Impact of bronchial colonization with Candida spp. on the risk of bacterial ventilator-associated pneumonia in the ICU: the FUNGIBACT prospective cohort study.

INTRODUCTION: Respiratory tract Candida spp. colonization is associated with more frequent bacterial ventilator-associated pneumonia (VAP). However, this colonization could be causally related to VAP or simply reflect the immune paralysis associated with multiple organ failure. OBJECTIVE: To prospectively evaluate the relationship between Candida spp. colonization and bacterial VAP in mechanically ventilated patients with multiple organ failure. INCLUSION: Patients receiving mechanical ventilation for > 4 days and presenting multiple organ failure were included. Tracheal colonization with Candida spp. was evaluated at inclusion (day 0, D0) and every 4 days until extubation. Quantitative proximal and tracheal cultures were performed at each VAP episode. Monocyte human leukocyte antigen-DR isotype (mHLA-DR) expression and the ratio of polymononuclear leukocytes to lymphocytes were used to evaluate immunoparalysis at D0 and D7. The relationship between fungal colonization and VAP was modelled using cause-specific models for repeated events with adjustment for time-dependent confounders and immune factors. RESULTS: A total of 213 patients, with a median age of 64, simplified acute physiology score II (SAPS II) score 55 and sequential organ failure assessment (SOFA) score 10, mainly admitted for medical reasons (n = 197, 92%), were enrolled in 2012-2015. The median ICU stay was 24 days and the mortality rate was 32% (69 cases). Median mHLA-DR was 5916 Ab-bound/cell [3863-8934]; median lymphocyte count, 0.9Giga/L [0.6-1.3]; neutrophil-to-lymphocyte ratio, 10.9 [6.5-19.7]. Overall, 146 cases (68.5%) had tracheal colonization with Candida spp. An episode of VAP occurred (either for the first or only time) in 62 (29.1%) cases 5.5 days (median) after D0; a second episode occurred in 12 (5.6%) cases, 15.5 days (median) after D0. After adjustment, bronchial colonization with Candida was not associated with VAP [adjusted cause-specific hazard ratio = 0.98 (0.59-1.65), p = 0.95]. CONCLUSION: In patients with mechanical ventilation for more than 4 days and multiple organ failure, bronchial colonization with Candida spp. was not associated with VAP, even after adjustment for immune function.