationship betweenPAP and 90-day mortality. However, only 8.3% ofpatients with SLT had mPAP 40 mm Hg. Thedecision as to which procedure (ie, SLT or BSLT)should be done depends on the expertise of thelocal transplant program. 158 However, given theimproved survival with BSLT among patients withidiopathic PAH, 293 most centers perform BSLT forIPF patients with secondary PAH.Adjunctive Therapy: Supplemental oxygen iscritical to optimize quality of life and enhanceexercise capacity in patients with IPF 296 ; impact onsurvival has not been studied. However, continuousoxygen ameliorates pulmonary vasoconstrictionand may delay the clinical development of corpulmonale. Judicious use of diuretics may be necessaryto control peripheral edema in patients withcor pulmonale. The role of vasodilators in patientswith secondary PAH is controversial. 192,297What Is the Relationship Between NSIP and UIP?Controversy exists as to whether NSIP is aseparate disease entity or a continuum of diseaserelated to IPF/UIP. 5,13,209,220 Histologic features ofNSIP overlap, and distinguishing fibrotic NSIPfrom UIP is difficult. 98,204,206 More importantly, bothNSIP and UIP may coexist in individual patients.<strong>Review</strong> of SLB from patients with IIPs observedboth NSIP and UIP (ie, discordant UIP) in 13 to 26%of patients. 97,205 Further, both NSIP and UIP may beobserved in PF complicating CTDs. 298−301 NSIP andUIP may represent different points in the progressionof a single disease or may represent distinct(albeit overlapping) processes. It is possible thatcellular NSIP represents an early phase that mayprogress to fibrotic NSIP and ultimately UIP. If so,the survival advantage of NSIP may simply representlead-time bias. 155,172 Until these questions areanswered, I approach NSIP as a potential reversiblelesion, with both inflammatory and fibrotic components.Unless specific contraindications exist,I treat NSIP with a combination of corticosteroids,40 mg/d for 1 month, with gradual taper, andazathioprine, 2 mg/kg/d, for a 6-month trial.Responders are continued on medical therapy,often for prolonged periods. Patients in whomtherapy fails and who exhibit a deterioratingcourse are listed for LT. (Criteria for listing aresimilar to IPF.)SarcoidosisSarcoidosis is a poorly understood granulomatousdisease that involves the lung and intrathoraciclymph nodes in 90% of patients. 302−306However, virtually any organ can be affected. 307Skin involvement, peripheral lymphadenopathy,and eye involvement each occur in 20 to 30% ofpatients. 304 Clinically significant involvement ofspleen, 308 liver, 308 heart, 309 CNS, 310−312 or bone 313occurs in 2 to 7% of patients. 304 The clinicalexpression and course are heterogeneous. Onethird or more of patients are asymptomatic, withincidental findings of bilateral hilar lymphadenopathy(BHL) on chest radiographs. Symptomsare protean, reflecting the site of organ involvement.The natural history is usually favorable.Spontaneous remissions occur in nearly twothirds of patients, and a waxing and waningcourse is common. 303,305,306 The presence of acuteinflammatory manifestations (ie, erythema nodosum,polyarthritis, and fever), termed Löfgrensyndrome, portends an excellent prognosis, withhigh rates ( 85%) of spontaneous remission.302,314−317 Factors associated with poor prognosisand more aggressive disease include blackrace, osseous involvement, lupus pernio (disfiguringnasolabial cutaneous lesions), chronichypercalcemia, and chronic pulmonary sarcoidosis.303−306,314 Ethnic, geographic, and geneticfactors influence prognosis. 317−319Löfgren syndrome is three to six times morecommon in women. 317,320 Black race is associatedwith a greater rate of chronic progressive disease,worse long-term prognosis, extrapulmonaryinvolvement, and greater risk of relapses. 314,316,321−323In 10 to 15% of patients with sarcoidosis, the courseis chronic and progressive, resulting in permanentdamage and fibrosis of affected organs; 1 to 6% ofpatients die of sarcoidosis. 303−306,314 Lower mortalityrates (0 to 0.5%) have been cited in nonreferral<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 657ACC-PULM-09-0302-038.indd 6577/10/09 10:25:02 PM


settings, when the diagnosis was made as part ofroutine radiographic screening. 324,325 An epidemiologicstudy 31 in the United Kingdom identified1,019 cases of sarcoidosis between 1991 and 2003.Mortality rates at 3 years and 5 years for patientswith sarcoidosis were 5% and 7%, respectively,compared with 2% and 4% among age- and gendermatchedcontrols without sarcoidosis. Causes ofdeath were not reported.EpidemiologySarcoidosis is worldwide in distribution, butits prevalence varies according to racial and geographicfactors. Sarcoidosis is four to eight timesmore common in black than in white patients. 326−328In North America and Europe, prevalence rates of10 to 20 cases per 100,000 persons have beencited. 327,329 In Scandinavia and certain parts of theBritish Isles, prevalence rates exceed 80 cases per100,000. 329,330 The incidence is much lower ( 2 per100,000) in southern Europe. 329,331 Sarcoidosis hasinfrequently been reported in Central or SouthAmerica or Africa, but whether this representsunderrecognition or reduced prevalence of thedisease is not known. More than two thirds ofpatients present between the ages of 20 and 40years. 332 There is a slight female predominance. 306,327Sporadic cases within families are well recognized.332,333 Familial sarcoidosis (defined as havingfirst- or second-degree relatives with sarcoidosis)occurs in 17% of African-American patients withsarcoidosis, compared with 6% among whitepatients. 327Data from the multicenter ACCESS study 326noted that the familial relative risk of sarcoidosiswas greatest among siblings, followed by grandparents,and then parents. A specific genetic defecthas not been identified, but genes of the majorhistocompatibility complex locus on chromosome6p are believed to be involved. 334 The inheritancepattern is likely complex, with heterogeneousalleles, polymorphisms, and linkages. 335,336 Thecause of sarcoidosis remains elusive, but genetic,environmental, and infectious causes have beensuggested. 305 Specific genetic polymorphisms mayinfluence prognosis. Human leukocyte antigenDQB1*0201 is a marker for a good prognosis inDutch and British patients with sarcoidosis. 337Polymorphisms of the C-C chemokine receptorgene (CCR2 haplotype 2) are associated with Löfgrensyndrome and a good prognosis in Dutchpatients. 338HistopathologyThe histologic hallmark of sarcoidosis is thenoncaseating (nonnecrotizing) granuloma composedof epithelioid cells and multinucleated giantcells, surrounded by a cuff of lymphocytes andplasma cells (Fig 11). 306,339 Fibrosis is present invarying degrees (Fig 12). Disruption and destructionof parenchyma may be prominent. Becausemycobacterial and fungal granulomas can causenonnecrotizing granulomas, special stains for acidfastbacilli and fungi should be performed toexclude these infectious etiologies. In the respiratorytract, sarcoid granulomas are often situated inthe submucosa of bronchioles and along bronchovascularbundles (Fig 13). 306,339 Coalescent granulomatamay give rise to confluent mass lesions,nodules, or consolidation of lung parenchyma. 303Exuberant granulomatous inflammation may infiltrateand destroy affected organs, leading to significantand irreparable loss of function. In thelung, progression to end-stage fibrosis (honeycomblung) can occur.Fiberoptic bronchoscopy with TBB is the preferreddiagnostic procedure to diagnose pulmonarysarcoidosis. Diagnostic yields are 60 to 95%. 303To avoid sampling error, I obtain several biopsyFigure 11. Sarcoidosis. Photomicrograph of TBB specimen.Multinucleated giant-cell in the center of a granuloma, surroundedby lymphocytes and mononuclear cells in the periphery(hematoxylin-eosin) [reproduced with permissionfrom Lynch JP III, Strieter RM. In: Internal medicine for thespecialist. Montvale, NJ: Medical Economics, 1994; 38–62].658 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6587/10/09 10:25:02 PM


Figure 12. Sarcoidosis. Photomicrograph of TBB specimen,showing “burned out” granuloma demonstrating nonnecrotizinggranuloma containing multinucleated giant cells andepithelioid cells in the center. In the periphery, no significantinflammatory cells are present, but concentric rings of collagenencircle the granuloma (hematoxylin-eosin) [reproducedwith permission from Lynch JP III, Strieter RM. In: Internalmedicine for the specialist. Montvale, NJ: Medical Economics,1994; 38–62].Figure 13. Sarcoidosis. Photomicrograph of endobronchiallung biopsy specimen showing prominent multinucleatedgiant cells within the submucosa underlying the bronchioles(hematoxylin-eosin) [reproduced with permission fromLynch JP III, Strieter RM. In: Immunologically mediated pulmonarydisease. Philadelphia, PA: Lippincott-Williams andWilkins, 1991; 189–216].specimens from both the upper and lower lobesat the time of initial diagnostic bronchoscopy.When TBBs are nondiagnostic, mediastinoscopiclymph node biopsies may substantiate the diagnosis,provided that enlarged mediastinal lymphnodes are present. However, mediastinoscopy asa routine, initial diagnostic procedure for hilarlymphadenopathy is not cost-effective and has apotential for patient morbidity. 340Percutaneous needle aspiration may be analterative to surgical biopsies. In one study 341 of 19patients with suspected sarcoidosis, endosonography-guidedfine-needle aspiration (FNA) ofenlarged mediastinal lymph nodes demonstratednonnecrotizing granulomas in 100%; 1 patient hadtuberculosis (sensitivity and specificity rates of100% and 94%, respectively). Experience with thistechnique is limited. In another series, 342 transbronchialFNA (TBFNA) with a 26-gauge needlerevealed cytological features consistent with sarcoidosisin 88 of 116 patients (76%) with mediastinalor hilar adenopathy. Sensitivity of TBFNAcytology in this context was 79%; specificity was92%. Complications of TBFNA include the following:pneumothoraces (10 to 60%) or hemoptysis(5 to 10%). 343 Tambouret et al 344 performed 32 percutaneousFNA biopsies from a variety of sites (eg,lymph nodes, salivary glands, lung, liver) in 28patients with sarcoidosis; all aspirates showedgranulomatous inflammation. Simultaneous orsubsequent excisional biopsies confirmed the diagnosisin 17 patients. 344 Thus, FNA biopsy can be analternative to surgical biopsies in selected patients.The optimal approach to diagnosing mediastinallymph nodes (ie, TBFNA or CT-guided FNA)depends on the expertise and preference of thelocal institution. SLB (ie, VATS or open) is rarelynecessary to diagnose sarcoidosis. Biopsy ofextrapulmonary sites may be appropriate whenspecific lesions or abnormalities are identified (eg,lymphadenopathy, skin lesions, abnormal liverenzymes). 304,307Laboratory FeaturesLaboratory features are nonspecific. Hypercalcemiaoccurs in 1 to 4% of patients, and hypercalciuriain 15 to 40%. 304−306,345 These derangements incalcium metabolism reflect enhanced productionof 1,2-dihydroxycalciferol by mononuclear phagocytesfrom sarcoid granulomas. 304 Polyclonalhypergammaglobulinemia occurs in 30 to 80% ofpatients with chronic disease. 304−306,314,345 Serumangiotensin-converting enzyme (ACE) levels areelevated in 30 to 80% of patients with sarcoidosis.303,306,345 False-positive results are uncommon( 10%), but increased serum ACE can occur in<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 659ACC-PULM-09-0302-038.indd 6597/10/09 10:25:03 PM


active histoplasmosis and other granulomatousprocesses. The use of serum ACE as a surrogatemarker of disease activity is discussed later.Chest RadiographsAbnormalities are present on chest radiographsin 90% of patients with sarcoidosis. 303 BHL, oftenassociated with enlargement of right paratracheallymph nodes, is the classic radiographic feature,observed in more than two thirds of patients (Fig14). 143,303 <strong>Pulmonary</strong> parenchymal infiltrates arepresent in 25 to 50% of patients. 303 <strong>Pulmonary</strong>parenchymal infiltrates are typically bilateral andpatchy, with a predilection for the mid- and upperlung zones (Fig 15). Multiple focal nodular oralveolar opacities in the upper lobes may mimictuberculosis and fungal pneumonia. Other disordersthat preferentially involve the upper lobesinclude LCG, cystic fibrosis, silicosis, and chroniceosinophilic pneumonia. Cavitation is rare. 346 Pleuraleffusions occur in 2% of patients. 303 Pleuralthickening on plain chest radiographs is rare andusually reflects longstanding chronic disease. Diffusereticulonodular or miliary infiltrates may beindistinguishable from other chronic ILDs (eg, IPF,pneumoconiosis, pulmonary alveolar proteinosis,lymphangitic carcinomatosis; Fig 16). Progressionof the lung lesion results in distortion and destructionof the lung architecture, with upward retractionof the hila, broad coarse septal bands, bullae,Figure 15. Stage II sarcoidosis. Chest radiograph demonstratingextensive cystic radiolucencies (honeycombing) andfibrosis preferentially affecting upper lobes of both lungs.Linear fibrotic strands and volume loss are noted throughout.In addition, BHL is present.Figure 14. Sarcoidosis (stage I). Chest radiograph demonstratingextensive bilateral hilar and right paratracheallymphadenopathy.Figure 16. Stage III sarcoidosis. Chest radiograph demonstratespatchy reticulonodular densities throughout bothlung fields. Lung volumes are well preserved. There is nodefinite lymphadenopathy.660 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6607/10/09 10:25:04 PM


and end-stage honeycomb lung (Fig 17, top). Withfar advanced disease, mycetomas, bullous emphysema,and pulmonary hypertension may beobserved (Fig 17, bottom). Calcification of hilar ormediastinal nodes may be seen in long-standingsarcoidosis.Radiographic Classification SystemThe radiographic staging system developedmore 4 decades ago continues to be useful prognostically.This schema applies the followingcriteria on plain chest radiographs: stage 0, normal;stage I, BHL without parenchymal infiltrates; stageII, BHL plus parenchymal infiltrates; and stage III,parenchymal infiltrates without intrathoraciclymphadenopathy. 303 Some investigators advocatedefining patients with extensive destruction, fibrosis,and volume loss as stage IV. This radiographicstaging system may be useful as a prognosticguide, although significant variability exists. Spontaneousremissions occur in 60 to 90% of patientswith stage I disease, 40 to 70% with stage II, and10 to 20% with stage III. 302,303,347 Serious sequela arerare with stage I sarcoidosis but may be appreciablein other stages. Virtually all fatalities causedby pulmonary sarcoidosis occur in patients withradiographic stage II, III, or IV disease. 303 Thecourse of the disease is usually dictated withinthe first 18 to 24 months of onset. 303 Spontaneousremissions occur in up to 40% within the first 6months 347,348 ; 80% of spontaneous remissionsoccur within the first 2 years. 349 Persistence ofradiographic infiltrates beyond 2 years suggeststhat spontaneous remissions are unlikely to occur,and strongly warrants consideration of corticosteroidtherapy. 349 In some patients, the course ischronic, with multiple exacerbations or diseaseprogression over the course of years. 303,322,323Chest CTFigure 17. Top: Stage IV sarcoidosis. Posteroanterior chestradiograph demonstrates extensive cystic, bullous, andfibrotic changes; upward retraction of the hilum; andextensive pleural thickening in the left apex. Note surgicalclips on the left lung from a previous left upper lobe lobectomyfor a mycetoma. Bottom: HRCT demonstrating a mycetoma(arrow) in the superior segment of the left lower lobe in thesame patient as in Fig 17, top, 8 months after resection a leftupper lobe mycetoma. Note the bullae and cystic changes inthe remaining lung parenchyma.Chest CT is far more sensitive than conventionalchest radiography in delineating parenchymaldetails or detecting the extent of intrathoraciclymphadenopathy. 143,350 Hilar or mediastinal lymphadenopathy is present in 47 to 94% of patients withsarcoidosis. 351 The most commonly involved nodalstations (in order of decreasing frequency) are asfollows: right lower paratracheal, right hilar, subcarinal,and aortopulmonary window. 351 RoutineCT scanning is not required to diagnose or stagesarcoidosis and is not cost-effective. 350 However,CT may be helpful in patients with atypical manifestationsor with normal chest radiographs butwith clinical suspicion of disease. 143,350 The use of1- to 2-mm slice HRCT scans is superior to conventionalCT in depicting parenchymal lesions anddiscriminating alveolitis from fibrosis. 143,352 CharacteristicHRCT features of sarcoidosis include thefollowing: nodular opacities and micronodules( 3 mm in diameter) along bronchovascular<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 661ACC-PULM-09-0302-038.indd 6617/10/09 10:25:05 PM


undles, central bronchovascular thickening andnodularity, confluent nodular opacities with airbronchograms, GGOs, crowding and central retractionof bronchi and vessels near the hilae, andpleural or subpleural nodules. 143Nodules are present in 80% of patients withsarcoidosis and represent aggregates of granulomas.143,351,353 Irregularity or thickening of bronchovascularbundles, occasionally with a beadedappearance, is a cardinal sign of pulmonary sarcoidosis.351 GGOs were noted in 16 to 83% ofpatients with sarcoidosis 351,353 ; granulomatous andfibrotic lesions may give rise to this CT feature. 351Progression of the sarcoid lung lesion may formconglomerate masses, architectural distortion, andcystic destruction (Fig 18). Cavitation was notedon CT in 2.7% of patients with sarcoidosis, typicallyin patients with advanced disease. 346 In a retrospectivestudy 346 of 23 patients with cavitary sarcoidosis,52 were in radiographic stage IV); 48% hadsevere extrapulmonary involvement. Active lesions(eg, consolidation or GGO) were present concomitantlyin 19 patients (83%). 346 Wall thickening duringfollow-up suggests an infectious complication(particularly aspergillomas). 346 Other nonspecificfeatures include irregular interfaces; thickenedalveolar septae or pleural surfaces; traction bronchiectasis;and distortion or displacement of vessels,bronchi, or interlobar tissues.Figure 18. HRCT scan from a 48-year-old man with severe,chronic pulmonary sarcoidosis. Cuts at the level of the maincarina show areas of bronchial thickening and consolidationcentrally, in an axial distribution. Large cystic, bullouschanges are evident in the posterior lung fields. The lungarchitecture is distorted.In sarcoidosis, the preferential distribution ofparenchymal lesions along bronchovascular bundlesand lymphatics with an upper-lobe predominance143 contrasts sharply with IPF/UIP, which hasa predilection for the basilar and subpleural(peripheral) regions of the lungs. 253 Emphysemamay be observed in advanced stage sarcoidosis(typically stage IV) but is more extensive in smokers.354 CT features of airflow obstruction mayinclude a mosaic pattern of attenuation; bronchialwall or peribronchiolar thickening; bronchial distortion,compression, or stenosis; and bronchiectasis.355 Fibrosis and conglomerate masses may bepresent concomitantly, particularly in patients withbronchial distortion. 355 Specific CT features haveprognostic significance. Focal nodular, alveolar, orGGOs suggest an active inflammatory component.In contrast, distortion of lung parenchyma, volumeloss, linear bands, bronchiectasis, cystic radiolucencies,and bullae are characteristic of end-stagefibrosis. 143Because of its expense, I do not advocate routineHRCT in the diagnosis or staging of sarcoidosis.However, HRCT has a role in selected patientsin assessing prognosis and determining the likelihoodof response to therapy. Patients with extensivefibrosis, honeycombing, and lung distortionare not likely to respond to therapy with corticosteroidsor immunosuppressive agents. By contrast,patients with large focal alveolar opacitiesand a ground-glass pattern are better candidatesfor therapy. CT patterns may evolve over time.Serial CT in 40 patients with pulmonary sarcoidosisshowed several distinctive evolutionary patterns. 356Macroscopic nodules often disappeared ordecreased in size at follow-up. Consolidation andGGO resolved in some patients but evolved intohoneycombing in other patients (associated witha decrease in FVC). A conglomeration patternshrank and evolved into bronchial distortion anda decline in FEV 1/FVC ratio. Given the imprecisecorrelations between CT and physiologic parameters,direct measurement of PFTs is critical toassess the extent and degree of pulmonary functionalimpairment.PFTsRestrictive ventilatory defects are obser -ved in 30 to 60% of patients with pulmonary662 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6627/10/09 10:25:06 PM


sarcoidosis. 303 Airway obstruction, with reducedFEV 1and expiratory flow rates, occurs in 9 to 40%of patients. 355,357 A prospective study 358 noted airflowlimitation (defined as 70% FEV 1/FVC ratio)in 20 of 228 (8.8%) consecutive Japanese patientswith sarcoidosis. Airflow limitation was morecommon in men, smokers, and those patients classifiedas radiographic stage IV. Airflow obstructionmay reflect submucosal or endobronchial inflammation,parenchymal distortion, bronchostenosis,or exaggerated bronchial reactivity. 303,355,359,360Increased airway hyperreactivity in response tomethacholine is common in patients with sarcoidosis.359,361 The Dlco is the most sensitive of the PFTparameters, 359 but the degree of impairment is lesssevere in sarcoidosis than in IPF. 302 Hypoxemia isunusual but may be present in patients withadvanced disease. CPET findings are abnormal inup to 50% of patients with sarcoidosis, even whenstatic PFT results are normal. 362 In one study, significantassociations were noted between chestradiographic stage and TLC, Dlco, P(A-a)O 2, andchange in Po 2with exercise. 363 However, CPETsare logistically cumbersome and have limitedpractical value.PFTs at a single time point cannot discriminateactive from inactive disease and do not correlatewith histology. However, serial PFTs are importantto follow the course of the disease and assessresponse to therapy. VC improves more frequentlythan Dlco, TLC, or arterial oxygenation. 302 Changesin VC and Dlco are concordant in 90 ofpatients. 302,316 A prospective study 316 in the UnitedStates of 193 sarcoid patients cited excellent concordancebetween changes in FVC and FEV 1.Given the variability of Dlco and the expenseof obtaining lung volumes, spirometry andflow-volume loops are the most useful and costeffectiveparameters to follow the course of pulmonarysarcoidosis. FVC is the most sensitiveindicator of disease progression (or resolution).Spirometry should be repeated at 6-month intervalsfor the first 2 years, even in patients withminimal or no pulmonary symptoms. Patientswith pulmonary symptoms or derangements inpulmonary function require more frequent studies.Deteriorating pulmonary function or persistenceof severe derangements warrants a trial ofcorticosteroid therapy, regardless of whetherradiographic abnormalities are present.Ancillary StudiesSerum ACE levels, radionuclide scans, andBAL cell profiles have been used to more accuratelydetermine the extent and activity of the disease,but their role remains controversial. Serum ACElevels, 67 Ga scans, and BAL may reflect differentstages or components of the disease process. Further,serum ACE levels and 67 Ga scans may reflectthe macrophage component, which may not necessarilycorrelate with BAL lymphocytosis or lymphokinerelease.ACE: ACE, an enzyme normally produced byendothelial cells, is produced in exaggeratedamounts by sarcoid granuloma macrophages. 303Serum ACE levels may reflect total body granulomaburden but are not sensitive to active localizeddisease. Changes in serum ACE often parallelthe course of the disease and may be used to guidetreatment decisions, provided clinical informationis taken into account. Isolated elevations of serumACE in the absence of clinical symptoms do notrequire therapy. Importantly, serum ACE levelsmay be normal in patients with active disease. 302 Iobtain a baseline serum ACE level in patients withsarcoidosis and obtain serial levels in selectedpatients when clinical criteria are inadequate tojudge disease activity.67Ga Citrate Scans: 67 Ga citrate scans were usedas a surrogate marker of disease activity in sarcoidosisand other inflammatory pulmonary disorders.67 Ga is taken up avidly by activatedalveolar macrophages, and intrapulmonary uptakeof 67 Ga may be a marker of active alveolar inflammation(alveolitis). Characteristic patt erns ofuptake have been noted in sarcoidosis (eg,increased uptake in lacrimal, salivary, and parotidglands and hilar and mediastinal lymph nodes). 364,365However, 67 Ga scans are expensive, inconvenient,difficult to quantitate, and lack prognosticvalue. 364,365 I see no role for 67 Ga scanning in eitherthe initial staging or longitudinal follow-up ofpatients with sarcoidosis. However, 67 Ga scanningmay have a limited role in patients with normalchest radiographs and suspected sarcoidosis todetect clinically silent extrathoracic sites of 67 Gauptake and provide sites for biopsies. 365,366 PETscans may demonstrate increased metabolicactivity in patients with pulmonary 365 or extrapulmonarysarcoidosis (eg, bone, 367 cardiac, 368 or<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 663ACC-PULM-09-0302-038.indd 6637/10/09 10:25:06 PM


neural sites 369 ). However, the clinical value of PETis uncertain. 351BAL: BAL has provided significant insights intothe pathogenesis of sarcoidosis. 370 BAL in sarcoidosisdemonstrates increased numbers of activatedlymphocytes (typically CD4 + T cells), alveolarmacrophages, and myriad proinflammatory cytokinesand mediators. 370,371 BAL lymphocytosis ispresent in 85% of patients with pulmonary sarcoidosis;granulocytes are normal or low. 237,302,370BAL reveals increased numbers of activated Tlymphocytes (predominantly CD4 + ) and activatedmacrophages, with reduced T suppressor cells(CD8 + ), in patients with active pulmonary sarcoidosis.370−372 In late phases of sarcoidosis, neutrophilsand/or mast cells may be increased. 373 BAL cellprofiles are not specific for sarcoidosis but narrowthe differential diagnosis. 237,370 Importantly, BALcell profiles fail to predict prognosis or responsivenessto corticosteroid therapy. 302,370 High levels ofT lymphocytes or lymphokines may reflect activealveolitis but do not imply that deterioration willinevitably occur. 302 Because of the waxing and waningnature of sarcoidosis, BAL has limited prognosticvalue. Marked increases in CD4 + lymphocytes,consistent with an intensive alveolitis, are characteristicof Löfgren syndrome, yet spontaneousremissions occur in 85% of patients in this setting.302,315 Serial bronchoscopy is invasive, expensive,and impractical for sequential evaluation ofsarcoidosis.PathogenesisThe inciting signals responsible for the exuberantgranulomatous response and its subsequentprogression to fibrosis (or resolution) have notbeen identified. Interactions between activatedmononuclear phagocytes (eg, monocytes andmacrophages) and activated T-helper/inducerlymphocytes are responsible for the induction andevolution of the granulomatous process. 371,372 Atsites of active disease, striking increases in helper/inducer (CD4 + ) lymphocytes and increased CD4 + /CD8 + ratios and increased release of diverse lymphokines(consistent with a T-helper type 1 [Th1]response), monokines, and biochemical markershave been noted. 371,372 Early in the course, Th1cytokines (eg, -IFN and IL-2) predominate at sitesof disease activity; this compartmentalization ofthe immune response (with a Th1 cytokine bias)may abate as the disease remits. 371,372 Lung T cellsfrom patients with active pulmonary sarcoidosisrelease several cytokines that contribute to theinduction of the immune response and facilitateT-cell replication and mononuclear phagocyteactivation. 371,372 Alveolar macrophages frompatients with pulmonary sarcoidosis are activated,enhance antigen presentation, express receptorsfor IL-2, and release growth factors that promotefibroblast recruitment and proliferation. 371 Macrophagesplay a dual role in orchestrating the sarcoidlung lesion by producing monokines that mayeither amplify or down-regulate the inflammatoryprocess.TherapyCorticosteriods are the mainstay of therapy andmay produce dramatic remissions in patients withsevere or progressive sarcoidosis. 302,374 Favorableresponses to corticosteroid therapy are achieved in60 to 90% of patients with symptomatic sarcoidosis,but relapses occur in 16 to 74% of patients after thecessation of therapy. 302,303,348,374,375 Long-term efficacyis less clear. Because of the potential for spontaneousresolution and the toxicities associated withcorticosteroid therapy, indications for treatmentare controversial. Several prospective studies performedin the 1970s failed to show a favorableeffect on long-term prognosis. Although severalstudies suggested short-term improvement withcorticosteroid therapy, relapses often occurred afterthe cessation of therapy. The lack of efficacy mayreflect the study designs because patients withstage I disease, minimal or no symptoms, andnormal lung function were often enrolled intotreatment trials. In most of these studies, high ratesof improvement or stabilization were noted in bothtreated and untreated patients.The long-term impact of corticosteroid therapyis difficult to assess because of the high rate ofspontaneous remissions as well as inclusion ofpatients with irreversible disease. These data cannotbe extrapolated to patients with severe orprogressive pulmonary disease who are the bestcandidates for therapy. A multicenter, randomizedtrial of stage II or III sarcoidosis sponsored by theBritish Thoracic Society found that corticosteroidtherapy in patients with persistent radiographic664 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6647/10/09 10:25:07 PM


infiltrates after an initial 6-month observationperiod was associated with improved lung function,chest radiographs, and symptoms comparedwith untreated controls. 347Among 149 patients with stage II or III pulmonarysarcoidosis eligible for the study, 33 patientsrequired immediate corticosteroid therapy forcontrol of symptoms and were excluded. In addition,58 patients whose chest radiographs improvedspontaneously within 6 months of entry served asan untreated observation group. The remaining58 untreated patients with persistent radiographicinfiltrates after 6 months were randomly assignedto be administered routine corticosteroids for18 months (n 27) or selective therapy only (tocontrol symptoms or deteriorating pulmonarysymptoms; n 31). No placebo group was included.The dose of corticosteroids in the treated group wasmodest (ie, prednisolone, 30 mg/d for 1 month,tapered to 10 mg by 4 months; maintenance therapywith 10 mg/d was continued for 9 months). Incrementaldoses were permitted as necessary tomaximize radiographic improvement.The results affirmed the benefit of regular treatmentwith corticosteroid therapy in patients withradiographic stage II or III disease who failed tospontaneously remit during the 6-month period ofobservation. These data support the use of corticosteroidtherapy for patients with stage II or IIIdisease whose condition does not improve spontaneouslywithin a reasonable period. The optimalobservation period before deciding on therapy hasnot been determined. A period of not more than 6to 18 months is appropriate because spontaneousremission beyond this point is unlikely. Moreimmediate institution of therapy is warranted forpatients with severe or deteriorating symptoms orpulmonary dysfunction.Corticosteroids have potential toxicity, androutine therapy for patients with mild or no symptomsis inappropriate. Indications for therapyshould be circumscribed and focused. Despite thelack of consensus regarding efficacy of therapy,there is little doubt that corticosteroids dramaticallysuppress or reverse the disease in some cases. Corticosteroidsare warranted for myocardial, 309 CNS, 312or ophthalmologic (eg, uveitis) complications ofsarcoidosis. 304,307 Corticosteroids are also indicatedfor chronic hypercalcemia (to avert late complicationsof nephrocalcinosis and nephrolithiasis) andfor severe or progressive pulmonary or extrapulmonarydysfunction.The dose and duration of therapy need to beindividualized. Chronic disease of mild-to-moderateseverity can be treated with modest dosages(eg, 40 mg every other day for 3 to 4 months, withsubsequent taper). Greater doses (eg, prednisone1 mg/kg/d for 4 weeks, with a gradual taper) areappropriate for myocardial or CNS manifestations.The duration of therapy and the rate of taper needto be individualized according to clinical responseand presence or absence of adverse effects. A3-month trial of corticosteroids is usually adequateto judge efficacy. If no objective response has beenshown within this time, corticosteroids can betapered and discontinued. In contrast, amongresponders, corticosteroids should be continued,albeit in a tapering regimen, for 12 to 18 months.Relapses may occur as the steroid dosage isreduced or discontinued. In such cases, escalationof the dose may be efficacious. Long-term corticosteroidsare indicated only for patients who haveshown unequivocal response to therapy yet relapsewith attempts to stop therapy.Inhaled Corticosteroids: Inhaled corticosteroidswere associated with anecdotal responses inpatients with mild pulmonary sarcoidosis, but firmdata supporting their efficacy are lacking. In onedouble-blind, placebo-controlled trial in Finland, 376189 patients with pulmonary sarcoidosis (stage I,II, or III) were randomly assigned to receive eitheroral prednisone for 3 months followed by inhaledbudesonide or placebo. Among patients with stageII disease, chest radiographic findings did not differbetween groups. Differences in PFTs (FVC,Dlco) between groups were minimal. A trendtoward improvement in Dlco was noted in corticosteroid-treatedpatients with stage II disease.These differences were small and did not achievestatistical significance. Changes in FVC over timewere similar between groups.Long-term follow-up from the Finnish study, 377which randomized patients with stage I and IInewly diagnosed ( 3 months) sarcoidosis toimmediate treatment with placebo or prednisolonefor 3 months followed by inhaled budesonide for15 months found no differences in treated or placebogroups in chest radiographs at 1 year. However,at 5 years, patients in the placebo cohorthad lower mean FVC and Dlco and worse chest<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 665ACC-PULM-09-0302-038.indd 6657/10/09 10:25:07 PM


adiograph findings compared with treatedpatients. Differences between groups were small,and the significance of these observations is notclear.In another double-blind trial, British investigators378 randomly assigned 44 adults with stage IIor III pulmonary sarcoidosis to inhaled fluticasone(1,000 µg bid) or placebo for 6 months. Approximately75% were administered oral corticosteroidsat the start of the study. No significant differencesin PFTs, mean dose of oral corticosteroid, or chestradiograph findings were detected at any timepoint. These studies 376−378 suggest that inhaled corticosteroidshave minimal value as the primarytherapy for pulmonary sarcoidosis. However,given their low toxicity, a trial of an inhaled corticosteroidis reasonable in patients with significantendobronchial inflammation, cough, or persistentpulmonary symptoms.Immunosuppressive or Cytotoxic Agents: Immunosuppressiveor cytotoxic agents (eg, methotrexate,azathioprine, cyclosporine, chlorambucil,cyclophosphamide, and leflunomide) have beenused, with anecdotal successes, in patients inwhom treatment is failing or who are experiencingside effects from corticosteroids. However, randomizedtrials evaluating these agents are lacking,and the best agent has not been determined.Immunosuppressive Agents: Both azathioprineand methotrexate may be useful for sarcoidosisrefractory to corticosteroids or as steroid-sparingagents in patients requiring high-dose corticosteroidsfor control of the disease.Azathioprine: azathioprine (2 to 3 mg/kg/d),alone or combined with corticosteroids, has beenassociated with anecdotal successes in sarcoidosis,even in patients in whom corticosteroid therapy isfailing. However, randomized trials have not beenperformed. Further, studies directly comparingazathioprine with alternative agents for sarcoidosisare lacking. In two early studies, 379,380 10 of 20patients in whom corticosteroid therapy is failingresponded to azathioprine. In another retrospectivestudy, 381 8 of 14 patients with neurosarcoidosisresponded to azathioprine. Diab et al 382 treatedseven patients with a combination of prednisoneand azathioprine; all patients improved. In contrast,investigators from South Africa found azathioprineto be of marginal benefit in a retrospectivestudy 383 of 10 patients with pulmonary sarcoidosis.All had shown only partial (n 6) or no (n 4)response to corticosteroids. The addition of azathioprine(100 to 150 mg/d) plus low-dose corticosteroidswas associated with significant andsustained improvement in lung function in onlytwo patients; two additional patients had transientimprovement and a steroid-sparing effect. Twopatients in whom azathioprine therapy failed subsequentlyresponded to cyclosporine A (CsA) andhigh-dose corticosteroids. No patient who did notrespond to high-dose corticosteroids responded toazathioprine. German investigators 384 treated 11patients with chronic sarcoidosis with azathioprineplus prednisolone. All had symptomatic improvement;chest radiographs improved in nine patients,and PFT results improved in seven patients. Laterelapses (8 to 22 months) occurred in three patients.Although these data are limited, I believe azathioprinemay be useful as a steroid-sparing agent orin selected patients with severe or progressivesarcoidosis refractory to corticosteroids. Extensiveclinical experience with azathioprine in organtransplant recipients and other immune disorderssuggests that serious late sequela associated withchronic azathioprine use are uncommon. 265Methotrexate: Methotrexate, a folic acid antagonistwith both immunosuppressive and antiinflammatoryeffects, has been efficacious for bothpulmonary and extrapulmonary sarcoidosis. 385,386Methotrexate can be administered parenterally(intramuscularly) or orally as a pulse once weekly.Dosages ranging from 10 to 25 mg/wk have beenused. In uncontrolled studies 387,88 by investigatorsfrom the University of Cincinnati who evaluated 230 patients, favorable responses to methotrexatewere cited in 52 to 66% of patients. Relapses werefrequent with cessation of therapy but usuallyresponded to reinstitution of methotrexate.Although these studies are not definitive,methotrexate has a role as a steroid-sparing agentin sarcoidosis. Because of potential toxicities,methotrexate should be restricted to patientsrequiring unacceptably high doses of corticosteroids( 20 mg/d prednisolone or equivalent) orexperiencing serious side effects from corticosteroids.In this context, a 4- to 6-month trial ofmethotrexate is reasonable. I initiate therapy with7.5 mg once weekly and escalate the dose by 2.5-mgincrements weekly (to a maximal dose of 20 mg)until a clinical response or toxicity has occurred.666 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6667/10/09 10:25:07 PM


Serious adverse effects are rare ( 1% of patients),but side effects requiring cessation of therapy occurin 3 to 15% of cases. 265Adverse effects of methotrexate are dosedependent and may be minimized by the additionof folic acid (1 mg/d). Adverse effects associatedwith methotrexate include GI effects (eg, nausea,vomiting, and diarrhea), stomatitis (eg, mucosal ororal ulcers), and rash. Less common complications,each occurring in 1 to 4% of patients, includemegaloblastic anemia, leukopenia, or thrombocytopenia;opportunistic infections; and interstitialpneumonitis. 265,389In contrast to the alkylating agents, methotrexateis not carcinogenic. Methotrexate is teratogenicand cannot be used in patients at risk of conception.The most worrisome side effect is hepatic cirrhosis,which may occur in up to 2% of patients with longtermuse ( 2 years). 265 The risk of permanent liverdisease with long-term use of methotrexate ( 2years) has not been defined in patients with sarcoidosis.A series 390 of 68 patients with chronicsarcoidosis treated with long-term ( 2 years)methotrexate therapy cited histologic featuresconsistent with MTX toxicity in 14 patients (21%).Importantly, serial liver function tests were notuseful in determining which patients would havemethotrexate toxicities.Contraindications to methotrexate includeethanol abuse, concomitant liver disease, historyof hepatitis, patients unable to adhere to dosingschedules, renal failure, active infection, anemia,and leukopenia. Serial transaminase levels, CBCcounts, and platelet counts should be monitoredevery 4 to 8 weeks in patients receiving methotrexate.Persistent or progressive rises in hepaticenzymes, thrombocytopenia, or leukopenia warrantdiscontinuation of therapy or reduction of thedose. Given its potential for late, irreversible hepatotoxicity,methotrexate should be continuedbeyond 6 months only in patients demonstratingunequivocal objective improvement. Enthusiasmfor its use needs to be tempered by the high ratesof relapse noted on cessation of therapy. I preferazathioprine for patients with chronic, progressivesarcoidosis requiring long-term treatment ( 1year).Leflunomide: Leflunomide has been used to treatsarcoidosis, but data are sparse. One nonrandomizedtrial 391 cited responses in 25 of 32 sarcoidpatients (78%) treated with leflunomide (alone orcombined with methotrexate). Leflunomideappears to be less toxic than methotrexate and maybe an alternative to methotrexate, but data aresparse.Mycophenolate Mofetil: Responses to mycophenolatemofetil were noted in anecdotal cases ofextrapulmonary sarcoidosis, 392,393,394 but data arelacking in pulmonary sarcoidosis.CsA: CsA, a fungal decapeptide that exhibitsrelatively selective inhibitory effects on T-cellactivation, proliferation, and lymphokine release,would be expected to be an ideal agent for treatingsarcoidosis. However, despite anecdotal responsesin some corticosteroid-resistant patients, 383 CsAhas been disappointing as therapy for pulmonarysarcoidosis. In a study 395 from the National Institutesof Health, eight patients with symptomaticpulmonary sarcoidosis received oral CsA, 10mg/kg/d, for 6 months. Despite inhibitory effectson T-cell proliferation and cytokine release invitro, pulmonary function or BAL results did notchange.In one study, 396 six patients with corticosteroid-refractoryCNS sarcoidosis were treatedwith CsA plus corticosteroids for 6 months. Nopatient achieved complete remission, but a steroid-sparingeffect was suggested. A subsequentstudy 381 by these investigators cited favorableresponses in 11 of 14 patients with neurosarcoidosistreated with CsA. A randomized controlledtrial by Wyser and coworkers 397 showed no benefitwith oral CsA, 5 to 7 mg/kg/d, combinedwith prednisone as compared with prednisonealone in a cohort of 37 patients with pulmonarysarcoidosis. Adverse effects (particularly renalinsufficiency and infections) and relapses weregreater in the patients receiving combined therapy.397 CsA is very expensive and is associatedwith numerous toxicities (eg, hypertension, renalinsufficiency, hirsutism, neuropathies, disordersin lipids, heightened susceptibilities to infections,and lymphoproliferative disorders). 265 In light ofthe myriad complications associated with itsuse and the lack of demonstrated efficacy, cyclosporinehas at best a marginal role as salvagetherapy for patients with severe, progressivesarcoidosis refractory to corticosteroids. Topicalcyclosporine eye drops may be effective for sarcoidconjunctivitis. 398<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 667ACC-PULM-09-0302-038.indd 6677/10/09 10:25:07 PM


