Treatment of invasive candidiasis in non-haematological patients

EpidemiologyMainly data from the USA or from individual EuropeancountriesTrick Clin Infect Dis 2002

Candidemia incidence in LeuvenLagrou, EJCMID 2007

Survival in candidemia, species relatedParapsilosisOther speciesAlbicansGlabrataTropicalisKruseiHorn Clin Infect Dis 2009

Question1 : Which Candida species is mostfrequently cultured from blood next to C.albicans (in Belgium)?001 - C. glabrata2 - C. krusei3 - C. parapsilosis4 - C. tropicalis

Question1 : Which Candida species is mostfrequently cultured from blood next to C.albicans (in Belgium)?1 - C. glabrata0%2 - C. krusei0%3 - C. parapsilosis0%4 - C. tropicalis0%

| Advisory board Anidulafungin, Nov 8 th 2012Candida blood stream infectionsMean 22% C. glabrata candidiaemiaWhat would you start as primary therapy?Lagrou, EJCMID 2007

Incidence and mortality of Candida Speciesin EuropePercentage patients80%70%60%50%40%30%20%10%0%56%37%50%14% 13%28%Candida speciesHospital mortality7%43%59%2% 1%25%C. albicans C. glabrata C. C. tropicalis C. krusei C. lusitaniaeparapsilosisTortorano, EJCMID 2004;23:317-22Wispinghoff, CID 2004;39:309-17

Time to start antifungal therapy& hospital mortalityMorrell M, et al. AAC 2005157 patients – 2001-2004Independent determinants of mortality:‣ APACHE II score (one-point increments) (p 12 hoursafter the first positive blood culture (AOR, 2.09; p= 0.018)4035Crude mortality rate is high,the attributable mortality ratevaries between 5-71%!Falagas EJCMID 2006Hospital Mortality (%)Garey KW et al, CID 2006230 patientsIndependent determinants of mortality:‣ APACHE II score (∆1-pt.; p

HOW TO IDENTIFY PATIENTSAT RISK FOR CANDIDEMIA?

Question 2: Which factor has the greatestindependent association with developmentof invasive candidiasis ?001 - Abdominal surgery at baseline2 - Multifocal colonization3 - Severe sepsis at baseline4 - Total parenteral nutrition

Question 2: Which factor has the greatestindependent association with developmentof invasive candidiasis ?1 - Abdominal surgery at baseline0%2 - Multifocal colonization0%3 - Severe sepsis at baseline0%4 - Total parenteral nutrition0%

Prediction rules- Paphitou: TPN, hemodialysis, diabetes, broadspectrum antibiotics- Ostrosky-Zeichner:- CVC +/- broad spectrum antibioticsand at least 2 of:- TPN, hemodialysis, major surgery, pancreatitis,corticosteroids, immunosuppression- Candida score: Multivariate analysis of risk factorsthat predispose for invasive candidiasis

Identification of patients“Ostrosky rules”≥ 4 days on ICUAND-Antibiotics (d 1-3)OR-CVC (d 1-3)And at least 2 of theseTPN (d 1-3)Dialysis (d 1-3)Major surgery (d -7-0)Use steroids (d -7-3)Immunosupp. (d -7-0)Ostrosky-Zeichner, 2007;26:271

Identify patients for empirical/pre-emptive therapyDevelopment of “Candida score”1699 patients, ≥ 7 days on ICMultivariate analysis of risk factors for candidemiaweighto Colonization on multiple locations 1.112 1o TPN 0.908 1o Surgery at admission 0.997 1o Severe sepsis 2.038 2pointsLeón, CCM 2006;34:730

Question3: What’s the problem withthe prediction rules to start empiricaltreatment?1 - High sensitivity, low specificity002 - Low sensitivity, high specificity3 - Only useful in high incidence areas4 - 1+35 - 2+3

Question3: What’s the problem withthe prediction rules to start empiricaltreatment?1 - High sensitivity, low specificity0%2 - Low sensitivity, high specificity0%3 - Only useful in high incidence areas 0%4 - 1+35 - 2+30%0%

Identification of patients“Ostrosky rules”≥ 4 days on ICUAND-Antibiotics (d 1-3)OR-CVC (d 1-3)And at least 2 of theseTPN (d 1-3)Dialysis (d 1-3)Major surgery (d -7-0)Use steroids (d -7-3)Immunosupp. (d -7-0)Sensitivity 34%, specificity 90%,PPV 1%, NPV 97%Prevalence 7% invasive candidiasisOstrosky-Zeichner, 2007;26:271

