PCR-Based Detection of Pathogen DNA in Critical Illness: Septifast ...

PCR-Based Detection of Pathogen
DNA in Critical Illness:
Septifast and other……
Geoffrey Warhurst
1Infection, Infection, Injury & Inflammation Research Group, Salford Royal
Hospital NHS Trust; 2School School of Translational Medicine,
University of Manchester; 3Institute Institute of Biomedicine, University of
Salford UK

Aims
• The diagnostic challenge
- Clinical impact of sepsis in critical care
- Diagnosis – the need for speed
• PCR approaches to pathogen DNA detection
• Uncertainties
- Biology
- Broad range PCR – High Resolution Melt
- Septifast – clinical validity studies
- Clinical Meaning
- Clinical / Cost Efficacy

Background
Research collaboration
- responses to severe injury and infection
- new diagnostic/therapeutic approaches to help
patients survive and recover from intensive care
Near-patient research laboratories
Level 3 intensive care (17-bed, ~1000 patients)
Patients requiring multi-organ support and
treatment.
Regional centre for neuroinjury, complex
bowel surgery

INFECTION
Endothelial Damage
Tissue Injury
Susceptibility to secondary infection
Multi-organ Dysfunction / Death
SIRS = Systemic Inflammatory Response Syndrome

SEPSIS – Systemic Inflammatory Response
to Infection
Sepsis associated with increases in.…
ICU and hospital length of stay
frequency and duration of organ failures
mortality (20-50%; 135,000 European deaths)
costs €7.6 billion per annum
A growing problem (double mortality since 1990)
- aging population
- complex invasive procedures
- resistant pathogens

The Sepsis Continuum
SIRS
Infection-25% Injury-75%
A clinical response
arising from a
nonspecific insult, with
≥2 2 of the following:
T >38 oC C or 90 beats/min
RR >20/min
WBC >12,000/mm 3
or

Early and appropriate antibiotic treatment
is key to survival
Intravenous broad spectrum antibiotics to
be given within first hour of recognition of
clinical signs
Mortality increases by 8% for every hour of delay in antibiotic therapy
Kumar A et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant
of survival in human septic shock. Crit Care Med (2006)
Focus antibiotics to specific organism (de-escalate)
as quickly as possible
18% mortality rate in de-escalated patients vs 43% in non-de-escalated;
Rello J, et al. Crit Care Med 2004;32:2183–2190

Injury
(75%)
Infection
(25%)
Culture delays prevent effective diagnosis and
management of sepsis
CULTURE Day 1 Day 2 Day 3 Day 4
BC
Patient with >
2 SIRS
criteria
Broad spectrum Ab
Culture Gram
Positivity
BC BC BC
CULTURE
Other
samples
Species
Antibiotic
Susceptibility
Focus therapy

Detection of pathogen DNA in blood – what
could it offer?
• A rapid “rule-out” diagnosis ie. differentiate sepsis from SIRS?
- Reduce unnecessary antibiotic use
- Impact on emergence of antibiotic-resistant organisms
- Change clinical management
• A reliable and rapid “rule in” diagnosis of sepsis?
- Shorten time to positive diagnosis to ~6-8 hr
- Rapid identification of pathogen species involved
- Focussing of therapy

PCR Detection of Pathogen DNA
Bacteria -16S ITS
Amplification of hyper-variable regions
in ribosomal DNA.
Detect broad range of pathogens
Gram type Speciation [Resistance]
23S
Fungi - 18S 5.8S
Cursons et al “Use of PCR to detect septicemia in critically ill patients”
Crit Care Med 1999.