Alkylating AgentsPublished data regarding alkylating agents (eg,cyclophosphamide, chlorambucil) for the treatmentof sarcoidosis are limited to sporadic casereports and small retrospective series, Chlorambucil,an alkylating agent of the nitrogen mustardclass, was associated with favorable responses innonrandomized studies (primarily in patientsexperiencing adverse effects from corticosteroids).399−401 However, chlorambucil is oncogenicand has numerous toxicities. 265 Given the availabilityof alternative therapeutic agents, chlorambucilhas no role in the treatment of sarcoidosis.Cyclophosphamide: Cyclophosphamide (oral orpulse) has rarely been used to treat sarcoidosis.Anecdotal responses have been noted, but data arelimited to a few case reports and small series. 312,374,402,403Lower and colleagues 312 cited favorable responsesin 8 of 10 patients treated with IV pulse cyclophosphamidefor neurosarcoidosis, 8 of whom hadconcomitant lung involvement. Previous therapyincluded corticosteroids in all 10 and methotrexatein 8. Thus, cyclophosphamide has a role for severesarcoidosis refractory to corticosteroids or otherimmunosuppressive agents. However, in light of itsoncogenic potential, particularly in relatively youngpatients, I reserve cyclophosphamide for severesarcoidosis refractory to corticosteroids and azathioprineor methotrexate.Other Antiinflammatory andImmunomodulatory AgentsNonsteroidal antiinflammatory agents (eg,indomethacin and phenylbutazone) have beenused to treat the acute articular manifestations ofsarcoidosis, but have no role in treating moresevere cases of pulmonary or extrapulmonarysarcoidosis.Antimalarial Drugs: Antimalarial drugs (eg,chloroquine and hydroxychloroquine) have immunomodulatingproperties and are efficacious intreating rheumatoid arthritis, systemic lupus erythematosus,and diverse immune-mediated diseases.Antimalarials concentrate in cells of thereticuloendothelial system and melanin-containingtissues (eg, skin, spleen, leukocytes, kidney) andare preferentially concentrated in epithelioid,mononuclear, and giant cells comprising sarcoidgranulomata. Anecdotal responses have beennoted with the antimalarials for cutaneous, 404 osseous,374 and neurologic 405 sarcoidosis and sarcoidinducedhypercalcemia. 406,407The role of antimalarials in treating pulmonarysarcoidosis is less well established. However, arandomized trial 408 of 23 patients with symptomaticpulmonary sarcoidosis (radiographic stage IIor III) suggested benefit with chloroquine. Sixteenof the 23 patients had previously been treated withhigh-dose corticosteroids, without sustainedimprovement. All 23 patients were treated withhigh-dose chloroquine (750 mg/d) for 6 months.Five patients withdrew from the study before the6-month acute phase (three as the result of intolerableside effects). After 6 months, the dose of chloroquinewas tapered by 250 mg every 2 months.Eighteen patients were then randomly assignedto a maintenance group (chloroquine, 250 mg/d;n 10) or observation group (no chloroquine;n 8).After the initial 6-month treatment with highdosechloroquine, symptoms, PFT results, serumACE levels, and Ga 67 scan findings improved. Afterrandomization, the rate of decline in PFTs wasslower, and there were fewer relapses in patientsreceiving maintenance chloroquine vs no chloroquinetherapy. Studies comparing chloroquine withchloroquine therapy have not been performed. Themajor toxicity of antimalarials is ocular (particularlychloroquine); dizziness and nausea may alsooccur. 408 Unfortunately, prolonged administrationof chloroquine can lead to irreversible retinopathyand blindness. Given the potential serious oculartoxicity associated with chloroquine, hydroxychloroquine,which is less toxic, is preferred. The doseof hydroxychloroquine is 200 mg once or twicedaily for a 6-month trial. Long-term maintenancetherapy (100 to 400 mg/d) is reserved for patientsmanifesting unequivocal responses. Slit-lampexaminations should be done by an ophthalmologistevery 9 months to rule out ocular toxicity. Thecombination of hydroxychloroquine with corticosteroidsor immunosuppressive agents mayenhance immunomodulatory effects comparedwith any of these agents alone.Alternative Therapies: Anecdotal responseshave been cited with thalidomide (a sedativewith teratogenic properties), 409,410 melatonin, 411 and668 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6687/10/09 10:25:08 PM


pentoxifylline 412 in patients with sarcoidosis, butthe value of these agents is unproven. Pentoxifyllineand dexamethasone suppressed release ofTNF- from sarcoid alveolar macrophages in vitro,but in vivo data were lacking. 413TNF-: TNF- inhibitors (particularly infliximab)have been used, with anecdotal success, totreat refractory sarcoidosis (particularly lupus pernioand extrapulmonary cases). 414−427 Data in pulmonarysarcoidosis are limited. 422,428,429 In a recentmulticenter trial, 422 138 patients with chronic pulmonarysarcoidosis were randomized to placebo orinfliximab, 3 mg/kg, or infliximab, 5 mg/kg, atweeks 0, 2, 6, 12, 18, and 24. At 24 weeks, the primaryend point change in FVC percentage of predictedwas slightly greater among infliximab-treatedpatients (2.5% above baseline) compared with placebo-treatedpatients (no change). This difference,although statistically significant, is of doubtfulclinical significance. Treatment with TNF inhibitorsis very expensive and has been associated with aheightened risk of tuberculosis, 430 other opportunisticinfections, and other adverse effects, 431−434including lymphoma development. 435 Currently,additional studies are required to establish the roleof TNF- inhibitors (eg, infliximab, etanercept, oradalimumab) as therapy for sarcoidosis.<strong>Pulmonary</strong> Hypertension<strong>Pulmonary</strong> hypertension is a rare complicationof sarcoidosis but may occur in patients withadvanced radiographic stages. 195,436−439 Althoughdata are limited, favorable responses to short- andlong-term vasodilators have been cited (eg, inhalednitric oxide, IV epoprostenol, and oral calciumchannelblockers). 437 Supplemental oxygen shouldbe administered for patients with hypoxemia.LTSLT or BLT has been successfully accomplishedin patients with severe pulmonary sarcoidosisrefractory to aggressive medical therapy. 158,438,440−442From January 1995 to June 2006, 438 adults worldwidehad received LT for sarcoidosis. 293 Long-termsurvival rates after LT for sarcoidosis are generallysimilar to other indications. 441 Recurrence of sarcoidgranulomas within the lung allografts is commonbut rarely causes clinical symptoms. 443 Decidingwhen to list patients for LT is difficult becausepredicting life expectancy in individual patients isdifficult. 438,444 However, waiting time for organs canbe prolonged, and 27 to 53% of patients with endstagepulmonary sarcoidosis die while awaitingtransplantation. 438,440,445 Typically, patients withsarcoidosis referred for LT have severe impairmentin PFTs (mean percentage of predicted FVC 44%;mean percentage of predicted FEV 1 38%). 440,445In one study 438 of patients with sarcoidosisawaiting LT, the following factors were associatedwith increased mortality on the waiting list: Pao 2 60 mm Hg, mean PAP 35 mm Hg, cardiac index 2 L/min/m 2 , and right atrial pressure 15 mmHg. A retrospective review 445 of 405 sarcoid patientslisted for LT in the United States noted three variablesthat were independently associated withincreased mortality: black race, amount of supplementaloxygen needed, and mPAP. <strong>Pulmonary</strong>arterial hypertension is an ominous sign in sarcoidosis,and warrants referral for LT. Invasive aspergillosismay complicate LT among sarcoid patientswith preexisting mycetomas. 158References1. Lynch J III, Fishbein MC, Saggar R, et al. Idiopathicpulmonary fibrosis. Exp Rev Respir Med2007; 1:377–3902. Lynch JP 3rd, Saggar R, Weigt SS, et al. Usualinterstitial pneumonia. Semin Respir Crit CareMed 2006; 27:634–6513. American Thoracic Society. Idiopathic pulmonaryfibrosis: diagnosis and treatment. Internationalconsensus statement. American Thoracic Society(ATS), and the European Respiratory Society (ERS).Am J Respir Crit Care Med 2000; 161:646–6644. American Thoracic Society/European RespiratorySociety International Multidisciplinary ConsensusClassification of the Idiopathic InterstitialPneumonias. This joint statement of the AmericanThoracic Society (ATS), and the EuropeanRespiratory Society (ERS) was adopted by theATS board of directors, June 2001 and by the ERSExecutive Committee, June 2001. Am J RespirCrit Care Med 2002; 165:277–3045. Katzenstein AL, Mukhopadhyay S, Myers JL.Diagnosis of usual interstitial pneumonia anddistinction from other fibrosing interstitial lungdiseases. Hum Pathol 2008; 39:1275–1294<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 669ACC-PULM-09-0302-038.indd 6697/10/09 10:25:08 PM


6. Katzenstein AL, Myers JL. Idiopathic pulmonaryfibrosis: clinical relevance of pathologic classification.Am J Respir Crit Care Med 1998; 157:1301–13157. Lai CK, Wallace WD, Fishbein MC. Histopathologyof pulmonary fibrotic disorders. SeminRespir Crit Care Med 2006; 27:613–6228. King TE Jr. Clinical advances in the diagnosisand therapy of the interstitial lung diseases. Am JRespir Crit Care Med 2005; 172:268–2799. Ryu JH, Colby TV, Hartman TE, et al. Smokingrelatedinterstitial lung diseases: a concise review.Eur Respir J 2001; 17:122–13210. Ryu JH, Myers JL, Capizzi SA, et al. Desquamativeinterstitial pneumonia and respiratory bronchiolitis-associatedinterstitial lung disease. Chest2005; 127:178–18411. Hartman TE, Swensen SJ, Hansell DM, et al. Nonspecificinterstitial pneumonia: variable appearanceat high-resolution chest CT. Radiology 2000;217:701–70512. Flaherty KR, Martinez FJ, Travis W, et al. Nonspecificinterstitial pneumonia (NSIP). Semin RespirCrit Care Med 2001; 22:423–43413. Martinez FJ. Idiopathic interstitial pneumonias:usual interstitial pneumonia versus nonspecificinterstitial pneumonia. Proc Am Thorac Soc 2006;3:81–9514. Travis WD, Hunninghake G, King TE Jr., et al.Idiopathic nonspecific interstitial pneumonia:report of an American Thoracic Society project.Am J Respir Crit Care Med 2008; 177:1338–134715. Vourlekis JS, Brown KK, Cool CD, et al. Acuteinterstitial pneumonitis: case series and reviewof the literature. <strong>Medicine</strong> (Baltimore) 2000; 79:369–37816. Bouros D, Nicholson AC, Polychronopoulos V,et al. Acute interstitial pneumonia. Eur Respir J2000; 15:412–41817. Koss MN, Shigemitsu H. Lymphocytic interstitialpneumonia (LIP) and other pulmonary lymphoproliferativedisorders. In: Lynch JP III, ed. Interstitialpulmonary and bronchiolar disorders (Vol227). New York: InformaUSA, 2008; 403–42818. Nicholson AG. Lymphocytic interstitial pneumoniaand other lymphoproliferative disorders inthe lung. Semin Respir Crit Care Med 2001; 22:409–42219. Cordier JF. Cryptogenic organising pneumonia.Eur Respir J 2006; 28:422–44620. Silva CI, Muller NL, Lynch DA, et al. Chronichypersensitivity pneumonitis: differentiationfrom idiopathic pulmonary fibrosis and nonspecificinterstitial pneumonia by using thin-sectionCT. Radiology 2008; 246:288–29721. Selman M, Mejia M, Pardo A. HypersensitivityPneumonitis. In: Lynch JP III, ed. Interstitialpulmonary and bronchiolar disorders (Vol 227).New York: InformaUSA, 2008; 267–28822. Flaherty KR, Thwaite EL, Kazerooni EA, et al.Radiological versus histological diagnosis in UIPand NSIP: survival implications. Thorax 2003;58:143–14823. Lynch DA, Godwin JD, Safrin S, et al. High-resolutioncomputed tomography in idiopathic pulmonaryfibrosis: diagnosis and prognosis. Am JRespir Crit Care Med 2005; 172:488–49324. Hunninghake GW, Lynch DA, Galvin JR, et al.Radiologic findings are strongly associated witha pathologic diagnosis of usual interstitial pneumonia.Chest 2003; 124:1215–122325. Antoniou KM, Hansell DM, Rubens MB, et al.Idiopathic pulmonary fibrosis: outcome in relationto smoking status. Am J Respir Crit CareMed 2008; 177:190–19426. Hubbard R, Johnston I, Britton J. Survival inpatients with cryptogenic fibrosing alveolitis:a population-based cohort study. Chest 1998;113:396–40027. Douglas WW, Ryu JH, Schroeder DR. Idiopathicpulmonary fibrosis: impact of oxygen and colchicine,prednisone, or no therapy on survival. Am JRespir Crit Care Med 2000; 161:1172–117828. King TE Jr, Tooze JA, Schwarz MI, et al. Predictingsurvival in idiopathic pulmonary fibrosis:scoring system and survival model. Am J RespirCrit Care Med 2001; 164:1171–118129. Coultas DB, Zumwalt RE, Black WC, et al. Theepidemiology of interstitial lung diseases. Am JRespir Crit Care Med 1994; 150:967–97230. Raghu G, Weycker D, Edelsberg J, et al. Incidenceand prevalence of idiopathic pulmonary fibrosis.Am J Respir Crit Care Med 2006; 174:810–81631. Gribbin J, Hubbard RB, Le Jeune I, et al. Incidenceand mortality of idiopathic pulmonaryfibrosis and sarcoidosis in the UK. Thorax 2006;61:980–98532. Hubbard R, Johnston I, Coultas DB, et al. Mortalityrates from cryptogenic fibrosing alveolitis inseven countries. Thorax 1996; 51:711–716670 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6707/10/09 10:25:08 PM


33. Harris JM, Cullinan P, McDonald JC. Occupationaldistribution and geographic clustering ofdeaths certified to be cryptogenic fibrosing alveolitisin England and Wales. Chest 2001; 119:428–43334. Iwai K, Mori T, Yamada N, et al. Idiopathic pulmonaryfibrosis: epidemiologic approaches tooccupational exposure. Am J Respir Crit CareMed 1994; 150:670–67535. Scott J, Johnston I, Britton J. What causes cryptogenicfibrosing alveolitis? A case-control studyof environmental exposure to dust. BMJ 1990;301:1015–101736. Olson AL, Swigris JJ, Lezotte DC, et al. Mortalityfrom pulmonary fibrosis increased in the UnitedStates from 1992 to 2003. Am J Respir Crit CareMed 2007; 176:277–28437. Mannino DM, Etzel RA, Parrish RG. <strong>Pulmonary</strong>fibrosis deaths in the United States, 1979–1991: ananalysis of multiple-cause mortality data. Am JRespir Crit Care Med 1996; 153:1548–155238. Thomas AQ, Lane K, Phillips J 3rd, et al. Heterozygosityfor a surfactant protein C gene mutationassociated with usual interstitial pneumonitisand cellular nonspecific interstitial pneumonitisin one kindred. Am J Respir Crit Care Med 2002;165:1322–132839. Fan LL, Kozinetz CA. Factors influencing survivalin children with chronic interstitial lungdisease. Am J Respir Crit Care Med 1997; 156:939–94240. Baumgartner KB, Samet JM, Coultas DB, et al.Occupational and environmental risk factorsfor idiopathic pulmonary fibrosis: a multicentercase-control study: collaborating centers. Am JEpidemiol 2000; 152:307–31541. Garantziotis S, Schwartz DA. Host-environmentinteractions in pulmonary fibrosis. Semin RespirCrit Care Med 2006; 27:574–58042. Taskar VS, Coultas DB. Is idiopathic pulmonaryfibrosis an environmental disease? Proc Am ThoracSoc 2006; 3:293–29843. Baumgartner KB, Samet JM, Stidley CA, et al.Cigarette smoking: a risk factor for idiopathicpulmonary fibrosis. Am J Respir Crit Care Med1997; 155:242–24844. Hubbard R, Lewis S, Richards K, et al. Occupationalexposure to metal or wood dust and aetiologyof cryptogenic fibrosing alveolitis. Lancet1996; 347:284–28945. Schwartz DA, Van Fossen DS, Davis CS, et al.Determinants of progression in idiopathic pulmonaryfibrosis. Am J Respir Crit Care Med 1994;149:444–44946. Steele MP, Speer MC, Loyd JE, et al. Clinical andpathologic features of familial interstitial pneumonia.Am J Respir Crit Care Med 2005; 172:1146–115247. Miyake Y, Sasaki S, Yokoyama T, et al. Occupationaland environmental factors and idiopathicpulmonary fibrosis in Japan. Ann Occup Hyg2005; 49:259–26548. Sweet MP, Patti MG, Leard LE, et al. Gastroesophagealreflux in patients with idiopathic pulmonaryfibrosis referred for lung transplantation.J Thorac Cardiovasc Surg 2007; 133:1078–108449. Raghu G, Freudenberger TD, Yang S, et al. Highprevalence of abnormal acid gastro- oesophagealreflux in idiopathic pulmonary fibrosis. EurRespir J 2006; 27:136–14250. Patti MG, Tedesco P, Golden J, et al. Idiopathic pulmonaryfibrosis: how often is it really idiopathic?J Gastrointest Surg 2005; 9:1053–1056, discussion6–851. Highland KB, Garin MC, Brown KK. The spectrumof scleroderma lung disease. Semin RespirCrit Care Med 2007; 28:418–42952. D’Ovidio F, Mura M, Tsang M, et al. Bile acidaspiration and the development of bronchiolitisobliterans after lung transplantation. J ThoracCardiovasc Surg 2005; 129:1144–115253. D’Ovidio F, Singer LG, Hadjiliadis D, et al. Prevalenceof gastroesophageal reflux in end-stagelung disease candidates for lung transplant. AnnThorac Surg 2005; 80:1254–126054. Blondeau K, Mertens V, Vanaudenaerde BA, et al.Gastro-oesophageal reflux and gastric aspirationin lung transplant patients with or withoutchronic rejection. Eur Respir J 2008; 31:707–71355. Belperio JA, Weigt SS, Fishbein MC, et al. Chroniclung allograft rejection: mechanisms and therapy.Proc Am Thorac Soc 2009; 6:108–12156. Cantu E 3rd, Appel JZ 3rd, Hartwig MG, et al.J. Maxwell Chamberlain Memorial Paper: earlyfundoplication prevents chronic allograft dysfunctionin patients with gastroesophageal refluxdisease. Ann Thorac Surg 2004; 78:1142–1151; discussion115157. Hartwig MG, Appel JZ, Davis RD. Antirefluxsurgery in the setting of lung transplantation:<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 671ACC-PULM-09-0302-038.indd 6717/10/09 10:25:09 PM


strategies for treating gastroesophageal refluxdisease in a high-risk population. Thorac SurgClin 2005; 15:417–42758. Davis RD Jr., Lau CL, Eubanks S, et al. Improvedlung allograft function after fundoplication inpatients with gastroesophageal reflux diseaseundergoing lung transplantation. J Thorac CardiovascSurg 2003; 125:533–54259. du Bois RM. Genetic factors in pulmonary fibroticdisorders. Semin Respir Crit Care Med 2006;27:581–58860. Grutters JC, du Bois RM. Genetics of fibrosinglung diseases. Eur Respir J 2005; 25:915–92761. Marshall RP, Puddicombe A, Cookson WO, et al.Adult familial cryptogenic fibrosing alveolitis inthe United Kingdom. Thorax 2000; 55:143–14662. Lawson WE, Loyd JE. The genetic approach inpulmonary fibrosis: can it provide clues to thiscomplex disease? Proc Am Thorac Soc 2006;3:345–34963. Garcia CK, Raghu G. Inherited interstitial lungdisease. Clin Chest Med 2004; 25:421–43364. Rosas IO, Ren P, Avila NA, et al. Early interstitiallung disease in familial pulmonary fibrosis. Am JRespir Crit Care Med 2007; 176:698–70565. Yang IV, Burch LH, Steele MP, et al. Gene expressionprofiling of familial and sporadic interstitialpneumonia. Am J Respir Crit Care Med 2007;175:45–5466. Kaminski N. Microarray analysis of idiopathicpulmonary fibrosis. Am J Respir Cell Mol Biol2003; 29(3 Suppl):S32–S3667. Selman M, Pardo A, Barrera L, et al. Gene expressionprofiles distinguish idiopathic pulmonaryfibrosis from hypersensitivity pneumonitis. Am JRespir Crit Care Med 2006; 173:188–19868. Whyte M, Hubbard R, Meliconi R, et al. Increasedrisk of fibrosing alveolitis associated with interleukin-1receptor antagonist and tumor necrosisfactor-alpha gene polymorphisms. Am J RespirCrit Care Med 2000; 162:755–75869. Zorzetto M, Ferrarotti I, Trisolini R, et al. Complementreceptor 1 gene polymorphisms are associatedwith idiopathic pulmonary fibrosis. Am JRespir Crit Care Med 2003; 168:330–33470. Xaubet A, Marin-Arguedas A, Lario S, et al.Transforming growth factor-beta1 gene polymorphismsare associated with disease progressionin idiopathic pulmonary fibrosis. Am J RespirCrit Care Med 2003; 168:431–43571. Molina-Molina M, Xaubet A, Li X, et al. Angiotensinogengene G-6A polymorphism influencesidiopathic pulmonary fibrosis disease progression.Eur Respir J 2008; 32:1004–100872. Nogee LM, Dunbar AE 3rd, Wert SE, et al. Amutation in the surfactant protein C gene associatedwith familial interstitial lung disease. NEngl J Med 2001; 344:573–57973. Cameron HS, Somaschini M, Carrera P, et al. Acommon mutation in the surfactant protein Cgene associated with lung disease. J Pediatr 2005;146:370–37574. Hamvas A, Nogee LM, White FV, et al. Progressivelung disease and surfactant dysfunctionwith a deletion in surfactant protein C gene. AmJ Respir Cell Mol Biol 2004; 30:771–77675. Whitsett JA. Genetic basis of familial interstitiallung disease: misfolding or function of surfactantprotein C? Am J Respir Crit Care Med 2002;165:1201–120276. Whitsett JA, Weaver TE. Hydrophobic surfactantproteins in lung function and disease. N Engl JMed 2002; 347:2141–214877. Armanios MY, Chen JJ, Cogan JD, et al. Telomerasemutations in families with idiopathicpulmonary fibrosis. N Engl J Med 2007; 356:1317–132678. Hao LY, Armanios M, Strong MA, et al. Short telomeres,even in the presence of telomerase, limittissue renewal capacity. Cell 2005; 123:1121–113179. Alder JK, Chen JJ, Lancaster L, et al. Short telomeresare a risk factor for idiopathic pulmonaryfibrosis. Proc Natl Acad Sci USA 2008; 105:13051–1305680. Cronkhite JT, Xing C, Raghu G, et al. Telomereshortening in familial and sporadic pulmonaryfibrosis. Am J Respir Crit Care Med 2008; 178:729–73781. Brantly M, Avila NA, Shotelersuk V, et al. <strong>Pulmonary</strong>function and high-resolution CT findingsin patients with an inherited form of pulmonaryfibrosis, Hermansky-Pudlak syndrome, due tomutations in HPS-1. Chest 2000; 117:129–13682. Auwerx J, Boogaerts M, Ceuppens JL, et al.Defective host defence mechanisms in a familywith hypocalciuric hypercalcaemia and coexistinginterstitial lung disease. Clin Exp Immunol1985; 62:57–6483. Riccardi VM. Von Recklinghausen neurofibromatosis.N Engl J Med 1981; 305:1617–1627672 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6727/10/09 10:25:09 PM


84. Young LM, Hopkins R, Wilsher ML. Lower occurrenceof idiopathic pulmonary fibrosis in Maoriand Pacific Islanders. Respirology 2006; 11:467–47085. Churg A. The uptake of mineral particles by pulmonaryepithelial cells. Am J Respir Crit CareMed 1996; 154:1124–114086. Ortiz LA, Lasky J, Hamilton RF Jr., et al. Expressionof TNF and the necessity of TNF receptorsin bleomycin-induced lung injury in mice. ExpLung Res 1998; 24:721–74387. Nunes H, Uzunhan Y, Valeyre D, et al. Connectivetissue disease-associated interstitial lung disease.In: Lynch JP III, ed. Interstitial pulmonaryand bronchiolar disorders (Vol 227). New York:Informa-USA, 2008; 429–48688. Hubbard RB, Smith C, Le Jeune I, et al. The associationbetween idiopathic pulmonary fibrosisand vascular disease: a population-based study.Am J Respir Crit Care Med 2008; 178:1257–126189. Kizer JR, Zisman DA, Blumenthal NP, et al. Associationbetween pulmonary fibrosis and coronaryartery disease. Arch Intern Med 2004; 164:551–55690. Ponnuswamy A, Manikandan R, Sabetpour A,et al. Association between ischaemic heart diseaseand interstitial lung disease: a case-controlstudy. Respir Med 2009; 103:503–50791. Gribbin J, Hubbard R, Smith C. Role of diabetesmellitus and gastro-oesophageal reflux inthe aetiology of idiopathic pulmonary fibrosis.Respir Med 2009; 103:927-93192. Enomoto T, Usuki J, Azuma A, et al. Diabetesmellitus may increase risk for idiopathic pulmonaryfibrosis. Chest 2003; 123:2007–201193. Martinez FJ, Safrin S, Weycker D, et al. The clinicalcourse of patients with idiopathic pulmo naryfibrosis. Ann Intern Med 2005; 142:963–96794. Selman M, Carrillo G, Estrada A, et al. Acceleratedvariant of idiopathic pulmonary fibrosis:clinical behavior and gene expression pattern.PLoS ONE 2007; 2:e48295. Collard HR, Moore BB, Flaherty KR, et al.Acute exacerbations of idiopathic pulmonaryfibrosis. Am J Respir Crit Care Med 2007; 176:636–64396. Bjoraker JA, Ryu JH, Edwin MK, et al. Prognosticsignificance of histopathologic subsets in idiopathicpulmonary fibrosis. Am J Respir Crit CareMed 1998; 157:199–20397. Flaherty KR, Travis WD, Colby TV, et al. Histopathologicvariability in usual and nonspecificinterstitial pneumonias. Am J Respir Crit CareMed 2001; 164:1722–172798. Flaherty KR, Toews GB, Travis WD, et al. Clinicalsignificance of histological classification of idiopathicinterstitial pneumonia. Eur Respir J 2002;19:275–28399. Daniels CE, Yi ES, Ryu JH. Autopsy findings in 42consecutive patients with idiopathic pulmonaryfibrosis. Eur Respir J 2008; 32:170–174100. Rudd RM, Prescott RJ, Chalmers JC, et al. BritishThoracic Society Study on cryptogenic fibrosingalveolitis: response to treatment and survival.Thorax 2007; 62:62–66101. Panos RJ, Mortenson RL, Niccoli SA, et al. Clinicaldeterioration in patients with idiopathic pulmonaryfibrosis: causes and assessment. Am JMed 1990; 88:396–404102. Le Jeune I, Gribbin J, West J, et al. The incidenceof cancer in patients with idiopathic pulmonaryfibrosis and sarcoidosis in the UK. Respir Med2007; 101:2534–2540103. Stern JB, Mal H, Groussard O, et al. Prognosis ofpatients with advanced idiopathic pulmonaryfibrosis requiring mechanical ventilation for acuterespiratory failure. Chest 2001; 120:213–219104. Parambil JG, Myers JL, Ryu JH. Histopathologicfeatures and outcome of patients with acuteexacerbation of idiopathic pulmonary fibrosisundergoing surgical lung biopsy. Chest 2005;128:3310–3315105. Saydain G, Islam A, Afessa B, et al. Outcomeof patients with idiopathic pulmonary fibrosisadmitted to the intensive care unit. Am J RespirCrit Care Med 2002; 166:839–842106. Mallick S. Outcome of patients with idiopathicpulmonary fibrosis (IPF) ventilated in intensivecare unit. Respir Med 2008; 102:1355–1359107. Hubbard R, Venn A, Lewis S, et al. Lung cancerand cryptogenic fibrosing alveolitis: a population-basedcohort study. Am J Respir Crit CareMed 2000; 161:5–8108. Daniels CE, Jett JR. Does interstitial lung diseasepredispose to lung cancer? Curr Opin Pulm Med2005; 11:431–437109. Aubry MC, Myers JL, Douglas WW, et al. Primarypulmonary carcinoma in patients with idiopathicpulmonary fibrosis. Mayo Clin Proc 2002; 77:763–770<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 673ACC-PULM-09-0302-038.indd 6737/10/09 10:25:09 PM


110. Kawasaki H, Nagai K, Yoshida J, et al. Postoperativemorbidity, mortality, and survival in lungcancer associated with idiopathic pulmonaryfibrosis. J Surg Oncol 2002; 81:33–37111. Kumar P, Goldstraw P, Yamada K, et al. <strong>Pulmonary</strong>fibrosis and lung cancer: risk and benefitanalysis of pulmonary resection. J Thorac CardiovascSurg 2003; 125:1321–1327112. Swigris JJ, Brown KK. Acute interstitial pneumoniaand acute exacerbations of idiopathic pulmonaryfibrosis. Semin Respir Crit Care Med 2006;27:659–667113. Kim DS, Park JH, Park BK, et al. Acute exacerbationof idiopathic pulmonary fibrosis: frequencyand clinical features. Eur Respir J 2006; 27:143–50114. Suh GY, Kang EH, Chung MP, et al. Early interventioncan improve clinical outcome of acuteinterstitial pneumonia. Chest 2006; 129:753–761115. Yokoyama A, Kohno N, Hamada H, et al. CirculatingKL-6 predicts the outcome of rapidlyprogressive idiopathic pulmonary fibrosis. Am JRespir Crit Care Med 1998; 158:1680–1684116. Ohnishi H, Yokoyama A, Kondo K, et al. Comparativestudy of KL-6, surfactant protein-A, surfactantprotein-D, and monocyte chemoattractantprotein-1 as serum markers for interstitial lungdiseases. Am J Respir Crit Care Med 2002;165:378–381117. Takahashi H, Shiratori M, Kanai A, et al. Monitoringmarkers of disease activity for interstitiallung diseases with serum surfactant proteins Aand D. Respirology 2006; 11 Suppl:S51–S54118. Takahashi H, Fujishima T, Koba H, et al. Serumsurfactant proteins A and D as prognostic factorsin idiopathic pulmonary fibrosis and their relationshipto disease extent. Am J Respir Crit CareMed 2000; 162:1109–1114119. Kocheril SV, Appleton BE, Somers EC, et al. Comparisonof disease progression and mortality ofconnective tissue disease-related interstitial lungdisease and idiopathic interstitial pneumonia.Arthritis Rheum 2005; 53:549–557120. Lynch JP III, Orens J, Kazerooni EA. Collagenvascular diseases. In: Sperber M, ed. Diffuse lungdiseases: a comprehensive clinical-radiologicaloverview. London: Springer-Verlag, 1999; 325–355121. Elliot TL, Lynch DA, Newell JD Jr., et al. Highresolutioncomputed tomography features ofnonspecific interstitial pneumonia and usualinterstitial pneumonia. J Comput Assist Tomogr2005; 29:339–345122. Franquet T, Gimenez A, Alegret X, Rodriguez-Arias JM. Mediastinal lymphadenopathy in cryptogenicfibrosing alveolitis: the effect of steroidtherapy on the prevalence of nodal enlargement.Clin Radiol 1998; 53:435–438123. Gay SE, Kazerooni EA, Toews GB, et al. Idiopathicpulmonary fibrosis: predicting response totherapy and survival. Am J Respir Crit Care Med1998; 157:1063–1072124. Mogulkoc N, Brutsche MH, Bishop PW, et al. <strong>Pulmonary</strong>function in idiopathic pulmonary fibrosisand referral for lung transplantation. Am J RespirCrit Care Med 2001; 164:103–108125. Hunninghake GW, Zimmerman MB, SchwartzDA, et al. Utility of a lung biopsy for the diagnosisof idiopathic pulmonary fibrosis. Am J RespirCrit Care Med 2001; 164:193–196126. Aziz ZA, Wells AU, Hansell DM, et al. HRCTdiagnosis of diffuse parenchymal lung disease:inter-observer variation. Thorax 2004; 59:506–511127. Sumikawa H, Johkoh T, Colby TV, et al. Computedtomography findings in pathological usualinterstitial pneumonia: relationship to survival.Am J Respir Crit Care Med 2008; 177:433–439128. Xaubet A, Agusti C, Luburich P, et al. <strong>Pulmonary</strong>function tests and CT scan in the management ofidiopathic pulmonary fibrosis. Am J Respir CritCare Med 1998; 158:431–436129. Doherty MJ, Pearson MG, O’Grady EA, et al.Cryptogenic fibrosing alveolitis with preservedlung volumes. Thorax 1997; 52:998–1002130. Wells AU, King AD, Rubens MB, et al. Lone cryptogenicfibrosing alveolitis: a functional-morphologiccorrelation based on extent of disease onthin-section computed tomography. Am J RespirCrit Care Med 1997; 155:1367–1375131. Cottin V, Nunes H, Brillet PY, et al. Combinedpulmonary fibrosis and emphysema: a distinctunderrecognised entity. Eur Respir J 2005; 26:586–593132. Souza CA, Muller NL, Lee KS, et al. Idiopathicinterstitial pneumonias: prevalence of mediastinallymph node enlargement in 206 patients. AJRAm J Roentgenol 2006; 186:995–999133. Attili AK, Kazerooni EA, Gross BH, et al. Thoraciclymph node enlargement in usual interstitialpneumonitis and nonspecific-interstitial674 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6747/10/09 10:25:09 PM


pneumonitis: prevalence, correlation with diseaseactivity and temporal evolution. J ThoracImaging 2006; 21:288–292134. Jung JI, Kim HH, Jung YJ, et al. Mediastinallymphadenopathy in pulmonary fibrosis: correlationwith disease severity. J Comput AssistTomogr 2000; 24:706–710135. Niimi H, Kang EY, Kwong JS, et al. CT of chronicinfiltrative lung disease: prevalence of mediastinallymphadenopathy. J Comput Assist Tomogr1996; 20:305–308136. Hartman TE, Primack SL, Swensen SJ, et al.Desquamative interstitial pneumonia: thin-sectionCT findings in 22 patients. Radiology 1993;187:787–790137. Hartman TE. The HRCT features of extrinsicallergic alveolitis. Semin Respir Crit Care Med2003; 24:419–426138. Lacasse Y, Selman M, Costabel U, et al. Clinicaldiagnosis of hypersensitivity pneumonitis. Am JRespir Crit Care Med 2003; 168:952–958139. MacDonald SL, Rubens MB, Hansell DM, et al.Nonspecific interstitial pneumonia and usualinterstitial pneumonia: comparative appearancesat and diagnostic accuracy of thin-section CT.Radiology 2001; 221:600–605140. Marchand E, Cordier JF. Idiopathic chronic eosinophilicpneumonia. Semin Respir Crit Care Med2006; 27:134–141141. Trapnell BC, Whitsett JA, Nakata K. <strong>Pulmonary</strong>alveolar proteinosis. N Engl J Med 2003;349:2527–2539142. Tazi A. Adult pulmonary Langerhans’ cell histiocytosis.Eur Respir J 2006; 27:1272–1285143. Lynch III JP. Computed tomographic scanningin sarcoidosis. Semin Respir Crit Care Med 2003;24:393–418144. Johnson SR. Lymphangioleiomyomatosis. EurRespir J 2006; 27:1056–1065145. Sahin H, Brown KK, Curran-Everett D, et al.Chronic hypersensitivity pneumonitis: CT featurescomparison with pathologic evidence offibrosis and survival. Radiology 2007; 244:591–598146. Zisman DA, Kazerooni E, Flaherty K, et al. Idiopathicpulmonary fibrosis: role of high-resolutionthin-section computed tomographic scanning. In:Lynch JP III, ed. Idiopathic pulmonary fibrosis.New York, NY: Marcel Dekker Press, 2004; 237–252147. Nagai S, Kitaichi M, Itoh H, et al. Idiopathic nonspecificinterstitial pneumonia/fibrosis: comparisonwith idiopathic pulmonary fibrosis andBOOP. Eur Respir J 1998; 12:1010–1019148. Daniil ZD, Gilchrist FC, Nicholson AG, et al. Ahistologic pattern of nonspecific interstitial pneumoniais associated with a better prognosis thanusual interstitial pneumonia in patients withcryptogenic fibrosing alveolitis. Am J Respir CritCare Med 1999; 160:899–905149. Akira M, Sakatani M, Ueda E. Idiopathic pulmonaryfibrosis: progression of honeycombing atthin-section CT. Radiology 1993; 189:687–691150. Wells AU, Hansell DM, Rubens MB, et al. Functionalimpairment in lone cryptogenic fibrosingalveolitis and fibrosing alveolitis associated withsystemic sclerosis: a comparison. Am J RespirCrit Care Med 1997; 155:1657–1664151. Wells AU, Rubens MB, du Bois RM, et al. SerialCT in fibrosing alveolitis: prognostic significanceof the initial pattern. AJR Am J Roentgenol 1993;161:1159–1165152. Wells AU, Cullinan P, Hansell DM, et al. Fibrosingalveolitis associated with systemic sclerosishas a better prognosis than lone cryptogenicfibrosing alveolitis. Am J Respir Crit Care Med1994; 149:1583–1590153. Wells AU, Hansell DM, Rubens MB, et al. Thepredictive value of appearances on thin-sectioncomputed tomography in fibrosing alveolitis.Am Rev Respir Dis 1993; 148:1076–1082154. Hartman TE, Primack SL, Kang EY, et al. Diseaseprogression in usual interstitial pneumonia comparedwith desquamative interstitial pneumonia:assessment with serial CT. Chest 1996; 110:378–382155. Flaherty KR, Mumford JA, Murray S, et al.Prognostic implications of physiologic andradiographic changes in idiopathic interstitialpneumonia. Am J Respir Crit Care Med 2003;168:543–548156. Misumi S, Lynch DA. Idiopathic pulmonaryfibrosis/usual interstitial pneumonia: imagingdiagnosis, spectrum of abnormalities, and temporalprogression. Proc Am Thorac Soc 2006;3:307–314157. Coxson HO, Hogg JC, Mayo JR, et al. Quantificationof idiopathic pulmonary fibrosis using computedtomography and histology. Am J RespirCrit Care Med 1997; 155:1649–1656<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 675ACC-PULM-09-0302-038.indd 6757/10/09 10:25:09 PM