Identify patients for empirical/pre-emptive therapyDevelopment of “Candida score”1699 patients, ≥ 7 days on ICMultivariate analysis of risk factors for candidemiao Colonization on multiple locations 1o TPN 1o Surgery at admission 1o Severe sepsis 2points3: sensitivity 81%, specificity 74%PPV 16%, NPV 98%Prevalence 5.8%León, CCM 2006;34:730

Are the risk predictive models generalizable?Geographical variability in epidemiology of IC, case-mix &medical practicesValidation in Australia:Playford Int Care Med 2009

Question 4: Empirical treatment is most ofteninappropriate in septic ICU patients infectedwith…001 - Staphylococcus aureus2 - Pseudomonas aeruginosa3 - Klebsiella species4 - Candida albicans

Question 4: Empirical treatment is most ofteninappropriate in septic ICU patients infectedwith…1 - Staphylococcus aureus0%2 - Pseudomonas aeruginosa0%3 - Klebsiella species0%4 - Candida albicans0%

Appropriate empirical treatment in ICUKumar Chest 2009

HOW TO PREVENT INVASIVECANDIDIASIS IN THE ICU?

Antifungal prophylaxis in the ICU?Pro:-Reduces the incidence ofinvasive candidiasis!with app. 50%-Again: identification oftarget patients difficult-High risk patients-In high incidence area-Necrotizing pancreatitis-Recurrent GI leakageCon:- Unselected ICUs: 1-2% invasivecandidiasis- In most hospitals;Ineffective strategy

Empirical antifungal therapy in ICURCT Prospective study in ICU patientsInclusion:•Unexplained fever > 4 d•Broadspectrum antibiotics ≥ 4 d•APACHE-II score ≥ 16•CVC ≥ 1 d(no colonization criteria!) 14d Fluconazole (800mg/d) or placeboSchuster, Ann Intern Med. 2008

Outcome empirical therapy in ICUOutcome:•resolution of fever (72h)•no invasive mycosis•no discontinuation for toxicity•no other antifungal drugsTotal failures:55% vs 57%No resolution of fever51% vs 54%Fluconazole (122) Placebo (127)Success 34% 38% (0.69-1.32)Inv. Mycoses 5% 9% (0.22-1.49)Mortality 24% 17% (0.23-1.67)No effect of empirical fluconazole! Why?

TREATMENT OF INVASIVECANDIDIASIS

Time to start antifungal therapy& hospital mortalityMorrell M, et al. AAC 2005157 patients – 2001-2004Independent determinants of mortality:‣ APACHE II score (one-point increments) (p 12 hoursafter the first positive blood culture (AOR, 2.09; p= 0.018)4035Crude mortality rate is high,the attributable mortality ratevaries between 5-71%!Falagas EJCMID 2006Hospital Mortality (%)Garey KW et al, CID 2006230 patientsIndependent determinants of mortality:‣ APACHE II score (∆1-pt.; p

Why do we choose whichantifungal drug?• Broad spectrum of antifungal activity• Efficacy• Patient characteristics• Safety / interactions

Antifungal agentsAmphotericin B• deoxycholate AmB• lipid formulationsEffective, broad spectrum,but toxicLipid: expensiveAzols• Fluconazole• voriconazoleEchinocandins• Caspofungin• Anidulafungin• MicafunginEffective, less toxic, less broadspectrumInteractionsEffective, broad spectrumLittle toxicity, few interactionsExpensive

Randomized Non-Inferior TrialsAmB vs fluconazole: equally effective (3 trials)AmB/fluc vs Fluc: equally effective Rex, 2003AmB -> fluc vs Vori: equally effective Kullberg, 2005Caspofungin vs AmB: equally effective Mora-Duarte, 2002Micafungin vs L-AmB: equally effective Kuse, 2007Mica vs Caspo: equally effective Pappas, 2007Anidulafungin vs fluc: favours Anidula Reboli, 2007

SusceptibilityAntfungal Albicans Glabrata Parapsilosis TropicalisKruseiAmB-d S S S S SL-AmB S S S S SFluconazole S (SDD)-R S S Rvoriconazole S SDD-R S S SAnidulafungin S S S/? S SCaspofungin S S S/? S Smicafungin S S S/? S S

Caspofungin vs. Amphotericin B trialTime to First Negative Blood Culture100%90Caspofungin70/50 mgAmphotericin B0.6-1.0 mg/kg8070605040302010Success (MITT) 73% 62% P=0.09n=224Crude Mortality 34% 30% P=0.53ToxicityDiscontinuations3% 23% P=0.03Caspofungin (N=92)Amphotericin B (N=94)00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24STUDY DAYMora-Duarte NEJM2002

