Faut-il traiter les malades colonisés à Candida sp. en réanimation? DESC Réanimation 6 fevrier 2011 Faut-il traiter les malades colonisés à Candida sp. en réanimation? Jean-François Timsit MD PhD Medical polyvalent ICU CHU Albert Michallon U 823 Grenoble, France

Questions Le traitement des candidoses invasives doit être précoce et efficace Faiblesse des tests diagnostiques La colonisation fongique est très fréquente et possède une VPP faible pour le diagnostic de candidémie Qui traiter avec quelles conséquences potentielles?

International Study of the Prevalence and Outcomes Of Infection in Intensive Care Units 17% Jean Louis Vincent et al – JAMA – 2009; 302(21):2323-2329

Candidémies en réanimation Raisin REACAT NNIS 2007 2000-2003 1990-1999 874 132 30701 939 % Bacteremia Colonization CVC SCN 22,2 34 37,3 35,5 Enterobacteria 28,2 16,3 10,6 27,8 S. aureus 13,6 28,8 12,6 9,1 Candida spp. 9,9 6,5 5 6,6 P. aeruginosa 9,4 8,5 3,8 12,2 Enterococcus 6,4 2,6 13,5 3,6 Other BGN 3,4 2 2,2 Streptococcus 2,4 1,3 0,8 Others 4,2 17,2 1,9

Nosocomial BSI in ICU OR=8.83 Background. Overall rates of bloodstream infection (BSI) are often used as quality indicators in intensive care units (ICUs). We investigated whether ICU-acquired BSI increased mortality (by 10%) after adjustment for severity of infection at ICU admission and during the pre-BSI stay. Methods. We conducted a matched, risk-adjusted (1:n), exposed-unexposed study of patients with stays longer than 72 h in 12 ICUs randomly selected from the Outcomerea database. Results. Patients with BSI after the third ICU day (exposed group) were matched on the basis of risk-exposure time and mortality predicted at admission using the Three-Day Recalibrated ICU Outcome (TRIO) score to patients without BSI (unexposed group). Severity was assessed daily using the Logistic Organ Dysfunction (LOD) score. Of 3247 patients with ICU stays of 13 days, 232 experienced BSI by day 30 (incidence, 6.8 cases per 100 admissions); among them, 226 patients were matched to 1023 unexposed patients. Crude hospital mortality was 61.5% among exposed and 36.7% among unexposed patients ( ). Attributable mortality P ! .0001 was 24.8%. The only variable associated with both BSI and hospital mortality was the LOD score determined 4 days before onset of BSI (odds ratio [OR], 1.10; 95% confidence interval [CI], 1.03–1.16; Pp.0025). The adjusted OR for hospital mortality among exposed patients (OR, 3.20; 95% CI, 2.30–4.43) decreased when the LOD score determined 4 days before onset of BSI was taken into account (OR, 3.02; 95% CI, 2.17–4.22; P ! .0001). The estimated risk of death from BSI varied considerably according to the source and resistance of organisms, time to onset, and appropriateness of treatment. Conclusions. When adjusted for risk-exposure time and severity at admission and during the ICU stay, BSI was associated with a 3-fold increase in mortality, but considerable variation occurred across BSI subgroups. Focusing on BSI subgroups may be valuable for assessing quality of care in ICUs. Garrouste-Orgeas et al – Clin Infect Dis – 2006; 42:1118

Delay in instauration of appropriate treatment is prejudicial to candidaemic ICU-patients early treatment (≤ 48 h) is associated with a better survival rate Nolla-Salas J et al. Intensive Care Med 1997; 23: 23-30 Late treatment is independently associated with death (odds ratio 1,52 ; p < 0,05) 50 45 40 35 Mortality % 30 25 20 15 10 5 Culture Day 1 Day 2 Day ≥ 3 day Days to start of fluconazole Garey KW et al. Clin Infect Dis 2006; 43: 25-31.

Early antifungal treatment: a mouse model Example of the 105 challenge Challenge with 2 X 104 and 105 C albicans intravenously Treatment at D-1, Day0, Day1, 2 and 3 AmB, Flu, CAS 12 mice per experiments CAS FIG. 2. Survival of mice infected i.v. with a challenge dose of 1 105 CFU of C. albicans/g and treated with (a) amphotericin B (1 mg/kg/day), (b) caspofungin (1 mg/kg/day), and (c) fluconazole (10 mg/kg/day). Filled circles, placebo treatment only; open circles, treatment started on day 1; filled squares, treatment started on day 0; open squares, treatment started on day 1; filled triangles, treatment started on day 2. When data points overlapped, they were adjusted by 1 or 2% from actual values to make all curves visible. Placebo  Start d-1  Start d0  Start D+1  Start D+2  FLU Mac-Callum AAC; 2004: 4911-4914

Patient en réanimation Patients en réanimation: plusieurs facteurs de risque Intervention chirurgicale Sepsis Antibiotiques Patient en réanimation Durée de séjour Sévérité Alimentation parentérale Patient identique à tout patient mais défaillances d ’organe Gravité = mortalité hsopitalière scores à l ’admission scores en cours d ’hospitalisation Type de pathologie : populations différentes Procédures : ventilation (50 à 60%) Autres facteurs de risque (Wey) : antibiotiques (60% à 70 %), hémodialyse, alimentation parentérale... MSJ Procédures “invasives“ - cathéters vasculaires - sonde vésicale - intubation Hémodialyse

Questions Le traitement des candidoses invasives doit être précoce et efficace Faiblesse des tests diagnostiques La colonisation fongique est très fréquente et possède une VPP faible pour le diagnostic de candidémie Qui traiter avec quelles conséquences potentielles?

Timing to positivity of Blood cultures in humans Eukariotic cells Slow growth, weak CO2 production time to positivity Se detection Usual aerobic. anaerobic bottles, automated systems Simulation 50 X 2 BC 103 CFU 10 ml We studied the ability of the BACTEC 9240 automated blood culture system to detect simulated candidemia, including both Candida albicans and non-albicans Candida species. Simulated blood cultures were produced using 50 Candida isolates and BACTEC Plus Aerobic/F and Anaerobic/F blood culture bottles. Ten milliliters of blood and a suspension of each isolate containing 1,000 CFU were introduced into each bottle and then incubated at 35°C in the BACTEC 9240 system. The system detected growth in 56 of 100 bottles. Four isolates did not have growth detected in either bottle after 21 days of incubation, resulting in four missed episodes of candidemia. If the blood culture bottles had been incubated for 5 days, an additional episode of candidemia would have remained undetected. If the bottles had been incubated for only 3 days, another episode would have been missed, resulting in up to six missed episodes of candidemia (four Candida glabrata isolates, one C. albicans isolate, and one Candida rugosa isolate). Terminal subculture of bottles without detected growth recovered yeast in 93% (41 of 44) of the bottles, representing 41 false negatives. In bottles where growth was detected, the time to detection was 24 h. However, the mean time to growth detection for C. glabrata isolates in anaerobic medium was 22.14 2.47 h, but it was 120.89 35.33 h in aerobic medium (P < 0.001). The BACTEC 9240 system detected growth of most Candida isolates; however, the delayed time to detection of C. glabrata is clinically significant. Given the high rate of false negatives, terminal subcultures may be helpful in certain situations. 4/50 missed Role of terminal subcultures of negative bottles? Horvarth et al - JCM 2003:4714