PCR Approaches
Detection (Y/N)
Blood
EDTA tube
Isolation of pathogen and human DNA
Amplification
Separation of pathogen DNA
Multi-pathogen PCR Broad range PCR
Pathogen Identity +
Septifast
VYOO
Sequencing
Microarray
Mass Spect
SepsiTest
Licensed for clinical use

Experimental Approaches
- Broad-range detection
BactScreen Toolset
- Rapid, cost-effective pathogen species identification
High Resolution Melting Analysis (HRMA)
- Clinical validity
SeptiFast

Broad Range PCR for Detection of Pathogen DNA
Broad Range 16S PCR primers +
Real time detection of 1ng – 25fg
E. coli DNA
Dual colour FRET probes
BACTSCREEN ToolSet
Melting
Analysis
Gr+ve
Gr-ve
Gram typing of amplified DNA

Ability to detect broad range of sepsis pathogens
7
6
5
Con B Con A
BACTSCREEN ToolSet; LightCycler480
4
1
2
3
Gram +ve Gram ‐ve
S. aureus (6) E. Coli (1)
CNS (7) Klebsiella pneum (2)
MRSA Pseudomonas aeruginosa
Streptococcus pneum. Proteus mirabilis
Enterococcus faecalis (5) Serratia marescens
Enterococcus faecium Hemophilus influenzae (3)
Acinetobacter baumanii (4)
Enterobacter cloacae
>97% of bacterial bloodstream
pathogens from Salford Royal ICU

Analytical Sensitivity
Limit of detection ~10 CFU/ml blood
Similar LOD for other
species
*70% of bacteremias are

High Resolution Melting Analysis (HRMA)
Rapid Detection and Identification of Clinically Important Bacteria by High
Resolution Melting Analysis after Broad Range Ribosomal PCR – Cheng et
al Clin. Chem. 52: 1997-2004 (2006)
2 nd broad range PCR run in presence of DNA saturating dye
(Resolight)
High resolution (16 bit) acquisition of melting data (LC480 Gene
Scanning software)
Detect single base pair differences
Sequences differentiated by Tm and/or melting shape
Low cost

HRMA Stage 1: Differentiation of pathogens by melting temperature
Group 1 Group 2 Group 3
Group Bacterial species Melting
temp (Tm)
Group 1 Coagulase negative staphylococci (CNS) 82.50-83.50
Staphylococcus epidermidis 82.50-83.50
MRSA 83.20-83.80
Staphylococcus aureus 83.00-84.20
Group 2 Klebsiella aerogenes 84.40-85.10
Enterobacter aerogenes 84.40-85.10
Serratia marcescens 84.40-85.30
Proteus Mirabilis 84.40-85.30
Klebsiella pneumoniae 84.40-85.45
Haemophilus influenzae 84.50-85.45
Group 3 Acinetobacter baumannii 85.00-86.20
Streptococcus agalactia 85.80-86.80
Escherichia coli 85.90-86.95
Enterobacter cloacae 85.90-86.90
Streptococcus pyogenes 86.00-86.90
Enterococcus faecalis 86.00-86.90
Streptococcus pneumoniae 86.20-87.00
Pseudomonas aeruginosa 86.60-87.15
Non-template control 87.30-87.95
16S PCR primers
modified from Cheng et
al, 2006
Species cover >97% of
ICU sepsis pathogens
Pathogens differentiate
into 3 groups according
to Tm

Stage 2: Difference Plots vs Reference Organisms
Reference strain = Klebsiella
Enterococcus faecalis ‐ A
Reference strain = Staph aureus
Enterococcus faecalis ‐ A
Different isolate of
Enterococcus faecalis
Different isolate of
Enterococcus faecalis
Highly reproducible pathogen “signatures”
Enterococcus faecalis ‐ B
Enterococcus faecalis ‐ B

Speciation of pathogens using difference plots
Ref: S. aureus Ref:Klebsiella
A. baum A. baum
S. agal
S. agal
E. coli E. coli
E.
faecalis E.
faecalis
4 species with similar Tm
2 reference strains
discriminates all 4 species
Discriminate 16/18
commonest ICU bacteria
Pattern recognition
software to automate
detection
Similar approach for fungal
pathogens