158. Lynch JP 3rd, Saggar R, Weigt SS, et al. Overviewof lung transplantation and criteria for selectionof candidates. Semin Respir Crit Care Med 2006;27:441–469159. Martinez FJ, Flaherty K. <strong>Pulmonary</strong> function testingin idiopathic interstitial pneumonias. ProcAm Thorac Soc 2006; 3:315–321160. Lama VN, Flaherty KR, Toews GB, et al. Prognosticvalue of desaturation during a 6-minute walktest in idiopathic interstitial pneumonia. Am JRespir Crit Care Med 2003; 168:1084–1090161. Fell CD, Liu LX, Motika C, et al. The prognosticvalue of cardiopulmonary exercise testing inidiopathic pulmonary fibrosis. Am J Respir CritCare Med 2009; 179:402–407162. Lama VN, Martinez FJ. Resting and exercisephysiology in interstitial lung diseases [review].Clin Chest Med 2004; 25:435–453, v163. Nicholson AG, Fulford LG, Colby TV, et al. Therelationship between individual histologic featuresand disease progression in idiopathic pulmonaryfibrosis. Am J Respir Crit Care Med 2002;166:173–177164. King TE Jr., Schwarz MI, Brown K, et al. Idiopathicpulmonary fibrosis: relationship betweenhistopathologic features and mortality. Am JRespir Crit Care Med 2001; 164:1025–1032165. Wells AU, Desai SR, Rubens MB, et al. Idiopathicpulmonary fibrosis: a composite physiologicindex derived from disease extent observed bycomputed tomography. Am J Respir Crit CareMed 2003; 167:962–969166. Mogulkoc N, Brutsche MH, Bishop PW, et al.<strong>Pulmonary</strong> (99m)Tc-DTPA aerosol clearance andsurvival in usual interstitial pneumonia (UIP).Thorax 2001; 56:916–923167. Timmer SJ, Karamzadeh AM, Yung GL, et al.Predicting survival of lung transplantation candidateswith idiopathic interstitial pneumonia:does Pao 2predict survival? Chest 2002; 122:779–784168. Lederer DJ, Arcasoy SM, Wilt JS, et al. Six-minutewalkdistance predicts waiting list survival inidiopathic pulmonary fibrosis. Am J Respir CritCare Med 2006; 174:659–664169. Kawut SM, O’Shea MK, Bartels MN, et al. Exercisetesting determines survival in patients withdiffuse parenchymal lung disease evaluated forlung transplantation. Respir Med 2005; 99:1431–1439170. Eaton T, Young P, Milne D, et al. Six-minute walk,maximal exercise tests: reproducibility in fibroticinterstitial pneumonia. Am J Respir Crit CareMed 2005; 171:1150–1157171. Miki K, Maekura R, Hiraga T, et al. Impairmentsand prognostic factors for survival in patientswith idiopathic pulmonary fibrosis. Respir Med2003; 97:482–490172. Latsi PI, du Bois RM, Nicholson AG, et al. Fibroticidiopathic interstitial pneumonia: the prognosticvalue of longitudinal functional trends. Am JRespir Crit Care Med 2003; 168:531–537173. Hallstrand TS, Boitano LJ, Johnson WC, et al. Thetimed walk test as a measure of severity and survivalin idiopathic pulmonary fibrosis. Eur RespirJ 2005; 25:96–103174. Caminati A, Bianchi A, Cassandro R, et al. Walkingdistance on 6-MWT is a prognostic factor inidiopathic pulmonary fibrosis. Respir Med 2009;103:117–123175. Flaherty KR, Andrei AC, Murray S, et al. Idiopathicpulmonary fibrosis: prognostic value ofchanges in physiology and six-minute-walk test.Am J Respir Crit Care Med 2006; 174:803–809176. Dal Corso S, Duarte SR, Neder JA, et al. A step testto assess exercise-related oxygen desaturation ininterstitial lung disease. Eur Respir J 2007; 29:330–336177. Watters LC, King TE, Schwarz MI, et al. A clinical,radiographic, and physiologic scoring systemfor the longitudinal assessment of patients withidiopathic pulmonary fibrosis. Am Rev RespirDis 1986; 133:97–103178. Hanson D, Winterbauer RH, Kirtland SH, et al.Changes in pulmonary function test resultsafter 1 year of therapy as predictors of survivalin patients with idiopathic pulmonary fibrosis.Chest 1995; 108:305–310179. Flaherty KR, Toews GB, Lynch JP 3rd, et al. Steroidsin idiopathic pulmonary fibrosis: a prospectiveassessment of adverse reactions, response to therapy,and survival. Am J Med 2001; 110:278–282180. Schwartz DA, Helmers RA, Galvin JR, et al.Determinants of survival in idiopathic pulmonaryfibrosis. Am J Respir Crit Care Med 1994;149:450–454181. Collard HR, King TE Jr., Bartelson BB, et al.Changes in clinical and physiologic variablespredict survival in idiopathic pulmonary fibrosis.Am J Respir Crit Care Med 2003; 168:538–542676 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6767/10/09 10:25:10 PM


182. Nathan SD, Noble PW, Tuder RM. Idiopathic pulmonaryfibrosis and pulmonary hypertension:connecting the dots. Am J Respir Crit Care Med2007; 175:875–880183. Patel NM, Lederer DJ, Borczuk AC, et al. <strong>Pulmonary</strong>hypertension in idiopathic pulmonaryfibrosis. Chest 2007; 132:998–1006184. Nathan SD, Shlobin OA, Ahmad S, et al. Serialdevelopment of pulmonary hypertension inpatients with idiopathic pulmonary fibrosis. Respiration2008; 76:288–294185. Colombat M, Mal H, Groussard O, et al. <strong>Pulmonary</strong>vascular lesions in end-stage idiopathicpulmonary fibrosis: histopathologic study onlung explant specimens and correlations withpulmonary hemodynamics. Hum Pathol 2007;38:60–65186. Hamada K, Nagai S, Tanaka S, et al. Significanceof pulmonary arterial pressure and diffusioncapa city of the lung as prognosticator in patientswith idiopathic pulmonary fibrosis. Chest 2007;131:650–656187. Lettieri CJ, Nathan SD, Barnett SD, et al. Prevalenceand outcomes of pulmonary arterial hypertensionin advanced idiopathic pulmonaryfibrosis. Chest 2006; 129:746–752188. Leuchte HH, Neurohr C, Baumgartner R, et al.Brain natriuretic peptide and exercise capacity inlung fibrosis and pulmonary hypertension. Am JRespir Crit Care Med 2004; 170:360–365189. Voelkel NF, Douglas IS, Nicolls M. Angiogenesisin chronic lung disease. Chest 2007; 131:874–879190. Nadrous HF, Pellikka PA, Krowka MJ, et al. <strong>Pulmonary</strong>hypertension in patients with idiopathicpulmonary fibrosis. Chest 2005; 128:2393–2399191. Nathan SD, Shlobin OA, Ahmad S, et al. <strong>Pulmonary</strong>hypertension and pulmonary function testingin idiopathic pulmonary fibrosis. Chest 2007;131:657–663192. Ghofrani HA, Wiedemann R, Rose F, et al. Sildenafilfor treatment of lung fibrosis and pulmonaryhypertension: a randomised controlled trial.Lancet 2002; 360:895–900193. Zisman DA, Ross DJ, Belperio JA, et al. Predictionof pulmonary hypertension in idiopathic pulmonaryfibrosis. Respir Med 2007; 101:2153–2159194. Song JW, Song JK, Kim DS. Echocardiographyand brain natriuretic peptide as prognostic indicatorsin idiopathic pulmonary fibrosis. RespirMed 2009; 103:180–186195. Arcasoy SM, Christie JD, Ferrari VA, et al. Echocardiographicassessment of pulmonary hypertensionin patients with advanced lung disease.Am J Respir Crit Care Med 2003; 167:735–740196. Collard HR, Anstrom KJ, Schwarz MI, et al. Sildenafilimproves walk distance in idiopathic pulmonaryfibrosis. Chest 2007; 131:897–899197. Dandel M, Lehmkuhl HB, Mulahasanovic S, et al.Survival of patients with idiopathic pulmonaryarterial hypertension after listing for transplantation:impact of iloprost and bosentan treatment.J Heart Lung Transplant 2007; 26:898–906198. King Jr TE, Behr J, Brown KK, et al. A randomizedplacebo-controlled trial of bosentan in patientswith idiopathic pulmonary fibrosis. Am J RespirCrit Care Med 2008; 177:75–81199. Daniels CE, Ryu JH. Treatment of idiopathic pulmonaryfibrosis. Semin Respir Crit Care Med2006; 27:668–676200. Agostini C, Gurrieri C. Chemokine/cytokinecocktail in idiopathic pulmonary fibrosis. ProcAm Thorac Soc 2006; 3:357–363201. Thannickal V, Flaherty K, Hyzy R, et al. Emergingdrugs for idiopathic pulmonary fibrosis. ExpertOpin Emerging Drugs 2005; 10:707–727202. Winterbauer RH, Hammar SP, Hallman KO,et al. Diffuse interstitial pneumonitis: clinicopathologiccorrelations in 20 patients treated with prednisone/azathioprine.Am J Med 1978; 65:661–672203. Travis WD, Matsui K, Moss J, Ferrans VJ. Idiopathicnonspecific interstitial pneumonia: prognosticsignificance of cellular and fibrosingpatterns: survival comparison with usual interstitialpneumonia and desquamative interstitialpneumonia. Am J Surg Pathol 2000; 24:19–33204. Nicholson AG, Colby TV, du Bois RM, et al. Theprognostic significance of the histologic patternof interstitial pneumonia in patients presentingwith the clinical entity of cryptogenic fibrosingalveolitis. Am J Respir Crit Care Med 2000;162:2213–2217205. Monaghan H, Wells AU, Colby TV, et al. Prognosticimplications of histologic patterns in multiplesurgical lung biopsies from patients withidiopathic interstitial pneumonias. Chest 2004;125:522–526206. Thomeer M, Demedts M, Behr J, et al. Multidisciplinaryinterobserver agreement in the diagnosisof idiopathic pulmonary fibrosis. Eur Respir J2008; 31:585–591<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 677ACC-PULM-09-0302-038.indd 6777/10/09 10:25:10 PM


207. Nicholson AG, Addis BJ, Bharucha H, et al. Interobservervariation between pathologists in diffuseparenchymal lung disease. Thorax 2004;59:500–505208. Portnoy J, Veraldi KL, Schwarz MI, et al. Respiratorybronchiolitis-interstitial lung disease: longtermoutcome. Chest 2007; 131:664–671209. Nicholson AG, Wells AU. Nonspecific interstitialpneumonia: nobody said it’s perfect. Am J RespirCrit Care Med 2001; 164:1553–1554210. Flaherty KR, Martinez FJ. Nonspecific interstitialpneumonia. Semin Respir Crit Care Med 2006;27:652–658211. Visscher DW, Myers JL. Histologic spectrum ofidiopathic interstitial pneumonias. Proc Am ThoracSoc 2006; 3:322–329212. Caminati A, Harari S. Smoking-related interstitialpneumonias and pulmonary Langerhans cell histiocytosis.Proc Am Thorac Soc 2006; 3:299–306213. Vassallo R, Jensen EA, Colby TV, et al. The overlapbetween respiratory bronchiolitis and desquamativeinterstitial pneumonia in pulmonaryLangerhans cell histiocytosis: high-resolutionCT, histologic, and functional correlations. Chest2003; 124:1199–1205214. Fraig M, Shreesha U, Savici D, et al. Respiratorybronchiolitis: a clinicopathologic study in currentsmokers, ex-smokers, and never-smokers. Am JSurg Pathol 2002; 26:647–653215. Moon J, du Bois RM, Colby TV, et al. Clinicalsignificance of respiratory bronchiolitis onopen lung biopsy and its relationship to smokingrelated interstitial lung disease. Thorax 1999;54:1009–1114216. Myers JL, Veal CF Jr., Shin MS, et al. Respiratorybronchiolitis causing interstitial lung disease:a clinicopathologic study of six cases. Am RevRespir Dis 1987; 135:880–884217. Yousem SA, Colby TV, Gaensler EA. Respiratorybronchiolitis-associated interstitial lung diseaseand its relationship to desquamative interstitialpneumonia. Mayo Clin Proc 1989; 64:1373–1380218. Akira M, Hamada H, Sakatani M, et al. CT findingsduring phase of accelerated deterioration inpatients with idiopathic pulmonary fibrosis. AJRAm J Roentgenol 1997; 168:79–83219. Katzenstein AL, Fiorelli RF. Nonspecific interstitialpneumonia/fibrosis: histologic featuresand clinical significance. Am J Surg Pathol 1994;18:136–147220. Flaherty K. Nonspecific Interstitial Pneumonitis(NSIP). In: Lynch JP III, ed. Interstitial pulmonaryand bronchiolar disorders (Vol 227). New York:InformaUSA; 2008, 365–378221. Jegal Y, Kim DS, Shim TS, et al. Physiology is astronger predictor of survival than pathology infibrotic interstitial pneumonia. Am J Respir CritCare Med 2005; 171:639–644222. Kim TS, Lee KS, Chung MP, et al. Nonspecificinterstitial pneumonia with fibrosis: high-resolutionCT and pathologic findings. AJR Am J Roentgenol1998; 171:1645–1650223. Veeraraghavan S, Latsi PI, Wells AU, et al. BALfindings in idiopathic nonspecific interstitialpneumonia and usual interstitial pneumonia.Eur Respir J 2003; 22:239–244224. Turner-Warwick M, Haslam PL. The value ofserial bronchoalveolar lavages in assessing theclinical progress of patients with cryptogenicfibrosing alveolitis. Am Rev Respir Dis 1987;135:26–34225. Watters LC, Schwarz MI, Cherniack RM, et al.Idiopathic pulmonary fibrosis: pretreatmentbronchoalveolar lavage cellular constituents andtheir relationships with lung histopathology andclinical response to therapy. Am Rev Respir Dis1987; 135:696–704226. Cottin V, Donsbeck AV, Revel D, et al. Nonspecificinterstitial pneumonia: individualization of aclinicopathologic entity in a series of 12 patients.Am J Respir Crit Care Med 1998; 158:1286–1293227. Park JS, Lee KS, Kim JS, et al. Nonspecific interstitialpneumonia with fibrosis: radiographic andCT findings in seven patients. Radiology 1995;195:645–648228. Tiitto L, Heiskanen U, Bloigu R, et al. Thoracoscopiclung biopsy is a safe procedure in diagnosingusual interstitial pneumonia. Chest 2005;128:2375–2380229. Lettieri CJ, Veerappan GR, Helman DL, et al. Outcomesand safety of surgical lung biopsy for interstitiallung disease. Chest 2005; 127:1600–1605230. Kondoh Y, Taniguchi H, Kitaichi M, et al. Acuteexacerbation of interstitial pneumonia followingsurgical lung biopsy. Respir Med 2006; 100:1753–1759231. Park IN, Kim DS, Shim TS, et al. Acute exacerbationof interstitial pneumonia other thanidiopathic pulmonary fibrosis. Chest 2007;132:214–220678 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6787/10/09 10:25:10 PM


232. The diagnosis, assessment and treatment of diffuseparenchymal lung disease in adults. Introduction.Thorax 1999; 54(Suppl 1):S1–S14233. Johnston ID, Gomm SA, Kalra S, et al. The managementof cryptogenic fibrosing alveolitis inthree regions of the United Kingdom. Eur RespirJ 1993; 6:891–893234. Chang AC, Yee J, Orringer MB, et al. Diagnosticthoracoscopic lung biopsy: an outpatient experience.Ann Thorac Surg 2002; 74:1942–1946, discussion6–7235. Flaherty KR, Colby TV, Travis WD, et al. Fibroblasticfoci in usual interstitial pneumonia: idiopathicversus collagen vascular disease. Am JRespir Crit Care Med 2003; 167:1410–1415236. Hanak V, Ryu JH, de Carvalho E, et al. Profusionof fibroblast foci in patients with idiopathic pulmonaryfibrosis does not predict outcome. RespirMed 2008; 102:852–856237. Welker L, Jorres RA, Costabel U, et al. Predictivevalue of BAL cell differentials in the diagnosisof interstitial lung diseases. Eur Respir J 2004;24:1000–1006238. Nagai S, Handa T, Ito Y, et al. Bronchoalveolarlavage in idiopathic interstitial lung diseases.Semin Respir Crit Care Med 2007; 28:496–503239. Wells AU, Hansell DM, Harrison NK, et al. Clearanceof inhaled 99mTc-DTPA predicts the clinicalcourse of fibrosing alveolitis. Eur Respir J 1993;6:797–802240. Singh S, Wells AU, du Bois R. Other imaging techniquesfor idiopathic interstitial pneumonias. In:Lynch JP III, ed. Idiopathic pulmonary fibrosis.New York, NY: Marcel Dekker, 2004; 237–252241. Thomeer MJ, Dehaes B, Mortelmans L, et al.Pertechnegas lung clearance in different formsof interstitial lung disease. Eur Respir J 2002; 19:31–36242. Lynch JP III, Madihara R, Fishbein MC, et al.Idiopathic pulmonary fibrosis. In: Lynch JP III,ed. Interstitial pulmonary and bronchiolar disorders.Vol. 227. New York: InformaUSA, 2008;333–364243. Selman M, King TE, Pardo A. Idiopathic pulmonaryfibrosis: prevailing and evolving hypothesesabout its pathogenesis and implications fortherapy. Ann Intern Med 2001; 134:136–151244. Maher TM, Wells AU, Laurent GJ. Idiopathic pulmonaryfibrosis: multiple causes and multiplemechanisms? Eur Respir J 2007; 30:835–839245. Selman M, Thannickal VJ, Pardo A, et al. Idiopathicpulmonary fibrosis: pathogenesis and therapeutic approaches. Drugs 2004; 64:405–430246. Uhal BD, Joshi I, Hughes WF, et al. Alveolar epithelialcell death adjacent to underlying myofibroblastsin advanced fibrotic human lung. Am JPhysiol 1998; 275:L1192–L1199247. Kuhn C, McDonald JA. The roles of the myofibroblastin idiopathic pulmonary fibrosis: ultrastructuraland immunohistochemical features ofsites of active extracellular matrix synthesis. AmJ Pathol 1991; 138:1257–1265248. Lama VN, Phan SH. The extrapulmonary originof fibroblasts: stem/progenitor cells and beyond.Proc Am Thorac Soc 2006; 3:373–376249. Keane MP, Belperio JA, Burdick MD, et al. ENA-78 is an important angiogenic factor in idiopathicpulmonary fibrosis. Am J Respir Crit Care Med2001; 164:2239–2242250. Keane MP, Arenberg DA, Lynch JP 3rd, et al. TheCXC chemokines, IL-8 and IP-10, regulate angiogenicactivity in idiopathic pulmonary fibrosis.J Immunol 1997; 159:1437–1443251. Keane MP, Belperio JA, Arenberg DA, et al.IFN-gamma-inducible protein-10 attenuates bleomycin-inducedpulmonary fibrosis via inhibitionof angiogenesis. J Immunol 1999; 163:5686–5692252. Williams TJ, Wilson JW. Challenges in pulmonaryfibrosis: 7–Novel therapies and lung transplantation.Thorax 2008; 63:277–284253. Lynch JP III, Wurfel M, Flaherty K, et al. Usualinterstitial pneumonia. Semin Respir Crit CareMed 2001; 22:357–387254. Nagai S, Kitaichi M, Hamada K, et al. Hospitalbasedhistorical cohort study of 234 histologicallyproven Japanese patients with IPF. SarcoidosisVasc Diffuse Lung Dis 1999; 16:209–214255. Tukiainen P, Taskineu E, Holsti P, et al. Prognosisof cryptogenic fibrosing alveolitis. Thorax 1983;38:349–355256. Douglas WW, Ryu JH, Swensen SJ, et al. Colchicineversus prednisone in the treatment of idiopathic pulmonary fibrosis: a randomized prospectivestudy. Members of the Lung Study Group.Am J Respir Crit Care Med 1998; 158:220–225257. Mapel DW, Samet JM, Coultas DB. Corticosteroidsand the treatment of idiopathic pulmonaryfibrosis: past, present, and future. Chest 1996;110:1058–1067<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 679ACC-PULM-09-0302-038.indd 6797/10/09 10:25:10 PM


258. Lynch JP 3rd, White E, Flaherty K. Corticosteroidsin idiopathic pulmonary fibrosis. Curr OpinPulm Med 2001; 7:298–308259. Collard HR, Ryu JH, Douglas WW, et al. Combinedcorticosteroid and cyclophosphamidetherapy does not alter survival in idiopathic pulmonaryfibrosis. Chest 2004; 125:2169–2174260. Zisman DA, Lynch JP 3rd, Toews GB, et al. Cyclophosphamidein the treatment of idiopathic pulmonaryfibrosis: a prospective study in patientswho failed to respond to corticosteroids. Chest2000; 117:1619–1626261. Johnson MA, Kwan S, Snell NJ, et al. Randomisedcontrolled trial comparing prednisolone alonewith cyclophosphamide and low dose prednisolonein combination in cryptogenic fibrosingalveolitis. Thorax 1989; 44:280–288262. Raghu G, Depaso WJ, Cain K, et al. Azathioprinecombined with prednisone in the treatment ofidiopathic pulmonary fibrosis: a prospective double-blind,randomized, placebo- controlled clinicaltrial. Am Rev Respir Dis 1991; 144:291–296263. Kolb M, Kirschner J, Riedel W, et al. Cyclophosphamidepulse therapy in idiopathic pulmonaryfibrosis. Eur Respir J 1998; 12:1409–1414264. Selman M, Carrillo G, Salas J, et al. Colchicine, D-penicillamine, and prednisone in the treatment ofidiopathic pulmonary fibrosis: a controlled clinicaltrial. Chest 1998; 114:507–512265. Lynch JP 3rd, McCune WJ. Immunosuppressiveand cytotoxic pharmacotherapy for pulmonarydisorders. Am J Respir Crit Care Med 1997;155:395–420266. Raghu G, Brown KK, Costabel U, et al. Treatmentof idiopathic pulmonary fibrosis with etanercept:an exploratory, placebo-controlled trial. Am JRespir Crit Care Med 2008; 178:948–955267. Bargagli E, Galeazzi M, Rottoli P. Infliximab treatmentin a patient with rheumatoid arthritis andpulmonary fibrosis. Eur Respir J 2004; 24:708268. Antoniou KM, Mamoulaki M, Malagari K, et al.Infliximab therapy in pulmonary fibrosis associatedwith collagen vascular disease. Clin ExpRheumatol 2007; 25:23–28269. Ostor AJ, Crisp AJ, Somerville MF, et al. Fatalexacerbation of rheumatoid arthritis associatedfibrosing alveolitis in patients given infliximab.BMJ 2004; 329:1266270. Ziesche R, Hofbauer E, Wittmann K, et al. Apreliminary study of long-term treatment withinterferon gamma-1b and low-dose prednisolonein patients with idiopathic pulmonary fibrosis.N Engl J Med 1999; 341:1264–1269271. Raghu G, Brown KK, Bradford WZ, et al. A placebo-controlledtrial of interferon gamma-1bin patients with idiopathic pulmonary fibrosis.N Engl J Med 2004; 350:125–133272. Thannickal VJ, Flaherty KR, Martinez FJ, et al.Idiopathic pulmonary fibrosis: emerging conceptson pharmacotherapy. Expert Opin Pharmacother2004; 5:1671–1686273. Behr J, Maier K, Degenkolb B, et al. Antioxidativeand clinical effects of high-dose N-acetylcysteinein fibrosing alveolitis: adjunctive therapy tomaintenance immunosuppression. Am J RespirCrit Care Med 1997; 156:1897–1901274. Demedts M, Behr J, Buhl R, et al. High-doseacetylcysteine in idiopathic pulmonary fibrosis.N Engl J Med 2005; 353:2229–2242275. Hunninghake GW. Antioxidant therapy for idiopathicpulmonary fibrosis. N Engl J Med 2005;353:2285–2287276. Oku H, Shimizu T, Kawabata T, et al. Antifibroticaction of pirfenidone and prednisolone:different effects on pulmonary cytokines andgrowth factors in bleomycin-induced murinepulmonary fibrosis. Eur J Pharmacol 2008; 590:400–408277. Raghu G, Johnson WC, Lockhart D, et al. Treatmentof idiopathic pulmonary fibrosis with a newantifibrotic agent, pirfenidone: results of a prospective,open-label Phase II study. Am J RespirCrit Care Med 1999; 159:1061–1069278. Nagai S, Hamada K, Shigematsu M, et al. Openlabelcompassionate use one year-treatment withpirfenidone to patients with chronic pulmonaryfibrosis. Intern Med 2002; 41:1118–1123279. Gahl WA, Brantly M, Troendle J, et al. Effect ofpirfenidone on the pulmonary fibrosis of Hermansky-Pudlaksyndrome. Mol Genet Metab2002; 76:234–242280. Azuma A, Nukiwa T, Tsuboi E, et al. Doubleblind,placebo-controlled trial of pirfenidone inpatients with idiopathic pulmonary fibrosis. AmJ Respir Crit Care Med 2005; 171:1040–1047281. Gross TJ, Hunninghake GW. Idiopathic pulmonaryfibrosis. N Engl J Med 2001; 345:517–525282. Park SH, Saleh D, Giaid A, et al. Increasedendothelin-1 in bleomycin-induced pulmonaryfibrosis and the effect of an endothelin receptor680 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6807/10/09 10:25:11 PM


antagonist. Am J Respir Crit Care Med 1997;156:600–608283. Galie N, Ghofrani HA, Torbicki A, et al. Sildenafilcitrate therapy for pulmonary arterial hypertension.N Engl J Med 2005; 353:2148–2157284. Madden BP, Sheth A, Wilde M, et al. Does sildenafilproduce a sustained benefit in patients withpulmonary hypertension associated with parenchymallung and cardiac disease? Vasc Pharmacol2007; 47:184–188285. Walter N, Collard HR, King TE Jr. Current perspectiveson the treatment of idiopathic pulmonaryfibrosis. Proc Am Thorac Soc 2006; 3:330–338286. Kubo H, Nakayama K, Yanai M, et al. Anticoagulanttherapy for idiopathic pulmonary fibrosis.Chest 2005; 128:1475–1482287. Noth I, Martinez FJ. Recent advances in idiopathicpulmonary fibrosis. Chest 2007; 132:637–650288. Nathan SD, Saggar R, Lynch JP III. Lung transplantationfor interstitial lung disorders. In:Lynch JP III, Ross D, eds. Lung and heart-lungtransplantation (Vol 227). New York: Taylor andFrancis, 2006; 165–204289. Meyer DM, Edwards LB, Torres F, et al. Impact ofrecipient age and procedure type on survival afterlung transplantation for pulmonary fibrosis. AnnThorac Surg 2005; 79:950–957, discussion 7–8290. Trulock EP, Edwards LB, Taylor DO, et al. Theregistry of the International Society for Heat andLung Transplantation: twenty-first official adultheart transplant report---2004. J Heart LungTransplant 2004; 23:804–815291. Meyers BF, Lynch JP, Trulock EP, et al. Single versusbilateral lung transplantation for idiopathicpulmonary fibrosis: a ten-year institutional experience.J Thorac Cardiovasc Surg 2000; 120:99–107292. Charman SC, Sharples LD, McNeil KD, et al.Assessment of survival benefit after lung transplantationby patient diagnosis. J Heart LungTransplant 2002; 21:226–232293. Trulock EP, Christie JD, Edwards LB, et al. Registryof the International Society for Heart andLung Transplantation: twenty-fourth officialadult lung and heart-lung transplantation report-2007. J Heart Lung Transplant 2007; 26:782–795294. Christie JD, Edwards LB, Aurora P, et al. Registryof the International Society for Heart and LungTransplantation: twenty-fifth official adult lungand heart/lung transplantation report–2008.zJ Heart Lung Transplant 2008; 27:957–969295. Whelan TP, Dunitz JM, Kelly RF, et al. Effect ofpreoperative pulmonary artery pressure on earlysurvival after lung transplantation for idiopathicpulmonary fibrosis. J Heart Lung Transplant2005; 24:1269–1274296. Harris-Eze AO, Sridhar G, Clemens RE, et al.Oxygen improves maximal exercise performancein interstitial lung disease. Am J Respir Crit CareMed 1994; 150:1616–1622297. Olschewski H, Ghofrani HA, Walmrath D,et al. Inhaled prostacyclin and iloprost in severepulmonary hypertension secondary to lung fibrosis.Am J Respir Crit Care Med 1999; 160:600–607298. Douglas WW, Tazelaar HD, Hartman TE, et al.Polymyositis-dermatomyositis-associated interstitiallung disease. Am J Respir Crit Care Med2001; 164:1182–1185299. Bouros D, Wells AU, Nicholson AG, et al. Histopathologicsubsets of fibrosing alveolitis inpatients with systemic sclerosis and their relationshipto outcome. Am J Respir Crit Care Med2002; 165:1581–1586300. Nakamura Y, Chida K, Suda T, et al. Nonspecificinterstitial pneumonia in collagen vascular diseases:comparison of the clinical characteristicsand prognostic significance with usual interstitialpneumonia. Sarcoidosis Vasc Diffuse LungDis 2003; 20:235–241301. Tansey D, Wells AU, Colby TV, et al. Variationsin histological patterns of interstitial pneumoniabetween connective tissue disorders and theirrelationship to prognosis. Histopathology 2004;44:585–596302. Lynch JP 3rd, Ma YL, Koss MN, et al. <strong>Pulmonary</strong>sarcoidosis. Semin Respir Crit Care Med 2007;28:53–74303. Lynch III JP, Kazerooni EA, Gay SE. <strong>Pulmonary</strong>sarcoidosis. Clin Chest Med 1997; 18:755–785304. Lynch III J, Baughman R, Sharma O. Extrapulmonarysarcoidosis. Semin Respir Infect 1998;13:229–254305. Newman LS, Rose CS, Maier LA. Sarcoidosis.N Engl J Med 1997; 336:1224–1234306. Statement on sarcoidosis. Joint Statement of theAmerican Thoracic Society (ATS), the EuropeanRespiratory Society (ERS), and the World Associationof Sarcoidosis and Other GranulomatousDisorders (WASOG) adopted by the ATS <strong>Board</strong> ofDirectors and by the ERS Executive Committee,<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 681ACC-PULM-09-0302-038.indd 6817/10/09 10:25:11 PM


February 1999 [review]. Am J Respir Crit CareMed 1999; 160:736–755307. Judson MA. Extrapulmonary sarcoidosis. SeminRespir Crit Care Med 2007; 28:83–101308. Judson MA. Hepatic, splenic, and gastrointestinalinvolvement with sarcoidosis. Semin RespirCrit Care Med 2002; 23:529–541309. Deng J, Baughman R, Lynch III JP. Cardiacinvolvement in sarcoidosis. Semin Respir CritCare Med 2002; 23:513–528310. Kidd D, Beynon HL. The neurological complicationsof systemic sarcoidosis. Sarcoidosis VascDiffuse Lung Dis 2003; 20:85–94311. Allen RK, Sellars RE, Sandstrom PA. A prospectivestudy of 32 patients with neurosarcoidosis.Sarcoidosis Vasc Diffuse Lung Dis 2003; 20:118–125312. Lower EE, Broderick JP, Brott TG, et al. Diagnosisand management of neurological sarcoidosis.Arch Intern Med 1997; 157:1864–1868313. Zisman DA, Shorr A, Lynch JP III. Sarcoidosisinvolving the musculoskeletal system. SeminResp Crit Care Med 2002; 23:555–570314. Neville E, Walker A, James DG. Prognostic factorspredicting the outcome of sarcoidosis: an analysisof 818 patients. Q J Med 1983; 208:525–533315. Mana J, Gomez-Vaquero C, Montero A, et al.Lofgren’s syndrome revisited: a study of 186patients. Am J Med 1999; 107:240–245316. Judson MA, Baughman RP, Thompson BW, et al.Two year prognosis of sarcoidosis: the ACCESSexperience. Sarcoidosis Vasc Diffuse Lung Dis2003; 20:204–211317. Grunewald J, Eklund A. Sex-specific manifestationsof Lofgren’s syndrome. Am J Respir CritCare Med 2007; 175:40–44318. Pietinalho A, Ohmichi M, Lofroos AB, et al. Theprognosis of pulmonary sarcoidosis in Finlandand Hokkaido, Japan: a comparative five-yearstudy of biopsy-proven cases. Sarcoidosis VascDiffuse Lung Dis 2000; 17:158–166319. Chappell AG, Cheung WY, Hutchings HA. Sarcoidosis:a long-term follow up study. SarcoidosisVasc Diffuse Lung Dis 2000; 17:167–173320. Baughman RP, Teirstein AS, Judson MA, et al.Clinical characteristics of patients in a case controlstudy of sarcoidosis. Am J Respir Crit CareMed 2001; 164:1885–1889321. Johns CJ, Schonfeld SA, Scott PP, et al. Longitudinalstudy of chronic sarcoidosis with low-dosemaintenance corticosteroid therapy: outcome andcomplications. Ann N Y Acad Sci 1986; 465:702–712322. Johns CJ, Michele TM. The clinical managementof sarcoidosis: a 50-year experience at the JohnsHopkins Hospital. <strong>Medicine</strong> (Baltimore) 1999;78:65–111323. Gottlieb JE, Israel HL, Steiner RM, et al. Outcomein sarcoidosis: the relationship of relapse to corticosteroidtherapy. Chest 1997; 111:623–631324. Reich JM. Mortality of intrathoracic sarcoidosis inreferral vs population-based settings: influence ofstage, ethnicity, and corticosteroid therapy. Chest2002; 121:32–39325. Huhti E, Poukkula A, Lilja M. Prognosis for sarcoidosisin a defined geographical area. Br J DisChest 1987; 81:381–390326. Rybicki BA, Iannuzzi MC, Frederick MM, et al.Familial aggregation of sarcoidosis: a case-controletiologic study of sarcoidosis (ACCESS). Am JRespir Crit Care Med 2001; 164:2085–2091327. Rybicki BA, Major M, Popovich J Jr., et al. Racialdifferences in sarcoidosis incidence: a 5-yearstudy in a health maintenance organization. Am JEpidemiol 1997; 145:234–241328. Rybicki BA, Iannuzzi MC. Epidemiology ofsarcoidosis: recent advances and future pros -pects. Semin Respir Crit Care Med 2007; 28:22–35329. Hosoda Y, Yamaguchi M, Hiraga Y. Global epidemiologyof sarcoidosis: what story do prevalenceand incidence tell us? Clin Chest Med 1997;18:681–694330. Hillerdal G, Nou E, Osterman K, et al. Sarcoidosis:epidemiology and prognosis: a 15-year Europeanstudy. Am Rev Respir Dis 1984; 130:29–32331. Mana J, Badrinas F, Morera J, et al. Sarcoidosis inSpain. Sarcoidosis 1992; 9:118–122332. Rybicki BA, Maliarik MJ, Major M, et al. Epidemiology,demographics, and genetics of sarcoidosis.Semin Respir Infect 1998; 13:166–173333. McGrath DS, Daniil Z, Foley P, et al. Epidemiologyof familial sarcoidosis in the UK. Thorax2000; 55:751–754334. Schurmann M, Lympany PA, Reichel P, et al.Familial sarcoidosis is linked to the major histocompatibilitycomplex region. Am J Respir CritCare Med 2000; 162:861–864335. Schurmann M, Reichel P, Muller-Myhsok B,et al. Results from a genome-wide search forpredisposing genes in sarcoidosis. Am J RespirCrit Care Med 2001; 164:840–846682 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6827/10/09 10:25:11 PM


336. Petrek M, Drabek J, Kolek V, et al. CC chemokinereceptor gene polymorphisms in Czech patientswith pulmonary sarcoidosis. Am J Respir CritCare Med 2000; 162:1000–1003337. Sato H, Grutters JC, Pantelidis P, et al. HLA-DQB1*0201: a marker for good prognosis in Britishand Dutch patients with sarcoidosis. Am JRespir Cell Mol Biol 2002; 27:406–412338. Spagnolo P, Renzoni EA, Wells AU, et al. C-C chemokinereceptor 2 and sarcoidosis: associationwith Lofgren’s syndrome. Am J Respir Crit CareMed 2003; 168:1162–1166339. Rosen Y. Pathology of sarcoidosis. Semin RespirCrit Care Med 2007; 28:36–52340. Reich JM, Brouns MC, O’Connor EA, et al. Mediastinoscopyin patients with presumptive stage Isarcoidosis: a risk/benefit, cost/benefit analysis.Chest 1998; 113:147–153341. Fritscher-Ravens A, Sriram PV, Topalidis T,et al. Diagnosing sarcoidosis using endosonography-guidedfine-needle aspiration. Chest 2000;118:928–935342. Smojver-Jezek S, Peros-Golubicic T, Tekavec-Trkanjec J, et al. Transbronchial fine needle aspirationcytology in the diagnosis of mediastinal/hilarsarcoidosis. Cytopathology 2007; 18:3–7343. Zwischenberger JB, Savage C, Alpard SK, et al.Mediastinal transthoracic needle and core lymphnode biopsy: should it replace mediastinoscopy?Chest 2002; 121:1165–1170344. Tambouret R, Geisinger KR, Powers CN, et al.The clinical application and cost analysis offine-needle aspiration biopsy in the diagnosisof mass lesions in sarcoidosis. Chest 2000; 117:1004–1011345. Lynch JP III. <strong>Pulmonary</strong> sarcoidosis: current concepts& controversies. Compr Ther 1997; 23:197–210346. Hours S, Nunes H, Kambouchner M, et al. <strong>Pulmonary</strong>cavitary sarcoidosis: clinico-radiologic characteristicsand natural history of a rare form ofsarcoidosis. <strong>Medicine</strong> (Baltimore) 2008; 87:142–151347. Gibson GJ, Prescott RJ, Muers MF, et al. British ThoracicSociety Sarcoidosis study: effects of long termcorticosteroid treatment. Thorax 1996; 51:238–247348. Hunninghake GW, Gilbert S, Pueringer R, et al.Outcome of the treatment for sarcoidosis. Am JRespir Crit Care Med 1994; 149:893–898349. Romer FK. Presentation of sarcoidosis and outcomeof pulmonary changes. Dan Med Bull 1982;29:27–32350. Mana J, Teirstein AS, Mendelson DS, et al. Excessivethoracic computed tomographic scanning insarcoidosis. Thorax 1995; 50:1264–1266351. Nunes H, Brillet PY, Valeyre D, et al. Imaging insarcoidosis. Semin Respir Crit Care Med 2007;28:102–120352. Drent M, De Vries J, Lenters M, et al. Sarcoidosis:assessment of disease severity using HRCT. EurRadiol 2003; 13:2462–2471353. Nishimura K, Itoh H, Kitaichi M, et al. CT andpathological correlation of pulmonary sarcoidosis.Semin Ultrasound CT MR 1995; 16:361–370354. Terasaki H, Fujimoto K, Muller NL, et al. <strong>Pulmonary</strong>sarcoidosis: comparison of findings ofinspiratory and expiratory high-resolution CTand pulmonary function tests between smokersand nonsmokers. AJR Am J Roentgenol 2005;185:333–338355. Naccache JM, Lavole A, Nunes H, et al. Highresolutioncomputed tomographic imaging ofairways in sarcoidosis patients with airflow obstruction.J Comput Assist Tomogr 2008; 32:905–912356. Akira M, Kozuka T, Inoue Y, et al. Long-termfollow-up CT scan evaluation in patients withpulmonary sarcoidosis. Chest 2005; 127:185–191357. Lavergne F, Clerici C, Sadoun D, et al. Airwayobstruction in bronchial sarcoidosis: outcomewith treatment. Chest 1999; 116:1194–1199358. Handa T, Nagai S, Fushimi Y, et al. Clinical andradiographic indices associated with airflow limitationin patients with sarcoidosis. Chest 2006;130:1851–1856359. Subramanian I, Flaherty K, Martinez F. <strong>Pulmonary</strong>function testing in sarcoidosis. In: BaughmanRP, ed. Sarcoidosis (vol 210). New York, NY:Taylor and Francis Group, 2006; 415–433360. Shorr AF, Torrington KG, Hnatiuk OW. Endobronchialinvolvement and airway hyperreactivityin patients with sarcoidosis. Chest 2001;120:881–886361. Mihailovic-Vucinic V, Zugic V, Videnovic- IvanovJ. New observations on pulmonary functionchanges in sarcoidosis. Curr Opin Pulm Med2003; 9:436–441362. Miller A, Brown LK, Sloane MF, et al. Cardiorespiratoryresponses to incremental exercise in sarcoidosispatients with normal spirometry. Chest 1995;107:323–329363. Medinger AE, Khouri S, Rohatgi PK. Sarcoidosis:the value of exercise testing. Chest 2001; 120:93–101<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 683ACC-PULM-09-0302-038.indd 6837/10/09 10:25:11 PM