Timing to positivity of blood cultures in humans: the role of selective bottles Eukariotic cells Slow growth, weak CO2 production  time to positivity  Se detection Selective bottles: Better sensitivity + 24% (92.5 vs 75.9%) Especially for C glabrata + 42% (100 vs 58.1) Or concomitant bacteria + 53% (79.9 vs 26.9%) Decrease in the mean time to positivity All yeasts: 28.9 h vs 36.5 h C glabrata: 17.8 h vs 61.5 h Fungemia is associated with a high mortality rate. We compared the performance of the Mycosis IC/F selective fungal medium and the Plus Aerobic/F standard bacteriological medium for the diagnosis of fungemia on the Bactec 9240 automatic system. We retrospectively analyzed 550 blood culture pairs composed of one Mycosis IC/F vial and one Plus Aerobic/F vial, drawn in 187 patients with fungemia. The positivity rate by vial was significantly higher on Mycosis IC/F medium than on Plus Aerobic/F medium (88.0% versus 74.9%, P < 0.0001). The positivity rate for fungus detection on Plus Aerobic/F medium fell to 26.9% when bacteria were present in the same vial. The positivity rate by patient was also significantly higher on Mycosis IC/F medium than on Plus Aerobic/F medium (92.5% versus 75.9%, P < 0.0001). A marked superiority of Mycosis IC/F medium was demonstrated for diagnosis of Candida glabrata fungemia (31 of 31, 100%, versus 18 of 31, 58.1%, P < 0.0001). The mean detection time was significantly shorter on Mycosis IC/F medium than on Plus Aerobic/F medium (28.9 22.2 h versus 36.5 24.6 h, P < 0.0001). The mean time saving was 8.8 h for Candida albicans and 43.7 h for C. glabrata. Mycosis IC/F medium enabled more sensitive and earlier diagnosis, particularly for the two strains most frequently responsible for fungemia, C. albicans and C. glabrata, and also in the event of the concomitant presence of both yeasts and bacteria. In patients with risk factors, it would thus appear to be sensible to draw a Mycosis IC/F vial in addition to the standard bacteriological vials. Meyer MH et al - JCM 2004:773

Diagnostic tests 10 3 5 10 Mannan 13/32 BD Glucan 17/32 PCR 18/32 10 3 5 Background: Candidemia is a major infectious complication of seriously immunocompromised patients. In the absence of specific signs and symptoms, there is a need to evolve an appropriate diagnostic approach. A number of methods based on the detection of Candida mannan, nucleic acid and (1,3)-beta- D- glucan (BDG) have been used with varying specificities and sensitivities. In this retrospective study, attention has been focused to evaluate the usefulness of two or more disease markers in the diagnosis of candidemia. Methods: Diagnostic usefulness of Platelia Candida Ag for the detection of mannan, Platelia Candida Ab for the detection of anti-mannan antibodies, Fungitell for the detection of BDG, and of a semi-nested PCR (snPCR) for the detection Candida species-specific DNA have been retrospectively evaluated using 32 sera from 27 patients with culture-proven candidemia, 51 sera from 39 patients with clinically suspected candidemia, sera of 10 women with C. albicans vaginitis, and sera of 16 healthy controls. Results: Using cut-off values recommended by the manufacturers, the sensitivity of the assays for candidemia patients were as follows: Candida snPCR 88%, BDG 47%, mannan 41%, anti-mannan antibodies 47%, respectively. snPCR detected 5 patients who had candidemia due to more than one Candida species. The sensitivities of the combined tests were as follows: Candida mannan and anti-mannan antibodies 75%, and Candida mannan and BDG 56%. Addition of snPCR data improved the sensitivity further to 88%, thus adding 10 sera that were negative by BDG and/or mannan. In clinically suspected, blood culture negative patients; the positivities of the tests were as follows: Candida DNA was positive in 53%, BDG in 29%, mannan in 16%, and anti-mannan antibodies in 29%. The combined detection of mannan and BDG, and mannan, BDG and Candida DNA enhanced the positivity to 36% and 54%, respectively. None of the sera from Candida vaginitis patients and healthy subjects were positive for Candida DNA and mannan. Conclusion: The observations made in this study reinforce the diagnostic value of snPCR in the sensitive and specific diagnosis of candidemia and detection of more than one Candida species in a given patient. Additionally, in the absence of a positive blood culture, snPCR detected Candida DNA in sera of more than half of the clinically suspected patients. While detection of BDG, mannan and anti-mannan antibodies singly or in combination could help enhancing sensitivity and eliminating false positive tests, a more extensive evaluation of these assays in sequentially collected serum samples is required to assess their value in the early diagnosis of candidemia. 10 PCR 18/32 Alam et al - BMC Infectious Diseases 2007, 7:103

CONTROLS: healthy volunteers (13)-β-D-GLUCAN CONCENTRATIONS _____ BG values Pg/ml _____ _____ _____ CBSI PCBSI NCBSI CONTROLS CBSI: proven Candida BSI PCBSI: possible Candida BSI NCBSI: no Candida BSI CONTROLS: healthy volunteers Horizontal bars indicate median values Del Bono V et al. 49th ICAAC, 2009

Questions Le traitement des candidoses invasives doit être précoce et efficace Faiblesse des tests diagnostiques La colonisation fongique est très fréquente et possède une VPP faible pour le diagnostic de candidémie Qui traiter avec quelles conséquences potentielles?

La colonisation à Candida Le tube digestif est la porte d’entrée principale des candidémies chez le neutropénique La peau est une importante source de candidémie chez le non-neutropénique Le nombre de sites et l’intensité de la colonisation augmente le risque d’IFI La colonisation trachéale est le reflet de la colonisation oropharyngée est n’est pas associée à la pneumonie à candida chez les patient non-neutropénique

Colonization index (preemptive treatment) Infected Colonized Cohort prospective surg. ICU - 5,3 sites /patient 29 patients/11 IFI (8 candidaemias) Colonization index N. site(s) colonized N. sites sampled 1,0 0,8 0,6 0,4 0,2 colonization Se Sp PPV NPV >2 sites 100 22 44 ≥3 sites 45 72 50 68 Index >0,5 69 66 Ann Surg 1994 Dec;220(6):751-8 Candida colonization and subsequent infections in critically ill surgical patients. Pittet D, Monod M, Suter PM, Frenk E, Auckenthaler R OBJECTIVE. The authors determined the role of Candida colonization in the development of subsequent infection in critically ill patients. DESIGN. A 6-month prospective cohort study was given to patients admitted to the surgical and neonatal intensive care units in a 1600-bed university medical center. METHODS. Patients having predetermined criteria for significant Candida colonization revealed by routine microbiologic surveillance cultures at different body sites were eligible for the study. Risk factors for Candida infection were recorded. A Candida colonization index was determined daily as the ratio of the number of distinct body sites (dbs) colonized with identical strains over the total number of dbs tested; a mean of 5.3 dbs per patient was obtained. All isolates (n = 322) sequentially recovered were characterized by genotyping using contour-clamped homogeneous electrical field gel electrophoresis that allowed strain delineation among Candida species. RESULTS. Twenty-nine patients met the criteria for inclusion; all were at high risk for Candida infection; 11 patients (38%) developed severe infections (8 candidemia); the remaining 18 patients were heavily colonized, but never required intravenous antifungal therapy. Among the potential risk factors for candida infection, three discriminated the colonized from the infected patients--i.e., length of previous antibiotic therapy (p < 0.02), severity of illness assessed by APACHE II score (p < 0.01), and the intensity of Candida spp colonization (p < 0.01). By logistic regression analysis, the latter two who were the independent factors that predicted subsequent candidal infection. Candida colonization always preceded infection with genotypically identical Candida spp strain. The proposed colonization indexes reached threshold values a mean of 6 days before Candida infection and demonstrated high positive predictive values (66 to 100%). CONCLUSIONS. The intensity of Candida colonization assessed by systematic screening helps predicting subsequent infections with identical strains in critically ill patients. Accurately identifying high-risk patients with Candida colonization offers opportunity for intervention strategies. 10 20 30 40 60 80 100 140 Pittet et al. Ann Surg 1994 Dec;220(6):751-8 18

