Article Summaries

PI-TB Article Summaries

Voila C’est Fini?  The French EMPIRICUS Trial and the Shrinking Market for Untargeted Antifungal Therapy in the ICU

Timsit JF, Azoulay E, Schwebel C, et al for the EMPIRICUS Trial Group.  Empirical Micafungin Treatment and Survival Without Invasive Fungal Infection in Adults With ICU-Acquired Sepsis, Candida Colonization, and Multiple Organ Failure: The EMPIRICUS Randomized Clinical Trial.


Despite support for the practice in guideline recommendations, data on the use of empiric antifungal administration for invasive candidiasis (IC) in the ICU for immunocompetent hosts is conflicting. Although a pioneering study of empiric antifungal treatment in surgical ICU patients yielded favorable results in the 1990s, subsequent studies in mixed ICU populations have consistently failed to demonstrate benefit. Enter EMPIRICUS.


This was a multi-center, placebo-controlled trial that enrolled 260 antifungal-naïve patients with ICU-acquired severe sepsis of unknown etiology who underwent permuted-block randomization to micafungin or placebo for 14 days. Additional eligibility requirements were mechanical ventilation ≥ 5 days with at least one other organ dysfunction, a central venous or arterial catheter, at least one Candida colonization site, and ongoing broad-spectrum antibacterial exposure of >4 days’ duration. Those with severe neutropenia, prior transplantation, and therapeutic immunosuppression were excluded. The primary end-point was 28-day survival free of proven IC according to the 2008 EORTC/MSG definition.


Nearly three-quarters of the subjects were medical admissions with 5% and 6% having undergone abdominal surgery or suffering from acute pancreatitis, respectively. Serum (1-3)-β-D-glucan was elevated (i.e., >80pg/ml) in 70% of patients. There was no difference in 28-day IC-free survival between the micafungin (68%) and placebo (60%) arms [HR 1.04 (0.64-1.67), p=0.88].  This result held up across all of the pre-defined subgroups. The only significant difference was in the rate of proven IC during the study follow-up period, which favored the micafungin arm (3% vs. 12%, p=0.008).       


In a severely ill but immunocompetent ICU population, this well-conducted trial represents yet another failure of empirical state-of-the-art antifungal therapy to improve meaningful outcomes. Disappointingly, this trial also failed to show a valuable discriminatory role for BDG, Candida score, or colonization index, casting doubt on the utility of colonization surveillance. Before concluding that empirical antifungals should be abandoned in the ICU, though, a few points are worth noting. First, the study may have been underpowered given that the actual 28-day IC-free survival among placebo recipients dramatically exceeded the assumed rate (60% as opposed to 37%). Additionally, in an ICU population with a high severity of illness, determinants of mortality are myriad, and so IC may not be the primary factor for most patients (i.e., the attributable mortality of IC in a high-acuity ICU is uncertain).  Also, the low numbers of abdominal surgery and pancreatitis subjects meant that two patient categories at particular risk for invasive infection were relatively under-represented. These considerations, among others, leave the door open for further investigation. In the meantime, for immunocompetent hosts, empirical coverage of IC in the ICU should be restricted to carefully selected cases chosen by multidisciplinary consensus.