A systematic review of fluconazole resistance in clinical isolates of<i>Cryptococcus</i>species

Fluconazole is the most commonly used antifungal agent for both the treatment of cryptococcal meningitis, and for prophylaxis against the disease. However, its prolonged use has the potential to exert selection pressure in favour of fluconazole-resistant strains. We evaluated the prevalence of fluconazole resistance in Cryptococcus spp. clinical isolates in 29 studies from 1988 to May 2017 included in EMBASE and MEDLINE databases. A total of 4995 Cryptococcus isolates from 3210 patients constituted this study; 248 (5.0%) of the isolates from relapsed episodes of cryptococcosis were included in this analysis. Eleven (38%) of the studies used minimum inhibitory concentrations (MICs) breakpoints of ≥64 μg/mL to define fluconazole resistance, 6 (21%) used ≥32 μg/mL, 11 (38%) used ≥16 μg/mL and 1 (3%) used ≤20 μg/mL. Overall, mean prevalence of fluconazole resistance was 12.1% (95% confidence interval [CI]: 6.7-17.6) for all isolates (n = 4995). Mean fluconazole resistance was 10.6% (95% CI: 5.5-15.6) for the incident isolates (n = 4747) and 24.1% (95% CI: -3.1-51.2) for the relapse isolates (n = 248). Of the 4995 isolates, 936 (18.7%) had MICs above the ecological cut-off value. Fluconazole resistance appears to be an issue in Cryptococcus isolates from patients with relapses. It remains unclear whether relapses occur due to resistance or other factors. There is an urgent need to establish antifungal breakpoints for Cryptococcus spp.