Group 1 Group 2 Group 3
Using Staphylococcus aureus
as reference bacterium against
group 1 for generating melting
shapes
No differentiation
Using Klebsiella pneumoniae as
reference bacterium against group
1 for generating melting shapes
Staphylococcus aureus
Coagulase negative
staphylococci (CNS)
Staphylococcus epidermidis
MRSA
Using Staphylococcus aureus as
a reference bacterium against
group 2 for generating melting
shapes
Haemophilus influenzae
Klebsiella aerogenes
Enterobacter aerogenes
Serratia marcescens
Proteus Mirabilis
Klebsiella pneumoniae
Using Klebsiella pneumoniae as
reference bacterium against the
remaining 5 species for generating
melting shapes
Proteus Mirabilis
Klebsiella aerogenes
Enterobacter aerogenes
Serratia marcescens
Klebsiella pneumoniae
Proteus Mirabilis as reference bacterium
Klebsiella pneumoniae
Klebsiella aerogenes
Enterobacter aerogenes
Serratia marcescens
Using Staphylococcus aureus
as reference bacterium against
group 3 for generating melting
shapes
Streptococcus agalactia
Acinetobacter baumannii
Streptococcus pneumoniae
Pseudomonas aeruginosa
Enterobacter cloacae)
Escherichia coli
Streptococcus pyogenes
Enterococcus faecalis
Using Klebsiella pneumoniae as
reference bacterium against the
remaining 3 species for generating
melting shapes
Escherichia coli
Streptococcus pyogenes
Enterococcus faecalis

Summary of BactScreen – HRMA Approach
• Potential for detection and species identification within 8 hours
~4 hr ~2hr
Clinical sample DNA Bactscreen No
HRMA
Yes
• Clinical utility? Limited clinical validity study in progress
• Preliminary data shows promise
2 hr
• Limitations: -polymicrobial samples
- currently LOD ~100 fg

Experimental Approaches
- Broad-range detection
BactScreen Toolset
- Rapid, cost-effective pathogen species identification
High Resolution Melting Analysis (HRMA)
- Clinical validity
SeptiFast

Septifast
• Multiplex PCR reporting a defined panel of pathogens causing bloodstream
infection in ICU
• DNA extracts from 1.5-3ml whole blood using ultrapure MGrade reagents
• Real time monitoring of PCR products using FRET hybridisation probes
• Closed system – reports presence/absence above pre-determined thresholds

Lehmann et al 2008
Laboratory analytical validity
Simultaneous detection and
speciation
Pathogen identification by melting
temperature
Limits of detection - 3-30 CFU/ml
Species dependent

SEPTIFAST – “RULE IN” STUDY OF BLOOD
STREAM INFECTION – SALFORD UK
Study: Diagnostic study of critically ill patients requiring blood
culture for suspected blood stream infection with SIRS
(i.e. sepsis)
Setting: Tertiary NHS-University NHS University intensive care (16 level 3 beds)
Infection prevalence: 12% (95% CI 6% - 16%) from recent study (n = 100)
Power calculation:
* 83 patient samples required to be 95% sure that
specificity is at least 95% Reporting 142 patient samples
Ethics: Service evaluation approved by LREC Chair

Clinical Evaluation of SeptiFast SeptiFast
vs Blood Culture
CULTURE +VE CULTURE –VE VE
PCR +VE 16 11
PCR –VE
VE 6 * 121
* Includes 3 assessed as BC
contaminants after review
Predictive value of a positive test – 59%
Specificity (“rule in” diagnosis) – 91%
Sensitivity (“rule out” diagnosis) – 84% (but not sufficient power)
First 90 samples reported in Dark PM, Chadwick P, Warhurst G Journal of Infection 59(4): 296-298; 2009

Micro-organism
Micro organism
Enterobacter
(cloacael cloacael / aerogenes)
aerogenes
Klebsiella
(pneumoniae
pneumoniae / oxytoca) oxytoca
Positive
BCulture and PCR
Positive
PCR only
Positive
BCulture only
4 4 0
3 1 0
Pseudomonas aeruginosa 2 3 1
Acinetobacter baumannii 1 0 1
E. coli 0 1 0
Staphylococcus aureus 1 2 0
CNS 2 0 3 *
S. penumoniae 1 0 0
G+ve bacillus sp. 0 0 1
Candida 2 0 0
Total 16 11 6 *