364. Mediwake R, Wells A, Desai D. Radiologicalimaging in sarcoidosis. In: Baughman RP, ed.Sarcoidosis (Vol 210). New York: Taylor andFrancis Group, 2006; 365–398365. Mana J, Vankroonenburgh M. Nuclear imagingtechniques in sarcoidosis. Eur Respir Monograph2005; 10:284–300366. Tanabe Y, Ohuchi Y, Ogawa T. Muscular andmyocardial involvement in sarcoidosis: the usefulnessof Ga-67 imaging. Clin Nucl Med 2002;27:749–752367. Aberg C, Ponzo F, Raphael B, et al. FDG positronemission tomography of bone involvement insarcoidosis. AJR Am J Roentgenol 2004; 182:975–977368. Takeda N, Yokoyama I, Hiroi Y, et al. Positronemission tomography predicted recovery ofcomplete A-V nodal dysfunction in a patientwith cardiac sarcoidosis. Circulation 2002; 105:1144–1145369. Dubey N, Miletich RS, Wasay M, et al. Role of fluorodeoxyglucosepositron emission tomographyin the diagnosis of neurosarcoidosis. J Neurol Sci2002; 205:77–81370. Costabel U, Guzman J, Albera C, et al. Bronchoalveolarlavage in sarcoidosis. In: Baughman RP,ed. Sarcoidosis (Vol 210). New York: Taylor andFrancis Group, 2006; 399–414371. Zissel G, Prasse A, Muller-Quernheim J. Sarcoidosis:immunopathogenetic concepts. SeminRespir Crit Care Med 2007; 28:3–14372. Semenzato G, Bortoli M, Brunetti E, et al. Immunologyand pathophysiology. In: In: BaughmanRP, ed. Sarcoidosis (Vol 210). New York: Taylorand Francis Group, 2006; 49–63373. Ziegenhagen MW, Rothe ME, Schlaak M, et al.Bronchoalveolar and serological parametersreflecting the severity of sarcoidosis. Eur RespirJ 2003; 21:407–413374. Baughman RP, Sharma OP, Lynch JP III. Sarcoidosis.is therapy effective? Semin Respir Infect1998; 13:255–273375. Sharma OP. <strong>Pulmonary</strong> sarcoidosis and corticosteroids.Am Rev Respir Dis 1993; 147:1598–1600376. Pietinalho A, Tukiainen P, Haahtela T, et al. Oralprednisolone followed by inhaled budesonide innewly diagnosed pulmonary sarcoidosis: a double-blind,placebo-controlled multicenter study:Finnish <strong>Pulmonary</strong> Sarcoidosis Study Group.Chest 1999; 116:424–431377. Pietinalho A, Tukiainen P, Haahtela T, et al. Earlytreatment of stage II sarcoidosis improves 5-yearpulmonary function. Chest 2002; 121:24–31378. du Bois RM, Greenhalgh PM, Southcott AM,et al. Randomized trial of inhaled fluticasone propionatein chronic stable pulmonary sarcoidosis:a pilot study. Eur Respir J 1999; 13:1345–1350379. Pacheco Y, Marechal C, Marechal F, et al. Azathioprinetreatment of chronic pulmonary sarcoidosis.Sarcoidosis 1985; 2:107–113380. Sharma O, Hughs DTD, James DG, et al. Immunosuppressivetherapy with azathioprine in sarcoidosis.In: Levnisky L, Macholoa F, eds. FifthInternational Conference on Sarcoidosis andOther Granulomatous Disorders, Prague, CzechRepublic, 1971; 635–637381. Agbogu BN, Stern BJ, Sewell C, et al. Therapeuticconsiderations in patients with refractory neurosarcoidosis.Arch Neurol 1995; 52:875–879382. Diab SM, Karnik AM, Ouda BA, et al. Sarcoidosisin Arabs: the clinical profile of 20 patients andreview of the literature. Sarcoidosis 1991; 8:56–62383. Lewis SJ, Ainslie GM, Bateman ED. Efficacy ofazathioprine as second-line treatment in pulmonarysarcoidosis. Sarcoidosis Vasc Diffuse LungDis 1999; 16:87–92384. Muller-Quernheim J, Kienast K, Held M, et al.Treatment of chronic sarcoidosis with an azathioprine/prednisoloneregimen. Eur Respir J 1999;14:1117–1122385. Baughman RP, Winget DB, Lower EE. Methotrexateis steroid sparing in acute sarcoidosis: resultsof a double blind, randomized trial. SarcoidosisVasc Diffuse Lung Dis 2000; 17:60–66386. Baughman RP, Lower EE. A clinical approach tothe use of methotrexate for sarcoidosis. Thorax1999; 54:742–746387. Lower EE, Baughman RP. Prolonged use of methotrexatefor sarcoidosis. Arch Intern Med 1995;155:846–851388. Baughman RP, Lower EE. Steroid-sparing alternativetreatments for sarcoidosis. Clin Chest Med1997; 18:853–864389. Zisman DA, McCune WJ, Tino G, et al. Druginducedpneumonitis: the role of methotrexate. SarcoidosisVasc Diffuse Lung Dis 2001; 18:243–252390. Baughman RP, Koehler A, Bejarano PA, et al. Roleof liver function tests in detecting methotrexateinducedliver damage in sarcoidosis. Arch InternMed 2003; 163:615–620684 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6847/10/09 10:25:12 PM


391. Baughman RP, Lower EE. Leflunomide for chronicsarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis2004; 21:43–48392. Moudgil A, Przygodzki RM, Kher KK. Successfulsteroid-sparing treatment of renal limitedsarcoidosis with mycophenolate mofetil. PediatrNephrol 2006; 21:281–285393. Choudhary A, Harding SP, Bucknall RC, etal. Mycophenolate mofetil as an immunosuppressiveagent in refractory inflammatory eyedisease. J Ocul Pharmacol Ther 2006; 22:168–175394. Stern BJ, Corbett J. Neuro-ophthalmologic manifestationsof sarcoidosis. Curr Treat Options Neurol2007; 9:63–71395. Martinet Y, Pinkston P, Saltini C, et al. Evaluationof the in vitro and in vivo effects of cyclosporineon the lung T-lymphocyte alveolitis of active pulmonarysarcoidosis. Am Rev Respir Dis 1988;138:1242–1248396. Stern BJ, Schonfeld SA, Sewell C, et al. The treatmentof neurosarcoidosis with cyclosporine. ArchNeurol 1992; 49:1065–1072397. Wyser CP, van Schalkwyk EM, Alheit B, et al.Treatment of progressive pulmonary sarcoidosiswith cyclosporin A: a randomized controlledtrial. Am J Respir Crit Care Med 1997; 156:1371–1376398. Akpek EK, Ilhan-Sarac O, Green WR. Topicalcyclosporin in the treatment of chronic sarcoidosisof the conjunctiva. Arch Ophthalmol 2003;121:1333–1335399. Kataria YP. Chlorambucil in sarcoidosis. Chest1980; 78:36–43400. Israel HL, McComb BL. Chlorambucil treatmentof sarcoidosis. Sarcoidosis 1991; 8:35–41401. Israel HL, Fouts DW, Beggs RA. A controlled trialof prednisone treatment of sarcoidosis. Am RevRespir Dis 1973; 107:609–614402. Demeter SL. Myocardial sarcoidosis unresponsiveto steroids: treatment with cyclophosphamide.Chest 1988; 94:202–203403. Doty JD, Mazur JE, Judson MA. Treatment ofcorticosteroid-resistant neurosarcoidosis with ashort-course cyclophosphamide regimen. Chest2003; 124:2023–2026404. Zic JA, Horowitz DH, Arzubiaga C, et al. Treatmentof cutaneous sarcoidosis with chloroquine:review of the literature. Arch Dermatol 1991;127:1034–1040405. Sharma OP. Neurosarcoidosis: a personal perspectivebased on the study of 37 patients. Chest1997; 112:220–228406. Adams JS. Hypercalcemia and hypercalciuria.Semin Respir Med 1992; 13:402–410407. Adams JS, Diz MM, Sharma OP. Effective reductionin the serum 1,25-dihydroxyvitamin D andcalcium concentration in sarcoidosis-associatedhypercalcemia with short-course chloroquinetherapy. Ann Intern Med 1989; 111:437–438408. Baltzan M, Mehta S, Kirkham TH, et al. Randomizedtrial of prolonged chloroquine therapy inadvanced pulmonary sarcoidosis. Am J RespirCrit Care Med 1999; 160:192–197409. Baughman RP, Judson MA, Teirstein AS, et al.Thalidomide for chronic sarcoidosis. Chest 2002;122:227–232410. Carlesimo M, Giustini S, Rossi A, et al. Treatmentof cutaneous and pulmonary sarcoidosis with thalidomide.J Am Acad Dermatol 1995; 32:866–869411. Cagnoni ML, Lombardi A, Cerinic MC, et al.Melatonin for treatment of chronic refractory sarcoidosis.Lancet 1995; 346:1229–1230412. Zabel P, Entzian P, Dalhoff K, et al. Pentoxifyllinein treatment of sarcoidosis. Am J Respir Crit CareMed 1997; 155:1665–1669413. Tong Z, Dai H, Chen B, et al. Inhibition of cyto -kine release from alveolar macrophages in pulmonarysarcoidosis by pentoxifylline: comparisonwith dexamethasone. Chest 2003; 124:1526–1532414. Meyerle JH, Shorr A. The use of infliximab incutaneous sarcoidosis. J Drugs Dermatol 2003;2:413–414415. Mallbris L, Ljungberg A, Hedblad MA, et al. Progressivecutaneous sarcoidosis responding toanti-tumor necrosis factor-alpha therapy. J AmAcad Dermatol 2003; 48:290–293416. Katz JM, Bruno MK, Winterkorn JM, et al. Thepathogenesis and treatment of optic disc swellingin neurosarcoidosis: a unique therapeuticresponse to infliximab. Arch Neurol 2003; 60:426–430417. Roberts SD, Wilkes DS, Burgett RA, et al. Refractorysarcoidosis responding to infliximab. Chest2003; 124:2028–2031418. Pettersen JA, Zochodne DW, Bell RB, et al. Refractoryneurosarcoidosis responding to infliximab.Neurology 2002; 59:1660–1661419. Yee AM, Pochapin MB. Treatment of complicatedsarcoidosis with infliximab anti-tumor necrosis<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 685ACC-PULM-09-0302-038.indd 6857/10/09 10:25:12 PM


factor-alpha therapy. Ann Intern Med 2001;135:27–31420. Baughman RP, Lower EE. Infliximab for refractorysarcoidosis. Sarcoidosis Vasc Diffuse LungDis 2001; 18:70–74421. Baughman RP, Iannuzzi M. Tumour necrosis factorin sarcoidosis and its potential for targetedtherapy. BioDrugs 2003; 17:425–431422. Baughman RP, Drent M, Kavuru M, et al. Infliximabtherapy in patients with chronic sarcoidosisand pulmonary involvement. Am J Respir CritCare Med 2006; 174:795–802423. Baughman RP, Lower EE. Novel therapies forsarcoidosis. Semin Respir Crit Care Med 2007;28:128–133424. Pritchard C, Nadarajah K. Tumour necrosis factoralpha inhibitor treatment for sarcoidosis refractoryto conventional treatments: a report of fivepatients. Ann Rheum Dis 2004; 63:318–320425. Utz JP, Limper AH, Kalra S, et al. Etanerceptfor the treatment of stage II and III progressivepulmonary sarcoidosis. Chest 2003; 124:177–185426. Haley H, Cantrell W, Smith K. Infliximab therapyfor sarcoidosis (lupus pernio). Br J Dermatol 2004;150:146–149427. Khanna D, Liebling MR, Louie JS. Etanerceptameliorates sarcoidosis arthritis and skin disease.J Rheumatol 2003; 30:1864–1867428. Doty JD, Mazur JE, Judson MA. Treatment of sarcoidosiswith infliximab. Chest 2005; 127:1064–1071429. Saleh S, Ghodsian S, Yakimova V, et al. Effectivenessof infliximab in treating selected patientswith sarcoidosis. Respir Med 2006; 100:2053–2059430. Keane J, Gershon S, Wise RP, et al. Tuberculosisassociated with infliximab, a tumor necrosis factoralpha-neutralizing agent. N Engl J Med 2001;345:1098–1104431. Ioannou Y, Isenberg DA. Current evidence for theinduction of autoimmune rheumatic manifestationsby cytokine therapy. Arthritis Rheum 2000;43:1431–1442432. Chung ES, Packer M, Lo KH, et al. Randomized,double-blind, placebo-controlled, pilot trialof infliximab, a chimeric monoclonal antibodyto tumor necrosis factor-alpha, in patients withmoderate-to-severe heart failure: results of theanti-TNF Therapy Against Congestive Heart Failure(ATTACH) trial. Circulation 2003; 107:3133–3140433. Mann DL, McMurray JJ, Packer M, et al. Targetedanticytokine therapy in patients with chronicheart failure: results of the Randomized EtanerceptWorldwide Evaluation (RENEWAL). Circulation2004; 109:1594–602434. Kavanaugh A, St Clair EW, McCune WJ, BraakmanT, Lipsky P. Chimeric anti-tumor necrosisfactor-alpha monoclonal antibody treatmentof patients with rheumatoid arthritis receivingmethotrexate therapy. J Rheumatol 2000; 27:841–850435. Brown SL, Greene MH, Gershon SK, et al. Tumornecrosis factor antagonist therapy and lymphomadevelopment: twenty-six cases reported to theFood and Drug Administration. Arthritis Rheum2002; 46:3151–3158436. Shorr AF, Helman DL, Davies DB, et al. <strong>Pulmonary</strong>hypertension in advanced sarcoidosis: epidemiologyand clinical characteristics. Eur RespirJ 2005; 25:783–788437. Preston IR, Klinger JR, Landzberg MJ, et al. Vasoresponsivenessof sarcoidosis-associated pulmonaryhypertension. Chest 2001; 120:866–872438. Arcasoy SM, Christie JD, Pochettino A, et al.Characteristics and outcomes of patients withsarcoidosis listed for lung transplantation. Chest2001; 120:873–880439. Nunes H, Humbert M, Capron F, et al. <strong>Pulmonary</strong>hypertension associated with sarcoidosis:mechanisms, haemodynamics and prognosis.Thorax 2006; 61:68–74440. Shorr AF, Davies DB, Nathan SD. Outcomes forpatients with sarcoidosis awaiting lung transplantation.Chest 2002; 122:233–238441. Nathan S, Saggar R, Lynch J. Lung transplantationfor interstitial lung disorders. In: Lynch JPIII, Ross DJ, eds. Lung and heart-lung transplantation(Vol 217). New York: Taylor and Francis,2006; 165–204442. Shah L. Lung transplantation in sarcoidosis.Semin Respir Crit Care Med 2007; 28:134–140443. Martinez FJ, Orens JB, Deeb M, et al. Recurrenceof sarcoidosis following bilateral allogeneic lungtransplantation. Chest 1994; 106:1597–1599444. Shorr AF, Helman DL, Davies DB, et al. Sarcoidosis,race, and short-term outcomes followinglung transplantation. Chest 2004; 125:990–996445. Shorr AF, Davies DB, Nathan SD. Predicting mortalityin patients with sarcoidosis awaiting lungtransplantation. Chest 2003; 124:922–928686 IPF, NSIP/Fibrosis, and Sarcoidosis (Lynch)ACC-PULM-09-0302-038.indd 6867/10/09 10:25:12 PM


<strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage SyndromesUlrich Specks, MDObjectives:• <strong>Review</strong> the spectrum of pulmonary vasculitides and theirclinical manifestations• Describe the clinical spectrum of antineutrophil cytoplasmicantibody (ANCA)-associated vasculitis and review treatmentprinciples• <strong>Review</strong> the clinical presentation and management ofChurg-Strauss syndrome (CSS)• Discuss the differential diagnosis of diffuse alveolarhemorrhage (DAH)• <strong>Review</strong> specific diseases that can cause DAHKey words: alveolar hemorrhage; antibasement membranedisease; antineutrophil cytoplasmic antibodies; Behçetdisease; Churg-Strauss syndrome; Goodpasture syndrome;Henoch-Schönlein purpura; idiopathic pulmonary hemosiderosis;microscopic polyangiitis; Takayasu arteritis; temporalarteritis; vasculitis; Wegener granulomatosis<strong>Pulmonary</strong> Vasculitis<strong>Pulmonary</strong> vasculitis is usually a manifestationof a systemic disorder leading to inflammation ofvessels of different sizes by a variety of immunologicmechanisms. Vasculitis can be separated intoprimary and secondary vasculitis. The primarysystemic vasculitides are a heterogeneous group ofsyndromes of unknown etiology that share a clinicalresponse to immunosuppressive therapy. Theirwide spectrum of frequently overlapping clinicalmanifestations is defined by the size and locationof the affected vessels as well as by the nature ofthe inflammatory infiltrate. Secondary vasculitismay occur in the context of a well-defined underlyingdisorder or have a specific etiology, such ascollagen vascular disease, infection, therapeutic,or illicit drug use.Most classification schemes and definitions ofthe vasculitides are based on the size of the mostprominently affected vessels. Definitions, nomenclature,and classification schemes are currentlyunder review and may change. Physicians shouldbe familiar with two sets of criteria and definitions.First, the 1990 criteria 1,2 of the American College ofRheumatology for the classification of vasculitiswere developed to allow the separation of patientswith one form of vasculitis from those with anotherand not to serve as diagnostic criteria. Second, thenomenclature and definitions put forth by the ChapelHill consensus conference 3 in 1992 acknowledgethe presence of ANCA and separate classicpolyarteritis nodosa from microscopic polyangiitisand ANCA-associated vasculitis. It is important torecognize that not all respiratory symptoms occurringin patients with vasculitis are caused by theinflammation of pulmonary vessels.Urticarial VasculitisUrticarial vasculitis (UV) is a systemic disordercharacterized by urticaria associated witharthralgias in 60% of patients, arthritis in 28% ofpatients, abdominal pain in 25% of patients, andglomerulonephritis in 5% of patients. Angioedema,fever, uveitis, episcleritis, and seizures may alsooccur. 4,5 Hypocomplementemia is present in 38%of patients, and this hypocomplementemic form ofUV has been associated with pulmonary complicationsnot directly caused by vasculitis. Obstructivepulmonary disease, which occurs in up totwo thirds of these patients, is thought to resultfrom a combination of smoking and an unknownimmunologic process. Emphysematous changesin the basal regions may occur in some patients. 6UV with or without hypocomplementemia canalso occur in patients with systemic lupus erythematosus(SLE).Giant Cell ArteritisGiant-cell arteritis (GCA), also called temporalarteritis, cranial arteritis, and granulomatous arteritis,is a generalized inflammatory disorder involvinglarge- and medium-sized arteries. 1 It is the mostcommon form of vasculitis in the Northern hemisphere,and it appears to affect predominantlyelderly patients. The typical features of the disease<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 687ACC-PULM-09-0302-039.indd 6877/10/09 10:26:11 PM


include new-onset headache, a palpably tender ornodular temporal artery with decreased pulsation,and an increased erythrocyte sedimentation rate(ESR). Granulomatosus inflammation of the vesselwall is found in 60% of temporal artery biopsyspecimens. GCA and polymyalgia rheumatica arethought to be part of a disease spectrum. The clinicalillness appears gradually with the developmentof nonspecific systemic symptoms such as lowgradefever, malaise, and weight loss. Headache,which is variable but often severe, is the mostcommon symptom in GCA. Amaurosis fugax isobserved in 20% of patients and visual loss in 10%.It is now recognized that the aorta may also beaffected by the disease. 7Respiratory symptoms have been reportedin up to 25% of patients, 8 but they rarely requiremanagement by a pulmonologist. Cough, hoarseness,and throat pain usually resolve promptlywith glucocorticoid therapy. However, respiratorysymptoms may be the initial presentation of GCA.Therefore, this possibility should be consideredin any elderly patient with new onset of cough,hoarseness, or throat pain without other identifiablecause, and it is reasonable to measure theESR in such patients. 9 Isolated cases with pleuraleffusion or multinodular pulmonary lesions havealso been reported in GCA. Such cases are difficultto interpret. Particularly in the later situation,Wegener granulomatosis (WG) should be consideredin the differential diagnosis because it mayalso present with temporal arteritis. 10 Therapy ofGCA continues to rely heavily on the use of glucocorticoidswithout a proven alternative. 11 Theglucocorticoid-sparing role of methotrexate forGCA remains controversial. 12Takayasu ArteritisTakayasu arteritis (TA) is a large-vessel vasculitisaffecting predominantly the aorta and itsmajor branches in young patients. 13 TA has alsobeen called pulseless disease, aortic arch syndrome,or reversed coarctation. The disease affects mostlyyoung adult women. It is not limited to patients ofAsian decent. Early disease manifestations includeconstitutional symptoms, low-grade fever, andarthralgias. Variable pulses of the extremities andclaudication of affected vascular territories aretypical. Renovascular hypertension, pulmonaryhypertension, and ischemia of affected organs maybe the more disabling complications of this chronicallyrelapsing disease.<strong>Pulmonary</strong> complications are the result ofa unique arteriopathy predominantly of thelarge- and medium-size pulmonary vessels.Progressive defects in the outer media of thearteries and ingrowth of granulation tissue-likecapillaries associated with thickened intima andsubendothelial smooth-muscle proliferation leadto pulmonary artery stenoses and occlusion aswell as pulmonary hypertension in up to onehalf of all patients. The inflammatory infiltrateof the vessel wall is predominantly lymphoplasmacyticwith variable amounts of giantcells. The involvement of pulmonary arteries iscommon but often asymptomatic. It is detectableby conventional angiography, perfusionscan, or magnetic resonance angiography. Chestroentgenograms are often normal, but CT scansmay show areas of low attenuation as a resultof regional hypoperfusion, subpleural reticulolinearchanges, and pleural thickening. Fistulaformation between pulmonary artery branchesand bronchial arteries, as well as nonspecificinflammatory interstitial lung disease, have alsobeen reported.Therapy for TA consists primarily of immunosuppressionwith glucocorticoids. Other immunosuppressiveagents, including methotrexate,are used in conjunction with glucocorticoids forremission induction and as glucocorticoid-sparingagents for remission maintenance. Unfortunately,many patients relapse when the glucocorticoiddose is reduced to 15 mg/d. The use of antitumornecrosis factor- agents has been reportedas beneficial in patients refractory to standardtherapy. 14 Vascular bypass procedures may bebeneficial in severe disease, but beneficial resultsare only temporary. 15Classic Polyarteritis NodosaSince its formal separation from microscopicpolyangiitis, this form of vasculitis, which affectspredominantly medium-sized vessels, is diagnosedrarely. 16 Polyarteritis nodosa (PAN) doesnot affect capillaries. Therefore, it does not causeglomerulonephritis or alveolar hemorrhage (AH).However, classic PAN can on rare occasions affect688 <strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage Syndromes (Specks)ACC-PULM-09-0302-039.indd 6887/10/09 10:26:11 PM


the bronchial or bronchiolar arteries. Most casesof classic PAN diagnosed are associated with viralinfections, specifically hepatitis B and C. Consequently,antiviral therapy plays a prominent rolein the management of such cases in addition toimmunosuppression. Classic PAN is far less likelyto reoccur than microscopic polyangiitis and thereforecan generally be treated with a shorter courseof immunosuppression. 17ANCA-Associated VasculitisThe primary systemic small vessel vasculitidesare WG, microscopic polyangiitis (MPA), and theCSS. In contrast to the other forms of systemicvasculitis with predilections for larger vessels,most patients with active WG and MPA and morethan half of patients with active CSS have ANCA.Of all systemic vasculitides, the ANCA-associatedvasculitides are most likely to cause respiratorymanifestations. WG and MPA will be discussedtogether here because the same diagnostic andtherapeutic principles apply; CSS will be discussedseparately.The Chapel Hill Consensus Nomenclature hasdefined WG as “granulomatous inflammationinvolving the respiratory tract, and necrotizingvasculitis affecting small to medium-sized vessels(ie, capillaries, venules, arterioles, and arteries).” 3Histopathologic documentation of granulomatousinvolvement of the respiratory tract is not explicitlyrequired. Roentgenographic evidence or clinicalexamination findings highly predictive of granulomatouspathology are sufficient. Consequently,the diagnosis of WG depends on a correlation ofclinical, pathologic, and serologic features.MPA is defined as a “necrotizing vasculitis withfew or no immune deposits, affecting small vessels(ie, capillaries, venules, or arterioles). Necrotizingarteritis involving small and medium-sized arteriesmay be present. Necrotizing glomerulonephritisis very common; pulmonary capillaritis resultingin AH occurs frequently.” 3 The vasculitis of MPAis indistinguishable from that of WG, and theremay be substantial clinical overlap. The literatureneeds to be interpreted with this in mind. In fact,the therapeutic approach to patients with WGand MPA is governed by the same principles, andmost ongoing clinical studies and therapeutic trialscombine both diseases.For treatment stratification, a patient’s diseaseactivity is categorized as limited disease or severedisease. Severe disease is either life threateningor threatening an organ with irreversible loss offunction. Consequently, patients with any of thefollowing disease manifestations should be labeledas having severe disease: AH, glomerulonephritis,eye involvement (except mere episcleritis), and nervoussystem involvement, including sensorineuralhearing loss. Limited disease includes essentiallyall patients who have nonsevere disease.The term limited disease used in the UnitedStates comprises what European investigatorshave referred to as early systemic disease as wellas localized disease. Although this separation isnot based on well-defined biological distinctions,most disease manifestations leading to the categorizationas severe disease are caused by capillaritis.In contrast, most symptoms leading to theclassification as limited disease are the result ofnecrotizing granulomatous inflammation. Patientswith limited WG have a longer disease duration, agreater likelihood of experiencing an exacerbationof previous disease after a period of remission, anda greater prevalence of destructive upper respiratorytract disorders (eg, saddle-nose deformity). 18Tracheobronchial involvement also distinguishesWG from MPA; it may represent unique treatmentchallenges (discussed later in this section).Histopathologic features that define WG andseparate it from MPA include discrete or confluentnecrotizing granulomatous inflammation withvasculitis. Fibrinoid necrosis, microabscesses, focalvasculitis, thrombosis, and fibrous obliteration ofthe vascular lumen may be observed. Giant cellsare a hallmark of the necrotizing granulomatousinflammation of WG. Other histopathologic findings,including bronchiolitis obliterans organizingpneumonia, bronchocentric inflammation, and amarked number of eosinophil infiltrates, are atypicalfeatures that have also been described in WG.The etiology of ANCA-associated vasculitisremains unknown. Occupational exposures havebeen suggested as possible etiologic factors, includinginhalation of silica-containing compounds,grain dust, and heavy metals. 19–22 Infection hasbeen linked to the onset of the disease as well as torelapses. 23 A multifactorial genetic predispositionmay also contribute to the development of ANCAassociatedvasculitis. 24 Genetic polymorphisms in<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 689ACC-PULM-09-0302-039.indd 6897/10/09 10:26:11 PM


the PR3 molecule as well as its expression on neutrophilsmay influence the induction and courseof WG. Along the same line, heterozygotes for theP I* Z variant of the α 1-antitrypsin gene are reportedto have a sixfold-greater risk for developing thedisease than the general population. Last not least,ANCAs seem to play a pathogenic role for thedevelopment of vasculitis. 25The characteristic disease manifestation of WGaffecting the respiratory tract is caused by necrotizinggranulomatous inflammation. Ear, nose, and throatsymptoms are initially noted by 85% of patients.These may include rhinorrhea, purulent or bloodynasal discharge, nasal mucosal drying and crustformation, epistaxis, and otitis media. Deep facialpain from paranasal sinus involvement, nasalseptal perforation, and ulceration of the vomer areimportant signs. Staphylococcus aureus is frequentlydetected in the nose and sinuses and has been linkedto relapses of the disease. 26 Other signs includeaphthous lesions of the nasal and oral mucosa andinflammation and destruction of the nasal cartilageleading to a saddle-nose deformity. Ulceratedlesions of the larynx and trachea are present in 30%of untreated cases and may produce hemoptysis.The median age of patients is 45 years, themale/female ratio is about equal, and 90% ofpatients are white. The clinical presentation variesfrom subacute nonspecific respiratory illnessto rapidly progressive AH syndrome. <strong>Pulmonary</strong>symptoms are usually associated with unilateralor bilateral abnormalities on the chest radiograph,including infiltrates, nodules, or mass lesions thatmay or may not cavitate. The nodules range froma few millimeters to several centimeters in size.Solitary nodules may also occur. Unusual manifestationsinclude lymphadenopathy, lobar consolidation,and large pleural effusions. Tracheobronchiallesions are common, and they may be asymptomaticor mistaken for asthma. 27,28 <strong>Pulmonary</strong> functiontests and inspiratory and expiratory flow-volumeloops may aid in follow-up.Patients with ANCA-associated vasculitis havevariable degrees of elevated of ESR or C-reactiveprotein. If WG or MPA are suspected, it is crucial toobtain a urine analysis and microscopy and measureserum creatinine because early renal involvementmay be clinically silent yet progress rapidly.In WG and MPA, ANCAs that cause a cytoplasmicimmunofluorescence pattern (C-ANCA) onethanol-fixed neutrophils are caused by antibodiesreacting with proteinase 3 (PR3; PR3-ANCA). Incontrast, a variety of antibodies can cause a perinuclearimmunofluorescence pattern (P-ANCA)on ethanol-fixed neutrophils. Only those that alsoreact with myeloperoxidase are of interest in thecontext of ANCA-associated vasculitis. Maximaldiagnostic accuracy of ANCA testing requires corroborationof a positive target antigen specific testresult (PR3-ANCA or myeloperoxidase-ANCA)by immunofluorescence, and vice versa. 29 Only thePR3-ANCA with C-ANCA combination and themyeloperoxidase-ANCA with P-ANCA combinationare sensitive and specific for ANCA-associatedvasculitis. 30–32The clinical utility, ie, positive and negativepredictive values of ANCA testing for WG andMPA, are critically dependent on the pretest probabilityof the disease in the patient tested as well ason the analytical accuracy of the test method. 33 Ifapplied in patients with clinical features indicatinga high pretest probability of WG or MPA, ANCAtesting has a very high positive predictive value.However, occasional false-positive ANCA testresults have been reported in a variety of infections.Particularly, subacute bacterial endocarditismay represent a diagnostic dilemma because itmay mimic small-vessel vasculitis clinically andhas been reported with C-ANCA/PR3-ANCA. 34Other infections reported with ANCA either donot have the right pairing of immunofluorescencetest results with its corresponding antigen specificityor their clinical features are distinct fromWG or MPA. 23 In patients undergoing evaluationfor necrotizing glomerulonephritis with or withoutAH, ANCA may occur in conjunction withanti-glomerular basement membrane antibodies(anti-GBMs). 35 These ANCAs are usually of themyeloperoxidase-ANCA variety. The presenceof anti-GBM seems to determine the prognosis insuch double-positive patients. 35Most patients with severe WG or MPA havea positive ANCA test (sensitivity 95%), but upto 30% of patients with limited WG may not havedetectable ANCA. 36 Although there is an associationbetween ANCA titers and disease activity,changes in ANCA levels do not reliably predict thedisease activity in individual patients. Therefore,serial titers of ANCA should not be used to planlong-term therapy. 37690 <strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage Syndromes (Specks)ACC-PULM-09-0302-039.indd 6907/10/09 10:26:12 PM


Standard therapy for WG and MPA currentlyfollows the same basic principles. Methotrexate,25 mg, once a week in combination with oralprednisone is now the standard of care for patientswith limited WG. 38 However, there is only one prospectiverandomized trial 39 in which the authorscompared methotrexate with cyclophosphamidefor remission induction in such patients. The trialconducted by the European Vasculitis Study Group(EUVAS) showed that methotrexate is indeed noninferiorto cyclophosphamide for remission induction,but the side effects were less. The trial 39 alsodocumented that early discontinuation of immunosuppressionin patients with ANCA-associatedvasculitis is fraught with a high rate of relapse.The largest reported group of patients with limitedWG was treated with methotrexate for remissioninduction in the context of the Wegener’s GranulomatosisEtanercept Trial (WGET). 40 More than 90%of patients achieved remission with this regimen,and 70% achieved a sustained remission (lasting 6 months). These rates are equivalent to thoseachieved with cyclophosphamide in severe disease(discussed to follow).Cyclophosphamide at a dose of 2 mg/kg/din combination with prednisone remains the standardof care for patients with severe WG or MPA.In contrast to the original regimen introduced byFauci, 40a the consensus now is to limit the durationof cyclophosphamide therapy to the first 3 to6 months. Once remission has been induced andthe prednisone taper is well under way, cyclophosphamideshould be switched to either azathioprine,which is preferred in patients with renalinvolvement and any degree of renal insufficiency,or methotrexate. The first option is supported bythe results of a randomized trial conducted by theEUVAS that showed azathioprine is as good ascyclophosphamide for remission maintenance to18 months. 41 A randomized controlled trial 42 hasshown that methotrexate and azathioprine areequivalent for remission maintenance.Mycophenolate mofetil (MMF) representsanother alternative to methotrexate and azathioprinefor remission maintenance. 43 However,this agent is significantly more expensive, andthe available published reports indicate that asubstantial number of patients also relapse whilereceiving this agent. 43 In the absence of randomizedtrials comparing the different remissionmaintenance agents, it remains unclear to whatextent the reported results with MMF are taintedby patient selection. The use of MMF for remissionmaintenance can only be supported for patients inwhom methotrexate and azathioprine has failed orwho have contraindications for both agents. TheWGET trial, 40 in which methotrexate was usedfor remission maintenance, confirmed that longtermremission remains an elusive goal for manypatients; remission was maintained in less thanhalf of the patients.Whenever cyclophosphamide is used forremission induction, consideration shouldbe given to the patient’s fertility. Young menshould be offered sperm banking before therapyis initiated. If time allows, ovarian protectionshould be offered to young women in additionto minimizing the cumulative exposure as muchas possible. 44Recent data from the EUVAS group 44a indicatethat the IV application of cyclophosphamide isas effective as oral application and maybe safer.Earlier studies 44b had indicated that safety wasimproved because of a lower cumulative dose withIV cyclophosphamide but that a greater relapse ratewas observed after discontinuation. In the author’sexperience-based opinion, IV cyclophosphamideshould be avoided in the ICU setting. However,its use is preferred over oral cyclophosphamide inpatients with questionable compliance, in youngwomen with fertility issues, and in patients whohave GI problems with oral cyclophosphamideapplication.In patients with MPA who have myeloperoxidase-ANCAand mild renal disease (creatinine 3to 5 mg/d) and no other life- or organ-threateningdisease manifestations, MMF may representan alternative to cyclophosphamide providedthat careful observation is guaranteed. 45 Thisstatement is based on results from a randomizedcontrolled trial in 35 patients from China in whichthe authors compared oral MMF (1.5 to 2 g/d)with IV cyclophosphamide (0.75 to 1.0 g/m 2 oncemonthly). 45 In addition, all patients received IVmethylprednisolone bolus therapy, 0.5 g/d for 3days, followed by oral prednisone, 0.6 to 0.8 mg/kgfor 4 weeks tapered by 5 mg/wk to 10 mg/d. Thisstudy showed that the efficacy of these regimenswas equivalent, but MMF was better tolerated thancyclophosphamide. 45<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 691ACC-PULM-09-0302-039.indd 6917/10/09 10:26:12 PM


For some patients with WG and MPA, the combinationof glucocorticoids and cyclophosphamidemay not be sufficient to induce a remission quickly.Plasma exchange (PLEX) should be considered earlyin patients who present with rapidly progressiveglomerulonephritis and renal failure as well as inpatients who present with DAH. PLEX is currentlysupported by two studies. 46,47 The EUVAS investigatorsconducted the Methylprednisolone VersusPlasma Exchange Trial in patients who presentedwith a serum creatinine level of 5.5 mg/dL. 46In addition to standard therapy for severe disease(oral prednisone and cyclophosphamide), patientswere randomized to either receive three pulses ofIV methyl-prednisolone or 2 weeks of PLEX (7 60mL/kg). PLEX was clearly superior to methylprednisolonewith respect to renal recovery. 46 A singlecentercohort study 47 of 20 patients presenting withAH described 100% survival of the patients whenPLEX was added to standard immunosuppressivetherapy. If AH is uncontrolled despite aggressiveimmunosuppressive therapy and PLEX, theuse of recombinant activated factor VII may beconsidered. 48,49The term refractory disease is commonly usedto describe patients who have persistent diseaseactivity on the maximal tolerated dose of cyclophosphamideor who have contraindications forthe use of cyclophosphamide. A variety of agentshave been proposed for use in addition or insteadof the failing regimen in such patients. Small caseseries and uncontrolled trials have suggested someefficacy of infliximab in such patients. 50,51 However,infliximab therapy was associated with a highfrequency of infections with bad outcomes. 50,51The authors of WGET 40 showed that etanercepthad no effect for remission induction or maintenancewhen used in addition to standard therapy.Moreover, significantly more malignancies wereobserved in the etanercept arm among patientswho had received cyclophosphamide. 52 As a consequence,the use of etanercept in patients who havereceived or are receiving cyclophosphamide iscontraindicated. Given all the available data, thereis no convincing rationale for the use of anti-tumornecrosis factor therapy in patients with WG.A novel promising treatment approach hasrecently emerged. B lymphocytes are thought to beinstrumental in the pathogenesis of autoimmunedisease including ANCA-associated vasculitis.B lymphocytes can be depleted with the chimericmonoclonal antibody targeting CD20, rituximab,that is selectively expressed on the surface ofB lymphocytes. A retrospective analysis of thecompassionate use of rituximab and a small prospectivepilot trial have shown that this agent wassuccessful in inducing remission and allowingdiscontinuation of glucocorticoids in patients withrefractory disease. 53,54 Other single case reports andsmall series with short-term follow-up supportthese observations. 54a This experience has led toan ongoing large randomized placebo-controlledtrial that compares rituximab to cyclophosphamidefor remission induction in patients with severeANCA-associated vasculitis (www.clinicaltrials.gov). There are several reports on the use of otheragents in patients with refractory disease. Antithymoglobulinhas some efficacy but significant sideeffects. 54b Dispergualine is an agent available inJapan and Europe but not in the United States. 54cThe management of large airway involvementin WG may represent unique challenges. Subglotticstenosis may require dilation procedures pairedwith local injection of long-acting glucocorticoidswith or without mitomycin C. 55 Stenosis of thelarge airways may require bronchoscopic interventions,including dilation by rigid bronchoscope,Nd:YAG treatment, and the placement of siliconeairway stents. 27,28Complications are responsible for the significantrate of morbidity and mortality seen inpatients with WG and MPA. Both the disease andthe treatment lead to organ damage. In the NationalInstitutes of Health report on 158 patients, permanentcomplications developed in 86% of patientsfrom WG itself, including end-stage renal disease,chronic pulmonary dysfunction, diminishedhearing, destructive sinus disease, saddle-nosedeformities, proptosis, and blindness. Among thetreated patients, 42% experienced permanent treatment-relatedproblems, including chemical (druginduced)cystitis, osteoporotic fracture, bladdercancer, myelodysplasia, and avascular necrosis. 56Reflecting more recent standard practice, in 180patients in WGET, 57 damage that occurred despite(or because of) therapy included visual impairment,hearing loss, nasal blockade, pulmonaryfibrosis, hypertension, renal insufficiency, peripheralneuropathy, gonadal failure, and diabetesmellitus. Only 11% of the enrolled patients did not692 <strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage Syndromes (Specks)ACC-PULM-09-0302-039.indd 6927/10/09 10:26:12 PM


exhibit a single point on the vascular damage index(which was developed from 61 various items)after 1 year of study enrollment. 57 The WGET alsoconcluded that patients with limited WG are at agreater risk of WG-related damage than are thosewith severe disease. 57In the same cohort of 180 patients with WG,the incidence rate of venous thromboembolismwas found to be high when compared with availablerates in the general population, patients withlupus, and patients with rheumatoid arthritis. 58This increased risk of thromboembolic disease hasalso been documented for the other ANCA-associatedvasculitides. 59A high frequency of echocardiographic abnormalitiesattributable to WG and associatedincreased mortality was observed in a study 60 of85 patients with confirmed WG. In 26 of these 73patients with echocardiographic abnormalities(36%), the lesions appeared directly related to WGand consisted of regional wall motion abnormalitiesin 17 patients (65%), left ventricular systolicdysfunction with decreased ejection fraction in13 patients (50%), and pericardial effusion in5 patients (19%). Other findings included valvulitis,left ventricular aneurysm, and a large intracardiacmass. 60Churg-Strauss SyndromeThe Chapel Hill Consensus definition for CSSis “eosinophil-rich and granulomatous inflammationinvolving the respiratory, and necrotizingvasculitis affecting small to medium-sized vessels,and associated with asthma and eosinophilia.” 3CSS is included among the ANCA-associatedvasculitides, but only 40 to 70% of patients withactive CSS have detectable ANCA. 61–63 CSS is primarilydistinguished from WG and MPA by a highprevalence of asthma and peripheral blood andtissue eosinophilia. Three distinct disease phasesof the disease have been described. 64 The firstis a prodromal allergic phase with asthma. Thisphase may last for a number of years. Second isan eosinophilic phase with prominent peripheraland tissue eosinophilia. This phase may also lasta number of years, and the manifestations mayremit and recur during this time period. The differentialdiagnosis for patients in this phase of thedisease includes parasitic infection and chroniceosinophilic pneumonia. The third vasculiticphase consists of systemic vasculitis and may belife threatening. The three phases are not seen inall patients, do not necessarily occur in this order,and may even concur. However, asthma usuallypredates vasculitic symptoms by a mean of 7 years(range 0 to 61 years). 61 Formes frustes of CSS havealso been described with eosinophilic vasculitisand/or eosinophilic granulomas in isolated organswithout evidence of systemic disease. 65<strong>Pulmonary</strong> parenchymal involvement occurs in38% of patients. Transient alveolar-type infiltratesare most common. These have a predominantlyperipheral distribution and are indistinguishablefrom infiltrates found in chronic eosinophilic pneumonia.Occasionally, nodular lesions may be seenin CSS. In contrast to WG and MPA, AH is exceedinglyrare. 61 Renal involvement in CSS is less prominentthan in WG on MPA and does not generallylead to renal failure. In contrast, peripheral nerveinvolvement, typically in the form of mononeuritismultiplex, is more frequent. Skin, heart, CNS, andabdominal viscera may also be involved.The classic histopathologic picture consists ofnecrotizing vasculitis, eosinophilic tissue infiltration,and extravascular granulomas. However, notall features are found in every case, and they arenot pathognomonic of the condition. Particularlythe finding of a Churg-Strauss granuloma on skinbiopsy should not be confused with the diagnosisof CSS. This type of necrotizing extravasculargranuloma may be seen in CSS as well as in othersystemic autoimmune diseases, including WG andrheumatoid arthritis.If ANCAs are present, they are usually P-ANCAreacting with myeloperoxidase. 61 The ANCA statusappears to correlate with disease activity. 61 Recentstudies suggest a more vasculitic disease phenotypein the presence of ANCA, but not all studieshave found this, and there remains substantialoverlap of organ manifestations between patientswith CSS who are ANCA positive and those whoare ANCA negative. 61–63In recent years, significant attention has beendevoted to CSS detected in patients by the use of leukotrienereceptor antagonists. Available case studiesand limited population-based incidence estimatessuggest that these agents may lead to unmaskingof vasculitic symptoms in asthmatic patients byallowing dose reductions or discontinuation of<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 693ACC-PULM-09-0302-039.indd 6937/10/09 10:26:13 PM


oral glucocorticoid therapy. There is no evidencesuggesting that these agents cause CSS.The prognosis of CSS is better than that ofWG or MPA because the overall rate of mortalityis lower and not significantly different from thenormal population. 61 Most deaths are secondaryto cardiac involvement.Systemic glucocorticoids remain the mainstayof therapy for CSS. There are no clinical trials thatprovide clear guidance. It seems most appropriateto treat CSS according to the principles applied tothe management of ANCA associated vasculitis.Accordingly, cyclophosphamide should be addedto glucocorticoids for remission induction in allpatients with disease manifestations that threatenthe patient’s life or the function of a vital organ, ie,particularly those with central- or peripheral-nerveinvolvement, glomerulonephritis, heart involvement,or AH. Methotrexate, azathioprine, or MMFhave been used as glucocorticoid-sparing agents inless-severe disease and for remission maintenance.Refractory disease and disease dominated by difficult-to-controleosinophilic inflammation havebeen reported to respond to interferon- therapy. 66However, continued long-term interferon- therapymay be necessary, and this treatment carries therisk of substantial toxicity. Rituximab has also beenused successfully in CSS. 67,68Idiopathic Pauci-Immune <strong>Pulmonary</strong> CapillaritisThis is a diagnosis of exclusion in patientspresenting with diffuse AH as a result of capillaritisin the absence of symptoms or serologicevidence of any detectable underlying systemicdisorder. 69 Direct immunofluorescence studies ofthe lung tissue reveal no immune deposits. Thisisolated pauci-immune pulmonary capillaritis ishistopathologically indistinguishable from thatof ANCA-associated vasculitis. These patientsare best treated with an immunosuppressive regimenaccording to the guidelines for severe WG ormicroscopic polyangiitis.Behçet DiseaseBehçet disease (BD) is a rare chronically relapsingsystemic inflammatory disorder characterizedby aphthous oral ulcers and at least two or moreof the following: aphthous genital ulcers, uveitis,cutaneous nodules or pustules, or meningoencephalitis.Reported prevalence of the disease is 1:16,000in Japan and 1:20,0000 in the United States. Thereis a strong association with the major histocompatibilitycomplex antigen HLA-B51. 70 The mean age ofpatients at the onset of BD is 35 years; most studieshave reported a predominance of men with thedisease over women. Respiratory manifestationsare common in BD and include cough, hemoptysis,chest pain, and dyspnea. 71 Hemoptysis is often massiveand fatal. The vasculitis of BD is immune complexmediated and may affect vessels of all sizes.Secondary thrombosis with major venous occlusioncan occur. Thrombosis may not be preventable inBD by anticoagulation, but aspirin 80 mg/d, hasbeen advocated. Destruction of the elastic laminaof pulmonary arteries causing aneurysm formation,secondary erosion of bronchi, and arterial-bronchialfistulae may result in massive hemoptysis. CT orMR angiography are used to detect pulmonaryartery aneurysms. Recurrent pneumonia as well asbronchial obstruction as a consequence of mucosalinflammation have also been described.Therapy of the underlying disease consists ofimmunosuppression. 72 Prednisone alone may notbe sufficient to control the vasculitis. The additionof other drugs, such as colchicine, chlorambucil,methotrexate, cyclosporin, or azathioprine, isrecommended. The use of biological agents, inparticular anti-tumor necrosis factors agents, hasalso been reported. The addition of azathioprine orcyclophosphamide to glucocorticoids may resultin the resolution of pulmonary aneurysms. Oncepulmonary arteritis has been identified, anticoagulationshould be avoided. The prognosis ofpulmonary involvement is poor. Approximatelyone third of patients die within 2 years of pulmonaryinvolvement, most from fatal pulmonaryhemorrhage. Embolization therapy may be usedas treatment and prevention of hemorrhage frompulmonary artery aneurysms.Henoch-Schönlein PurpuraHenoch-Schönlein purpura (HSP), also knownas anaphylactoid purpura or allergic purpura, is asyndrome that is characterized by acute purpura,arthritis, colicky abdominal pain, and nephritis.Pathologic findings include acute arteriolitis andvenulitis in the superficial dermis and the bowel.694 <strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage Syndromes (Specks)ACC-PULM-09-0302-039.indd 6947/10/09 10:26:13 PM