Colonization index in medical ICU 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 baseline (n=92) day 7 (n=92) day 14 (n=49) day 21 (n=26) day 28 (n=15) day 35 (n=12) day 42 (n=5) day 49 (n=4) * * Indicate statistical difference as compared with baseline value Candida spp. colonization significance in critically ill medical patients: a prospective study CI ≥ 0,5 = 39% No candidaemia P.E. Charles et al., Intensive Care Med (2005) 31:393-400

77 patients with Candida in LRT 58 patients without Candida in LRT Significance of the isolation of Candida species from airway samples in critically ill patients: a prospective, autopsy study 1587 admissions 301 (19%) died 232 autopsies 135 (58%) with pneumonia 97 (42%) without pneumonia 77 patients with Candida in LRT 0 Candida pneumonia 58 patients without Candida in LRT 0 Candida pneumonia W. Meersseman et al.; Intensive Care Med. (2009) 35:1526-1531

Clinical prediction rule for candidiasis in the ICU 2,890 patients (> 4 days in nine hospitals). Incidence of candidiasis was 3% (88 cases). The best performing rule was as follows: Any systemic antibiotic OR presence of a CVC AND at least TWO of the following: total parenteral nutrition (days 1-3), any dialysis (days 1-3), any major surgery (days -7-0), pancreatitis (days -7-0), any use of steroids (days -7-3), or use of other immunosuppressive agents (days -7-0). Ostrosky-Zeichner L, Eur J Clin Microbiol Infect Dis. 2007

Clinical prediction rule validation on a retrospective international cohort CPR CPR + Colo Se 86% 66% Sp 65% 88% PPV 3% 8% NPV 99% % patients with IC 4.7% 9.3% % patients treated 35.5% 12.7% % patients captured 83% 67% Ostrosky-Zeichner – 48th ICAAC- M 1853

Colonization index ≥0.5 (95% CI) Candida score Construction n =1699 pts (Leon CCM 2006) Parenteral nutrition 1pt OR = 2,48 IC95:1,16 - 5,31 Surgical admission 1pt OR = 2,71 IC95:1,45 - 5,06 Multiple Colonization 1pt OR = 3,04 IC95:1,45 - 6,39 Severe sepsis 2pts OR = 7,68 IC95:4,14 - 14,22 External Validation n =1107 pts (Leon CCM 2009) Candida score ≥3 (95% CI) Colonization index ≥0.5 (95% CI) Area under ROC curve 0.774 (0.715-0.832) 0.633 (0.557-0.709) Sensitivity 77.6 (66.9-88.3) 72.4 (60.0-83.9) Specificity 66.2 (63.0-69.4) 47.4 (44.0-50.8) Positive predictive value 13.8 (10.0-17.5) 8.7 (6.2-11.3) Negative predictive value 97.7 (96.4-98.9) 96.1 (94.2-98.0) Relative risk for invasive candidiasis 5.98 (3.28-10.92) 2.24 (1.28-3.93) Objective: To assess the usefulness of the “Candida score” (CS) for discriminating between Candida species colonization and invasive candidiasis (IC) in non-neutropenic critically ill patients. A rate of IC <5% in patients with CS <3 was the primary end point. Design: Prospective, cohort, observational study. Setting: Thirty-six medical-surgical intensive care units of Spain, Argentina, and France. Patients: A total of 1,107 non-neutropenic adult intensive care unit patients admitted for at least 7 days between April 2006 and June 2007. Measurements and Main Results: Clinical data, surveillance cultures for fungal growth, and serum levels of (1–3)-beta-Dglucan and anti-Candida antibodies (in a subset of patients) were recorded. The CS was calculated as follows (variables coded as absent 0, present 1): total parenteral nutrition 1, plus surgery 1, plus multifocal Candida colonization 1, plus severe sepsis 2. A CS >3 accurately selected patients at high risk for IC. The colonization index was registered if >0.5. The rate of IC was 2.3% (95% confidence interval [CI] 1.06 –3.54) among patients with CS <3, with a linear association between increasing values of CS and IC rate (p < 0.001). The area under the receiver operating characteristic curve for CS was 0.774 (95% CI 0.715–0.832) compared with 0.633 (95% CI 0.557–0.709) for CI. (1–3)-Beta-D-glucan was also an independent predictor of IC (odds ratio 1.004, 95% CI 1.0 –1.007). The relative risk for developing IC in colonized patients without antifungal treatment was 6.83 (95% CI 3.81–12.45). Conclusions: In this cohort of colonized patients staying >7 days, with a CS <3 and not receiving antifungal treatment, the rate of IC was <5%. Therefore, IC is highly improbable if a Candida-colonized non-neutropenic critically ill patient has a CS <3. (Crit Care Med 2009; 37:1624 –1633) KEY WORDS: algorithms; antifungal agents/therapeutic use; candidiasis/ diagnosis; critical illness; fungemia/prevention and control; risk assessment PPV (Candida score)=Proba (Dis+/CS +) = 13.8%… PPV (Colonization index)=Proba (Dis/CI+) = 8.7%… Leon et al - Crit Care Med 2006; 34:730–737 and Crit Care Med 2009; 37:1624 –1633

Questions Le traitement des candidoses invasives doit être précoce et efficace Faiblesse des tests diagnostiques Qui traiter avec quelles conséquences potentielles?