“False positives” - majority associated with organisms
cultured from other sites
Case type
ICU day
of
sample
Receiving
broad
spectrum
antibiotics
PCR
Blood
Cultures
Neuro-injury
Neuro injury 14 yes E. cloa/aero cloa/aero
negative
Neuro-injury
Neuro injury 7 yes E. cloa/aero cloa/aero
negative
Neuro-injury
Neuro injury 7 yes K. pneum/oxytoc negative
Neuro-injury
Neuro injury 13 yes E. cloa/aero cloa/aero
negative
Neuro-injury
Neuro injury 4 no P.aeruginosa negative
Pancreatitis 13 yes P. aeruginsoa negative
Pancreatitis 27 yes P.aeruginosa negative
Oesophagectomy 10 yes E. cloa/aero cloa/aero
negative
COPD 13 yes E. cloa/aero cloa/aero
negative
Other
supporting
culture
results
CVC tip positive
E. cloa
Nil (NBL and
urine negative)
CVC tip positive
K. pneum
CVC tip positive
E. cloa
Urine/NBL
negative
NBL positive for
P. aeroginosa
NBL positive for
P. aeroginsoa
NBL positive for
E. cloa
CVC tip positive
E. aero

Other Published Studies
[PCR vs BC]
Study Setting Pat/Samp Spec
Westh 2009 Suspected sepsis 359/558 76%
Multi-European ?
Bloos 2009 Severe sepsis 63/111 77%
France/Germany ICU
Wallet et al 2009 Suspected sepsis 72/100 88%
France ICU
Dierkes et al 2009 Suspected sepsis 77/101 82%
Germany ICU
Louie et al, 2008 Suspected sepsis 200/200 89%
US ICU, ER, general
Broadly similar findings

Culture and pathogen DNA are not equivalent measures of
infection
Live
pathogen
Dead
pathogen
Free DNA
PCR
Culture
Live pathogen
Lung
Urine
Catheters etc
CSF
= Phagocyte
Shedding of bacterial DNA from extracirculatory
sites into circulation

Is blood culture an appropriate “gold standard”?
Pathogen DNA probably better indicator of infection in general (ie ( ie sepsis) not
specifically bloodstream infection
“Constructed gold standard”
Blood
culture
positive
Lehmann et al, 2009 (453 patient samples) - Increased assay
specificity from 81% to 93%
Dark et al, 2009 (146 patient samples) Increased assay specificity
from 91% to 96%
+ Cultures from
other sites (urine,
lavage, CVC)

Where do we go from here?
Biology
Kinetics of pathogen DNAemia
Does it contribute to the immune inflammatory response
Technology Improvements
Faster, cheaper
Reduce false negatives
Clinical Value and Cost Efficacy
Clinical meaning of pathogen DNAemia
Intervention studies – change therapy based on pathogen DNA
Cost benefit analysis

Clinical diagnostic validity of rapid detection of healthcareassociated
blood stream infection in intensive care using
multi-pathogen real-time PCR technology
Stage 1: Septifast: >600 ICU patients with clinical suspicion of sepsis
Powered for “rule-in” and “rule-out” diagnosis
Standardised protocols for source identification etc
Sample archiving for assessment emerging technologies
Stage 2: Detailed Therapeutic Intervention / Cost Efficacy Studies
HTA project No:
08/13/16

Conclusions
Pathogen DNAemia a promising biomarker of sepsis
Rapid rule in detection/speciation (and potential rule out) offers promise
of more effective therapy and patient management
Pathogen DNA and culture tell us different but complementary things
about infection - we will need both!
More information on the biology of pathogen DNAemia
Technology adoption will not occur without robust clinical
intervention and cost effectiveness studies
Clinical confidence – the next 10 years???

Dr Paul Dark
Dr Paul Chadwick
Acknowledgements
Huda Al-griw Al griw – Bactscreen
Hani Ozbak – HRM
Pamela Davies – Septifast
Norman Higgs – Septifast
Thank you!