Proliferative and necrotizing glomerulonephritis isusually mild. A similar type of renal lesion is seen inpatients with infective subacute bacterial endocarditis,WG, SLE, PAN, and Goodpasture syndrome.Immunofluorescence microscopy shows largedeposits of IgA in the skin and kidney. AlthoughHSP is more common in children (mean age ofpatients, 17 years), adults may also be affected. Palpablepurpura, which are usually distributed overthe buttocks and lower extremities, and fever aregenerally the first signs. The purpura may precede,accompany, or follow arthralgias and abdominalcolic. The triad of purpura, arthritis, and abdominalpain is present in approximately 80% of patients.Joint involvement is typically monoarticular andtransient, involves the large joints, and causes painthat is out of proportion to the objective evidenceof synovitis. Peritonitis and melena are common.<strong>Pulmonary</strong> manifestations of HSP are rare. Only 26cases have been reported to date, and capillaritishas been documented histopathologically only ina minority of them. 73 IgA deposits along the pulmonarycapillary walls, analogous to those foundin vessels of the skin and glomeruli of affectedkidneys, are pathognomonic for HSP.Secondary VasculitisMany of the rheumatologic diseases exhibita secondary vasculitic process in the organsinvolved. Infectious processes, particularly secondaryto infection with Aspergillus and Mucor sp,invade vascular structures and produce secondaryvasculitis. Certain drugs and chemicals can inducevasculitis. Other uncommon secondary vasculiticentities include benign lymphocytic angiitis andgranulomatosis, bronchocentric granulomatosis,and necrotizing sarcoid angiitis.AH SyndromesDiffuse hemorrhage into the alveolar spaces issometimes called alveolar hemorrhage syndrome.The clinical course of AH is unpredictable andmay progress rapidly to respiratory failure; it isalways potentially life threatening. Consequently,its consideration needs to be an integral part of thedifferential diagnostic evaluation of patients withalveolar infiltrates on chest roentgenogram. Thesymptoms of AH are nonspecific. Patients usuallyseek care because of dyspnea and cough, possiblyassociated with fever. Diffuse alveolar infiltrateson chest roentgenogram are the first diagnostichallmark. Depending on the severity of the diseaseprocess at the time of evaluation, anemia andhypoxemia may be prominent. Hemoptysis is acommon presenting symptom of AH, but it may beabsent in up to onethird of patients with AH. 74AH can result from a variety of underlying orassociated conditions that cause a disruption of thealveolar-capillary basement membrane integrity.Mechanisms leading to AH include immunologicalinflammatory conditions causing immune-complexdeposition or capillaritis (eg, Goodpasture syndrome,SLE, ANCA-associated vasculitis); directchemical/toxic injury (eg, from toxic or chemicalinhalation, abciximab use, all-trans-retinoic acid,trimellitic anhydride, or smoking crack cocaine);physical trauma (eg, pulmonary contusion); andincreased vascular pressure within the capillaries(eg, mitral stenosis or severe left ventricular failure).AH is best confirmed by BAL. Progressivelymore bloody return indicates alveolar origin of theblood. 75 The presence of 20% hemosiderin-ladenmacrophages among the total number of alveolarmacrophages recovered by BAL is reportedto indicate AH, even in the absence of ongoingactive bleeding. 76 <strong>Pulmonary</strong> AH is significantlyassociated with the following: thrombocytopenia( 50,000 cells/L), other abnormal coagulationvariables, renal failure (creatinine concentration, 2.5 mg/dL), and a history of heavy smoking.The diagnostic approach to the patient presentingwith AH is aimed at the rapid identification of theunderlying cause and at prompt implementationof appropriate therapy.Exposure to inhalational toxins such as trimelliticanhydride or pyromellitic dianhydrate,drug abuse (including crack cocaine abuse), andsmoking should be identified. The medical historywill also uncover preexisting comorbiditiesthat can cause AH, including mitral stenosis,coagulation disorders, recent bone marrow orhematopoietic stem-cell transplantation, preexistingautoimmune disorders, and therapeutic drugs.Similarly, the initial physical examination shouldinclude a careful search for signs of comorbiditiesand possible systemic autoimmune disorders.Initial blood and urine testing should screenfor other organ involvement, particularly kidney<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 695ACC-PULM-09-0302-039.indd 6957/10/09 10:26:13 PM


involvement (CBC, chemistry group, urineanalysis and microscopy), and determine thecurrent coagulation status (activated partialthromboplastin time, international normalizedratio). Baseline markers of inflammation (ESRand C-reactive protein) are helpful to monitorsubsequent responses to therapy. At the sametime, specific serologic testing for potentialunderlying systemic disease processes should beinitiated. This includes testing for ANCA, anti-GBM antibodies, anti-nuclear antibodies, anti--double-stranded DNA antibodies, rheumatoidfactor, and anti-phospholipid antibodies, as wellas determination of complement and creatininekinase levels. The decision to obtain a biopsyspecimen needs to be considered carefully, withone weighing the risks of the biopsy procedure,the likelihood of obtaining a diagnostic pieceof tissue, the likelihood of the biopsy findingsto alter the therapeutic approach, and the risksassociated with the chosen therapy.The AH syndromes can be broadly separatedinto those in which AH is caused by or associatedwith pulmonary capillaritis and those thatlack pulmonary capillaritis (bland histology).<strong>Pulmonary</strong> capillaritis refers to the specific histopathologicfinding of alveolar wall infiltrationwith inflammatory cells centered on capillarywalls and small veins. 77 The inflammatory cellsare predominantly neutrophils, but eosinophils ormonocytes may also be encountered. Capillaritisusually causes fibrinoid necrosis of alveolar andvessel walls and may culminate in the destructionof the underlying lung architecture. 77 The interstitialinfiltration by neutrophils seen in the contextof capillaritis needs to be distinguished from thepredominant intra-alveolar neutrophilic infiltrationcommonly associated with active infections.Another hallmark of capillaritis is the presence ofpyknotic cells and nuclear fragments from neutrophilsundergoing apoptosis. This feature, referredto as leukocytoclasis, allows the distinction of truecapillaritis from neutrophil margination related tosurgical trauma, which can simulate capillaritis. 78Most of the syndromes associated with pulmonarycapillaritis leading to AH have been discussed inthe “<strong>Pulmonary</strong> Vasculitis” section of this chapter.The subsequent paragraphs will describe a fewunique syndromes or conditions that may also beassociated with AH.Anti-Basement Membrane Antibody Disease(Goodpasture Syndrome)Anti-basement membrane antibody disease(ABMA) , also known as anti-GBM disease or Goodpasturesyndrome, is a rare autoimmune diseasecharacterized by the presence of autoantibodiesdirected against the NC1-domain of the α3 chain ofbasement membrane collagen type IV. This epitopeis only accessible for autoantibodies in the basementmembranes of kidneys and lungs. 79 Diffuse AHoccurs in about one half of patients with ABMA.It is thought to require an additional inhalationalinjury, particularly smoking, for the developmentof the pulmonary manifestation of this disease. 80,81Isolated AH in the absence of renal disease is rarein ABMA. Circulating autoantibodies to basementmembrane are detectable in the serum, but the diagnosisof ABMA hinges on the histopathologic documentationof linear IgG deposits along the basementmembranes in lung or kidney. ABMA is arguablynot a vasculitis. Bland pulmonary hemorrhage is themost frequently described histopathologic patternin diffuse AH associated with ABMA. However,capillaritis as a secondary histopathologic featurehas been encountered in some patients. 82 Earlyimplementation of immunosuppressive therapyin conjunction with PLEX is the key to a favorableoutcome in patients with ABMA. 83 The pulmonaryoutcome of ABMA is generally favorable, whereaschronic renal failure is common. 81VasculitidesThe various primary vasculitis syndromesthat may cause AH have been discussed in detailpreviously. It is worth reiterating here is that themost common cause of pulmonary renal syndromedefined as AH with glomerulonephritis is ANCAassociatedvasculitis, either WG or MPA. In contrast,AH is rare in the context of CSS.SLE and Other Collagen Vascular DisordersThe disease manifestations of SLE are highlyvariable. <strong>Pulmonary</strong> capillaritis leading to diffuseAH is rare in SLE but is usually a severe complicationof the disease. Direct immunofluorescencemicroscopy reveals prominent immune complexdeposits in the affected tissue of patients with696 <strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage Syndromes (Specks)ACC-PULM-09-0302-039.indd 6967/10/09 10:26:14 PM


SLE, including the lungs. Hence, the developmentof pulmonary capillaritis in SLE is thought to beimmune complex mediated. The onset of diffuseAH in patients with SLE is usually abrupt, andit is seldom the first sign of the disease. In theoverwhelming majority of patients, the rapiddevelopment of pulmonary infiltrates is associatedwith fever. Hemoptysis may be absent inup to one half of the patients. Consequently, thedifferentiation of diffuse AH from infection maybe difficult in SLE and may require a diagnosticBAL. Mechanical ventilation, infection, and cyclophosphamidetherapy were identified by univariateanalysis as negative prognostic factors in onecohort. 74 The reported mortality of diffuse AH inSLE ranges from 0 to 90%. 84–87 Treatment consistsof glucocorticoids and cyclophosphamide. Theuse of PLEX has been suggested, but its benefitremains unproven.In all other types of collagen vascular or connectivetissue disorders, respiratory complicationsare very common. However, pulmonary capillaritispresenting as diffuse AH is rare. Isolated caseshave been reported with polymyositis, rheumatoidarthritis, and mixed connective tissue disease. 88,89Consequently, serologic testing performed as partof an evaluation of diffuse AH should include studiesaimed at the identification of these potentialunderlying disease entities.Antiphospholipid SyndromeAntiphospholipid syndrome (APS) is definedby arterial and venous thromboses or recurrentmiscarriages occurring in patients with anti-phospholipidantibodies (anticardiolipin antibodies,lupus anticoagulant, or both). If APS occurs in thecontext of another autoimmune disease, malignancy,or drug exposure, it is labeled secondaryAPS. In the absence of other coexisting disorders,it is considered primary. 90 Hypercoagulability cancause pulmonary embolism and infarction, pulmonarymicrothrombosis, and pulmonary arterialthrombosis with secondary pulmonary hypertensionas consequence. 91–93 However, primary pulmonaryhypertension as well as ARDS have also beenreported as complications of APS.Diffuse AH is rare in APS. The clinical presentationis nonspecific and consists of cough,dyspnea, fever, and bilateral pulmonary infiltrates.Diffuse AH can also occur in the context of ARDS,and hemoptysis is absent in more than half of thereported patients with APS and AH. Therefore, anearly BAL may help in the differential diagnosis.Tissue necrosis from microthrombosis as well as pulmonarycapillaritis have been implicated as causeof AH in APS. Like in SLE, the capillaritis of APS isimmune-complex mediated. Most patients respondto glucocorticoids. 91 Yet, the coexistence of thrombosisand capillaritis with AH represents a therapeuticdilemma because anticoagulation may need to beinterrupted to control the hemorrhage. Early PLEXin addition to immunosuppressive therapy shouldbe considered in patients with APS and AH.Mitral Valve DiseaseDiffuse AH is a well-known feature of mitralstenosis, even though the possibility is rarely consideredin clinical practice. 94 Severe mitral insufficiencycan also produce AH. Hemoptysis can bethe presenting feature. It is caused either by therupture of dilated and varicose bronchial veinsearly in the course of mitral stenosis or as a resultof the stress failure of pulmonary capillaries. Insurgically untreated patients, recurrent episodesof AH may lead to chronic hemosiderosis of thelungs, fibrosis, and punctuate calcification/ossificationof the lung parenchyma.Idiopathic <strong>Pulmonary</strong> HemosiderosisIdiopathic pulmonary hemosiderosis (IPH) isa rare disorder of unknown cause. It is a diagnosisof exclusion. Many cases reported before thediscovery of ANCA probably represented ANCAassociatedvasculitis. Most cases diagnosed as IPHtoday are probably associated with gluten-sensitivesprue (celiac disease). The combination of AH andceliac disease has also been called Lane-Hamiltonsyndrome. 95 In contrast to diffuse AH of otheretiologies, the AH of IPH may be subtle and notnecessarily associated with hemoptysis. Iron deficiencyanemia is a hallmark of IPH. It is thoughtto be caused by recurrent lung hemorrhage ormalabsorption of iron in the GI tract or both. IPHcan occur in children or young adults. Patientssuspected of IPH should be tested for anti-gliadinantibodies, particularly because GI symptoms maybe absent or very subtle. Most patients with IPH<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 697ACC-PULM-09-0302-039.indd 6977/10/09 10:26:14 PM


attributable to celiac disease respond to a glutenfreediet and do not need immunosuppressivetherapy. 95 Repeated AH in IPH has been reported toprogress to lung fibrosis; early appropriate therapymay prevent this outcome. 96Toxic AHAH can result from toxic fume inhalation orfrom blood-borne toxins. Fumes or dust of trimelliticanhydride (a component of certain plastics,paints, and epoxy resins) cause acute rhinitis andasthmatic symptoms if exposure is minor; withgreater exposure, AH occurs. 97, 98 The trimelliticanhydride-hemoptysis anemia syndrome occursafter high-dose exposure to fumes. 99, 100 Isocyanateshave caused lung hemorrhage. Illicit drug use,particularly crack cocaine inhalation, has also beenlinked to AH. 101Many medications have been associated withthe development of AH, 102 including drugs withantiplatelet or anticoagulant effects such as abciximab;immunosuppressive and chemotherapeuticagents such as D-penicillamine, sirolimus, 103 andall-trans-retinoic acid 104 ; antithyroid drugs such aspropyl-thiouracil; or continuous IV infusion of epoprostenolfor primary pulmonary hypertension. 100AH After Bone Marrow TransplantationDAH occurs in both allogeneic and autologoushematopoietic stem-cell transplantation recipients(bone marrow transplantation [BMT]), usuallyduring the first 30 days, with most episodes occurringaround day 12 after BMT. 105–107 Recipients ofautologous BMT are at greater risk than recipientsof allogeneic BMT. AH and diffuse alveolar damageare the two major complications after BMT. Theoverall incidence of diffuse alveolar damage is 5to 7%. An autopsy series 108 has found diffuse AHin 24% of recipients after BMT. Risk factors for thedevelopment of AH include intensive conditioningchemotherapy and older age ( 40 years).Post-BMT AH is a form of noninfectious pneumonitischaracterized by sudden onset of dyspnea,nonproductive cough, fever, and hypoxemia.Hemoptysis is rare, and its absence may lead toincorrect diagnosis. Many cases are identifiedonly at autopsy. 109 The clinical presentation is nonspecific.Chest imaging shows diffuse but patchyalveolar densities with central predominance.Lung biopsy is rarely indicated. BAL is importantto detect AH and to exclude infections. DiffuseAH is a potentially fatal respiratory complicationwith a reported mortality in excess of 50%. 105,110Standard therapy consists of high-dose parenteralcorticosteroids.Miscellaneous Causes of AH<strong>Pulmonary</strong> lymphangioleiomyomatosis is anuncommon cause of AH syndrome. A significantnumber of patients with pulmonary veno-occlusivedisease, the development of AH may occur. 111,112 andpulmonary capillary hemangiomatosis has alsobeen described as a possible cause of AH. 113 Tumoremboli have also been associated with AH. 114References1. Hunder GG, Arend WP, Bloch DA, et al. The AmericanCollege of Rheumatology 1990 criteria for theclassification of vasculitis: introduction. ArthritisRheum 1990; 33:1065–10672. Hunder GG. The use and misuse of classificationand diagnostic criteria for complex diseases. AnnIntern Med 1998; 129:417–4183. Jennette JC, Falk RJ, Andrassy K, et al. Nomenclatureof systemic vasculitides: the proposal ofan international consensus conference. ArthritisRheum 1994; 37:187–1924. Schwartz HR, McDuffie FC, Black LF, et al. Hypocomplementemicurticarial vasculitis: associationwith chronic obstructive pulmonary disease. MayoClin Proc 1982; 57:231–2385. Wisnieski JJ, Baer AN, Christensen J, et al. Hypocomplementemicurticarial vasculitis syndrome:clinical and serologic findings in 18 patients. <strong>Medicine</strong>(Baltimore) 1995; 74:24–416. Ghamra Z, Stoller JK. Basilar hyperlucency in apatient with emphysema due to hypocomplementemicurticarial vasculitis syndrome. RespirCare 2003; 48:697–6997. Bongartz T, Matteson EL. Large-vessel involvementin giant-cell arteritis. Curr Opin Rheumatol2006; 18:10–178. Machado EB, Michet CJ, Ballard DJ, et al. Trendsin incidence and clinical presentation of temporalarteritis in Olmsted County, Minnesota, 1950–1985.Arthritis Rheum 1988; 31:745–749698 <strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage Syndromes (Specks)ACC-PULM-09-0302-039.indd 6987/10/09 10:26:14 PM


9. Larson TS, Hall S, Hepper NGG, et al. Respiratorytract symptoms as a clue to giant cell arteritis. AnnIntern Med 1984; 101:594–59710. Nishino H, DeRemee RA, Rubino FA, et al. Wegener’sgranulomatosis associated with vasculitis ofthe temporal artery: report of five cases. Mayo ClinProc 1993; 68:115–12111. Mazlumzadeh M, Hunder GG, Easley KA, et al.Treatment of giant cell arteritis using inductiontherapy with high-dose glucocorticoids: a doubleblind,placebo-controlled, randomized prospectiveclinical trial. Arthritis Rheum 2006; 54:3310–331812. Mahr AD, Jover JA, Spiera RF, et al. Adjunctivemethotrexate for treatment of giant cell arteritis:an individual patient data meta-analysis. ArthritisRheum 2007; 56:2789–279713. Seo P, Stone JH. Large-vessel vasculitis. ArthritisRheum 2004; 51:128–13914. Hoffman GS, Merkel PA, Brasington RD, et al.Anti-tumor necrosis factor therapy in patientswith difficult to treat Takayasu arteritis. ArthritisRheum 2004; 50:2296–230415. Maksimowicz-McKinnon K, Clark TM, HoffmanGS. Limitations of therapy and a guarded prognosisin an American cohort of Takayasu arteritispatients. Arthritis Rheum 2007; 56:1000–100916. Watts RA, Jolliffe VA, Carruthers DM, et al. Effectof classification on the incidence of polyarteritisnodosa and microscopic polyangiitis. ArthritisRheum 1996; 39:1208–121217. Guillevin L, Lhote F. Distinguishing polyarteritisnodosa from microscopic polyangiitis and implicationsfor treatment. Curr Opin Rheumatol 1995;7:20–2418. Stone JH. Limited versus severe Wegener’s granulomatosis:baseline data on patients in the Wegener’sgranulomatosis etanercept trial. ArthritisRheum 2003; 48:2299–230919. Duna GF, Cotch MF, Galperin C, et al. Wegener’sgranulomatosis: role of environmental exposures.Clin Exp Rheumatol 1998; 16:669–67420. Lane SE, Watts RA, Bentham G, et al. Are environmentalfactors important in primary systemicvasculitis? A case-control study. Arthritis Rheum2003; 48:814–82321. Albert D, Clarkin C, Komoroski J, et al. Wegener’sgranulomatosis: possible role of environmentalagents in its pathogenesis. Arthritis Rheum 2004;51:656–66422. Hogan SL, Cooper GS, Savitz DA, et al. Associationof silica exposure with anti-neutrophil cytoplasmicautoantibody small-vessel vasculitis: apopulation-based, case-control study. Clin J AmSoc Nephrol 2007; 2:290–29923. Capizzi SA, Specks U. Does infection play a role inthe pathogenesis of pulmonary vasculitis? SeminRespir Infect 2003; 18:17–2224. Kallenberg CG, Rarok A, Stegeman CA. Geneticsof ANCA-associated vasculitides. Cleve Clin JMed 2002; 69:SII61–SII6325. Jennette JC, Falk RJ. New insight into the pathogenesisof vasculitis associated with antineutrophilcytoplasmic autoantibodies. Curr Opin Rheumatol2008; 20:55–6026. Stegeman CA, Cohen Tervaert JW, Sluiter WJ, et al.Association of chronic nasal carriage of Staphylococcusaureus and higher relapse rates in Wegenergranulomatosis. Ann Intern Med 1994; 120:12–1727. Daum DE, Specks U, Colby TV, et al. Tracheobronchialinvolvement in Wegener’s granulomatosis.Am J Respir Crit Care Med 1995; 151:522–52628. Polychronopoulos VS, Prakash UB, Golbin JM,et al. Airway involvement in Wegener’s granulomatosis.Rheum Dis Clin North Am 2007; 33:755–77529. Savige J, Gillis D, Benson E, et al. Internationalconsensus statement on testing and reporting ofantineutrophil cytoplasmic antibodies (ANCA).Am J Clin Pathol 1999; 111:507–51330. Merkel PA, Polisson RP, Chang Y, et al. Prevalenceof antineutrophil cytoplasmic antibodies in a largeinception cohort of patients with connective tissuedisease. Ann Intern Med 1997; 126:866–87331. Russell KA, Wiegert E, Schroeder DR, et al. Detectionof anti-neutrophil cytoplasmic antibodiesunder actual clinical testing conditions. ClinImmunol 2002; 103:196–20332. Vassilopoulos D, Niles JL, Villa-Forte A, et al. Prevalenceof antineutrophil cytoplasmic antibodies inpatients with various pulmonary diseases or multiorgandysfunction. Arthritis Rheum 2003; 49:151–15533. Mandl LA, Solomon DH, Smith EL, et al. Usingantineutrophil cytoplasmic antibody testing todiagnose vasculitis: can test-ordering guidelinesimprove diagnostic accuracy? Arch Intern Med2002; 162:1509–151434. Choi HK, Lamprecht P, Niles JL, et al. Subacute bacterialendocarditis with positive cytoplasmic antineutrophilcytoplasmic antibodies and anti-proteinase3 antibodies. Arthritis Rheum 2000; 43:226–231<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 699ACC-PULM-09-0302-039.indd 6997/10/09 10:26:14 PM


35. Rutgers A, Slot M, van Paassen P, et al. Coexistenceof anti-glomerular basement membraneantibodies and myeloperoxidase-ANCAs in crescenticglomerulonephritis. Am J Kidney Dis 2005;46:253–26236. Finkielman JD, Lee AS, Hummel AM, et al. ANCAare detectable in nearly all patients with activesevere Wegener’s granulomatosis. Am J Med 2007;120:643 e649–e61437. Finkielman JD, Merkel PA, Schroeder D, et al. Antiproteinase3 antineutrophil cytoplasmic antibodiesand disease activity in Wegener granulomatosis.Ann Intern Med 2007; 147:611–61938. Specks U. Methotrexate for Wegener’s granulomatosis:what is the evidence? Arthritis Rheum 2005;52:2237–224239. De Groot K, Rasmussen N, Bacon PA, et al. Randomizedtrial of cyclophosphamide versus methotrexatefor induction of remission in early systemicantineutrophil cytoplasmic antibody-associatedvasculitis. Arthritis Rheum 2005; 52:2461–246940. The WGET Research Group. Etanercept plus standardtherapy for Wegener’s granulomatosis. NEngl J Med 2005; 352:351–36140a. Fauci AS, Haynes BF, Katz P, et al. Wegener’sgranulomatosis: prospective clinical and therapeuticexperience with 85 patients for 21 years. AnnIntern Med 1983; 98:76–8541. Jayne D, Rasmussen N, Andrassy K, et al. A randomizedtrial of maintenance therapy for vasculitisassociated with antineutrophil cytoplasmicautoantibodies. N Engl J Med 2003; 349:36–4442. Pagnoux C, Mahr A, Hamidou MA, et al. Azathioprineor methotrexate maintenance for ANCAassociatedvasculitis. N Engl J Med 2008; 359:2790–280343. Koukoulaki M, Jayne DR. Mycophenolate mofetilin anti-neutrophil cytoplasm antibodies-associatedsystemic vasculitis. Nephron Clin Pract 2006; 102:c100–c10744. Somers EC, Marder W, Christman GM, et al. Use ofa gonadotropin-releasing hormone analog for protectionagainst premature ovarian failure duringcyclophosphamide therapy in women with severelupus. Arthritis Rheum 2005; 52:2761–276744a. de Groot K, Harper L, Jayne DR, et al. Pulseversus daily oral cyclophosphamide for inductionof remission in antineutrophil cytoplasmicantibody-associated vasculitis: a randomized trial.Ann Intern Med 2009; 150:670–68044b. de Groot K, Adu D, Savage CO. The value of pulsecyclophosphamide in ANCA-associated vasculitis:meta-analysis and critical review. Nephrol DialTransplant 2001; 16:2018–202745. Hu W, Liu C, Xie H, et al. Mycophenolate mofetilversus cyclophosphamide for inducing remissionof ANCA vasculitis with moderate renal involvement.Nephrol Dial Transplant 2008; 23:1307–131246. Jayne DR, Gaskin G, Rasmussen N, et al. Randomizedtrial of plasma exchange or high-dosagemethylprednisolone as adjunctive therapy forsevere renal vasculitis. J Am Soc Nephrol 2007;18:2180–218847. Klemmer PJ, Chalermskulrat W, Reif MS, et al.Plasmapheresis therapy for diffuse alveolar hemorrhagein patients with small-vessel vasculitis.Am J Kidney Dis 2003; 42:1149–115348. Betensley AD, Yankaskas JR. Factor VIIa for alveolarhemorrhage in microscopic polyangiitis. Am JRespir Crit Care Med 2002; 166:1291–129249. Henke D, Falk RJ, Gabriel DA. Successful treatmentof diffuse alveolar hemorrhage with activated factorVII. Ann Intern Med 2004; 140:493–49450. Lamprecht P, Voswinkel J, Lilienthal T, et al. Effectivenessof TNF-alpha blockade with infliximab inrefractory Wegener’s granulomatosis. Rheumatology(Oxford) 2002; 41:1303–130751. Booth A, Harper L, Hammad T, et al. Prospectivestudy of TNFalpha blockade with infliximab inanti-neutrophil cytoplasmic antibody- associatedsystemic vasculitis. J Am Soc Nephrol 2004; 15:717–72152. Stone JH, Holbrook JT, Marriott MA, et al. Solidmalignancies among patients in the Wegener’sGranulomatosis Etanercept Trial. Arthritis Rheum2006; 54:1608–161853. Keogh KA, Wylam ME, Stone JH, et al. Inductionof remission by B lymphocyte depletion in elevenpatients with refractory antineutrophil cytoplasmicantibody-associated vasculitis. Arthritis Rheum2005; 52:262–26854. Keogh KA, Ytterberg SR, Fervenza FC, et al. Rituximabfor refractory Wegener’s granulomatosis:report of a prospective, open-label pilot trial. AmJ Respir Crit Care Med 2006; 173:180–18754a. Golbin JM, Specks U. Targeting B lymphocytes astherapy for ANCA-associated vasculitis. RheumDis Clin North Am 2007; 33:741–75454b. Schmitt WH, Hagen EC, Neumann I, et al. Treatmentof refractory Wegener’s granulomatosis with700 <strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage Syndromes (Specks)ACC-PULM-09-0302-039.indd 7007/10/09 10:26:15 PM


antithymocyte globulin (ATG): an open study in15 patients. Kidney Int 2004; 65:1440–144854c. Flossmann O, Baslund B, Bruchfeld A, et al.Deoxyspergualin in relapsing and refractoryWegener’s granulomatosis. Ann Rheum Dis 2009;68:1125–113055. Lebovics RS, Hoffman GS, Leavitt RY, et al. Themanagement of subglottic stenosis in patients withWegener’s granulomatosis. Laryngoscope 1992;102:1341–134556. Hoffman GS, Kerr GS, Leavitt RY, et al. Wegenergranulomatosis: an analysis of 158 patients. AnnIntern Med 1992; 116:488–49857. Seo P, Min YI, Holbrook JT, et al. Damage causedby Wegener’s granulomatosis and its treatment:prospective data from the Wegener’s GranulomatosisEtanercept Trial (WGET). Arthritis Rheum2005; 52:2168–217858. Merkel PA, Lo GH, Holbrook JT, et al. Highincidence of venous thrombotic events amongpatients with Wegener granulomatosis: theWegener’s Clinical Occurrence of Thrombosis(WeCLOT) study. Ann Intern Med 2005; 142:620–62659. Allenbach Y, Seror R, Pagnoux C, et al. Highfrequency of venous thromboembolic events inChurg-Strauss syndrome, Wegener’s granulomatosisand microscopic polyangiitis but not polyarteritisnodosa: a systematic retrospective studyon 1130 patients. Ann Rheum Dis 2009; 68:564–56760. Oliveira GH, Seward JB, Tsang TS, et al. Echocardiographicfindings in patients with Wegener granulomatosis.Mayo Clin Proc 2005; 80:1435–144061. Keogh KA, Specks U. Churg-Strauss syndrome:clinical presentation, antineutrophil cytoplasmicantibodies, and leukotriene receptor antagonists.Am J Med 2003; 115:284–29062. Sinico RA, Di Toma L, Maggiore U, et al. Prevalenceand clinical significance of antineutrophilcytoplasmic antibodies in Churg-Strauss syndrome.Arthritis Rheum 2005; 52:2926–293563. Sable-Fourtassou R, Cohen P, Mahr A, et al. Antineutrophilcytoplasmic antibodies and the Churg-Strauss syndrome. Ann Intern Med 2005; 143:632–63864. Lanham JG, Elkon KB, Pusey CD, et al. Systemicvasculitis with asthma and eosinophilia: a clinicalapproach to the Churg-Strauss syndrome. <strong>Medicine</strong>1984; 63:65–8165. Churg A, Brallas M, Cronin SR, et al. Formesfrustes of Churg-Strauss syndrome. Chest 1995;108:320–32366. Tatsis E, Schnabel A, Gross WL. Interferon-a treatmentof four patients with the Churg-Strauss syndrome.Ann Intern Med 1998; 129:370–37467. Koukoulaki M, Smith KG, Jayne DR. Rituximab inChurg-Strauss syndrome. Ann Rheum Dis 2006;65:557–55968. Pepper RJ, Fabre MA, Pavesio C, et al. Rituximabis effective in the treatment of refractory Churg-Strauss syndrome and is associated with diminishedT-cell interleukin-5 production. Rheumatology(Oxford) 2008; 47:1104–110569. Jennings CA, King TE Jr., Tuder R, et al. Diffusealveolar hemorrhage with underlying isolated,pauciimmune pulmonary capillaritis. Am J RespirCrit Care Med 1997; 155:1101–110970. Verity DH, Marr JE, Ohno S, et al. Behçet’s disease,the Silk Road and HLA-B51: historical andgeographical perspectives. Tissue Antigens 1999;54:213–22071. Uzun O, Akpolat T, Erkan L. <strong>Pulmonary</strong> vasculitisin Behçet disease: a cumulative analysis. Chest2005; 127:2243–225372. Kurokawa MS, Yoshikawa H, Suzuki N. Behcet’sdisease. Semin Respir Crit Care Med 2004; 25:557–56873. Nadrous HF, Yu AC, Specks U, et al. <strong>Pulmonary</strong>involvement in Henoch-Schönlein purpura. MayoClin Proc 2004; 79:1151–115774. Zamora MR, Warner ML, Tuder R, et al. Diffusealveolar hemorrhage and systemic lupus erythematosus:clinical presentation, histology, survival,and outcome. <strong>Medicine</strong> (Baltimore) 1997; 76:192–20275. Robbins RA, Linder J, Stahl MG, et al. Diffusealveolar hemorrhage in autologous bone marrowtransplant recipients. Am J Med 1989; 87:511–51876. De Lassence A, Fleury-Feith J, Escudier E, et al.Alveolar hemorrhage: diagnostic criteria andresults in 194 immunocompromised hosts. Am JRespir Crit Care Med 1995; 151:157–16377. Colby TV, Fukuoka J, Ewaskow SP, et al. Pathologicapproach to pulmonary hemorrhage. AnnDiagn Pathol 2001; 5:309–31978. Kadokura M, Colby TV, Myers JL, et al. Pathologiccomparison of video-assisted thoracic surgicallung biopsy with traditional open lung biopsy. JThorac Cardiovasc Surg 1995; 109:494–498<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 701ACC-PULM-09-0302-039.indd 7017/10/09 10:26:15 PM


79. Salama AD, Pusey CD. Immunology of anti-glomerularbasement membrane disease. Curr OpinNephrol Hypertens 2002; 11:279–28680. Donaghy M, Rees AJ. Cigarette smoking andlung haemorrhage in glomerulonephritis causedby autoantibodies to glomerular basement membrane.Lancet 1983; 2:1390–139381. Lazor R, Bigay-Game L, Cottin V, et al. Alveolarhemorrhage in anti-basement membrane antibodydisease: a series of 28 cases. <strong>Medicine</strong> (Baltimore)2007; 86:181–19382. Lombard CM, Colby TV, Elliott CG. Surgicalpathology of the lung in anti-basement membraneantibody-associated Goodpasture’s syndrome.Hum Pathol 1989; 20:445–45183. Levy JB, Turner AN, Rees AJ, et al. Long-termoutcome of anti-glomerular basement membraneantibody disease treated with plasma exchangeand immunosuppression. Ann Intern Med 2001;134:1033–104284. Schwab EP, Schumacher HR, Freundlich B, et al.<strong>Pulmonary</strong> alveolar hemorrhage in systemic lupuserythematosus. Semin Arthritis Rheum 1993; 23:8–1585. Koh WH, Thumboo J, Boey ML. <strong>Pulmonary</strong> haemorrhagein Oriental patients with systemic lupuserythematosus. Lupus 1997; 6:713–71686. Santos-Ocampo AS, Mandell BF, Fessler BJ. Alveolarhemorrhage in systemic lupus erythematosus:presentation and management. Chest 2000;118:1083–109087. Chang MY, Fang JT, Chen YC, et al. Diffuse alveolarhemorrhage in systemic lupus erythematosus:a single center retrospective study in Taiwan. RenFail 2002; 24:791–80288. Schwarz MI, Sutarik JM, Nick JA, et al. <strong>Pulmonary</strong>capillaritis and diffuse alveolar hemorrhage: a primarymanifestation of polymyositis. Am J RespirCrit Care Med 1995; 151:2037–204089. Schwarz MI, Zamora MR, Hodges TN, et al. Isolatedpulmonary capillaritis and diffuse alveolar hemorrhagein rheumatoid arthritis and mixed connectivetissue disease. Chest 1998; 113:1609–161590. Wilson WA, Gharavi AE, Koike T, et al. Internationalconsensus statement on preliminary classificationcriteria for definite antiphospholipidsyndrome: report of an international workshop.Arthritis Rheum 1999; 42:1309–131191. Gertner E. Diffuse alveolar hemorrhage in theantiphospolipid syndrome: spectrum of diseaseand treatment. J Rheumatol 1999; 26:805–80792. Asherson RA, Cervera R, Piette JC, et al. Catastrophicantiphospholipid syndrome; clinical andlaboratory features of 50 patients. <strong>Medicine</strong> (Baltimore)1998; 77:195–20793. Espinosa G, Cervera R, Font J, et al. The lung inthe antiphospholipid syndrome. Ann Rheum Dis2002; 61:195–19894. Woolley K, Stark P. <strong>Pulmonary</strong> parenchymal manifestationsof mitral valve disease. Radiographics1999; 19:965–97295. Agarwal R, Aggarwal AN, Gupta D. Lane-Hamiltonsyndrome: simultaneous occurrence of coeliacdisease and idiopathic pulmonary haemosiderosis.Intern Med J 2007; 37:65–6796. Buschman DL, Ballard R. Progressive massivefibrosis associated with idiopathic pulmonaryhemosiderosis. Chest 1993; 104:293–29597. Grammer LC, Shaughnessy MA, Zeiss CR, et al.<strong>Review</strong> of trimellitic anhydride (TMA) inducedrespiratory response. Allergy Asthma Proc 1997;18:235–23798. Grammer LC, Ditto AM, Tripathi A, et al. Prevalenceand onset of rhinitis and conjunctivitis insubjects with occupational asthma caused by trimelliticanhydride (TMA). J Occup Environ Med2002; 44:1179–118199. Herbert FA, Orford R. <strong>Pulmonary</strong> hemorrhageand edema due to inhalation of resins containingtrimellitic anhydride. Chest 1979; 76:546–551100. Ogawa A, Matsubara H, Fujio H, et al. Risk ofalveolar hemorrhage in patients with primarypulmonary hypertension-anticoagulation andepoprostenol therapy. Circ J 2005; 69:216–220101. Garcia-Rostan y Perez GM, Garcia BragadoF, Puras Gil AM. <strong>Pulmonary</strong> hemorrhage andantiglomerular basement membrane antibodymediatedglomerulonephritis after exposureto smoked cocaine (crack): a case report andreview of the literature. Pathol Int 1997; 47:692–697102. Schwarz MI, Fontenot AP. Drug-induced diffusealveolar hemorrhage syndromes and vasculitis.Clin Chest Med 2004; 25:133–140103. Vlahakis NE, Rickman OB, Morgenthaler T. Sirolimus-associateddiffuse alveolar hemorrhage.Mayo Clin Proc 2004; 79:541–545104. Nicolls MR, Terada LS, Ruder RM, et al. Diffusealveolar hemorrhage with underlying pulmonarycapillaritis in the retinoic acid syndrome. Am JRespir Crit Care Med 1998; 158:1302–1305702 <strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage Syndromes (Specks)ACC-PULM-09-0302-039.indd 7027/10/09 10:26:15 PM