Who needs to be treated (MV patients >4 days)? Length ICU stay >7days 1,205 patients Short life expectancy (APACHE II > 35) 13 patients Inadequate data collection 85 patients Study population 1,107 patients Objective: To assess the usefulness of the “Candida score” (CS) for discriminating between Candida species colonization and invasive candidiasis (IC) in non-neutropenic critically ill patients. A rate of IC <5% in patients with CS <3 was the primary end point. Design: Prospective, cohort, observational study. Setting: Thirty-six medical-surgical intensive care units of Spain, Argentina, and France. Patients: A total of 1,107 non-neutropenic adult intensive care unit patients admitted for at least 7 days between April 2006 and June 2007. Measurements and Main Results: Clinical data, surveillance cultures for fungal growth, and serum levels of (1–3)-beta-Dglucan and anti-Candida antibodies (in a subset of patients) were recorded. The CS was calculated as follows (variables coded as absent 0, present 1): total parenteral nutrition 1, plus surgery 1, plus multifocal Candida colonization 1, plus severe sepsis 2. A CS >3 accurately selected patients at high risk for IC. The colonization index was registered if >0.5. The rate of IC was 2.3% (95% confidence interval [CI] 1.06 –3.54) among patients with CS <3, with a linear association between increasing values of CS and IC rate (p < 0.001). The area under the receiver operating characteristic curve for CS was 0.774 (95% CI 0.715–0.832) compared with 0.633 (95% CI 0.557–0.709) for CI. (1–3)-Beta-D-glucan was also an independent predictor of IC (odds ratio 1.004, 95% CI 1.0 –1.007). The relative risk for developing IC in colonized patients without antifungal treatment was 6.83 (95% CI 3.81–12.45). Conclusions: In this cohort of colonized patients staying >7 days, with a CS <3 and not receiving antifungal treatment, the rate of IC was <5%. Therefore, IC is highly improbable if a Candida-colonized non-neutropenic critically ill patient has a CS <3. (Crit Care Med 2009; 37:1624 –1633) KEY WORDS: algorithms; antifungal agents/therapeutic use; candidiasis/ diagnosis; critical illness; fungemia/prevention and control; risk assessment 834/1107= 75% Neither colonized nor infected 215 patients Candida spp. Colonization 834 patients 4.8% pts>7d 6.9% colo+ Proven Candida infection 58 patients Leon et al - Crit Care Med 2009; 37:1624 –1633

1893 without systemic antifungals (SAT) Fongiday 169 centres 2047 patients 1893 without systemic antifungals (SAT) 154 (7.5%) with SAT d28 follow-up ok 2032 patients (99.3%) Timsit et al – ICAAC 2009 Azoulay et al – Crit Care Med submitted

Reason for SAT at fongiday (16 mould excluded) b Curative documented infection 27.5% Febrile neutropenia 10.9% Prophylactic (only risk factors) 17.4% Empirical (sepsis + Risk Factors) 18.1% Early or pre-emptive based on Candida colonization 34.8% Because of patient’s hemodynamic instability 13% Other 5.1% Timsit et al – ICAAC 2009 Azoulay et al – Crit Care Med submitted

Who needs to be treated (MV patients >4 days)? Risk factors + colonization +- 30-80% Prophylactic Colonization + Risk factors + Sepsis +- (>30%) Preemptive Risk factors + Sepsis + Colonization+- Empirical (Probabilistic) Invasive candidasis 1-3% Curative British Society for Antimicrobial Chemotherapy Working Party. Int Care Med 1994; 20: 522-8.

SAT decreased colonization index Garbino, Intensive Care Med 2002

Prophylactic treatment: Meta-analyses N studies Candidaemia IFI Death F (Res.) Ho 7 (F) 0.2 (0.06-0.72) 0.39 (0.24-0.65) 0.82 (0.62-1.08) 0.66 (0.22-1.96) Playford 12 (keto +F) 0.46 (0.31-0.64) 0.76 (0.57-0.97) Shorr 4 (F) (2.2%!!) 0.44 (0.27-0.72) 0.87 (0.59-1.28) Vardakas 6 (F) 0.28 (0.09-0.86) 0.26 (0.12-0.53) 0.74 (0.52-1.05) Cruciani 9 (Keto+F) 0.3 (0.1-0.82) 0.25 (0.08-0.8) 0.6 (0.45-0.8) Playford G et al – JAC 2006; 57:628-638; Vardakas KZ et al – Crit Care Med 2006; 34:1216-1224; Ho KM et al – Crit Care 2005;9:R710 ; Cruciani M et al – Intensive care Med 2005; 1477-1485; Shorr A et al Crit Care Med 2005; 33:1928-35;

Traitement prophylactique Meta-analyses Fluconazole chez des malades à haut risque chirurgicaux Diminue l’infection fongique invasive Diminue la mortalité (2 Meta-analyses/ 5) Dans les RCT peu ou pas d’effets sur la résistance au fluconazole Playford G et al – JAC 2006; 57:628-638; Vardakas KZ et al – Crit Care Med 2006; 34:1216-1224; Ho KM et al – Crit Care 2005;9:R710 ; Cruciani M et al – Intensive care Med 2005; 1477-1485; Shorr A et al Crit Care Med 2005; 33:1928-35;

Restriction of fluconazole use for prophylaxis Med-surg ICU (~500 adm./an) 108 months (Jan. 99-Dec. 2007) Overall prevention of NI unchanged 213 candidaemia (1.42/10 000 patient-days) albicans (46%), parapsillosis (22%), glabrata 13% Intervention: Jan. 1999-Jan. 2003: Extensive Prophylaxis Jan. 2003-Dec. 2007:Incitation not to do Statistical analysis: Segmented linear regression Objectives: Candida spp. are the most important non-bacterial pathogens in critically ill patients. The aim of this study was to evaluate trends in the incidence of candidaemia and the distribution of Candida albicans and non-albicans over a 9 year period (1999–2007), and to assess their relationship with fluconazole use. Methods: This was an interventional cross-over study. Patients admitted to the intensive care unit (ICU) who developed a clinically and microbiologically documented candidaemia were analysed. Fluconazole was used as prophylaxis in critically ill patients until 2002; from January 2003 infectious disease consultants strongly discouraged its use. Fluconazole use, measured as defined daily dose per 1000 patient-days, was calculated. The main outcome of the study is the evaluation of the restriction policy in terms of change in fluconazole use and in incidence of candidaemia. Results: During the 108 month period (January 1999–December 2007), a total of 213 episodes of candidaemia (average incidence 1.42 episodes/10000 patient-days/year, range 0.36–3.02 episodes) were recorded in a mixed medical and surgical ICU in Italy. C. albicans was the most prevalent isolated species (n598, 46%); non-albicans (n5115, 54%) were mainly represented by Candida parapsilosis (n546, 22%) and by Candida glabrata (n528, 13%). Segmented regression analysis of the interrupted time series showed that a change in the fluconazole prophylactic strategy resulted in a significant reduction in fluconazole use from the second semester of 2002. A dramatic decrease in the incidence of fungaemia due to C. non-albicans was observed from the second semester of 2003 (intervention effect in the second semester of 2007: 22.31/10000 patient-days); minor changes in the incidence of C. albicans fungaemia emerged (intervention effect in the second semester of 2007: 20.23/10000 patient-days). Conclusions: The study showed a clear correlation between fluconazole use control and decreasing incidence of non-albicans candidaemia. Even if fluconazole remains a first-line treatment option in several cases of invasive candidiasis, its prophylactic use should be carefully evaluated. Bassetti et al – JAC 2009; 64:625-629

