Early de-escalation of antifungal treatment in critically ill immunocompromised patients with proven or probable invasive candidiasis: A retrospective multicentre study.

BACKGROUND AND OBJECTIVES: Antifungal treatment (AFT) for proven or probable invasive candidiasis (IC) is common in critically ill immunocompromised patients. Early AFT de-escalation appears safe in unselected ICU patients, but has not been studied specifically amongst immunocompromised hosts. METHODS: This retrospective study was conducted in 19 ICUs in France between January 2019 and December 2024. All immunocompromised patients hospitalized in ICU for ≥5 days and treated with antifungals for a first proven or probable episode of IC were included. Early AFT de-escalation was defined as either spectrum reduction, i.e., a switch from an initial echinocandin or liposomal amphotericin B to fluconazole, or complete discontinuation of the initial antifungal, occurring within 5 days (day 1-5) of treatment initiation. The primary objective was to determine the incidence of early AFT de-escalation (censored at day 28 after initiation of AFT). RESULTS: 242 patients were analysed (median age 64 years, 64% males). Most patients (n=175, 72.3%) were treated empirically or pre-emptively for probable IC. Early AFT de-escalation occurred in 92 patients (38%, 95% confidence interval [CI] 31.9%-44.2%). AFT de-escalation consisted of spectrum reduction in 35.9%, and AFT discontinuation in 64.1% of cases. There was no difference in ICU mortality, ICU length-of-stay and the number of ventilator-free days between patients with and without AFT de-escalation. AFT was re-initiated within 7 days of de-escalation in 6 of the 92 patients (6.5%) who had undergone early AFT de-escalation. CONCLUSIONS: Early AFT de-escalation occurred in 38% of critically ill immunocompromised patients, and was not associated with adverse prognostic outcomes.