105. Afessa B, Tefferi A, Litzow MR, et al. Diffusealveolar hemorrhage in hematopoietic stem celltransplant recipients. Am J Respir Crit Care Med2002; 166:641–645106. Yen KT, Lee AS, Krowka MJ, et al. <strong>Pulmonary</strong>complications in bone marrow transplantation:a practical approach to diagnosis and treatment.Clin Chest Med 2004; 25:189–201107. Afessa B, Peters SG. Major complications followinghematopoietic stem cell transplantation.Semin Respir Crit Care Med 2006; 27:297–309108. Roychowdhury M, Pambuccian SE, Aslan DL,et al. <strong>Pulmonary</strong> complications after bone marrowtransplantation: an autopsy study from alarge transplantation center. Arch Pathol LabMed 2005; 129:366–371109. Sharma S, Nadrous HF, Peters SG, et al. <strong>Pulmonary</strong>complications in adult blood and marrowtransplant recipients: autopsy findings. Chest2005; 128:1385–1392110. Lewis ID, DeFor T, Weisdorf DJ. Increasing incidenceof diffuse alveolar hemorrhage followingallogeneic bone marrow transplantation: crypticetiology and uncertain therapy. Bone MarrowTransplant 2000; 26:539–543111. Cohn RC, Wong R, Spohn WA, et al. Death dueto diffuse alveolar hemorrhage in a child withpulmonary veno-occlusive disease. Chest 1991;100:1456–1458112. Rabiller A, Jais X, Hamid A, et al. Occult alveolarhaemorrhage in pulmonary veno-occlusive disease.Eur Respir J 2006; 27:108–113113. Tron V, Magee F, Wright JL, et al. <strong>Pulmonary</strong>capillary hemangiomatosis. Hum Pathol 1986;17:1144–1150114. Almagro P, Julia J, Sanjaume M, et al. <strong>Pulmonary</strong>capillary hemangiomatosis associated with primarypulmonary hypertension: report of 2 newcases and review of 35 cases from the literature.<strong>Medicine</strong> (Baltimore) 2002; 81:417–424<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 703ACC-PULM-09-0302-039.indd 7037/10/09 10:26:15 PM


Notes704 <strong>Pulmonary</strong> Vasculitis and Alveolar Hemorrhage Syndromes (Specks)ACC-PULM-09-0302-039.indd 7047/10/09 10:26:15 PM


Women’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong>Stephanie M. Levine, MD, FCCPObjectives:• <strong>Review</strong> the normal respiratory and cardiovascular physiologyof pregnancy• <strong>Review</strong> the management of asthma in pregnancy• <strong>Review</strong> the management of venous thromboembolism inpregnancy• <strong>Review</strong> the management of tuberculosis and other respiratoryinfections in pregnancy• <strong>Review</strong> the causes and management of acute respiratoryfailure in pregnancy• Discuss the statistics regarding smoking and the epidemiologyof lung cancer in women• <strong>Review</strong> disorders unique to women, eg, catamenial diseasesand lymphangioleiomyomatosisKey words: acute respiratory failure; asthma; lung cancer;pregnancy; venous thromboembolism; womenThis chapter will focus on pulmonary diseasesunique to women, including issues surroundingpregnancy, as well as diseases that may have differentepidemiology and/or prognosis in women.Common pulmonary problems with different treatmentplans in the pregnant or lactating woman andcauses and management of acute respiratory failurein the pregnant patient will be reviewed. Other diseasesthat occur exclusively or are more prevalentin women, such as lymphangioleiomyomatosis(LAM), are discussed, but they are primarily coveredelsewhere in this course.Physiology of PregnancyTable 1. Normal Respiratory Physiologic Changes in PregnancyVariables<strong>Pulmonary</strong> functionExpiratory reservevolumeResidual volumeFunctional residualcapacityTotal lung capacityInspiratory capacityVital capacityTidal volumeRespiratory rateMinute ventilationPeak flowFEV 1Lung complianceTotal respiratorycomplianceDiffusion capacityGas exchangePaco 2ChangesDecreasedDecreasedDecreasedMildly decreasedIncreasedNo changeIncreasedNo change, mild increaseIncreasedNo changeNo changeNo changeDecreasedIncrease followed bydecreaseDecreased to 28 to32 mm HgPao 2Increase followedby decreasepH Increased to 7.40 to 7.45Serum bicarbonateDecreased to 18 to21 mEq/LAlveolar-arterial gradient Mildly increasedOxygen consumptionIncreasedCarbon dioxide production IncreasedTable 2. Normal Cardiovascular Physiologic Changesin PregnancyDuring pregnancy, the body undergoes anatomicand physiologic changes affecting both therespiratory and cardiovascular systems. It is importantthat the pulmonologist be familiar with thesenormal physiologic changes (Tables 1, 2).RespiratoryThe upper respiratory tract undergoes changesduring pregnancy, including hyperemia of theairway mucosa, increased secretion production,VariablesCardiac outputHeart rateStroke volumeSystemic vascular resistance<strong>Pulmonary</strong> vascular resistanceBPBlood volumeRBC volumeHematocritSerum protein levelsChangeIncreasedIncreasedIncreasedDecreasedDecreasedDecreasedIncreasedIncreased, but lessthan blood volumeDilutional decreaseDecreased<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 705ACC-PULM-09-0302-040.indd 7057/10/09 10:26:55 PM


and mucosal edema, particularly accentuatedin the third trimester. Patients can have upperrespiratory tract symptoms secondary to theaforementioned, including epistaxis, nasal stuffiness,hoarseness, and voice changes. Nasal polyposisand allergic rhinitis can begin or becomeexacerbated during pregnancy. These changes areprimarily caused by the production of estrogen.The health-care provider should be aware of thesechanges when performing procedures such asendotracheal intubation and/or nasopharyngealtube placement.The thoracic cage also undergoes anatomicchanges during pregnancy. The diaphragm becomeselevated up to an average of 4 cm above normal atfull-term pregnancy. However, diaphragmatic dysfunctioncaused by pregnancy is unusual because,concurrent with these changes, there is a wideningof the lower rib cage and alteration of the abdominalmuscles; thus, diaphragmatic excursion is actuallymaintained or increased.<strong>Pulmonary</strong> Function Tests: The major change inpulmonary function testing in pregnancy is a progressivedecrease in the expiratory reserve volumeby 8 to 40% and a decrease in residual volume by 7to 22%, both changes attributable to the enlarginguterus and diaphragmatic elevation. Because ofthese changes, functional residual capacity is, inturn, decreased by 10 to 25% by the third trimesterof pregnancy. Total lung capacity may decreaseslightly, inspiratory capacity increases, and vitalcapacity does not change significantly duringpregnancy. The decrease in residual volume with arelatively maintained total lung capacity results ina low ratio of residual volume to total lung capacity.Closure of the small airways during normaltidal breathing caused by the reduction in functionalresidual capacity can result in changes inventilation/perfusion matching and gas exchangediscussed below.Tidal volume increases significantly duringpregnancy by 150 mL to a final value of 450 to 600mL, or a 30 to 50% increase in the average patient.This increase is most likely to the result of a directprogesterone-mediated increase in central respiratorydrive and enhancement of the hypercapnicventilatory drive. There is little or no change inrespiratory rate during pregnancy, and tachypneais an unusual finding. Because of the increase intidal volume, there is a significant increase in restingminute ventilation from 6 L in the nonpregnant stateto 9 L at full term (or 20 to 50% above baseline).There are no significant changes in peak flowrates, FEV 1, airway resistance, or maximum voluntaryventilation during pregnancy. Althoughlung compliance does not change significantlyduring pregnancy, there is a reduction in totalrespiratory compliance caused by a reductionin chest wall compliance because of an elevateddiaphragm caused by uterine enlargement by thethird trimester. Diffusion capacity of the lung forcarbon monoxide may increase slightly in the firsttrimester, followed by a slight decrease later inpregnancy, likely the result of alterations in pulmonaryvascular volume.Gas Exchange: Gas exchange in pregnancy ischaracterized by a mild compensated respiratoryalkalosis secondary to an increase in minute ventilationout of proportion to maternal needs. Thus,normal Paco 2values are lower, 28 to 32 mm Hg, andserum bicarbonate is compensatorily decreased to18 to 21 mEq/L. Normal pH is slightly alkalemic at7.40 to 7.45. Pao 2is slightly elevated, 100 to 105 mmHg, with a slight decrease at term to 100 mm Hg.There is normally an increase of 5 to 10 mm Hg inin alveolar-arterial gradient above baseline, especiallyin the supine position. Oxygen consumptionincreases by 20 to 30% during pregnancy, with aconcomitant increase in carbon dioxide productionby 30 to 35%. These changes are attributable to increasedmaternal and fetal metabolic requirementsand increases in work of breathing and cardiacoutput. During delivery, oxygen consumption canincrease to 40 to 100% above baseline.Sleep-Disordered Breathing: Because of the abovenormalphysiologic changes during pregnancy, inparticular the estrogen-related changes in upperairway patency, the incidence of snoring increasesduring pregnancy. Most studies also suggest thatthere is an increased incidence and severity of sleepdisorderedbreathing during pregnancy, which increasesas pregnancy progresses. Obstructive sleepapnea during pregnancy may develop in patientspredisposed to sleep-disordered breathing, and patientswith preexisting obstructive sleep apnea mayhave worsening severity of disease. Untreated, thisworsening may lead to complications of maternalhypertension and fetal growth retardation. Continuouspositive airway pressure has been used fortreatment, same as in the nonpregnant individual.706 Women’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong> (Levine)ACC-PULM-09-0302-040.indd 7067/10/09 10:26:56 PM


CardiovascularThe cardiovascular system probably undergoesthe most significant changes during pregnancy,including an increase in cardiac output, beginningin the first trimester and peaking at about the <strong>25th</strong>to 32nd week at 30 to 50% above normal. Thisincreased is attributable to both a change in heartrate as well as stroke volume and a decrease insystemic vascular resistance, partially as the resultof shunting of blood to the low-resistance placentalbed and perhaps increased levels of vasodilatormediators. <strong>Pulmonary</strong> vascular resistance alsodecreases. Both systolic and particularly diastolicpressures are reduced. Postural hypotension maybe apparent, particularly in the third trimesterwhen the uterus can compress the inferior venacava (IVC), impeding venous return to the heart.During delivery, cardiac output may be furtheraugmented by an additional 10 to 15% as the resultof catecholamine release.Total blood and plasma volume increases upto 35 to 50% of normal, peaking in the mid-thirdtrimester with a lesser increase in RBC volume(approximately 20 to 40%). The result is hemodilutionwith a relative decrease in hemoglobin,hematocrit (approximately a 12% decrease), andrelative RBC volume, ie, the anemia of pregnancy.There is also a dilutional decrease in serum proteinlevels, resulting in a decrease of 5 mm Hg in plasmaoncotic pressures, predisposing patients to edemaformation, including pulmonary edema. There isan approximate total body fluid expansion of 6 to8 L of water divided among the fetus, amnioticfluid, and intracellular and extracellular spaces,and an increase in plasma volume of 1 to 1.5 L.These changes in fluid status are likely the resultof increased mineralocorticoid production and/orhormonally mediated vasodilation.Dyspnea During PregnancyUp to twothirds of pregnant women reportdyspnea during pregnancy. Although the etiologyof this normal physiologic dyspnea is not clearlydefined, it is most commonly reported in the firstand second trimesters, with improvement towardthe end of the third trimester. These findings supportthe fact that dyspnea is not purely related tomechanical enlargement of the uterus. Markedprogesterone elevations early in pregnancy, resultingin direct brainstem stimulation and hyperventilationand an alteration in the sensitivity tocarbon dioxide, likely contribute to the sensationof dyspnea. The normal anemia of pregnancy mayalso contribute to the sensation of dyspnea. Onemust also be aware of concomitant heart diseasemanifested in pregnant women. Mitral stenosis andother stenotic lesions may become apparent duringpregnancy, particularly in the third trimester whenblood volume is at its maximum, and right-to-leftshunts tend to worsen.Treatment of Specific <strong>Pulmonary</strong>Diseases in the Pregnant PatientAsthma, pulmonary embolism (PE), and infectionssuch as tuberculosis (TB) are can be presentthe pregnant patient. It is important that the pulmonologistbe aware of the management of theseconditions in pregnancy, including the acceptablepharmacologic treatments.US Food and Drug Administration DrugClassificationAcceptable pharmaceutical agents in pregnancyhave been classified by the US Food andDrug Administration. A category A drug meansthat well-controlled drug studies in pregnantwomen have failed to demonstrate any risk to thefetus, and the possibility of fetal harm appearsremote. A category B drug indicates that animaldrug studies have shown that there was no demonstratedfetal risk, but controlled drug studies havenot been performed in pregnant women, or animalreproduction studies have shown an adverse effectthat was not confirmed in studies conduced inpregnant women. A category C agent is a drug forwhich studies in animals have revealed adverseeffects on the fetus, including teratogenicity, andthere are no controlled studies in women; or thatstudies in women and animals are not available.These drugs can be used if the potential benefit ofthe drug outweighs the potential risk to the fetus. Acategory D agent is a drug for which studies haveshown positive evidence of human fetal risks, butin certain situations, these drugs may be of benefitif the risk is outweighed by potential gain. CategoryX agents are those drugs in which animals<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 707ACC-PULM-09-0302-040.indd 7077/10/09 10:26:57 PM


or humans have demonstrated fetal abnormalitiesand/or there is evidence of fetal risks based onhuman experience and clearly the risk outweighsthe benefit of the drug (eg, thalidomide).AsthmaOf the approximate 10 to 12 million asthmaticpatients in the United States, a significant numberare of childbearing age. Asthma affects up to 7%of pregnant women. 1 Poorly controlled asthmacan result in and small but increased risks of pretermlabor and birth, preeclampsia, congenitalanomalies, intrauterine growth retardation, andlow birth weight, as well as increased maternalmorbidity, such as placenta previa. These effectsare likely attributable to hypoxemia, resulting indecreased placental blood flow. General medicalteaching states that asthma gets worse in onethird of patients during pregnancy, onethird ofpatients improve and have fewer frequent asthmaexacerbations, and approximately one third haveno change in the frequency or severity of asthma.The best predictor of asthma severity in pregnancyis the severity of asthma in the nonpregnant stateand the course of asthma in previous pregnancies.Asthma tends to be less severe in the first andlatter third trimesters and worse in the middle ofpregnancy (ie, gestational weeks 17 to 36). Asthmamay also worsen during delivery and in the postpartumperiod. Gastroesophageal reflux, sinusitis,and allergic rhinitis leading to worsening asthmashould be carefully evaluated in the pregnantasthmatic.The treatment of chronic asthma in pregnancyis similar to that of the nongravid asthmatic. In1993, the National Asthma Education Programsponsored by the National Heart, Lung, and BloodInstitute developed guidelines for the treatmentof the pregnant asthmatic patient, subsequentlyrevised in 2004. 2 In 2008, the Global Initiative forAsthma was updated and outlines treatment on aspectrum of control from controlled to partly controlledto uncontrolled asthma. 3 In general, patientswhose asthma is well controlled on an appropriatemedical regimen before pregnancy can remain onthis regimen if it continues to be well controlledduring pregnancy.Short-acting selective β 2-agonists are indicatedfor intermittent (well-controlled) asthma (albuterolhas been the drug most widely studied in pregnancyand is the preferred agent). These agentsare category C. Low-dose inhaled corticosteroidsare the preferred agent for initiating treatment formild persistent asthma (partly controlled) in thepregnant patient. Alternative treatments includeleukotriene receptor antagonists (not zileuton) orlow-dose theophylline.Budesonide (category B) and beclomethasone(category C) are the best-studied inhaled corticosteroids,and budesonide is the preferred agent.Patients with moderate persistent asthma (partlycontrolled) should be treated with low-dose inhaledbudesonide or beclomethasone with a long-actingβ 2-agonist (such as salmeterol). Medium-doseinhaled corticosteroids are another alternativefor the treatment of moderate persistent asthma(partly controlled), and a long-acting β 2-agonist canbe added if needed. Patients already maintainedon other inhaled corticosteroids can be continuedon those agents. Other than budesonide, all otherinhaled corticosteroids are category C agents.Patients maintained on salmeterol for control ofmoderate persistent asthma prior to pregnancycan remain on this agent. Salmeterol has not beenwell studied in pregnancy and is listed as categoryC. Oral theophylline or a leukotriene receptorantagonist can be added if the asthma remainspoorly controlled.Severe persistent asthma (poorly controlled)should be managed with high-dose inhaled corticosteroidsand a long-acting β 2-agonist. Systemic corticosteroids(category C) can also be used for controlof severe and acute exacerbations of asthma. Earlierstudies raised the concern of increased risk of prematurityand low-birth-weight infants with the useof systemic corticosteroids, but this may have beenrelated to the severity of asthma in these patientsrather than the medications used. Other studieshave shown a 0.3% incidence of congenital malformations,an increased incidence of cleft palate,and increased risk of preeclampsia with the use ofsystemic corticosteroids. Patients who are receivingsystemic corticosteroids require careful evaluationfor gestational diabetes. In the steroid-dependentasthmatic, stress-dose corticosteroids should beadministered during labor and delivery.Theophylline (category C) has been used safelyin pregnancy, although clearance in the third trimestermay be reduced, and there is decreased708 Women’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong> (Levine)ACC-PULM-09-0302-040.indd 7087/10/09 10:26:57 PM


protein binding of the drug. There are few dataexamining the use of leukotriene receptor antagonistsduring pregnancy, but montelukast andzafirlukast can be used in patients demonstratinga previous response to these agents. Both arelisted as category B agents. Zileuton (category C)has been associated with teratogenicity in animalsand should be avoided. Terbutaline and otherparenteral β-agonists administered near term cancause tocolytic pulmonary edema and are oftenavoided, although terbutaline is a category B agent.Prostaglandin F 2α used for uterine atony should beavoided in asthmatics because it can cause bronchoconstriction.Omalizumab, an IgE inhibitor, is ratedpregnancy category B, and can be considered forimproved control of poorly controlled asthma.The management of acute asthma and statusasthmaticus is the same as in the nonpregnant asthmatic.For acute asthma, anticholinergic agents arealso permissible (category B). On the basis of thenormal compensated respiratory alkalosis of pregnancy,when evaluating the pregnant asthmatic inthe midst of an acute episode, a Paco 2of 35 mmHg could indicate fatigue and impending ventilatoryfailure. Intubation should be considered witha Paco 2of approximately 42 mm Hg consistentlybecause this relative acidosis places the fetus atrisk. Heliox has been used for the treatment of statusasthmaticus. Epinephrine is not recommendedfor use during pregnancy because of concern foruteroplacental vasoconstriction.Venous Thromboembolism in PregnancyPE is a leading cause of nonobstetric maternalmortality. Some studies have suggested thatPE can account for 20 to 50% of maternal deaths.The spectrum of venous thromboembolism (VTE)can affect 0.05 to 0.1% of pregnancies. Althoughearlier studies suggested an increased incidenceof deep-venous thrombosis (DVT) in the thirdtrimester and postpartum period, subsequentstudies have supported an equal distribution ofDVT throughout pregnancy and suggest that DVTmay be as common or more common antepartumas postpartum.Pregnant women are at increased risk of VTEduring pregnancy (some reports suggest as muchas five times greater) because of a combination ofall three componentsthe Virchow triad. There isincreased venous stasis, hypercoagulability, andendothelial disruption of pelvic and uteroplacentalvessels during delivery. The hypercoagulabilityin pregnancy is caused by alterations in levelsof clotting factors (increases in fibrinogen, factorsI, II, VII, VIII, IX, X, and XII); possibly decreasedfibrinolytic activity; decreases in protein S levels;and a progressive increase throughout pregnancyin activated protein-C resistance. The increase invenous stasis is mechanical as the result of reducedvenous flow in the lower-extremity and pelvic vesselsfrom the compression of the IVC by the graviduterus, and there is an interesting predominanceof left lower-extremity DVT (approximately 90%),likely attributable to greater compression of theleft iliac vein by the right iliac artery. There is alsoa predominance of iliofemoral DVT that is morelikely to embolize. Other risk factors for VTE inpregnancy include prolonged bed rest; cesareandelivery (a twofold risk increase as comparedwith vaginal delivery); preeclampsia; advancedmaternal age; obesity; multiparity; and previousVTE and thrombophilia. Despite the increasedrisks of VTE, an episode of thrombosis duringpregnancy should still prompt a thrombophilicevaluation.The most commonly used and best diagnostictest for determining DVT in pregnancy is Dopplerultrasound, although test results may be impairedand false-positive results can occur. Additionally,problems such as difficulty in detecting iliac, pelvic,and calf thromboses exist. The sensitivity andspecificity of this study may be improved whenperformed in the left lateral decubitus position.Venography ( 50 millirad fetal exposure) is rarelyused for the diagnosis of DVT. The utility of thed-dimer assay in pregnancy is limited becauseelevations may occur normally in pregnancy, butit is likely of some utility in the case of low clinicalsuspicion and in earlier stages of pregnancy.The diagnosis of suspected PE in pregnancyis similar to that in the nonpregnant individual.Ventilation/perfusion scanning can be used (6- to30-millirad fetal exposure); however, a smallerdose of radioisotope is recommended, and whenpossible, the perfusion scan should be performedwithout the ventilation scan to further minimizeradiation exposure. There are limited data evaluatingthe role of CT angiography for the diagnosis ofPE in the pregnant patient, but it appears safe, and<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 709ACC-PULM-09-0302-040.indd 7097/10/09 10:26:58 PM


it is often performed if Doppler study results arenegative. <strong>Pulmonary</strong> angiography should be usedwhen indicated as the “gold standard” test for thediagnosis of PE ( 50 millirad fetal exposure fora brachial approach, and 220 to 375 millirad for afemoral approach).There are several important caveats to treatingVTE during pregnancy, and this topic is includedin the eighth American College of Chest Physiciansevidenced-based clinical practice guidelineson anticoagulation. 4 Heparin is the anticoagulantagent of choice for both treatment and prophylaxisduring pregnancy because it does not crossthe placenta, and low-molecular-weight heparin(LMWH; category B) or unfractionated heparin(UFH; category C) can be safely used (level ofevidence Grade 1A). LMWH is the preferred agentwith (Grade 2C). For treatment of VTE, LMWHat weight-adjusted doses or UFH IV for 5 days,followed by adjusted-dose UFH subcutaneouslyto maintain the activated partial thromboplastintime (APTT) in the therapeutic range or adjustedLMWH is recommended (Grade 1A).Some authors suggest that anti-Xa levelsshould be used to monitor LMWH therapy and insome cases of full-dose heparin anticoagulation,rather than APTT because of alterations in fibrinogenand Factor VIII that can make APTT resultsless reliable. High doses of UFH may be required.Anticoagulation should be continued throughoutpregnancy (grade 1B) and 6 weeks postpartum tocomplete a total of 6 months of anticoagulation(grade 2C). Warfarin can be safely used in the postpartumperiod. In some unique situations, such asprophylaxis for mechanical heart valves, warfarincan be used after the 13th week of pregnancy untilthe mid-third trimester without complications.There is a paucity of data on the use of the pentasaccharideagents in pregnancy.One should be aware of the significant complicationswith prolonged heparin use, includingheparin-induced thrombocytopenia. Osteoporosiscan also be a complication in patients receivinglong-term therapy with heparin. LMWH may beassociated with a lower incidence of osteoporosisand heparin-induced thrombocytopenia. Heparinoidscan be used in cases of heparin-inducedthrombocytopenia.Warfarin (category X) crosses the placentalbarrier, particularly between 6 to 10 gestationalweeks, and is associated with poor fetal epiphysealcartilage formation, chondrodysplasia, and fetalnasal hypoplasia. CNS abnormalities have alsobeen described, including malformations, opticatrophy, and microhemorrhages with warfarin useat any time during pregnancy. The use of warfarinis permissible during lactation. After an episodeof VTE, anticoagulation should be continued withwarfarin for 4 to 6 weeks after delivery or to complete3 months of anticoagulation.Thrombolytic agents are relatively contraindicatedduring pregnancy, especially near term andwithin 2 weeks postpartum because of the risk ofhemorrhage. However, published case reports ofsuccessful use exist. IVC filter placement can beperformed, although suprarenal placement is recommendedbecause the left ovarian vein emptiesinto the left renal vein.The management of labor and delivery in thefully anticoagulated patient can be difficult. Epiduralanesthesia should be used with caution in fullyanticoagulated patients because epidural hematomasmay result. Most patients receiving LMWHor UFH should have anticoagulation stopped24 h before elective induction (grade 1C). Patientsat high risk for recurrent VTE should be switchedto UFH at least 24 h before expected induced deliveryto safely discontinue anticoagulation 6 to 8 hbefore the expected delivery.The indications for prophylaxis for VTE inpregnancy are the same as those in the nonpregnantindividual, with some of the followingspecific recommendations. 4 Patients with a historyof a single previous VTE related to a cleartransient risk factor, but not with additional riskfactors other than pregnancy currently, shouldundergo close surveillance during pregnancy andprophylactic anticoagulation in the 4- to 6-weekpostpartum period (grade 1C). Patients with a historyof idiopathic VTE should receive prophylaxiswith LMWH or UFH or close surveillance duringpregnancy and anticoagulation in the 4- to 6-weekpostpartum period (grade 1C). Patients with anytype of thrombophilia are at increased risk for VTEin pregnancy, but patients with antithrombin deficiencyare particularly at risk for VTE and shouldreceive antepartum prophylaxis and postpartumprophylactic anticoagulation (grade 2C).Most authors, and the guidelines, suggest thatpatients with a history of VTE during a previous710 Women’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong> (Levine)ACC-PULM-09-0302-040.indd 7107/10/09 10:26:58 PM


pregnancy should undergo surveillance or receiveprophylaxis for VTE in subsequent pregnanciesin addition to postpartum anticoagulation prophylaxis(grade 2C) because the recurrence ratecan be as high as 12%. Patients with a history oftwo or more episodes of VTE should receive anticoagulationduring pregnancy and be convertedto long-term anticoagulation postpartum (grade2C). Patients requiring long-term warfarin anticoagulationbefore pregnancy should be convertedto UFH or LMWH when pregnancy is planned, oras early in pregnancy as possible if not planned.The article by Bates et al 4 contains specific recommendationsfor unique situations, such as patientswith mechanical heart valves.LMWH heparin or UFH are the agents of choicefor prophylaxis, used at prophylactic doses of prophylacticUFH (5,000 U subcutaneously q12h), orintermediate-dose UFH (subcutaneously q12h afteranti-Xa levels), or prophylactic LMWH (subcutaneouslyq24h following peak anti-Xa levels). LMWHis the preferred prophylactic agent (grade 2C).TB in PregnancyWith increases in the rates of TB in the UnitedStates, largely resulting from increases in HIVinfection and increases in foreign-born immigrants,there has been an increase in TB cases in childbearingadults and children. The actual incidence ofTB in pregnancy is not completely clear but canbe in the range of 1% in poverty-stricken innercityareas with poor access to prenatal health care.The prognosis of TB in pregnant women has beendebated. At one time, it was thought that TB hadan increased incidence of dissemination in thepregnant woman. Most data support that thereis no difference in the susceptibility to infection,course of disease, obstetrical outcome, prognosis,and incidence of TB in nonpregnant or pregnantwomen unless immunosuppression coexists.Treatment of active TB in pregnancy is similarto that of treatment in the nonpregnant individualwith a few caveats. 5 Drugs approved in pregnantpatients include isoniazid (INH), rifampin, andethambutol. These drugs all cross the placenta buthave not been shown to have teratogenic effects.In general, pyrazinamide, streptomycin, and ethionamideshould be avoided in pregnancy. Thereare situations in which treatment with these otherdrugs may outweigh the potential risks of usingthese agents. The teratogenicity of streptomycinis primarily fetal ototoxicity as the result of nervedamage, congenital auditory malformations, and/or congenital deafness. Pyrazinamide has beenavoided traditionally in the United States becauseit has not been well studied in pregnancy, but it isused routinely by the World Health Organizationand the International Union Against Tuberculosisand Lung Disease.Thus, standard treatment for active TB shouldinclude INH, rifampin, and ethambutol for 9months because the regimen does not containpyrazinamide. These same drugs can be continuedduring lactation without compromise to theinfant. It the patient has infection with a potentiallydrug-resistant organism, then the addition of pyrazinamideshould be considered pending resultsof susceptibility testing. Patients with INH-resistantTB can be treated rifampin and ethambutol.Therapeutic abortion may have to be consideredin patients with multidrug-resistant TB. It is astandard recommendation that pyridoxine alwaysbe administered with INH during pregnancy todecrease the incidence of INH neurotoxicity (opticand peripheral neuropathy).Chemoprophylaxis for latent TB infection(LTBI) is somewhat more controversial. 6 Because itappears that pregnancy does not increase the riskfor active TB in patients with LTBI, many health-careworkers would delay prophylactic treatment untilafter delivery in an HIV-negative patient with clearchest radiograph findings who is not a recent converter,and who is not in another high-risk group.However, most authorities recommend 9 months ofINH prophylaxis if the patient is in a high-risk LTBIgroup. Routine purified protein derivative (PPD)testing during pregnancy has fallen out of favorat most medical centers and is not used unlessin an endemic TB area or unless the patient isin a high-risk group, such as those with HIVinfection.Children 5 years of age with LTBI are athigh risk of progression to disease and/or moresevere and/or disseminated disease; they require9 months of INH prophylaxis. If a child is exposedto a mother with active TB, he or she should beseparated from the mother (only until the motheris deemed not contagious), undergo PPD skintesting and a chest radiograph, and be placed on<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 711ACC-PULM-09-0302-040.indd 7117/10/09 10:26:59 PM


INH therapy for 8 to 12 weeks even if the skin testresult is negative. At that time, the PPD shouldbe repeated. If the PPD result remains negative,INH can be discontinued; if results are positive,9 months of INH prophylaxis should be administered.Pneumonia in PregnancyIn general, the incidence of bacterial pneumoniain the pregnant woman is similar to that in thenonpregnant woman. Pneumonia may complicate1.2 to 2.7 per 1,000 deliveries and may be morefrequent in inner cities. Preterm labor, pretermdelivery, and respiratory failure can result. TheMaternal mortality rate can be as great as 4%.Several small epidemiologic studies examiningthe bacterial organisms responsible for community-acquiredpneumonia in the pregnant patientshow the spectrum to be similar to that in the nonpregnantwoman, and similar empiric therapiescan be used. The cephalosporins and penicillinscan be safely used in pregnancy (all are category B).Quinolones are category C agents. The macrolidesare category B and C agents. The tetracyclines arecategory D agents and should be avoided. Thesulfonamide drugs are category C agents and areusually reserved for infections such as Pneumocystispneumonia.A few specific organisms responsible for pneumoniain pregnancy deserve mention. In pregnancy,cell-mediated immunity may be altered, andthus, infections with fungal and viral organismscan be more severe and life threatening. The riskof dissemination with coccidioidomycosis, as wellas mortality, is likely increased during pregnancy(20% vs 0.2% in the nonpregnant individual),although this has not been borne out consistently inall studies. 7 It is generally thought that disseminationis more likely if the infection is acquired in thethird trimester. Treatment includes amphotericin ora liposomal derivative. Both agents are category Bagents. The majority of the azoles are classified ascategory C.Of the viral agents, Varicella zoster is the mostfeared during pregnancy. Although this illness isusually self-limited and benign in children, in thenonexposed adult pregnant patient, the mortalityrate associated with Varicella pneumonia mayapproach 35%. Infection is most severe in the thirdtrimester. Patients present with fever and rashand, and pneumonia can develop rapidly. A typicalchest radiograph shows miliary and nodularinfiltrates, usually resolving within 14 days. Theend radiographic result of such pneumonia is oftencalcified, but physiologically insignificant, nodules.Acyclovir (category B) can be safely used duringpregnancy and should be begun with the first signof Varicella infection.Influenza virus can also be more severe duringpregnancy, although some data suggest that therate of mortality associated with influenza duringpregnancy may be similar to that in the nonpregnantwoman. The antiinfluenza drugs are categoryC agents. Influenza vaccine is made from inactivatedvirus. Women at high risk should be immunizedfor influenza during pregnancy regardless ofthe stage of pregnancy. All other pregnant womenshould be vaccinated after the first trimester.Acute Respiratory Failure in PregnancyHemodynamicsAs pregnancy progresses, cardiac output canbecome positionally dependent because of thegravid uterus potentially obstructing the IVC andreducing venous return. This is particularly true inthe third trimester and is most notable in the supineposition. Severe hypotension can result. This effectis reduced in the full or partial left lateral decubitusposition, allowing the uterus to be displaced fromthe IVC. This can be an important factor in theresuscitation of the hypotensive pregnant patient.Because of low colloid oncotic pressure, invasivemonitoring with pulmonary artery pressures mayhave to be interpreted with caution, and pulmonaryedema can result at a lower pulmonary arteryocclusion pressure.SepsisIn addition to non—obstetric-related causesof sepsis in the pregnant patient, sepsis can resultfrom endometritis, pelvic thrombophlebitis, andseptic abortion and from procedures such asamniocentesis and/or infection of cesarean orepisiotomy incisions. It is important to recall thatbaseline hemodynamic parameters in a nonsepticpregnant individual, ie, high cardiac output and712 Women’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong> (Levine)ACC-PULM-09-0302-040.indd 7127/10/09 10:26:59 PM


low systemic vascular resistance, can be confusedwith the hemodynamics of sepsis.Mechanical VentilationThe principles of mechanical ventilation in thepregnant patient are similar to those in the nonpregnantwoman. Intubation may be more difficultbecause of edema of the upper airway, a reducedairway caliber, and an increased risk for aspirationand bleeding; therefore, smaller endotrachealtubes may be required. Tidal volumes may haveto be reduced as the result of reduced chest wallcompliance from the gravid uterus, and greaterpeak pressures may be required to overcome chestwall stiffness. Gas exchange goals should be tomaintain Paco 2in the pregnant eucapnic range of28 to 32 mm Hg with a Pao 2 90 mm Hg to preventfetal hypoxemia. A further reduction in Paco 2can lead to reduced uterine blood flow and fetalhypoxemia. The fetus is very sensitive to hypoxemia,and attempts to compensate for maternalhypoxia by divergence of maternal blood flow toessential tissues, a leftward shift in the oxyhemoglobindissociation curve, and high avidity of fetalhemoglobin for oxygen. Noninvasive mechanicalventilation has not been well studied in pregnancy,but its utility may be limited by the increased riskof aspiration and narrow airway caliber in thispopulation.Amniotic Fluid EmbolismAmniotic fluid embolism (AFE) is a rare butcatastrophic complication of pregnancy. Theincidence of AFE ranges between 1 in 8,000 and1 in 80,000 pregnancies, with an associated 80 to90% mortality rate responsible for 10 to 20% ofmaternal deaths in the peripartum period. Riskfactors for AFE include advanced maternal age,multiparity, premature rupture of membranes,and meconium staining of amniotic fluid. The useof uterine stimulants and tumultuous labor mayalso be risk factors. The risk of AFE extends 48 hinto the immediate postpartum period and has alsobeen reported to develop during abortions andplacental abruption. The pathophysiology of AFEis not completely known, but postulated mechanismsinclude true mechanical obstruction of thepulmonary vasculature with fetal squamous cellsand other debris, alveolar capillary leak caused bya release of mediators during delivery, pulmonaryedema caused by left ventricular failure, and ananaphylactic reaction to exposure of the motherto fetal antigens.AFE presents catastrophically with the acuteonset of tachypnea, dyspnea, tachycardia, cyanosis,hypotension, hypoxemia likely caused by ventilation/perfusionabnormalities, and hemodynamiccollapse. Seizures can occur. Disseminated intravascularcoagulation can develop in 40 to 80% ofpatients, and hemorrhage can be the initial presentation.In the majority (70%) of those patients whosurvive the initial event, ARDS will develop. AFEis often followed by some transient left ventriculardysfunction, as supported by studies using pulmonaryartery catheters.The diagnoses of AFE is one of exclusion butcan be supported in the appropriate clinical settingwith the presence of fetal squamous cells andlanugo hairs in the maternal circulation, althoughthese can also be present under normal conditionsand are not pathognomonic for this diagnosis.Treatment is supportive with intubation, mechanicalventilation, vasopressors, sedation, neuromuscularblockade, and pulmonary artery catheterplacement. Factor replacement may be requiredfor hemorrhage.Venous Air EmbolismVenous air embolism can occur during normaldelivery, with placenta previa, and during abortion.It has also been reported during oral genitalsex and during gynecologic procedures using airinsufflation. One percent of maternal deaths arethought to be from venous air embolism. The usualsites of air entry are at the subplacental venoussinuses, during the antepartum or peripartumperiod, followed by entry into the venous circulation.In the right ventricle, air can obstruct the pulmonaryblood flow, resulting in cardiopulmonarycollapse. In general, it is thought that 100 mL of aircan lead to mortality. Recruitment and activationof neutrophils, protein aggregation at the turbulentair blood interface and obstruction of pulmonaryarterial vessels by microemboli also contribute tothe pathophysiology of venous air embolism.Patients with venous air embolism presentwith profound hypotension, as well as nonspecific<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 713ACC-PULM-09-0302-040.indd 7137/10/09 10:27:00 PM


signs of coughing, dizziness, tachypnea, dyspnea,tachycardia, and diaphoresis. Respiratory arrestsoon follows, and the rate of mortality can increaseto 90%. The classic but rarely heard cardiac millwheelmurmur audible over the precordium is supportiveof the diagnosis of venous air embolism.ARDS may develop. Other findings include mentalstatus changes, coma, seizures, stroke, myocardialinfarction, and thrombocytopenia. Bubbles may bevisualized in the retinal arterioles, and subdermalair may be present. Air in the heart or great vesselsis occasionally seen on the chest radiograph.Treatment includes recognition of the syndrome,followed by placing the patient in the leftlateral decubitus position so that the air bubble isremoved from the entrance to the right ventricularoutflow tract. Cases of aspiration of air from theright heart using a pulmonary artery or centralvenous catheter have been reported. Patientsshould receive ventilation with 100% oxygen tofacilitate removal of nitrogen, which comprisesa significant (up to 80%) of gas content in theembolus. The use of hyperbaric oxygen therapyshould be considered. There are anecdotal reportsof using heparin to treat microemboli and corticosteroidsto decrease pulmonary edema in thissyndrome.Tocolytic <strong>Pulmonary</strong> EdemaUntil recently, β-adrenergic agents were widelyused in obstetrics for inhibition of preterm labor,often administered in combination with corticosteroidsto promote fetal lung development. Themost common agents used were β 2-selective agentssuch as terbutaline, ritodrine, and isoxsuprine;and tocolytic pulmonary edema developed in asmany as 4 to 5% of patients receiving these agents.Currently, many obstetricians use magnesium fortreatment of preterm labor, which has resulted ina decrease in this entity.Usually, symptoms of tocolytic pulmonaryedema develop within 24 h but occur more commonly48 h after initiation of therapy. They canalso develop within 24 h after discontinuation ofthe drug. Those patients who receive prolongedtocolytic therapy with concomitant infusions ofcrystalloid volume, those with multiple gestations,and those with preeclampsia are more at risk fortocolytic pulmonary edema. The mechanisms oftocolytic pulmonary edema include possible fluidoverload, direct cardiac toxicity, alterations, andreductions in colloid oncotic pressure and/orincreased pulmonary capillary permeability.Patients with tocolytic pulmonary edema typicallypresent with dyspnea, tachycardia, tachypnea,chest pain, crackles, and the presence of pulmonaryedema on chest radiograph. This syndromereverses quickly, usually 12 to 24 h after recognitionand discontinuation of the offending agent. Theprognosis is excellent. Transient use of oxygen anddiuretics may be needed.AspirationAspiration historically has been a significantproblem in obstetrics and is estimated to accountfor 2% of maternal mortality in the United States.The classic description was made by Mendelson (in1946), who described large volumes of gastric contentsentering the tracheobronchial tree in womenundergoing labor and delivery. Because of thislarge volume of aspiration of low pH-containinggastric contents, ARDS and chemical pneumonitissubsequently developed. Immediate asphyxia wasalso described.The obstetric patient is at risk for aspirationfor many reasons, including progesterone-inducedrelaxation of lower esophageal sphincter tone,an increase in intragastric pressure caused bymechanical compression by the gravid uterus, aswell as by frequent examinations, a decrease in gastricemptying during parturition, and being in thesupine position. In some cases, alterations in mentalstatus caused by sedation and a reduction in vocalchord closure possibly related to analgesia usedduring labor may also contribute to an increasedrisk of aspiration.There is a correlation between the volume ofgastric contents aspirated, the acidity of the aspirate,the presence of particulate matter, the bacterialload, and the host resistance on the progressionand severity of clinical symptoms. It is thoughtthat the low pH ( 2.5) of the aspirate is the majorinciting pathogenic process for disease because ofa chemical pneumonitis, although large volumes,particularly those containing food particles, can beclinically significant even with greater pH levels.ARDS can result. A small subgroup of patients willhave immediate respiratory arrest and death after714 Women’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong> (Levine)ACC-PULM-09-0302-040.indd 7147/10/09 10:27:01 PM


aspiration as the result of uncorrectable hypoxemia.In those cases in which small volumes of gastriccontents are aspirated, symptoms may be delayeduntil 6 to 24 h after the event. Bacterial pneumoniacan develop 24 to 72 h after the aspiration.Treatment is supportive. As in the nonpregnantpatient, there is no role for prophylactic antibioticsor corticosteroids when treating this aspirationsyndrome. Resolution usually occurs over thenext 4 to 5 days unless secondary superinfectiondevelops. Bronchoscopy may be indicated whenwitnessed aspiration with large food particles hasoccurred. The chances of aspiration can be reducedby the use of regional anesthesia, restricting oralintake at the time of delivery and cricoid pressureif endotracheal intubation is required.ARDSARDS is defined similarly in the pregnant asin the nonpregnant individual. In addition to thecauses of ARDS discussed elsewhere in this course,there is a long list of ARDS etiologies unique to orassociated with pregnancy. These include placentalabruption, air embolism, amniotic fluid embolism,aspiration, eclampsia, septic abortion, and the deadfetus syndrome.<strong>Pulmonary</strong> Edema<strong>Pulmonary</strong> edema can accompany preeclampsia/eclampsiain approximately 3% of cases. <strong>Pulmonary</strong>edema may develop more frequently in theimmediate postpartum period. Risk factors includeadvanced age and multigravity. Mechanisms forpreeclampsia-related pulmonary edema includeincreased left ventricular afterload, myocardialdysfunction, the alterations in colloid oncoticpressure discussed earlier, as well as fluid overload.There may also be a component of increasedpulmonary capillary permeability. Management isapproached in the standard manner with oxygen,diuresis, control of hypertension, and mechanicalventilation, if required. Invasive hemodynamicmonitoring may be required.Peripartum CardiomyopathyPregnancy-related cardiomyopathy can developin 1 in 1,300 to 1 in 4,000 deliveries, and usuallypresents in the third trimester or up to 6 monthspostpartum. Risk factors include advanced age,multiple gestations, preeclampsia, and African-American race. Patients typically present withdyspnea, orthopnea, peripheral edema, pulmonaryedema, tachycardia, and a cardiac gallop. The chestradiograph shows cardiomegaly and pulmonaryedema, and echocardiography demonstratesglobal hypokinesis. Prognosis is variable, andapproximately 30% of these patients recover, 30%may have residual cardiac damage, and 30% mayrequire heart transplantation. The cause of deathis often thromboembolism from left ventricularthrombus. The recurrence of cardiomyopathy withsubsequent pregnancies is common.Primary <strong>Pulmonary</strong> HypertensionPatients with pulmonary hypertension, eithersecondary or primary, are at increased risk formaternal (35 to 50%) and fetal mortality duringpregnancy. The risk appears to be greatest inthe immediate peripartum period, when a largeamount of blood volume is “autotransfused”from the uteroplacental bed back to the maternalcirculation. Acute right-heart failure can result.Patients with primary pulmonary hypertensionshould be counseled against becoming pregnantand encouraged to seek termination of pregnancyand permanent forms of birth control. Thosepatients wishing to continue with their pregnancyshould be managed with medications such asinhaled nitric oxide and/or IV or inhaled prostacyclinduring the peripartum period. Placementof a pulmonary artery catheter should be stronglyconsidered during delivery. Long-term managementcould include epoprostenol (prostacyclin,a category B agent) or sildenafil (category B), butthe oral endothelin receptor blocker agents suchas bosentan are contraindicated as the result ofteratogenicity (category X).Tobacco and Lung Disease in WomenTobaccoAt the end of the 19th century and beginning ofthe 20th century, cigarette smoking and tobacco usewere limited primarily to male subjects and werethought to be socially unacceptable for women.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 715ACC-PULM-09-0302-040.indd 7157/10/09 10:27:01 PM