Restriction of fluconazole use for prophylaxis Non-albicans candidaemia C. albicans candidaemia X Objectives: Candida spp. are the most important non-bacterial pathogens in critically ill patients. The aim of this study was to evaluate trends in the incidence of candidaemia and the distribution of Candida albicans and non-albicans over a 9 year period (1999–2007), and to assess their relationship with fluconazole use. Methods: This was an interventional cross-over study. Patients admitted to the intensive care unit (ICU) who developed a clinically and microbiologically documented candidaemia were analysed. Fluconazole was used as prophylaxis in critically ill patients until 2002; from January 2003 infectious disease consultants strongly discouraged its use. Fluconazole use, measured as defined daily dose per 1000 patient-days, was calculated. The main outcome of the study is the evaluation of the restriction policy in terms of change in fluconazole use and in incidence of candidaemia. Results: During the 108 month period (January 1999–December 2007), a total of 213 episodes of candidaemia (average incidence 1.42 episodes/10000 patient-days/year, range 0.36–3.02 episodes) were recorded in a mixed medical and surgical ICU in Italy. C. albicans was the most prevalent isolated species (n598, 46%); non-albicans (n5115, 54%) were mainly represented by Candida parapsilosis (n546, 22%) and by Candida glabrata (n528, 13%). Segmented regression analysis of the interrupted time series showed that a change in the fluconazole prophylactic strategy resulted in a significant reduction in fluconazole use from the second semester of 2002. A dramatic decrease in the incidence of fungaemia due to C. non-albicans was observed from the second semester of 2003 (intervention effect in the second semester of 2007: 22.31/10000 patient-days); minor changes in the incidence of C. albicans fungaemia emerged (intervention effect in the second semester of 2007: 20.23/10000 patient-days). Conclusions: The study showed a clear correlation between fluconazole use control and decreasing incidence of non-albicans candidaemia. Even if fluconazole remains a first-line treatment option in several cases of invasive candidiasis, its prophylactic use should be carefully evaluated. Fluco use Bassetti et al – JAC 2009; 64:625-629

Antifungals infection pressure DDD/1000 pt-days C. parapsilosis rate Forrest GN et al – J infect 2008; 56:126

Caspo and fluco exposure influenced the epidemiology of candidaemia A prospective multicenter surveillance program on yeast bloodstream infections was 3 implemented in the Paris area without restrictions on ward of hospitalization (intensive care 4 unit, haematology, surgery) or age (adults and children). The present analysis concerns 2618 5 isolates collected over 7 years in 2441 patients. Centralized species identification and 6 antifungal susceptibility testing using EUCAST methodology were performed. 7 Almost 10% (232/2441) of the patients were recently (≤30 days) treated with antifungal 8 drugs. We analyzed the effect of recent exposure to fluconazole (n=159) or caspofungin 9 (n=61) on the proportion of the 5 major Candida species. For both drugs, preexposure was 10 associated with a decreased prevalence of Candida albicans in favor of less drug11 susceptible species (C. glabrata and C. krusei for the former, C. parapsilosis and to a lesser 12 extend C. glabrata and C. krusei for the latter, P= 0.001). In the multivariate analysis, the 13 risk of being infected with an isolate with decreased susceptibility to fluconazole was 14 independently associated with an age ≥15 years (OR [95%CI]= 2.45 [1.39-4.31], P=0.002) 15 and with recent preexposure to fluconazole (OR=2.17 [1.51-3.13], P<0.001), while the risk 16 of being infected with an isolate with decreased susceptibility to caspofungin was 17 independently associated with an age <15 years (OR=2.53 [1.43-4.48], P=0.001) and with 18 recent preexposure to caspofungin (OR= 4.79 [2.47-9.28], P<0.001). 19 These findings could influence future recommendations for the management of candidemia. Lortholary O et al - AAC Accepts, published online ahead of print on 15 November 2010

Caspo and fluco exposure influenced the epidemiology of candidaemia A prospective multicenter surveillance program on yeast bloodstream infections was 3 implemented in the Paris area without restrictions on ward of hospitalization (intensive care 4 unit, haematology, surgery) or age (adults and children). The present analysis concerns 2618 5 isolates collected over 7 years in 2441 patients. Centralized species identification and 6 antifungal susceptibility testing using EUCAST methodology were performed. 7 Almost 10% (232/2441) of the patients were recently (≤30 days) treated with antifungal 8 drugs. We analyzed the effect of recent exposure to fluconazole (n=159) or caspofungin 9 (n=61) on the proportion of the 5 major Candida species. For both drugs, preexposure was 10 associated with a decreased prevalence of Candida albicans in favor of less drug11 susceptible species (C. glabrata and C. krusei for the former, C. parapsilosis and to a lesser 12 extend C. glabrata and C. krusei for the latter, P= 0.001). In the multivariate analysis, the 13 risk of being infected with an isolate with decreased susceptibility to fluconazole was 14 independently associated with an age ≥15 years (OR [95%CI]= 2.45 [1.39-4.31], P=0.002) 15 and with recent preexposure to fluconazole (OR=2.17 [1.51-3.13], P<0.001), while the risk 16 of being infected with an isolate with decreased susceptibility to caspofungin was 17 independently associated with an age <15 years (OR=2.53 [1.43-4.48], P=0.001) and with 18 recent preexposure to caspofungin (OR= 4.79 [2.47-9.28], P<0.001). 19 These findings could influence future recommendations for the management of candidemia. Lortholary O et al - AAC Accepts, published online ahead of print on 15 November 2010