Caspofungin vs. AmB in ICU Patients (n=97)68%56%Response Rates45%40%5%11%DiNubile Crit Care 2007

Micafungin vs. L-AmB in ICU Patients (n=230)Dupont Crit Care 2009

Micafungin vs. L-AmB in ICU Patients (n=230)63% 66% 38% p=n.s.34%Dupont Crit Care 2009

Anidulafungin vs. fluconazole studyMITTpopulationAnidulafungin Fluconazole DifferenceN=127 (%) N=118 (%) 95% CISuccess EIV 96 (75.6) 71 (60.2) 15.4 (3.9-27.0)(p

Survival in ICU patients:anidulafugin vs fluconazoleKett, Int J Antimicr Ag 2008

Success rates different echinocandinsGlöckner Mycoses2009

Indications per echinocandinWagner, Pharmacology 2006

Question 5: Should the intravascularcatheters be removed < 24h? 001 - Yes, when > 24h mortality is higher2 - No, but < 48h; > 48h mortality is higher3 - No, but it should be removed4 - Removal of the catheter has no effect on mortality

Question 5: Should the intravascularcatheters be removed < 24h?1 - Yes, when > 24h mortality is higher0%2 - No, but < 48h; > 48h mortality is higher 0%3 - No, but it should be removed0%4 - Removal of the catheter has noeffect on mortality0%

Antifungal agents and survivalCandidemia and invasive candidiasis•7 randomized controlled trials•7 antifungal agents (flu, vori, amB, L-amB, caspo,anidula, mica)•New analysis of all individual patients•Primary endpoint: 30 day survival•Excluding: combination therapy, unknown candidaspecies, multiple species infection•1915 patients!Andes Clin Infect Dis 2012

Andes pooled data analysis1915 patients, two polyenes, two triazoles, three candinsLogistic regression using 30-day mortality as primary outcomeIncreased mortality• Age• APACHE II• Immunosuppressive therapy• C. tropicalisDecreased mortality• CVC removal during therapy• Echinocandin antifungal"These results support first-line treatment withan echinocandin to the majority of patients”Similar risk factors by subgroup: C. albicans, C. glabrata,all non-albicansAndes et al., Clin Infect Dis 2012

In conclusion• Invasive candidiasis still has a high mortality rate• Antifungal prophylaxis in ICU is effective butinappropriate• Empirical antifungal therapy in ICU is not proveneffective• Prediction rules to start pre-emptive therapy have ahign NPV, and a reasonable specificity. Especiallyuseful in high incidence areas/patients• Intravascular catheters should be removed/replacedwhenever feasible• Treatment with an echinocandin is preferred, for thehigher response rate

Treatment of invasive candidiasis innon-haematological patientsAML Oude Lashof, MD PhDInternist-infectiologist

Treatment of candidemia-Candida species and susceptibility-Localization of infection- Candidemia only- Disseminated disease-Mode of administration-Duration of candidemia-…

Empirical treatment is most ofteninappropriate in septic ICU patientsinfected with…1.Staphylococcus aureus2.Pseudomonas aeruginosa3.Klebsiella species4.Candida albicans

Success at EIV treatment by pathogenDifference driven by C. albicans infections!% Success75.6%∆15.4%P=0.00960.2%81.1%∆18.8%P=0.01562.3%AnidulafunginFluconazole71.1%∆11.1%P=0.2660.0%All patientsn=60 n=38 n=32 n=27C. albicans(N=74)(N=61)Non-albicans(N=45)(N=45)* Patients with a single baseline pathogenReboli NEJM 2007

Question 6: Should the intravascularcatheters be removed < 24h? 001 - Yes, when > 24h mortality is higher2 - No, but < 48h; > 48h mortality is higher3 - No, but it should be removed4 - Removal of the catheter has no effect on mortality

Question 6: Should the intravascularcatheters be removed < 24h?1 - Yes, when > 24h mortality is higher0%2 - No, but < 48h; > 48h mortality is higher 0%3 - No, but it should be removed0%4 - Removal of the catheter has noeffect on mortality0%

MortalityCrude mortality rate is high,the attributable mortality ratevaries between 5-71% dependingon the studyMorrell AAC 2005, Kumar Chest 2009

Candida Scorein practiceProspective observational multicenter study•≥7 days IC with multifocal colonization (1)•TPN(1)•Surgery (1)•Sepsis(2)Goal:

Results Candida Scorein practice1107 patients892: colonized / infected• 565: CS 13 (2.3%) invasive Candida infection• 327: CS ≥3 -> 45 (13.8%) invasive Candida infectionLeón, CCM 2009;37:1624