However, in the early 20th century, with the adventof the suffrage movement, numerous ad campaignspromoted the use of tobacco by women, includingwell-known advertisements with the famousslogan “You’ve come a long way, baby.” Tobaccosmoking began to be associated with the educatedand/or liberated American woman, a concept furtherpromoted by the advertising of the tobaccoindustry. The use of tobacco products by womenincreased dramatically after World War II andpeaked in 1965, when the prevalence of smokingamong women approached 35%. Currently, 18.1%of women and 23.9% of men smoke.The initial 1964 landmark US Public Health SurgeonGeneral’s Report on Smoking and Health primarilyaddressed smoking and respiratory healthin men. However, in 1980, the Surgeon General’sReport on the Health Consequences for Smokingfor Women was released. This report concludedthat women demonstrate the same dose—responserelationships with cigarette smoking as do men,and that the risk of lung cancer increases with theincreasing number of cigarettes smoked per day,earlier ages of cigarette smoking, longer durationof smoking, greater tar and nicotine content ofcigarettes, and inhalation of cigarette smoke.The study predicted that rates of mortalityin women from lung cancer may be comparablewith that experienced by men, that there wouldbe a rapid increase in lung cancer rates in womensimilar to that seen in men 25 years previously,and that lung cancer death rates in women wouldsurpass those of breast cancer by the early 1980s.The study also found that the incidence of otherrespiratory conditions such as influenza was 20%greater in ever-smoking women than in nonsmokingwomen, that there was an increase in deathrates attributable to COPD in smoking womencompared with nonsmoking women, and that thisdeath rate correlated with the number of cigarettessmoked. Finally, the study concluded that womensmokers had worse pulmonary function thanex-smokers or never-smokers and that the severityof decrease in pulmonary function was dose relatedto the number of cigarettes smoked.The follow-up 2001 Surgeon General Report onWomen and Smoking 8 found the following: (1) in1998, 22% of women smoked cigarettes; (2) lung canceris the leading cause of cancer deaths in women,and 90% of these deaths are related to smoking; and(3) in 2000, 25% of cancer deaths in women werecaused by cancer of the lung. The study also statedthat the risk of lung cancer increases with quantity,duration, and intensity of smoking.Lung CancerLung cancer has become the leading causeof cancer deaths in both men and women in theUnited States, surpassing death rates in womenas the result of breast cancer in 1987. More than80 to 90% of lung cancers in women are related totobacco use and, thus, are preventable. In 2007, 26%of cancer deaths in women were caused by cancerof the lung. Although breast cancer affects a largenumber of women per year, lung cancer has a significantlyworse prognosis, with greater mortalityrates. In 2007, 71,000 women died of cancer of thelung and bronchus as compared with 40,480 deathscaused by breast cancer. In 2007, 114,600 womenwere diagnosed with lung cancer. Lung cancerincidence rates in women began to level off fromthe mid-1990s to early 2000 and are now decreasing;however, death rates in women from lung cancercontinue to increase and are just now reaching theplateau seen in men in the 1980s. It is expected thatthe death rates from lung cancer in men and womenwill approach unity in the next 5 to 10 years.Of the different cell types, adenocarcinoma isthe most common cell type of lung cancer in bothsmoking and nonsmoking women, regardless ofage and is the most common cell type in youngpeople and in nonsmokers of both sexes. A possibleincrease in the susceptibility of tobacco carcinogensin women has been a highly debated topic. Somestudies have indicated that there is an increasein the relative risk for lung cancer developing insmoking women than in men; other studies havenot been able to support this. Yet other studieshave shown that there is an increased odds ratioof tobacco-related lung cancer in women only forsome certain types of lung cancer. For example, itappears that smoking women have a greater oddsratio for small cell carcinoma developing than domen. There are many postulated reasons as to whywomen may have an increased susceptibility totobacco carcinogens, including the fact that theremay be an increased frequency of mutations in theP53 and other tumor suppressor genes in womenthan in men; in addition, a greater promutagenicity716 Women’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong> (Levine)ACC-PULM-09-0302-040.indd 7167/10/09 10:27:02 PM


DNA level (CP450), greater levels of DNA adducts,and decreased capacity for DNA repair have beenobserved in women.In addition, hormones (estrogens) and hormonalreplacement may play a role in the variablesusceptibility to tobacco carcinogens, particularlywith adenocarcinoma. In never-smokers, ageadjusted incidence rate is greater for women 14.4to 20.8/100,000 person years as compared withmen 4.8 to 13.7/100,000. Other risk factors for lungcancer in women include a family history of lungcancer; occupational exposures to compounds suchas asbestos, cadmium, beryllium, silicosis, radon;and previous lung disease as are found in men. Ingeneral, women with lung cancer have a better prognosisand equal or better survival with treatmentthan do men, regardless of cell type and stage.COPDCigarette smoking is the major cause of COPDin women, and the risk increases with the amountand duration smoked. Ninety percent of COPDmortality among women in the United States isattributable to smoking. Mortality rates attributableto COPD among men now appear to be havereached a plateau, whereas COPD mortality ratesamong women continue to increase. A total of 1.5to 3.6 times more women are hospitalized withCOPD than are men. The current prevalence ofCOPD in women is greater than or equal to thatin men, and this may be underestimated becauseseveral studies have suggested that COPD is lessfrequently diagnosed in women.This finding parallels the trends previouslydiscussed in lung cancer. Because the slope of thiscurve is steeper than that seen in men, it is possiblethat women are more vulnerable to tobaccoinducedCOPD. Studies in these regards havebeen conflicting, but most indicate that womenwith exposure to tobacco smoke have more severeCOPD and more lung function impairment withfewer pack-years of tobacco use than do men. Itdoes seem clear that COPD develops in women atan earlier age than it does men. Some of the postulatedmechanisms for this increased prevalenceinclude small air airway size, which alters thedistribution of toxins contained in tobacco, hormonalmechanisms, and variations in cytochromeP450 levels.In children who have prenatal exposure toenvironmental tobacco smoke from a smokingmother, studies have shown that fetal lung developmentis affected. These children have moreairway obstruction, increased airway hyperresponsiveness,and alterations in lung maturationand lung growth. There is a small decrement inbirth weight and increased risk of intrauterinegrowth retardation and other obstetric complications.Those children exposed to environmentaltobacco smoke in the postnatal period have anincreased incidence of cough, wheezes, respiratoryillnesses, and infection. <strong>Pulmonary</strong> functionis decreased slightly, and there is an increase in airwayresponsiveness. These children tend to havean increase in childhood asthma, earlier developmentof asthma, and more severe asthma. Otherstudies have shown that atopy tends to developin children exposed to environmental tobaccosmoke, which can result in worsening asthma.There have also been correlations found betweenenvironmental tobacco smoke and obstructiveapnea in children and sudden infant death syndromein infants.Catamenial Pneumothorax andHemoptysisCatamenial complications by definitiondevelop during menstruation. Catamenial pneumothoraxoccurs very rarely and is usually recurrent.Patients are usually in their late 20s to 30years of age when they initially present. Symptomsdevelop within 24 to 48 h of the onset of menstrualflow. The pneumothoraces are often on the rightside and are often associated with pelvic endometriosis.The mechanisms of air entry into the pleuralspace may be to the result of a defect at the site ofpleural or diaphragmatic endometriosis or may beby air gaining access to the peritoneal cavity duringmenstruation and then subsequently enteringthe pleural cavity through a diaphragmatic defect.The diagnosis is fairly straightforward when apneumothorax develops during the first 48 h ofmenstrual flow. Treatment includes ovulationsuppressingdrugs. Patients wishing to conceiveor who do not want ovulation suppressed shouldundergo thoracotomy with repair of diaphragmaticdefects, if present, followed by pleurodesis. Catamenialhemothorax has also been described.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 717ACC-PULM-09-0302-040.indd 7177/10/09 10:27:02 PM


Catamenial hemoptysis is thought to occuras the result of endometriosis in the lung parenchyma.It is usually a cause of scant hemoptysis butcan, in some cases, be severe. Treatment includeshormonal suppression and/or resection of lungparenchyma involved with endometriosis.Hormone-Replacement Therapy andthe Risk of Venous ThromboembolicDiseaseThe cardiovascular disease risks of hormonereplacementtherapy (HRT) in women have beenaddressed by several large studies. 9,10 In theWomen’s Health Initiative Study, the risk of VTE(both DVT and PE) was significantly increased inthose women receiving HRT. There were 34 VTEevents per 10,000 person-years in the HRT groupvs 16 VTE events/10,000 person-years in the non-HRT group. The hazard ratio was 2.11 (confidenceinterval, 1.58 to 2.82). In the Heart and Estrogen/Progestin Replacement Study, there was a twofoldincrease in VTE risk in the women receiving HRT.One of the possible explanations for these findingsmay be an increase in activated protein C resistancewith the use of HRT. The results at longer followup,at 3 years after discontinuation of HRT, suggestthat cardiovascular risks return to a level comparablewith those patients who did not receive HRT,but an increased incidence of malignancy is beingfound in those women who had received HRT. 11LAMLAM is covered extensively in other portionsof this syllabus but will be reviewed briefly here.LAM is a rare disease (prevalence 1 to 2/million)affecting premenopausal women with a meanage of 35 years. The disorder is characterized byproliferation of atypical smooth muscle in thebronchovasculature, lymphatics, and interstitiumof the lung, as well as the abdomen and pelvis.Typically, these smooth-muscle cells stain positivewith the monoclonal antibody HMB-45 (a melanoma-relatedtumor marker).Clinically, patients present with symptoms ofdyspnea, cough, chest pain, and hemoptysis. Thephysical examination may reveal decreased breathsounds, crackles, and sometimes ascites and/orabdominal masses. Spontaneous pneumothoracesand/or chylothoraces may also be presenting findings.Radiographically, LAM is characterized bycyst formation and chronic interstitial infiltrates.Hyperinflation develops as the disease progresses.Pleural effusions and pneumothoraces may alsobe present. High-resolution CT can reveal diffuse,homogenous cystic disease with cysts rangingfrom a few millimeters to ≥ 1 cm in size (6-cm cystshave been reported); characteristically, nodularity,as seen in eosinophilic granuloma, is absent.<strong>Pulmonary</strong> function testing most commonly showsmixed obstructive and restrictive physiology withan elevated total lung capacity and residual volumeand a reduced diffusion capacity, but up toone third of patients can have normal pulmonaryfunction.Complications of LAM include recurrent pneumothoracesin 50% of patients, chylous effusionsin approximately onethird of patients, chylousascites, and renal angiomyolipomas in up to 50%of patients. These latter tumors can develop in thekidneys or in other solid organs (uterus, ovaries,liver, spleen) and are composed of blood vessels,smooth muscle, and fat.The diagnosis of LAM can be suspected bythe clinical presentation and characteristic radiographicappearance and confirmed by lung biopsywith HMB-45 staining, either transbronchial, thoracoscopic,or surgical. The major entities in thedifferential diagnosis are other causes of cystic lungdisease with preserved lung volume, particularlyeosinophilic granuloma and cystic sarcoidosis.The prognosis with LAM is variable, but mediansurvival is often reported to be 8 to 10 years. Thecause of death is usually respiratory failure. Treatmentfor LAM has included hormonal manipulationwith the use of antiestrogen agents (such as tamoxifen),progesterone agents, luteinizing hormonereleasinghormone analogs, and/or oophorectomywith variable results. Patients should avoid estrogensand pregnancy. Lung transplantation has beenperformed for LAM, although disease recurrencehas been documented.Recently, sirolimus, acting via suppression ofmammalian target of rapamycin signaling, hasbeen studied with a primary end point of decreasingangiomyolipoma volume at 12 months. 12 Inthis small study, there was a 50% reduction inangiomyolipoma size in patients receiving therapywith sirolimus, but the tumor size increased after718 Women’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong> (Levine)ACC-PULM-09-0302-040.indd 7187/10/09 10:27:03 PM


discontinuation of treatment. An interesting secondaryfinding in the study was an improvementin spirometry of 5 to 10% in FEV 1and FVC, whichwas only partially sustained after discontinuationof sirolimus.References1. Schatz M. The efficacy and safety of asthma medicationsduring pregnancy. Semin Perinatol 2001;25:145–1522. National Asthma Education and Prevention Program.Report of the Working Group on Asthmaand Pregnancy: management of asthma duringpregnancy; update 2004. Bethesda, MD; NationalInstitutes of Health, 2005; 1–57; NIH publication05–32793. GINA Report: Global Strategy for Asthma Managementand Prevention. Available at: http://www.ginasthma.com/Guidelineitem.asp??l1=2&l2=1&intId=60.Accessed February 16, 20094. Bates SM, Greer I, Pabinger I, et al. Venous thromboembolism,thrombophilia, antithrombotic therapy,and pregnancy: American College of Chest PhysiciansEvidence-based clinical practice guidelines(8th edition). Chest 2008; 133:844S–886S5. American Thoracic Society/Centers for DiseaseControl and Prevention/Infectious Diseases Societyof America. Treatment of tuberculosis. Am J RespirCrit Care Med 2003; 167:603–6626. ATS Official Statement. Targeted tuberculin testingand treatment of latent tuberculosis infection. Am JRespir Crit Care Med 2000; 161:S221–S2477. Ramsey PS, Ramin KD. Pneumonia in pregnancy.Obstet Gynecol Clin North Am 2001; 28:553–5698. Surgeon General’s report: women and smoking; a reportof the Surgeon General, 2001. Available at: www.cdc.gov/tobacco/sgr. Accessed February 16, 20099. Rossouw JE, Anderson GL, Prentice RL, et al. Risksand benefits of estrogen plus progestin in healthypostmenopausal women: principal results from theWomen’s Health Initiative randomized controlledtrial. JAMA 2002; 288:321–33310. Hulley S, Furberg C, Barrett-Connor E, et al. Noncardiovasculardisease outcomes during 6.8 yearsof hormone therapy: heart and Estrogen/ProgestinReplacement Study follow-up (HERS II). JAMA2002; 288:58–6611. Heiss G, Wallace R, Anderson GL, et al. Health risksand benefits 3 years after stopping randomizedtreatment with estrogen and progestin. JAMA 2008;299:1036–104512. Bissler JJ, McCormack FX, Young LR, et al. Sirolimusfor angiomyolipoma in tuberous sclerosis complexor lymphangioleiomyomatosis. N Engl J Med 2008;358:140–51Annotated BibliographyAndres RL, Miles A. Venous thromboembolism and pregnancy.Obstet Gynecol Clin North Am 2001; 28:613–630This is a review of this subject.American Thoracic Society/Centers for Disease Controland Prevention/Infectious Diseases Society of America.Treatment of tuberculosis. Am J Respir Crit Care Med2003; 167:603–662This statement includes a section on the treatment of TB inpregnancy.ATS official statement. Targeted tuberculin testing andtreatment of latent tuberculosis infection. Am J RespirCrit Care Med 2000; 161:S221–S247This statement includes a section on the treatment of LTBI inpregnancy.Bain C, Feskanich D, Speizer FE, et al. Lung cancer ratesin men and women with comparable histories of smoking.J Natl Cancer Inst 2004; 96:826–834An article suggesting that there is no sex difference in rates oflung cancer with comparable smoking histories.Bates SM, Greer I, Pabinger I, et al. Venous thromboembolism,thrombophilia, antithrombotic therapy, andpreg nancy: American College of Chest Physicians Evidence-basedclinical practice guidelines (8th edition).Chest 2008; 133:844S–886SThis is the most recent evidence-graded consensus conferenceregarding antithrombotic agents in pregnancy.Bissler JJ, McCormack FX, Young LR, et al. Sirolimus forangiomyolipoma in tuberous sclerosis complex orlymphangioleiomyomatosis. N Engl J Med 2008; 358:140–151This article describes the results of the trial using sirolimus.Chapman KR. Chronic obstructive pulmonary disease:are women more susceptible than men? Clin Chest Med2004; 25:331–341An article presenting evidence to support the question.Chapman KR, Tashkin DP, Pye DJ. Gender bias in thediagnosis of COPD. Chest 2001; 119:1691–1695A discussion of sex differences in the diagnosis of COPD.Collop NA, Adkins D, Phillips BA. Gender differencesin sleep and sleep-disordered breathing. Clin Chest Med2004; 25:257–268<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 719ACC-PULM-09-0302-040.indd 7197/10/09 10:27:03 PM


This article includes a section on sleep-disordered breathingin pregnancy.Deblieux PM, Summer WR. Acute respiratory failure inpregnancy. Clin Obstet Gynecol 1996; 39:143–152This is a review of causes of respiratory failure in pregnancy.Elkus R, Popovich JL. Respiratory physiology in pregnancy.Clin Chest Med 1992; 13:555–565This is an overview of the normal physiologic adaptation of therespiratory system to pregnancy.Fu JB, Kau TY, Severson RK, et al. Lung cancer in women:analysis of the national surveillance, epidemiology, andend results database. Chest 2005; 127:768–777More information is provided on this subject.Gazdar AF, Thun MJ. Lung cancer, smoke exposure, andsex. J Clinical Oncol 2007; 25:469–471An interesting editorial discussing the issues of tobacco exposure,lung cancer, and sex.GINA Report: Global Strategy for Asthma Managementand Prevention. Available at: http://www.ginasthma.com/Guidelineitem.asp??l1=2&l2=1&intId=60. AccessedFebruary 16, 2009The most recent guidelines for asthma classification and treatment.Greer IA. Thrombosis in pregnancy: maternal and fetalissues. Lancet 1999; 353:1258–1265This is a thorough review in this area, with a detailed discussionof the thrombophilias.Greer IA. Prevention and management of venous thromboembolismin pregnancy. Clin Chest Med 2003; 24:123–137This is a review article on this subject.Heiss G, Wallace R, Anderson GL, et al. Health risks andbenefits 3 years after stopping randomized treatment withestrogen and progestin. JAMA 2008; 299:1036–1045This article describes the health risks and benefits found at3-year follow-up of women who had formerly received HRT.Hulley S, Furberg C, Barrett-Connor E, et al. Noncardiovasculardisease outcomes during 6.8 years of hormonetherapy: Heart and Estrogen/Progestin ReplacementStudy follow-up (HERS II). JAMA 2002; 288:58–66Results of the Heart and Estrogen/Progestin ReplacementStudy are provided, including risks of VTE with HRT.Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2008.CA Cancer J Clin 2008; 58:71–96An update on 2008 cancer statistics including incidence andmortality.Lapinsky SE, Kruczynski K, Slutsky AS. Critical care inthe pregnant patient. Am J Respir Crit Care Med 1995;152:427–455This is an excellent comprehensive review of this subject.McCormack FX. Lymphangioleiomyomatosis: a clinicalupdate. Chest 2008; 133:507–516This is a good review article on this topic.Moore J, Baldisseri MR. Amniotic fluid embolism. CritCare Med 2005; 33:279s–285sThis is a good review of this topic.Murphy VE, Gibson PG, Smith R, Clifton VL. Asthmaduring pregnancy: mechanisms and treatment implications.Eur Respir J 2005; 25:731–750National Asthma Education and Prevention Program.Report of the Working Group on Asthma and Pregnancy:management of asthma during pregnancy;update 2004. Bethesda, MD: National Institutes ofHealth, 2005; 1–57This is an updated report on this subject from the NationalInstitutes of Health.Nelson-Piercy C. Asthma in pregnancy. Thorax 2001;56:325–328This is a review of this topic.Patel JD, Bach PB, Kris MG. Lung cancer in US women:a contemporary epidemic. JAMA 2004; 291:1763–1768This is a review article on this subject.Pereira A, Krieger BP. <strong>Pulmonary</strong> complications of pregnancy.Clin Chest Med 2004; 25:299–310This is a review of this general topic.Position statement: the use of newer asthma and allergymedications during pregnancy. Ann Allergy AsthmaImmunol 2000; 84:475–480This is another review on this topic.Ramsey PS, Ramin KD. Pneumonia in pregnancy. ObsetGynecol Clin North Am 2001; 28:553–569This is a good review in this area.Rossouw JE, Anderson GL, Prentice RL, et al. Risks andbenefits of estrogen plus progestin in healthy postmenopausalwomen: principal results from the Women’sHealth Initiative randomized controlled trial. JAMA2002; 288:321–333Results of the Women’s Health Initiative are presented, includingrisks of VTE with HRT.Schatz M. The efficacy and safety of asthma medicationsduring pregnancy. Semin Perinatol 2001; 25:145–152This article describes published consensus recommendationsregarding the pharmacologic management of asthma duringpregnancy.Schatz MM, Zeiger RS, Harden K, et al. The safetyof asthma and allergy medications during pregnancy.J Allergy Clin Immunol 1997; 100:301–306This is a review of the use and safety of asthma medicationsused during pregnancy.Surgeon General’s report: women and smoking; a reportof the Surgeon General, 2001. Available at: www.cdc.gov/tobacco/sgr. Accessed February 16, 2009720 Women’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong> (Levine)ACC-PULM-09-0302-040.indd 7207/10/09 10:27:04 PM


Tan KS, Thomson NC. Asthma in pregnancy. Am J Med2000; 109:727–733This is a good review of this subject.Tanoue LT. Cigarette smoking and women’s respiratoryhealth. Clin Chest Med 2000; 21:47–65This article reviews epidemiology of smoking and smokingrelateddiseases in women.Thomas L, Doyle LA, Edelman MJ. Lung cancer inwomen: emerging differences in epidemiology, biology,and therapy. Chest 2005; 128:370–381An update on the epidemiology, biology, and therapy of lungcancer in women.Unterborn J. <strong>Pulmonary</strong> function testing in obesity,pregnancy, and extremes of body habits. Clin Chest Med2001; 22:759–768This chapter includes a subsection on pulmonary functiontesting in pregnancy.Wendel PJ. Asthma in pregnancy. Obstet Gynecol ClinNorth Am 2001; 28:537–551This article discusses the pathophysiology of asthma and theeffects of asthma on pregnancy and vice versa, and reviews theNational Asthma Education Program guidelines for the treatmentof asthma in pregnancy.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 721ACC-PULM-09-0302-040.indd 7217/10/09 10:27:04 PM


Notes722 Women’s Issues in <strong>Pulmonary</strong> <strong>Medicine</strong> (Levine)ACC-PULM-09-0302-040.indd 7227/10/09 10:27:05 PM


Occupational AsthmaWilliam S. Beckett, MD, FCCPObjectives:• Provide a clinical definition of occupational asthma• Show the relationship of occupational asthma to the largerrealm of adult asthma• Outline mechanisms by which asthma develops fromexposures in the work setting• Discuss the evaluation of a patient with possible occupationalasthma• <strong>Review</strong> the management of a patient with confirmedoccupational asthmaKey words: asthma; irritant asthma; occupational asthma;occupational lung disease; reactive airways dysfunctionsyndrome; work-related asthmaThis chapter expands on the introduction to occupationalasthma by Dr. Sidney Braman in the chapter“Asthma” earlier in this publication. It addressesthe American <strong>Board</strong> of Internal <strong>Medicine</strong> <strong>Pulmonary</strong>Disease Certification “Occupational andEnvironmental Disease” content category topics of“Occupational Asthma,” “Byssinosis,” “ReactiveAirways Dysfunction Syndrome,” and “Work/DisabilityEvaluation.”Clinical Definition of OccupationalAsthmaFrom Dr. Braman’s chapter on “Asthma,” weread that: “Allergens and occupational factors areconsidered the most important causes of asthma. . . .Asthma that is precipitated by a particular occupationalenvironment and not by stimuli outsidethe workplace is called occupational asthma. Twotypes of asthma have been described: asthma thatfollows a latent period of exposure to either a highorlow-molecular-weight sensitizing antigen, andasthma that follows a latent period of exposure toworkplace irritants. One form of irritant asthma iscalled reactive airways dysfunction syndrome, a conditionthat usually results from the sudden inhalationof a large dose of a highly irritating substance.” 1–3In addition, individuals with a history ofasthma may experience a recurrence of asthmaafter years of being asymptomatic when theyare challenged with an irritant or allergen in theworkplace. This has been termed work-aggravatedasthma and, from the point of view of medical management,is similar to other forms of occupationalasthma.Because the workplace environment contributesto the severity of asthma in these patients,the chest physician’s challenge is to determinewhether a change in the workplace environmentwill improve or cure the patient’s occupationalasthma or whether, conversely, in the patient withnonoccupational asthma, it will have no effect. Thisdetermination usually requires a clear statementin the medical record as to whether the patient’sasthma is or is not occupational. A diagnosis ofoccupational asthma often presents problems to thepatient and employer, requiring a major change inemployment. If nonoccupational, it is safe to allowthe patient to continue in the same job, focusing onbetter medical management.Frequency of CasesReliable data show that the proportion of newcases of occupational asthma in North Americanadults is approximately 9 to 15%, or 1 in 6 to 1 in10 new-onset cases of asthma. 4 Any patient whois employed and who presents with new-onsetasthma should thus be questioned about occupationalexposures.Natural History of OccupationalAsthmaNumerous useful reviews, both brief and comprehensive,are available. 5–11 Typically, asthma willbegin weeks to years after working with a new substancethat can cause asthma. Symptoms may beminor or intermittent at first and gradually increaseas the patient’s sensitization increases, sometimesto the point where the patient has nightly wheezingand orthopnea.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 723ACC-PULM-09-0302-041.indd 7237/10/09 10:15:45 PM


Table 1. Most Frequently Reported Specific Causes or Contributorsto Work-Related Asthma in the United States*Di-isocyanatesStainless steel welding plumeFormaldehydePaintPesticidesNatural rubber latexChlorineGlutaraldehydeDiesel exhaustEpoxy resinsAcrylatesWood dust*From Jajosky et al. MMWR Morb Mortal Wkly Rep 1999;48(suppl): No SS-3.In some cases, patients will have had childhoodasthma that remitted in adolescence, but for others,this onset of asthma will be the first. The cardinalfeature is the onset of asthma while working withinhalational exposure to a substance that can causeasthma, although there is emerging evidence thatrepeated skin contact with some substances canresult in respiratory sensitization. Although there issome literature to suggest that the early diagnosisof occupational asthma leads to less severity andimproved chances for complete remission, thepublished literature in this area is uncertain.Causes of Occupational AsthmaMore than 400 substances have been identifiedas causes of occupational asthma. 12,13 Examples of afew of the most commonly reported causes in theUnited States are shown in Table 1.Other Risk Factors for OccupationalAsthmaSpecific gene polymorphisms are associatedwith an increased risk for the development ofasthma for a few causes of occupational asthma,but testing for these is not clinically useful. A historyof atopy or cigarette smoking is a risk factor forsensitization to high-molecular-weight substances(eg, biologically derived antigens, such as proteinsfrom laboratory animals or wheat flour) but not forsensitization to others (eg, low-molecular-weightchemicals like isocyanates).Pathophysiology, Histopathology, andPhysiologic MechanismsIn these respects, occupational asthma does not differ,or differs only in minor ways, from other typesof asthma. 14,15 Please refer to the “Asthma” chapterin this publication by Dr. Braman for a comprehensivediscussion.Clinical Features of OccupationalAsthmaUseful Diagnostic MeasuresHistory: The combination of the following fourelements of the patient’s medical history has a 64%positive predictive value in the diagnosis of occupationalasthma 16 : current diagnosis of asthma;and onset of asthma after entering the workplace;association between symptoms of asthma andwork; and workplace exposure to an agent knownto give rise to occupational asthma. Allergic rhinitisto the offending substance often precedes the onsetof lower respiratory tract symptoms.Physical Examination: Wheeze is usually presentminutes to hours after a significant exposure and isabsent between exacerbations.<strong>Pulmonary</strong> Function Testing: Comprehensive,evidence-based reviews of diagnostic approaches tooccupational asthma 17,18 recommend the followingtests in addition to a medical history and physicalexamination:Methacholine challenge: improves the diagnosticaccuracy of an asthma diagnosis (84% sensitivity,48% specificity).• Immunology testing: tests for serum-specific IgE(ie, blood radioallergosorbent tests or skin- pricktests) are available through reference laboratories(eg, Mayo Clinic Laboratory; Rochester, MN) for alarge number of common high-molecular-weightsubstances. Some allergists also have skin-pricktesting reagents for some of the large -molecularweightsubstances. However, there is no specificIgE for many of the less common low-molecularweightsubstances. A positive test result for aspecific IgE to the suspected antigen increasesthe diagnostic certainty of occupational asthma.When preexisting or subclinical asthma is exacerbatedby work, a similar approach is taken. 19724 Occupational Asthma (Beckett)ACC-PULM-09-0302-041.indd 7247/10/09 10:15:45 PM


• Measurement of peak expiratory flow rate withthe use peak flowmeters: the recording of portablepeak expiratory flow rates (best of three blows)four times per day during days at work and daysaway from work may also help to distinguishoccupational from nonoccupational asthmabecause reversible airflow obstruction is necessaryto the diagnosis of asthma, and an occupationallyrelated pattern may be evident. Clinical interpretationby visual comparison of the patient’sexposure history has been found to be as accurateas computer-based interpretation.Treatment of Occupational AsthmaAs with all asthma, the treatment goal is toremove the patient from exposure to triggers andto minimize symptoms by controlling asthma withmedications that have the least adverse effects forthe patient. Once diagnosed, the main differencebetween treating occupational and nonoccupationalasthma is removing the patient from exposure. Thistreatment begins with notifying the patient and, withthe patient’s permission, the employer, and trying toidentify a specific exposure that can be successfullyavoided. It is usually much to the patient’s advantageto continue working for the same employer, butin circumstances in which the offending agent hasbeen removed, the reduction of air levels has beenachieved or, in some cases, respiratory protectionhas been added. The assistance of an occupationalmedicine specialist or an industrial hygienist maybe necessary at this point.Reduction or elimination of exposure alonemay be enough to bring resolution of asthma.However, in 50% of recognized cases, patientsrequire ongoing therapy with medications. Theguidelines for treatment follow the standardapproaches, as outlined in the chapter “Asthma”by Dr. Braman in this syllabus.ByssinosisByssinosis is a form of occupational asthmathat is caused by the inhalation of dust releasedwhen processing the cotton plant (including theleaves and Gram-negative bacterial endotoxinassociated with the cotton fibers but not the cottonfibers themselves). Symptoms are intermittent andtend to improve with continued exposure throughthe work week, then return after a weekend awayfrom exposure. Unlike other types of occupationalasthma, severe cases may be associated with episodicfever on exposure.Work/Disability EvaluationIn the case of occupational asthma, a mostdifficult initial decision may be in determiningwhether it is safe to allow the patient to return towork, or whether he or she can return with specificwork restrictions. Once a diagnosis of occupationalasthma is made, it is often important for the treatingphysician to determine whether disability ispresent and, if so, to what extent.For any pulmonary patient who presents witha request for disability evaluation, the physicianmust first decide whether the lung diseaseis nonoccupational (in which case federal SocialSecurity Administration disability criteria applyif the patient is totally disabled), or whether it isoccupational (ie, caused by the job) in which statespecificworker compensation disability rules mayapply.American Medical Association GuidesMany states require the physician to apply thecriteria of the Guides to the Evaluation of PermanentImpairment 20 in rating the degree of disability forall occupational conditions including lung disease.Although for most conditions the guides focus onpermanent pulmonary function loss, for asthmamore emphasis is placed on the frequency andseverity of exacerbations despite maximal medicaltherapy.Social Security DisabilityFor the patient who is unable to work at allbecause of lung disease, whether occupational ornonoccupational, Social Security Administrationdisability criteria may apply. These criteria areavailable on the Internet. 21 In brief, total disabi lityfrom asthma is considered whenAttacks (as defined in section 3.00C), in spiteof prescribed treatment and requiring physicianintervention, occur at least once every<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 725ACC-PULM-09-0302-041.indd 7257/10/09 10:15:46 PM


2 months or at least six times a year. Each inpatienthospitalization for longer than 24 h forcontrol of asthma counts as two attacks, andan evaluation period of at least 12 consecutivemonths must be used to determine thefrequency of attacks.Fatal Occupational AsthmaAlthough infrequently reported, fatal cases ofoccupational asthma 22 have occurred in patientsfor whom the association with the workplace wasnot made in a timely fashion or who returned to bereexposed at work against medical advice.Annotated ReferencesReactive Airways Dysfunction Syndrome1. Alberts WM, do Pico GA. Reactive airways dysfunctionsyndrome. Chest 1996; 109:1618–1626Comprehensive review that discusses reactive airwaysdysfunction syndrome as a distinct subset of IIA.2. Bardana EJ. Reactive airways dysfunction syndrome(RADS): guidelines for diagnosis and treatment andinsight into likely prognosis. Ann Allergy AsthmaImmunol 1999; 83:583–586Excellent short review with a balanced discussion of thecontroversial areas of this disorder.Irritant Asthma3. Tarlo SM. Workplace respiratory irritants andasthma. Occup Med 2000; 15:471–484Tarlo uses a more inclusive definition of IIA.Frequency of Cases of Occupational Asthma4. Balmes J, Becklake M, Blanc P, et al. American ThoracicSociety statement: occupational contributionto the burden of airway diseases. Am J Respir CritCare Med 2003; 167:787–797Brief reviews of occupational asthma are presented.5. Chan-Yeung M, Malo J-L. Occupational asthma.N Engl J Med 1995; 333:107–1126. Tarlo SM, Liss GM. Occupational asthma: an approachto diagnosis and management. Can MedAssoc J 2003; 168:867–871These short-but-complete reviews are recommendedfor those whose review time is limited. Each review takesa little different tack but covers essentially the samematerial.Comprehensive <strong>Review</strong>s: Occupational Asthma7. Tarlo SM, Boulet LP, Cartier A, et al. Canadian ThoracicSociety: guidelines for occupational asthma.Can Respir J 1998; 5:289–300Perhaps the clearest, most clinically useful review addressingdiagnosis.8. Mapp CE, Boschetto P, Maestrelli, et al. State of theart: occupational asthma. Am J Respir Crit Care Med2005;172:280–305A “state-of-the-art” article that is full of facts. This updatesthe classic Chan-Yeung and Lam state-of-the-art articlefrom 1986.9. Chan-Yeung M, Malo JL, Tarlo SM. Proceedings ofthe first Jack Pepys occupational asthma symposium.Am J Respir Crit Care Med 2003; 167:450–471A 2003 review of the subject is presented from luminariesin the field.10. Nicholson PJ, Cullinan P, Taylor AJ, et al. Evidencebased guidelines for the prevention, identification,and management of occupational asthma. OccupEnviron Med 2005; 62:290–29911. Rachiotis G, Savani R, Brant A, et al. Outcome ofoccupational asthma after cessation of exposure: asystematic review. Thorax 2007; 62:147–152Sources of information on causes are listed.To know whether the materials your patient works withare known to cause asthma, ask the patient to bring youmaterials safety data sheets from work or request theseyourself from the employer under the OccupationalSafety and Health Administration “Right to Know”law. The employer must provide them within 15 businessdays.12. SIRI MSDS index. Available at: http://siri.org/msds/index.php. Accessed April 22, 2009If you know only the brand names or product names of thesubstances, you can quickly get the chemical names fromthe material safety data sheets online at SIRI.org.13. Sommaire. Tables of agents and substances that cancause asthma. Available at: http://www.remcomp.fr/asmanet/asmapro/agents.htm#start. AccessedApril 22, 2009Once you know the chemical names, you can check themagainst the > 400 known causes of occupational asthma726 Occupational Asthma (Beckett)ACC-PULM-09-0302-041.indd 7267/10/09 10:15:46 PM


on-line at “ASMAPRO,” which includes links to PubMedand references to the medical literature.Mechanisms of Occupational Asthma14. Mapp CE, Boschetto P, Miottoi D, et al. Mechanismsof occupational asthma. Ann Allergy Asthma Immunol1999; 83:645–66415. Sastre J, Vandenplas O, Park HS. Pathogenesis ofoccupational asthma. Eur Respir J 2003; 22:364–373Both articles discuss the known facts and the proposedtheories of the pathogenetic mechanisms at work inoccupational asthma.Diagnosis, Evaluation, Management16. Malo J-L, Ghezzo H, L’Arqueveque J, et al. Is theclinical history a satisfactory means of diagnosingoccupational asthma? Am Rev Respir Dis 1991;143:528–53217. Tarlo S, Balmes J, Balkisoon R, et al. Managementof occupational asthma: an <strong>ACCP</strong> evidence-basedclinical guideline. Chest 2008; 134(3 suppl):1S–41SThe most detailed (very long) evidence-based approach todate. It uses an expert panel review of the current literature.It is recommended that methacholine challenge testsand specific IgE assays or skin-prick tests (when available)be used to improve diagnostic accuracy.18. Nicholson PJ, Cullinan P, Taylor AJ, et al. Evidencedbased guidelines for the prevention, identificationand management of occupational asthma. OccupEnviron Med 2005; 62:290–299Similar guidelines from the United Kingdom that placemore emphasis on the use of serial portable peak flowmeasurements four times daily both at work and awayfrom work in the confirmation of suspected diagnosis. Itemphasizes asking patients about the occurrence of allergicrhinitis preceding asthma.19. Friedman-Jimenez G, Petsonk L, Szeinuk J, et al. Workrelatedasthma. Am J Ind Med 2000; 37:121–141A detailed review that can be downloaded free fromhttp://www3.interscience.wiley.com/cgi-bin/fulltext/67501287/PDFSTARTWork/Disability Evaluation20. American Medical Association. The respiratory system.In: Robert Rondinelli, ed. Guides to the evaluationof permanent impairment. 6th ed. Chicago, IL:AMA Press, 2007This rating system is the one most widely required byindividual state worker compensation systems. You willneed to check with your state to determine whether theseor other medical guidelines apply.21. Disability Evaluation Under Social Security. Baltimore,MD: Social Security Administration, June 2006;Publication No. 64–039. Available at: http://www.ssa.gov/disability/professionals/bluebook/3.00-Respiratory-Adult.htm#3.03%20Asthma. AccessedMarch 29, 2009Fatal Occupational Asthma22. Ortega HG, Kreiss K, Schill DP, et al. Fatal asthmafrom powdering shark cartilage and review of fataloccupational asthma literature. Am J Ind Med 2002;42:50–54<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 727ACC-PULM-09-0302-041.indd 7277/10/09 10:15:46 PM