Évolution des écosystèmes fongiques et utilisation des antifongiques Impact de la consommation des antifongiques sur l’épidémiologie et la sensibilité des Candida sp. dans un service de réanimation français entre 2004 et 2009   P. Fournier1, C. Schwebel2, A. Vesin3, D. Maubon1, B. Lebeau1, L. Foroni4, M. Cornet1, J.F. Timsit 2,3 and H. Pelloux1 1 Parasitologie-Mycologie, 2 Réanimation Médicale, 3 INSERM U823, 4 Pô1e Pharmacie, Centre Hospitalier Universitaire, Institut Albert Bonniot, UFR de Pharmacie, UFR de Médecine, Université Joseph Fourier Grenoble, France Introduction: L’évolution actuelle de l’épidémiologie des candidoses montre une diminution des infections liées à Candida albicans et une augmentation de celles dues à des espèces non-albicans. Pour certains, l’utilisation du fluconazole (FCZ) en prophylaxie a contribué à cette modification. Depuis 2004, les échinocandines sont très utilisées, et la caspofungine (CAS) est maintenant le traitement de première ligne dans certaines situations. Les conséquences éventuelles de l’utilisation massive de celles-ci ne sont pas établies clairement. Notre étude analyse l’impact de l’utilisation des antifongiques sur l’épidémiologie et la sensibilité des Candida sp. dans un service de réanimation. Méthodes: Nous avons analysé la distribution des Candida sp. durant une période de 6 ans (2004-2009) et leur sensibilité durant une période de 3 ans (2007-2009) dans un service de réanimation médicale adulte de l’hôpital universitaire de Grenoble. Pour l’étude épidémiologique, un seul isolat par espèce et par patient a été pris en compte. Les CMI ont été déterminées par la méthode E-Test®. Les consommations d’antifongiques ont été exprimées en nombre de jours de traitement équivalent par 1000 jours d’hospitalisation (JTE/1000JH). Les coefficients de corrélation de Spearman et de Pearson ont été déterminés pour analyser les relations entre la consommation d’antifongiques, la distribution des Candida sp. et l’évolution des CMI. Résultats: Sur cette période de 6 ans, 1511 souches de Candida sp. ont été analysées à partir de 3391 patients. C. albicans représentait 52,5% des souches, C. glabrata 16,6%, C. parapsilosis 7,5%, C. tropicalis 7,3%, C. kefyr 5,4% et C. krusei 4,6% et les autres Candida sp 5,9%. La proportion de C. parapsilosis a augmenté significativement de 5,4% en 2004 à 11,6% en 2009 (p<0,001). La consommation de CAS a augmenté elle aussi significativement de 2004 (17.9 JTE/1000JH) à 2009 (69.9 JTE/1000JH) (p<0,001).Toutes les souches testées restaient sensibles à la CAS avec toutefois une tendance à l’augmentation des souches de C. parapsilosis ayant des CMI entre 0,25 et 1µg/ml (de 63,6% en 2007 à 82,8% en 2009). Il existe une corrélation entre l’augmentation des CMI des C. parapsilosis et l’augmentation de la consommation de CAS. Au total, 99,4%, 96,6% et 83,7% des C. albicans, C. parapsilosis et C. glabrata étaient sensibles au FCZ et nous avons mis en évidence une diminution des CMI des C. albicans entre 2007 et 2009 (p<0,001). La consommation de FCZ a diminué, mais de manière non significative. Les CMI du voriconazole sont restées stables avec 99,7%, 100% et 94,2% de C. albicans, C. parapsilosis et C. glabrata sensibles respectivement. Conclusion: Notre étude a montré que l’évolution de la consommation d’antifongiques peut avoir un impact sur la sensibilité des Candida sp. Nous avons retrouvé une corrélation entre l’augmentation des CMI des C. parapsilosis et l’augmentation de la consommation de CAS. La proportion de C. parapsilosis a augmenté significativement sans toutefois pouvoir être corrélée à la consommation de caspofungine. Grenoble Rea Med 1511 premières souches Fournier P et al – SFMM 2010

Evolution des écosystèmes fongiques et utilisation des antifongiques Impact de la consommation des antifongiques sur l’épidémiologie et la sensibilité des Candida sp. dans un service de réanimation français entre 2004 et 2009   P. Fournier1, C. Schwebel2, A. Vesin3, D. Maubon1, B. Lebeau1, L. Foroni4, M. Cornet1, J.F. Timsit 2,3 and H. Pelloux1 1 Parasitologie-Mycologie, 2 Réanimation Médicale, 3 INSERM U823, 4 Pô1e Pharmacie, Centre Hospitalier Universitaire, Institut Albert Bonniot, UFR de Pharmacie, UFR de Médecine, Université Joseph Fourier Grenoble, France Introduction: L’évolution actuelle de l’épidémiologie des candidoses montre une diminution des infections liées à Candida albicans et une augmentation de celles dues à des espèces non-albicans. Pour certains, l’utilisation du fluconazole (FCZ) en prophylaxie a contribué à cette modification. Depuis 2004, les échinocandines sont très utilisées, et la caspofungine (CAS) est maintenant le traitement de première ligne dans certaines situations. Les conséquences éventuelles de l’utilisation massive de celles-ci ne sont pas établies clairement. Notre étude analyse l’impact de l’utilisation des antifongiques sur l’épidémiologie et la sensibilité des Candida sp. dans un service de réanimation. Méthodes: Nous avons analysé la distribution des Candida sp. durant une période de 6 ans (2004-2009) et leur sensibilité durant une période de 3 ans (2007-2009) dans un service de réanimation médicale adulte de l’hôpital universitaire de Grenoble. Pour l’étude épidémiologique, un seul isolat par espèce et par patient a été pris en compte. Les CMI ont été déterminées par la méthode E-Test®. Les consommations d’antifongiques ont été exprimées en nombre de jours de traitement équivalent par 1000 jours d’hospitalisation (JTE/1000JH). Les coefficients de corrélation de Spearman et de Pearson ont été déterminés pour analyser les relations entre la consommation d’antifongiques, la distribution des Candida sp. et l’évolution des CMI. Résultats: Sur cette période de 6 ans, 1511 souches de Candida sp. ont été analysées à partir de 3391 patients. C. albicans représentait 52,5% des souches, C. glabrata 16,6%, C. parapsilosis 7,5%, C. tropicalis 7,3%, C. kefyr 5,4% et C. krusei 4,6% et les autres Candida sp 5,9%. La proportion de C. parapsilosis a augmenté significativement de 5,4% en 2004 à 11,6% en 2009 (p<0,001). La consommation de CAS a augmenté elle aussi significativement de 2004 (17.9 JTE/1000JH) à 2009 (69.9 JTE/1000JH) (p<0,001).Toutes les souches testées restaient sensibles à la CAS avec toutefois une tendance à l’augmentation des souches de C. parapsilosis ayant des CMI entre 0,25 et 1µg/ml (de 63,6% en 2007 à 82,8% en 2009). Il existe une corrélation entre l’augmentation des CMI des C. parapsilosis et l’augmentation de la consommation de CAS. Au total, 99,4%, 96,6% et 83,7% des C. albicans, C. parapsilosis et C. glabrata étaient sensibles au FCZ et nous avons mis en évidence une diminution des CMI des C. albicans entre 2007 et 2009 (p<0,001). La consommation de FCZ a diminué, mais de manière non significative. Les CMI du voriconazole sont restées stables avec 99,7%, 100% et 94,2% de C. albicans, C. parapsilosis et C. glabrata sensibles respectivement. Conclusion: Notre étude a montré que l’évolution de la consommation d’antifongiques peut avoir un impact sur la sensibilité des Candida sp. Nous avons retrouvé une corrélation entre l’augmentation des CMI des C. parapsilosis et l’augmentation de la consommation de CAS. La proportion de C. parapsilosis a augmenté significativement sans toutefois pouvoir être corrélée à la consommation de caspofungine. DDD/1000HD Fournier P et al – SFMM 2010

Treatment of colonized patients in ICU? No RCTs

Antifungals for CVC tip > 103 Cfu/ml (retrospective) 58 patients CVC > 103 cfu/ml Candida sp. and negative blood cultures Only one patient developed IC (detected as candidaemia). 12/33 patients (36.4%) with a clinical improvement 8/25 (32.0%) with a poor outcome received SAT RF of poor outcome: Ultimately fatal underlying disease OR 12; 95% CI, 1.4–105 P = 0.025 Severe sepsis, septic shock or MOF OR 6.2; 95% CI, 1.0–38; P = 0.05 BUT NOT Antifungal use: OR 0.82; 95% CI, 0.27–2.47; P = 0.73 Abstract Purpose: To assess the influence of antifungal therapy on the outcome of non-candidemic adult patients with central vascular catheter (CVC) tips colonized by Candida species. Methods: A retrospective analysis of the outcome of patients with Candida colonization of their CVC tip and no concurrent candidemia was made over a 4-year period. Patients who either died or developed candidemia-invasive candidiasis (poor outcome) were compared with those who improved. Results: We finally included 58 patients for analysis. Almost all (91.4%) had to be admitted to the ICU during their hospital stay. Independent predictors for outcome were a McCabe and Jackson score corresponding to ultimately fatal underlying disease [odds ratio (OR) 11.98; 95% confidence interval (CI), 1.37–104.97; P = 0.02], and maximum severity corresponding to severe sepsis, septic shock or multiorgan failure (OR: 6.16, CI 95%: 1.00–37.93; P = 0.05). We were unable to demonstrate that antifungal therapy was an independent variable influencing outcome (OR 0.82; 95% CI, 0.27–2.47; P = 0.73). Conclusions: Our data suggest that, in non-neutropenic critically ill patients with no concomitant candidemia and with CVC tips colonized by Candida, antifungal therapy does not seem to have a significant influence on clinical outcome. Perez-Parra Intensive Care Med (2009) 35:707–712