Notes728 Occupational Asthma (Beckett)ACC-PULM-09-0302-041.indd 7287/10/09 10:15:46 PM


Diseases Related to High Altitude and Diving andNear-Drowning: Basics of Hyperbaric <strong>Medicine</strong>Igor Aksenov, MD, PhD; and Michael Strauss, MDObjectives:• Understand the physiology and pathophysiology ofdiving and high-altitude pulmonary-related problems• Describe the main diseases related to high altitude, theirdiagnosis, management, and prevention• Discuss the different types of diving, contraindications todiving, and the evaluation and management of commonmedical problems of diving• Explain the pathophysiology of drowning and neardrowning• Appreciate the mechanisms of hyperbaric oxygen therapyand its indications and contraindicationsKey words: altitude; barotrauma; decompression; diving;drowning; high-altitude diseases; hyperbaric oxygen;medical problems of divingDiseases Related to High AltitudeHigh-Altitude PhysiologyAscent in altitude results in a decrease in barometricpressure and corresponding decreases in Po 2.The barometric pressure at sea level is 760 mm Hg,whereas on the summit of Mount Everest (altitude8,848 m or 29,029 feet) it is approximately 250 mmHg. This correlates with a 67% decrease in Po 2fromapproximately 159 mm Hg at sea level to 53 mmHg at the top of Mount Everest.Acclimatization is a process of adjustment tohigh altitude and low inspired oxygen that allowsminimizing effects of hypoxia. The initial responseto hypoxia is increased alveolar ventilation, whichis called the hypoxic ventilatory response. The hypoxicventilatory response causes a reduction in thealveolar Pco 2and an increase in the alveolar Po 2with subsequent increase in arterial oxygen content.Increased alveolar ventilation also causesrespiratory alkalosis, which leads to increasedrenal bicarbonate elimination and metabolic acidosis.However, the renal compensation is onlypartial and therefore respiratory alkalosis persists.The adjustments of the cardiovascular systemto high altitude include increases in cardiacoutput and pulmonary artery pressure, improvedventilation/perfusion matching, as well as selectivevasodilatation and vasodilatation that improvesoxygen delivery to the brain and heart. In the brain,hypoxia induces cerebral vasoconstriction, whichis attenuated by hypocapnia-induced cerebralvasoconstriction. The final result is small increasein cerebral blood flow in proportion to tissuehypoxia. Hypoxia also induces secretion of renalgeneratederythropoietin, which stimulates RBCproduction in bone marrow. However, it takes 10to 14 days to increase the mass of the RBCs fromthis effect. Changes at the tissue and cell levelinclude improvement in oxidative metabolism inmitochondria and hypoxia-induced factor-1 signalingof vascular endothelial growth factor. Vascularendothelial growth factor, in turn, stimulatesangiogenesis.High-Altitude IllnessThe term high-altitude illness encompasses threeconditions that occur as a result of acute exposureto hypobaric hypoxia during rapid ascent in altitude.These conditions include acute mountainsickness (AMS), high-altitude cerebral edema(HACE), and high-altitude pulmonary edema(HAPE). Long-term exposure to high altitude mayresult in the development of chronic mountainsickness (CMS).AMS: AMS represents the most common formof altitude illness. It usually occurs 6 to 48 h afterrapid ascents to attitudes 8,000 feet (2,440 m).The exact mechanism of AMS is unknown, but itappears to result from hypoxia-induced cerebralvasodilation. Some studies suggest that vasogenicbrain edema and increased blood-brain barrier permeabilitycontribute to the pathogenesis of AMS.Typical symptoms of AMS include head ache,dizziness, fatigue, malaise, anorexia, nausea,<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 729ACC-PULM-09-0302-042.indd 7297/13/09 6:46:23 PM


vomiting, and difficulty sleeping. Findings fromthe physical examination may include tachycardia,mild crackles in the chest with auscultation,and peripheral edema. These findings are nonspecificand are not necessary to make a diagnosisof AMS. The diagnosis of AMS is establishedfrom the clinical symptoms in the setting ofascent to altitude. AMS can be exacerbated bystrenuous physical activity, alcohol consumption,and use of sedative medications.A slow ascent that makes it possible to acclimatizegradually to altitude is the best way toprevent AMS. The general rule is that at altitudes 3,000 m (9,840 feet), the climber should spendthe night at an altitude 300 m (984 feet) greaterthan the previous night. Adequate hydration andrest are important as well. Medications that can beused for the prevention of AMS include acetazolamideand dexamethasone. Acetazolamide(125 mg po bid starting a day before ascent) promotesacclimatization by stimulating bicarbonatediuresis and respiration. Acetazolamide is contraindicatedin patients with known sulfa allergy.Dexamethasone, unlike acetazolamide, does nothave effects on acclimatization. Mechanisms toexplain how this drug benefits AMS are unknown.It may be related to reduction of cerebral bloodvolume or decreased capillary permeability. Dexamethasonehas a rapid onset of action. However,it is important to remember that cessation of thisdrug may result in a recurrence of symptoms.Other drugs that may help prevent AMS includetheophylline and temazepam.Appropriate treatment of AMS is importantbecause it can progress to HACE. First and mostimportant, climbers with AMS symptoms shouldstop their ascent, rest and, if possible, descend 300 m ( 984 feet). Other measures include oxygensupplementation and adequate fluid intake.Acetazolamide and dexamethasone both are effectivein treatment of AMS. In general, acetazolamideshould be considered as a first-line agent, anddexamethasone should be administered to climberswith sulfa allergy. Acetazolamide and dexamethasonecan also be used as a combination if one hasrapid progression of symptoms.HACE: HACE is a potentially fatal neurologiccondition. It is a more severe form of AMS. HACEusually occurs within hours or days after arrival ataltitude. It has been reported to occur at altitudesas low as 2,750 m (9,022 feet). It is manifested byneurologic symptoms and signs such as headache,loss of coordination, ataxia, confusion, hallucinations,stupor, and coma. Physical examinationfindings may demonstrate papilledema, truncalataxia, and cranial nerve palsies. Death usuallyoccurs from brain herniation. Symptoms ofHACE are nonspecific, and other diagnoses suchas meningitis, encephalitis, intoxications, electrolyteimbalances, and glucose disturbances shouldbe considered in the differential diagnosis.The best prevention of HACE is slow ascent.Acetazolamide prophylaxis is recommended forindividuals with a previous HACE history. HACEis a medical emergency and requires immediateinterventions. Descent to a lower altitude shouldbe initiated without delay. The patient should beadministered oxygen and dexamethasone (8 mgIV, IM, or po). If descent is not possible, hyperbaricoxygen therapy (HBOT) in a portable chamber canbe a life-saving treatment if this equipment is availableon the mountain.HAPE: HAPE is a form of noncardiogenic, hydrostaticpulmonary edema. It is potentially fataland accounts for more deaths than any otherhigh-altitude illness. HAPE occurs 2 to 3 daysafter rapid ascent to altitudes 2,590 m (8,500 feet).Some individuals are more predisposed to HAPEthan others. The most important pathophysiologicmechanism of HAPE is hypoxic pulmonary vasoconstriction.This causes acute pulmonary hypertension,pulmonary capillary stress fractures and,eventually, the development of pulmonary edema.The release of cytokines and inflammatory mediatorsmay play a role in the pathogenesis of HAPE,but it appears to be a secondary reaction to thehydrostatic injury. Other proposed mechanismsfor HAPE include decreased nitric oxide synthesisand increased sympathetic tone in the pulmonaryvasculature.Symptoms of HAPE may occur gradually afterAMS, but they can also present rapidly withoutany other signs. Clinical findings include nonproductivecough, fatigue, and dyspnea. Cough maybecome productive with pink, frothy sputum.Affected individuals may have mental statuschanges, lethargy, and coma secondary to hypoxemia.Physical examination findings include tachycardiaand crackles on auscultation of the chest.Chest radiograph may show enlarged pulmonary730 Diseases Related to High Altitude and Diving (Aksenov and Strauss)ACC-PULM-09-0302-042.indd 7307/13/09 6:46:24 PM


arteries and patchy or homogenous infiltrates thatcommonly involve the right middle lobe and bothlower lobes of the lungs.As with AMS and HACE, the best method toprevent HAPE is a slow ascent that providesadequate time for acclimatization. Medications thatcan be used for prophylaxis include nifedipine,dexamethasone, phosphodiesterase inhibitors, andsalmeterol. A placebo-controlled study of nifedipinein alpinists with previous manifestations ofHAPE showed a significant decrease in developmentof HAPE when compared with the use ofplacebo. Phosphodiesterase inhibitors promotepulmonary vasodilation and reduce pulmonaryartery pressure. One study showed that sildenafilimproved exercise performance at altitude. However,only tadalafil was shown to reduce the incidenceof HAPE. Prophylactic inhalation of highdoses of the -agonist salmeterol was associatedwith a decreased incidence of HAPE.Treatment of HAPE includes immediatedescent, high-flow oxygen, rest, and the administrationof medications such as nifedipine, sildenafil,tadalafil, and dexamethasone. Nifedipineshould be used carefully because it can lower systemicBP and reduce cerebral perfusion pressure.The use of HBOT while at altitude, if available, canbe a life-saving intervention.CMS (Monge Disease): CMS or Monge diseaseis observed in inhabitants who reside at highaltitudes. This disorder was originally describedin Andean mountain residents. It is caused byerythropoietin-induced increased production ofRBCs in response to hypoxia. Findings in CMSinclude headache, dizziness, lethargy, insomnia,memory problems, cyanosis, fluid retention, andparasthesias. <strong>Pulmonary</strong> hypertension and congestiveheart failure may develop. CBC typicallydemonstrates increased levels of hemoglobinsand hematocrits. Definite treatment of CMS isa return to a lower altitude. Symptoms can bereduced by oxygen administration, phlebotomy,and acetazolamide.Altitude and Air Travel in Hypoxemic PatientsCommercial airplanes usually fly at altitudesof 3,000 m (10,000 feet) to 12,200 m (40,000 feet).The Federal Aviation Administration requires thatcabin pressures be maintained at pressures equivalentto altitudes less than 2,438 m (8,000 feet),although this maximum may be breached inemergen cies. At 2,438 m, the Po 2is equivalent tobreathing 15.1% oxygen at sea level. Because ofthis, patients with underlying lung disease maybecome hypoxic, and patients who are alreadyreceiving oxygen may require increased oxygensupplementation. According to guidelines publishedby the British Thoracic Society, patients withthe following conditions/risk factors should beassessed prior to permitting commercial air travel:• Severe COPD or asthma;• Severe restrictive lung disease (including chestwall and respiratory muscle disease), especiallywith hypoxemia and/or hypercapnia;• Cystic fibrosis;• History of air travel intolerance with respiratorysymptoms (dyspnea, chest pain, confusionor syncope);• Comorbidity with other conditions worsenedby hypoxemia (cerebrovascular disease, coronaryartery disease, heart failure);• <strong>Pulmonary</strong> tuberculosis;• Within 6 weeks of hospital discharge for acuterespiratory illness;• Recent pneumothorax;• Risk of or previous venous thromboembolism;and• Preexisting requirement for oxygen or ventilatorsupport.The preflight assessment should include thefollowing: (1) history and physical examination;(2) spirometry, and (3) oxygen saturation by pulseoximetry (Spo 2) measurement. Blood gas tensionsare preferred in patients with known or suspectedhypercapnia. In-flight oxygen is not recommendedif the patient’s room-air Spo 2is 95%. In thosepatients who have a resting sea level Spo 2of 92 to95% with additional risk factors, hypoxic challengetesting is recommended. If the sea level Spo 2is 92%, then in-flight oxygen is recommended.Patients who receive supplemental oxygen at sealevel should increase the flow while at cruisingaltitude. Additional risk factors for flying in commercialaircraft include hypercapnia, FEV 1 50%of predicted, lung cancer, and the need for ventilatorsupport.<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 731ACC-PULM-09-0302-042.indd 7317/13/09 6:46:24 PM


Currently, there are several methods to predictin-flight Pao 2, including the following:1. Predicting hypoxemia from equations: Theseequations are helpful as a screening test. Theequations have been derived almost exclusivelyfrom patients with COPD, and they donot predict response to exercise.2. Measurement of Pao 2or Spo 2under hypobaricconditions: This method is the most preciseone, but it requires depressurization in ahypobaric chamber, which is not usually anoption in most clinical settings and pulmonaryfunction laboratories.3. The high-altitude stimulation test reproducesin-flight hypoxia by changing the fraction ofinspired oxygen (Fio 2). In general, an altitudeof 1,520 m (5,000 feet) is equivalent to Fio 2of17%; 2,438 m (8,000 feet) to Fio 2of 15%; and3,048 m (10,000 feet) to Fio 2of 14%. This is asimple and accurate test that can be performedin most pulmonary function laboratories. If thepredicted in-flight Pao 2decreased 50 mm Hg,then supplemental oxygen is recommended.Federal Aviation Administration rules do notallow passengers to carry their own oxygen tanksor liquid oxygen on commercial flights. However,patients can bring portable battery-powered oxygenconcentrators. Patients should have a letterfrom a physician with an explanation of theirmedical condition(s) and their oxygen requirementsfor commercial air travel.Diving <strong>Medicine</strong>Types of DivingSports diving includes a spectrum of divingtypes ranging from breath-hold snorkeling to useof closed-circuit self-contained underwaterbreathing apparatus (SCUBA) gear (Table 1). Deeptechnical diving, surface supply diving, and saturationdiving are not considered recreational typesof diving and are not included in Table 1. The distinctionbetween the sports diver and the commercialdiver is no longer as clear as it was in the pastwhen commercial divers used equipment thatwas unavailable to the sports diver. Today, in ouropinion, the main distinction between the commercialand the sports diver is the former gets paidfor his or her diving activity, whereas the sportsdiver does not.Medical Clearance and Contraindicationsfor DivingDiving requires good physical and mentalhealth. Medical standards are different for breathhold,recreational SCUBA diving, and commercialdivers.Breath-Hold/Snorkel Diving: Generally, thereare no medical requirements to participate in thistype of diving. Fitness is the key to safe breathholddiving. If there are any fitness issues, thediver should be screened with the use of standardssimilar to those for a sedentary person whodecides to start a conditioning/aerobic exerciseprogram.Recreational SCUBA Diving: In the UnitedStates, according to the Professional Associationof Dive Instructors, no mandatory medical examinationis required. However, Professional Associationof Dive Instructors as well as other divingcertification organizations require candidates whotake their SCUBA diving training programs completea medical history questionnaire. If any of thehistory questions are checked as being present,then an evaluation and clearance by a physicianis usually required before the diving candidate isallowed to continue recreational SCUBA divingtraining. For those 40 years of age, especially iffitness for diving is a concern, a prediving trainingphysical examination is recommended preferablywith a stress ECG study.The physical examination of a potential SCUBAdiver should be directed toward identification ofear, nose, and throat problems that could interferewith pressure equilibration during descents andascents, as well as cardiovascular and pulmonarydiseases that would contribute to heart problemscaused by the physical demands of diving andpulmonary overpressure syndromes with depthchanges, respectively. Obesity significantlyincreases the risks of decompression sickness (DCS)because nitrogen is lipid soluble. Mental healthissues must not be overlooked because the abilityto follow instructions, avoid panic, and remain732 Diseases Related to High Altitude and Diving (Aksenov and Strauss)ACC-PULM-09-0302-042.indd 7327/13/09 6:46:24 PM


Table 1. Types of Sports DivingType Breathing Equipment Advantages Disadvantages CommentsBreath-holding Snorkel Freedom Depth limited bybreath-hold timeSCUBAEnriched-airnitroxRegulator and tank;open circuit*Regulator and tank;open circuitExtends bottom timeReduced chances ofdecompression sicknessInefficient use of airsupply †Increased chances ofoxygen toxicityRebreather Closed circuit ‡ Prolongs underwater times Many hours of training;increased hazardsProbably the largestnumber of participantsThe majority of recreationalSCUBA divingLess fatigueEquipment costs about$5,000–$10,000*Open circuit: After inhaling compressed gas from the SCUBA tanks via the regulator, the diver exhales each breath into thewater†Efficiency: This is a very inefficient use of the gas supply because 79% of the contents of the tank is the inert gas nitrogen.Each inhalation contains 21% oxygen, whereas there is 16% oxygen in the exhaled gas. Consequently, only 5% of the air in theSCUBA tank is used for the body’s respiration and ventilation needs, whereas 95% is exhaled into the water.‡Closed circuit: The gas supply recycles through breathing bags while carbon dioxide is absorbed in the rebreather circuit.Oxygen is added (usually via sensor controls) at a rate sufficiently to meet the diver’s oxygen requirements.clearheaded are essential for safe diving. Individualswho abuse alcohol or who are using street drugsshould not be permitted to dive. Prescriptionmedications or conditions for which they are prescribedmay be relative or absolute constraindications.For example, if the patient with anxietyreceives sedative medications, then he or sheshould not dive because (1) anxiety during divingmay cause rapid ascent and result in serious problems(see next section) and (2) effects of sedativemedications may be increased under pressure.Conditions such as diabetes and asthma are relativecontraindications to diving, and clearance to recreationalSCUBA dive with these conditions iscontingent on the patient’s insight to their problemsand how well they manage these conditions.Commercial Diving: This type of diving has specificstandards. In the United States, these stan dardsare provided by the Association of Diving ContractorStandards (1994), the United States Navy, theOccupational Safety and Health Administration,the National Oceanographic and AtmosphericAdministration, and the American Academy ofUnderwater Sciences (for scientific divers). Diversworking under the auspices of these organizationsare required to pass regular medical examinations.The exact standards are based on the tasks thatdiver has to perform. In addition to the informationrequired for the recreational SCUBA diver, attentionhas to be given to the health of the patient’sback as well as the presence of deformities or othermusculoskeletal problems that limit mobility andstrength. For those divers who perform long and/or deep dives or have had recurrent episodes ofDCS ima ging studies for dysbaric osteonecrosisshould be performed periodically. These latter diversalso seem so be prone to plaque-like lesions intheir brains and spinal cords that resemble thoseseen in multiple sclerosis. Consequently, carefulneurologic screening of the commercial diver isessential, and if there is any question of neurologicimpairments, screening with magne tic resonanceshould be done.<strong>Pulmonary</strong> contraindications for all types ofdiving include symptomatic asthma requiringlong-term use of medications, COPD, pulmonaryinfections, history of spontaneous pneumothorax,pulmonary cysts and scars, and pulmonary fibrosis.These conditions significantly increase the riskof extraalveolar air (pulmonary overpressure)syndromes. Epilepsy is also an absolute contraindicationfor all types of compressed gas divingbecause of the risk of seizures from alterations inthe partial pressure of the breathing gases.Diving-Related Medical ProblemsEar and Sinus Barotrauma: Barotraumas are themost frequent problems of divers. Ear squeeze isthe common terminology used to describe thisproblem. The middle ear spaces and sinus cavitiesare embedded in the skull bones. The tympanic<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 733ACC-PULM-09-0302-042.indd 7337/13/09 6:46:25 PM


membrane separates the middle ear space fromthe external ear canal. The Eustachian tube providesa passage to the middle ear space from theback of the nasopharynx, whereas the ostia of thesinuses provide connections to the sinus cavitiesand make it possible to equilibrate pressures inthese structures. Problems arise when these channelsbecome occluded, and the diver is unable toequilibrate pressure in the middle-ear spaces andsinus cavities as the ambient pressures changewith descents and ascents. The Boyle law quantifiesthese pressure-volume effects, stating that aspressures of a confined gas increase, the volumesdecrease and vice versa.The ear and sinus cavities are lined with wellvascularizedrespiratory epithelium. As a pressuredifferential develops, the vessels dilate and the eardrum distends inward (stage 1 of ear barotrauma).This generates pain. The next stage (stage 2) in theprogression is leakage of fluid from the vessels intothe middle ear space. This leaking causes mutedhearing. In stage 3, rupture of blood vessels withbleeding into the middle-ear space occurs. Sputummay become blood tinged. Perforation of the eardrum is the final step (stage 4) in the progression.Because water may then fill the ear canal, pain mayresolve because the middle-ear space becomes afluid-filled cavity and the pressure differential isobliterated. However, cold water entering themiddle-ear space can cause vertigo, lead to disorientation,and generate uncontrollable panic. Duringascent, reverse ear squeezes may occur as theresult of expansion of the air in closed space. Dizzinessand vertigo may be associated with thisespecially if unilateral. This condition is termedalternobaric vertigo.The diver himself or herself should be responsiblefor first-line interventions for ear and sinussqueezes. At the first sign of ear pain, descentshould be halted. Ascending a few feet may helpto “clear” the ears. If equilibration is not achieved,the diver should return to the surface. Over-thecountervasoconstrictor medications (for example,pseudoephedrine) both orally and via nasal installationcan reduce congestion and may allowthe diver to continue the dive. In most cases,the aforementioned measures are sufficient tofacilitate equilibration of pressure in the middle-earspaces and sinus cavities and to continue divingactivities.If symptoms persist or the ear drum perforates,the diver should be evaluated by a physician. Fluidmay need to be drained from the middle-ear space(tympanotomy). If a perforation occurs, the divermust not re-enter the water until the ear drum hashealed (ie, approximately 2 to 3 weeks) because ofthe risk of inducing infec tion into the middle-earspace. If the drum is perforated, antibiotics areprescribed in conjunction with decongestants. Ifmiddle-ear equilibration problems recur, antihistaminesfor those with allergies may be helpful.Any persistent vertigo, dizziness, nausea, or hearingloss symptoms are serious and warrant immediatemedical attention. They may be caused byround or oval window ruptures or injury to theinner-ear structures from DCS.Extraalveolar Air (<strong>Pulmonary</strong> Overpressure) Syndromes:Problems arising from extraalveolar airare of three types: subcutaneous/mediastinalemphysema, pneumothorax, and arterial gasembolism (AGE). These occur infrequently. AGEis one of the most serious of all diving problemsbecause immediate recompression may berequired to prevent irreparable brain damage.AGE is not related to the depth or the durationof the dive. It has been reported after breathing acompressed gas from depths as shallow as eightfeet and then breath-holding while coming to thesurface. Extraalveolar air syndromes are causedby air retention in the lungs either as the resultof breath-holding or from pathological conditionssuch as asthma or emphysematous blebs. Asambient pressure decreases with ascent, gas in thelung expands, as explained by the Boyle law.Subcutaneous/mediastinal emphysema occurswhen alveoli rupture and air dissects into the tissuesof the mediastinum, upper trunk, and/orneck. If the air dissects into the pleural space, apneumothorax may result. Finally, if air expansionin the lungs occurs explosively as the result of anuncontrolled ascent, concomitant rupture of alveoliand adjacent blood vessels allow air to pass directlyinto the bloodstream. The air bubbles becomeemboli, are carried to the brain, and occlude itscirculation. Consequently, the symptoms of AGEcan be the same as any of those seen with a cerebralvascular accident.Cessation of diving activities until the signsand symptoms of subcutaneous/mediastinalemphysema resolve is usually all that is required734 Diseases Related to High Altitude and Diving (Aksenov and Strauss)ACC-PULM-09-0302-042.indd 7347/13/09 6:46:25 PM


to manage these problems. Rarely, these conditionsinterfere with breathing. If so, oxygen inhalationshould be instituted. For pneumothorax and AGE,breathing of pure oxygen should be initiated immediately.The symptoms of AGE remit completely inapproximately 50% of patients while breathingpure oxygen. After they are extricated from thewater, possibly with pneumothorax and AGE, andare awaiting transportation to a medical center,patients may require first-line treatment for shock,seizures, and cardiac arrest. Re-expansion of thelung with placement of a chest tube is required fora pneumothorax that has caused 20% collapse ofthe lung. Hyperbaric oxygen recompression therapyis the definitive treatment for AGE.Two measures prevent extraalveolar air syndromes.First, medical screenings before startingdiving should be performed to detect asthma andother chronic lung conditions that may trap airduring ascent. Second, dive training to teach buoyancycontrol and avoiding panic are essential toprevent unscheduled, uncontrolled ascents. Themajority of AGE events occur in inexperienceddivers.DCS: DCS, which is sometimes referred to asthe bends or caisson disease, is caused by bubbleformation in tissues. The Dalton law (ie, the totalpressure in a gas system is the sum of the partialpressures of the contained gases) explains howchanges of pressure with descent and ascent alterthe partial pressures of the gases in the breathingmedium. For example, with a dive to a depth of 33feet (10 m; equivalent to 2 absolute atmospheres[atm abs]) the partial pressures of the oxygen andnitrogen in the breathing medium double. Ordinarily,this has little effect on oxygenation of tissues,but if the pressure is increased 10-fold, theincreased Po 2could lead to a seizure as the resultof oxygen toxicity. An increase in the partial pressureof nitrogen has different effects. This inert gasis forced into all body tissues in direct proportionto its partial pressure in the breathing mediumHenry law. The rate of on-gassing (process of gasentering the tissue due to the increased pressure)depends on depth (pressure gradient), duration,gas mixture, and perfusion (blood flow). If thepressure of nitrogen is so great that it exceeds theability of the tissue to off-gas (process of gasleaving the tissue due to the decreased pressure)it in an orderly fashion, bubbles form in the tissue,and DCS results. As the diver ascends, the bubblesexpand (the Boyle law).Presentations of DCS are variable and notalways predictable. DCS has a wide variety ofpresentations, from skin itches to joint pains, andfrom paralysis to unconsciousness. Sometimes,the symptoms of DCS are so similar to those ofAGE that it is difficult to make the exact diagnosis.In such situations, the term decompression illnessis used to encompass both possibilities. The symptomsof DCS arise from the location in which thebubbles form. If bubbles form in the extremelyfast (with respect to on-gassing and off-gassing ofnitrogen) lung tissues where it is exchanged withthe ambient air, respiratory distress termed thechokes occurs. If bubbles form in the blood streamduring transport of the nitrogen released from thetissues to the lungs and block circulation, bendsshock or symptoms caused by the occlusion of theblood supply to critical organs such as the heart,brain, and spinal cord occur. These presentationsare similar to those of AGE; hence, the utilizationof the term decompression illness. If bubbles formin peripheral and/or noncritical tissues, itching,rash, joint pain, paresthesias, and/or fatigueoccur.As in AGE, the breathing of pure oxygen shouldbe instituted as soon as the diagnosis of DCS issuspected. If the patient is alert, fluid administrationand ingestion of a single dose of an antiplateletagent such as aspirin are recommended. Oxygenbreathing helps to “wash out” the nitrogen,whereas fluids and aspirin help to maintain thecirculation of blood. The diver should be kept atrest. In remote areas, returning to the water andbreathing pure oxygen at a depth of 33 feet andthen gradually ascending, although controversial,is recommended by some authorities in divingmedicine.Hyperbaric oxygen recompression is thedefinitive treatment for DCS. If there are residualsymptoms after the initial hyperbaric oxygenrecompression treatment, repetitive treatmentsshould be administered until the symptoms resolvecompletely or plateau over a 3- to 7-day period.DCS is prevented by the use of safe diving practices.Factors such as fitness, adequate hydration,and following dive computer (or dive tables) profilesare fundamental to safe diving. Other specificmeasures include the following: (1) limiting ascent<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 735ACC-PULM-09-0302-042.indd 7357/13/09 6:46:25 PM


ates to 1 foot every 2 s; (2) taking a 3-min rest stopat 15 feet; (3) avoiding repetitive dives at the end ofthe day, when fatigued and/or chilled; (4) taking a1-day break after 3 or 4 days of repetitive, consecutivedives each day; (5) not flying for at least 24 hafter the last SCUBA dive; (6) avoiding alcoholicbeverages during the dive activities; and (7) andusing common sense to avoid interfering with theorderly off-loading of nitrogen; for example, notwearing constrictive bands, falling asleep with jointsin hyperflexed positions, or exercising excessively.Other Indirect Effects of Pressure: Nitrogen Narcosis,Oxygen Toxicity, Carbon Dioxide Toxicity, CarbonMonoxide Poisoning, and High-Pressure NervousSystem Syndrome: These problems occur infrequently,are invariable associated with the bottomphase of the dive, and are usually associatedwith the use of special equipment such as closedcircuit(rebreather) SCUBA or special gas (pureoxygen, nitrox [enriched air nitrox] or heliumoxygenmixtures). A summary of these conditionsis presented in Table 2.Near-Drowning and Drowning: Any time consciousnessis lost in the water, the victim is atrisk of drowning. Aspiration often is associatedwith drowning because the breathing reflexes areusually the last to remain after a hypoxic injuryto the brain. If the victim is rescued and resuscitated,then the incident is labeled a near-drowning.Whereas aspiration and the resulting lung injurycan be managed effectively with appropriateinterventions, the consequences of hypoxic braininjury are usually irreversible and can range fromimperceptible to a persistent vegetative state.The brain is the most critical organ in the bodywith respect to injury from oxygen deprivation.After 4 min of anoxia, brain injury is likely to occur.Although breath-holding for several minutes isTable 2. Additional Problems Associated With Indirect Effects of PressureCondition Etiology Symptoms Management PreventionNitrogen narcosis,ie, “rapture of thedeep”Narcotic effects frombreathing this gasunder pressure. Each50 feet of descent isroughly equivalent tothe effects from drinkingone martini.Confusion, irrationalactions, stupor, syncope.Victim may panicand ascend in anuncontrolled fashion.Complications includeAGE, DCS, and/ordrowning.Anxiety, tunnel vision,tinnitus, muscletwitching, andseizures.Ascent; the dive buddyneeds to control thevictim’s ascent toavoid complications.Symptom resolveimmediately withascent.Adhere to safe depthlimits ( 130 feet).Use helium fordeep dives.Oxygen toxicity, ie,“oxygen hit”Breathing oxygen underincreased pressure.Seizure threshold isa function of depth,duration, activity, andthe environment, eg,cold water, poor visibility,currents.Increased carbon dioxidefrom problemswith carbon dioxideabsorber or work inunventilated spaces(tunnel and caissonworkers).Ascent, breathe gas withlower oxygen content,eg, air instead ofpure oxygen or nitroxmixtures.Adhere to strictdepth limits whendiving with pureoxygen (33 feet for30 min) and tablelimits with nitroxmixtures.Carbon dioxidetoxicityHeadache, increasedrespiratory rate, feelingsof suffocation,syncope. Burns inmouth from soda lime.Ventilate by breathingfresh air. Terminatedive if rebreather(with carbon dioxideabsorber) equipmentis used.Strict adherence tosafety and set-upprotocols whenusing rebreatherequipment. Ensureadequate ventilationwhen workingin closed spaces.Ensure gas supplyis free ofcontamination.Carbon monoxidepoisoningContamination of airsupply with exhaustfumes from internalcombustion enginesDiving deeper than 650feet with heliumoxygenmixturesHeadache, confusion,collapse, syncope,coma, and deathSurface; breathe pureoxygen; hyperbaricoxygen;High-pressurenervous systemsyndromeTremors, in-coordination,syncopeAscentAdd small amountsof nitrogen to thebreathing gas.736 Diseases Related to High Altitude and Diving (Aksenov and Strauss)ACC-PULM-09-0302-042.indd 7367/13/09 6:46:25 PM


possible, the interruption of the perfusion of thebrain for more than a moment or so, for example,in a third-degree heart block, leads to immediateloss of consciousness. Conversely, loss of consciousnessassociated with water immersion has“protective effects” in terms of preservation ofbrain function. This is ascribed to the oxygen-conservingeffects of the diving reflex and hypothermiafrom immersion in cold water. Remarkable recoverieshave been reported from near-drowning, evenafter the victim has been immersed for 30 min.Consequently, the adage “never say drowned”is very appropriate for two reasons: first, remarkablerecoveries that have been observed; and second,the primary cause of the patient’s presentationshould be identified. Once the primary cause isidentified, appropriate management becomes logical.For example, if the diver loses consciousnesswhile at the bottom phase of the dive, then hyperbaricoxygen recompression therapy is essential tooffload the inert gas present in the tissues and preventcomplications of decompression illness (ie,DCS and/or AGE ). In addition, concurrent optimalmanagement of the pulmonary and brain injuriesfrom near-drowning must be administered.Basics of HBOTMechanisms of Hyperbaric OxygenThe Undersea and Hyperbaric Medical Societygives the following definition of HBOT: The patientbreathes 100% oxygen intermittently while the pressureof the treatment chamber is increased to 1atm abs. Treatment pressures range from 1.4 to 3 atmabs. Treatments may be performed in a single-personchamber (monoplace) or multiplace chamber(ie, may hold two or more people). Breathing 100%oxygen at 1 atm abs or exposing isolated parts of thebody to 100% oxygen does not constitute HBOT.Hyperbaric oxygen has two primary mechanismsthat account for the therapeutic and theundesirable side effects of this treatment modality:(1) the mechanical effect caused by pressurization(bubble size reduction a la the Boyle law), and (2)hyperoxygenation. A hyperbaric oxygen exposureat 2 atm abs increases plasma and tissue oxygentensions 10-fold, the blood oxygen content by125%, and the diffusion of oxygen through tissuefluids by a threefold factor. The main side effectfrom the mechanical effect is barotrauma as discussedunder the section “Diving <strong>Medicine</strong>.” Ahyperbaric oxygen toxicity seizure is a very rarelyoccurring side effect from hyperoxygenation.Secondary mechanisms occur as a result of tissuehyperoxygenation but are clear and distinctfrom hyperoxygenation itself. The secondary effectsinclude, but are not limited, to the following: (1)vasoconstriction to reduce posttraumatic edema; (2)enhancement of host factor functions such as fibroblastactivity, angiogenesis, leukocyte oxidativekilling, and bone remodeling for wound healing thatwas unable to occur in the hypoxic environment; (3)microbiological effects such as bacteriostasis, cessationof toxin production, toxin inactivation, andaugmentation of certain antibiotics; (4) mitigationof the reperfusion injury by perturbing the neutrophil-capillaryendothelium adhesion mechanism;and (5) gas washout for management of DCS andcarbon monoxide poisoning.Indications for HBOTThe following 13 indications are approved usesof hyperbaric oxygen therapy as defined by theHyperbaric Oxygen Therapy Committee of theUndersea and Hyperbaric Medical Society:• Air or gas embolism;• Carbon monoxide poisoning, carbon monoxidepoisoning complicated by cyanide poisoning;• Clostridial myositis and myonecrosis (gasgangrene);• Crush injury, compartment syndrome, andother acute traumatic ischemia;• DCS;• Enhancement of healing in selected problemwounds;• Exceptional blood loss (anemia);• Intracranial abscess;• Necrotizing soft-tissue infections;• Osteomyelitis (refractory);• Delayed radiation injury (soft-tissue and bonynecrosis);• Skin grafts and flaps (compromised); and• Thermal burns.Contraindications for HBOTAbsolute contraindications for HBOT includethe following:<strong>ACCP</strong> <strong>Pulmonary</strong> <strong>Medicine</strong> <strong>Board</strong> <strong>Review</strong>: <strong>25th</strong> <strong>Edition</strong> 737ACC-PULM-09-0302-042.indd 7377/13/09 6:46:25 PM


1. Untreated pneumothorax: Air in the pleuralspace can increase in size as pressure decreasesduring ascent.2. The concomitant administration of doxorubicinor cis-platinum: HBOT can increase thecytotoxic effect of these medications.3. Disulfiram: This medication blocks the productionof superoxide dismutase, a protectiveantioxidant against oxygen toxicity.4. Bleomycin: This drug increase pulmonarytoxicity in patient undergoing HBOT.5. Mafenide acetate: A carbonic anhydrate inhibitorthat is used in burn wounds. It causes metabolicacidosis as the result of carbon dioxidebuildup, causing peripheral vasodilation.6. Premature infants: Oxygen in premature infantscan cause retrolental fibroplasia, leadingto blindness.Suggested ReadingHigh-Altitude--Related DisordersBarry PW, Pollard AJ. Altitude illness. BMJ 2003; 326:915–919Bärtsch P, Maggiorini M, Ritter M, et al. Prevention ofhigh-altitude pulmonary edema by nifedipine. N EnglJ Med 1991; 325:1284–1289Gallagher SA, Hackett PH. High-altitude illness.Emerg Med Clin North Am 2004; 22:329–355Ghofrani HA, Reichenberger F, Kohstall MG, et al.Sildenafil increased exercise capacity during hypoxiaat low altitudes and at Mount Everest base camp: arandomized, double-blind, placebo-controlled crossovertrial. Ann Intern Med 2004; 141:169–177Grocott MP, Martin DS, Levett DZ, et al. Arterial bloodgases and oxygen content in climbers on Mount Everest.N Engl J Med 2009; 360:140–149Hackett PH, Roach RC. High-altitude illness. N Engl JMed 2001; 345:107–114Sartori C, Allemann Y, Duplain H, et al. Salmeterolfor the prevention of high-altitude pulmonary edema.N Engl J Med 2002; 346:1631–1636Schoene RB. Illnesses at high altitude. Chest 2008;134:402–416Air Travel in Hypoxemic PatientsAerospace Medical Association, Air Transport <strong>Medicine</strong>Committee. Medical guidelines for air travel, secondedition. Aviat Space Environ Med 2003; 74:A1–A19British Thoracic Society Standards of Care Committee.Managing passengers with respiratory disease planningair travel: British Thoracic Society recommendations.Thorax 2002; 57:289–304Dine CJ, Kreider ME. Hypoxia altitude simulation test.Chest 2008; 133:1002–1005Mohr LC. Hypoxia during air travel in adults with pulmonarydisease. Am J Med Sci 2008; 335:71–79Mortazavi MJ, Eisenberg MJ, Langleben D, et al. Altitude-relatedhypoxia: risk assessment and managementfor passengers on commercial aircraft. AviatSpace Environ Med 2003; 74:922–927Diving-Related DisordersBrubakk AO, Neuman TS, eds. Bennett and Elliott’sphysiology and medicine of diving. Edinburgh: Saunders,2003Melamed Y, Shupak A, Bitterman H. Medical problemsassociated with underwater diving. N Engl J Med 1992;326:30–35Strauss MB, Aksenov IV. Diving medicine: questionsphysicians often ask. Consultant Live. Part I: 2004;44:961–963; and Part II: 44:1167–1171. Available at:http://www.consultantlive.com/display/article/10162/34671. Accessed April 23, 2009Strauss MB, Aksenov IV. Diving science: essentialphysiology and medicine for divers. Champaign, IL:Human Kinetics, 2004Strauss MB, Aksenov IV. Medical problems of divingand the primary care physician: Part I. Primary CareReports 2002; 8:164–171Strauss MB, Aksenov IV. Medical problems of divingand the primary care physician: Part II. Primary CareReports 2002; 8:172–188Strauss MB, Aksenov IV. Medical problems of divingand the primary care physician: Part III. Primary CareReports 2002; 8:185–191HBOTKindwall EP, Whelan H, eds. Hyperbaric <strong>Medicine</strong>Practice, 3rd ed. Flagstaff, AZ: Best Publishing Company,2008The Undersea and Hyperbaric Medical Society(UHMS). Hyperbaric Oxygen Therapy Committee.Guidelines: indications for hyperbaric oxygen. Kensington,MD: UHMS, 2000. Available at: http://www.uhms.org/ResourceLibrary/Indications/tabid/270/Default.aspx. Accessed March 19, 2009738 Diseases Related to High Altitude and Diving (Aksenov and Strauss)ACC-PULM-09-0302-042.indd 7387/13/09 6:46:26 PM


NotesACC-PULM-09-0302-042.indd 7397/13/09 6:46:26 PM


NotesACC-PULM-09-0302-042.indd 7407/13/09 6:46:26 PM

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!