# highly positive samples Assessment of preemptive treatment to prevent severe candidiasis in critically ill patients Before/after study SICU>4 days 8/98-7/00: no treatment, colonization sampled not known 12/00-11/02: screening and known results 5 samples (trachea, gastric, urine, oropharyngeal, rectal) Admission then 1/week + if « highly » positives: (>100 ou 105 CFU according to samples) Corrected colonization index: Patients colonised with CCI ≥ 0,4 20% of the cohort: fluconazole: 800 mg D1 and then 400 mg/j IV x 14days # highly positive samples # samples R. Piarroux et Al - Crit Care Med 2004; 32:2443–2449

Assessment of preemptive treatment to prevent severe candidiasis in critically ill patients R. Piarroux et Al - Crit Care Med 2004; 32:2443–2449

Fongiday – a SAT is associated with a lower day 28 mortality… Cox model with left troncature Takes into account the elapse time between admission and fongiday Univariate analysis HR = 0.81 [0.53-1.22], p=0.31 Adjusted and stratified HR = 0.38 [0.17-0.82], p=0.01 Adjusted on RF of SAT and % death predicted And stratified according to candida score and center (*) all the population without mould infection Timsit et al – ICAAC 2009 Azoulay et al – Crit Care Med submitted

Traitement empirique? La question se pose de plus en plus, pas de réponses… Risque d’un traitement retardé Pas de tests diagnostiques fiables HC positives tardivement (ou negative) Nouveaux tests pas encore convaincants 13 B –D glucane Platelia- Manane / Antimanane PCR

Empirical fluconazole vs placebo RCT double blind Fluconazole 200 mg (n=18) vs placebo (n=19) Septic shock with nosocomial pneumonia or intra-abdominal sepsis 30-days death: Fluco 22% vs Placebo 54% p = 0,015 Pneumonia Intra-abdominal Fluco Placebo 28% (ns) 42% 14% P=0,013 65% But only one candidemia! And parameters imbalanced Fluco Placebo Immunocompromised 4 8 Organ dysfunctions 1,7 P=0,01 2,8 Jacobs S et al. CCM 2003

Empirical Fluconazole vs Placebo for ICU Patients Double blind randomized placebo-controlled trial: 270 adults, 6 years Fluconazole: 800 mg vs placebo 2 weeks if: > 18 years old ICU duration > 96h Apache 2 > 16 Temperature > 38,3°within 72 hours Received large spectrum antibiotics for at least the 4 to 6 previous days Central venous catheter 50% medical ICU Assessment criteria = composite score Initial fever resolution No emerging IFI No toxicity-related trial stopping No use of another systemic AF Schuster et Al, Annals of Internal Medicine

Empirical Fluconazole vs Placebo for ICU Patients Patient Characteristics at Baseline characteristics Fluconazole recipients (n=122) Placebo recipients (n=127) Mean age (SD), y 53 (19) 51 (19) Women, n (%) 29 (24) 28 (22) Median previous ICU stay (range), d 9.5 (4-171) 9 (4-57) Median previous hospital stay (range), d 11 (5-173) 11 (4-58) Median baseline APACHE II score (range) 22 (9-28) 20 (11-42) Corticosteroids, n (%) 19 (16) 15 (12) Total parenteral nutrition, n (%) 70 (57) 65 (51) Renal insufficiency, n (%) † Colonized with yeast in ≥1 site, n (%) 28 (23) 24 (19) Diabetes mellitus, n (%) 23 (19) 26 (20) Cancer, n (%) 12 (10) 8 (6) Surgery within 7 d before study entry, n (%) 65 (53) Schuster et Al, Ann Intern Med. 2008 Jul 15;149(2):83-90

Empirical Fluconazole vs Placebo for ICU Patients Overall success: fluconazole: 36% vs PCB 38% (RR 0,95 IC95:,69-1,32) Per item F vs Placebo Initial fever resolution 49 vs 46% Candidaemia 0 vs 2 Emerging IFI 5 vs 9% Other systemic AF 10 vs 16% Death 24 vs 17% N.B.: all IFIs occur in colonized patients Schuster et Al, Annals of Internal Medicine

Quels traitements?

Candidose invasive tout confondu 1 étude sur les neutropéniques 2009; 8:23 11 RCTs Candidose invasive tout confondu 1 étude sur les neutropéniques Objectives: Invasive fungal infections are a major cause of mortality among patients at risk. Treatment guidelines vary on optimal treatment strategies. We aimed to determine the effects of different antifungal therapies on global response rates, mortality and safety. Methods: We searched independently and in duplicate 10 electronic databases from inception to May 2009. We selected any randomized trial assessing established antifungal therapies for confirmed cases of invasive candidiasis among predominantly adult populations. We performed a meta-analysis and then conducted a Bayesian mixed treatment comparison to differentiate treatment effectiveness. Sensitivity analyses included dosage forms of amphotericin B and fluconazole compared to other azoles. Results: Our analysis included 11 studies enrolling a total of 965 patients. For our primary analysis of global response rates, we pooled 7 trials comparing azoles to amphotericin B, Relative Risk [RR] 0.87 (95% Confidence Interval [CI], 0.78–0.96, P = 0.007, I2 = 43%, P = 0.09. We also pooled 2 trials of echinocandins versus amphotericin B and found a pooled RR of 1.10 (95% CI, 0.99–1.23, P = 0.08). One study compared anidulafungin to fluconazole and yielded a RR of 1.26 (95% CI, 1.06–1.51) in favor of anidulafungin. We pooled 7 trials assessing azoles versus amphotericin B for all-cause mortality, resulting in a pooled RR of 0.88 (95% CI, 0.74–1.05, P = 0.17, I2 = 0%, P = 0.96). Echinocandins versus amphotericin B (2 trials) for allcause mortality resulted in a pooled RR of 1.01 (95% CI, 0.84–1.20, P = 0.93). Anidulafungin versus fluconazole resulted in a RR of 0.73 (95% CI, 0.48–1.10, P = 0.34). Our mixed treatment comparison analysis found similar within-class effects across all interventions. Adverse event profiles differed, with amphotericin B exhibiting larger adverse event effects. Conclusion: Treatment options appear to offer preferential effects on response rates and mortality. When mycologic data are available, therapy should be tailored. Annals of Clinical Microbiology and Antimicrobials 2009, 8:23

Caterpillar diagram (taux de réponses) Vs Fluconazole Vs Voriconazole Vs AmB Objectives: Invasive fungal infections are a major cause of mortality among patients at risk. Treatment guidelines vary on optimal treatment strategies. We aimed to determine the effects of different antifungal therapies on global response rates, mortality and safety. Methods: We searched independently and in duplicate 10 electronic databases from inception to May 2009. We selected any randomized trial assessing established antifungal therapies for confirmed cases of invasive candidiasis among predominantly adult populations. We performed a meta-analysis and then conducted a Bayesian mixed treatment comparison to differentiate treatment effectiveness. Sensitivity analyses included dosage forms of amphotericin B and fluconazole compared to other azoles. Results: Our analysis included 11 studies enrolling a total of 965 patients. For our primary analysis of global response rates, we pooled 7 trials comparing azoles to amphotericin B, Relative Risk [RR] 0.87 (95% Confidence Interval [CI], 0.78–0.96, P = 0.007, I2 = 43%, P = 0.09. We also pooled 2 trials of echinocandins versus amphotericin B and found a pooled RR of 1.10 (95% CI, 0.99–1.23, P = 0.08). One study compared anidulafungin to fluconazole and yielded a RR of 1.26 (95% CI, 1.06–1.51) in favor of anidulafungin. We pooled 7 trials assessing azoles versus amphotericin B for all-cause mortality, resulting in a pooled RR of 0.88 (95% CI, 0.74–1.05, P = 0.17, I2 = 0%, P = 0.96). Echinocandins versus amphotericin B (2 trials) for allcause mortality resulted in a pooled RR of 1.01 (95% CI, 0.84–1.20, P = 0.93). Anidulafungin versus fluconazole resulted in a RR of 0.73 (95% CI, 0.48–1.10, P = 0.34). Our mixed treatment comparison analysis found similar within-class effects across all interventions. Adverse event profiles differed, with amphotericin B exhibiting larger adverse event effects. Conclusion: Treatment options appear to offer preferential effects on response rates and mortality. When mycologic data are available, therapy should be tailored. Annals of Clinical Microbiology and Antimicrobials 2009, 8:23 Vs AmB-L Favours comparator Mills et al - Annals of Clinical Microbiology and Antimicrobials 2009, 8:23

Caterpillar diagram (Mortalité toutes causes) Vs Fluconazole Vs Voriconazole Vs AmB Objectives: Invasive fungal infections are a major cause of mortality among patients at risk. Treatment guidelines vary on optimal treatment strategies. We aimed to determine the effects of different antifungal therapies on global response rates, mortality and safety. Methods: We searched independently and in duplicate 10 electronic databases from inception to May 2009. We selected any randomized trial assessing established antifungal therapies for confirmed cases of invasive candidiasis among predominantly adult populations. We performed a meta-analysis and then conducted a Bayesian mixed treatment comparison to differentiate treatment effectiveness. Sensitivity analyses included dosage forms of amphotericin B and fluconazole compared to other azoles. Results: Our analysis included 11 studies enrolling a total of 965 patients. For our primary analysis of global response rates, we pooled 7 trials comparing azoles to amphotericin B, Relative Risk [RR] 0.87 (95% Confidence Interval [CI], 0.78–0.96, P = 0.007, I2 = 43%, P = 0.09. We also pooled 2 trials of echinocandins versus amphotericin B and found a pooled RR of 1.10 (95% CI, 0.99–1.23, P = 0.08). One study compared anidulafungin to fluconazole and yielded a RR of 1.26 (95% CI, 1.06–1.51) in favor of anidulafungin. We pooled 7 trials assessing azoles versus amphotericin B for all-cause mortality, resulting in a pooled RR of 0.88 (95% CI, 0.74–1.05, P = 0.17, I2 = 0%, P = 0.96). Echinocandins versus amphotericin B (2 trials) for allcause mortality resulted in a pooled RR of 1.01 (95% CI, 0.84–1.20, P = 0.93). Anidulafungin versus fluconazole resulted in a RR of 0.73 (95% CI, 0.48–1.10, P = 0.34). Our mixed treatment comparison analysis found similar within-class effects across all interventions. Adverse event profiles differed, with amphotericin B exhibiting larger adverse event effects. Conclusion: Treatment options appear to offer preferential effects on response rates and mortality. When mycologic data are available, therapy should be tailored. Annals of Clinical Microbiology and Antimicrobials 2009, 8:23 Vs AmB-L Favours comparator Mills et al - Annals of Clinical Microbiology and Antimicrobials 2009, 8:23

Résumé efficacité May 2009 Optimisation? Associations? Fluco =Ampho B C krusei, Toxicité rénale Ampho B = AmbL sauf toxicité++ VRZ= AmphoB puis Fluco Cas>Amb? meilleure efficacité ou moindre toxicité?, C parapsilosis? Anidula>Fluco? En particulier sur C albicans et C tropicalis (parapsilosis idem) Micafungin = Caspofungin Micafungin = AmbL Plus de toxicité (impact pronostic?) Optimisation? Associations?

Traitement probabiliste des candidoses invasives - IDSA 2008 Clinical Infectious Diseases 2009; 48:503–35 Traitement probabiliste (IA) Oui (A-III) Non (A-III) fluconazole echinocandine Pré-exposition azolé ? Haut risque de C. glabrata ou krusei ? Ou sévère (A-III) Oui Non L’amphotéricine B et ses formulations lipidiques sont des alternatives en cas d’intolérance ou de non disponibilité des autres traitements

Sensibilité au fluconazole Adaptation thérapeutique/Candidoses invasives - IDSA 2008 Clinical Infectious Diseases 2009; 48:503–35 Stabilité clinique Oui Non Connaissance du germe Sensibilité au fluconazole C. parapsilosis C. Glabrata (B-III) C krusei (A-I) Oui (B-III) Non fluconazole AmB-L echinocandine Durée du traitement : 2 semaines après stérilisation de HC et résolution clinique(AIII) Ablation des cathéters systématiques si non neutropéniques (A-II) Voriconazole: relais oral ou tt de C. krusei (B-III)

Conclusion: Nous devons apprendre à mieux utiliser les antifongiques… Les patients de réanimation ont de plus en plus de facteurs de risque La candidémie est un événement rare difficile à diagnostiquer Nous disposons de traitements efficaces des candidémies qui doit être administré précocement Le traitement prophylactique diminue les IFI Le traitement des malades colonisés n’a jamais été testé dans une étude randomisée contrôlée Les traitements antifongiques entrepris ont des influences sur les écosystèmes fongiques. Les traitements empiriques (non centrés sur la colonisation préalable) n’ont pas prouvé leur efficacité

Kaplan-Meier en % : D5 : 19.54 [16.48 ; 22.60] 0% 5% 10% 15% 20% 25% 30% 35% 40% 5 10 15 20 25 30 Day of occurrence of candida % 1-KM (IC 95%) Kaplan-Meier en % : D5 : 19.54 [16.48 ; 22.60] D10 : 28.14 [24.21 ; 32.07] D15 : 31.49 [27.10 ; 35.88] D20 : 32.86 [28.16 ; 37.56] D30 : 33.66 [28.76 ; 38.56]