Fluconazole resistance is not a predictor of poor outcome in patients with cryptococcosis

Description

Background Cryptococcus isolates with high MICs to fluconazole are increasingly reported, and a potential clinical impact has been advocated. However, there are different methods to evaluate fluconazole MICs and comparative analysis among such techniques and their comprehensive correlation with clinical outcome are not available. Methods Over a 13-year period (2000-2013), fluconazole MICs were determined for 62 cryptococcal isolates recovered from 22 patients with cryptococcosis using CLSI M27-A3, EUCAST, E test and Sensititre YeastOne, simultaneously. The relationship between the fluconazole MICs and the clinical outcome at week 10 was assessed in patients who received fluconazole as induction or maintenance therapy (n = 16). Results The percentage of cryptococcal strains with MIC values >= 16 mu g/mL according to different methods was CLSI 1.6%, EUCAST 16.1%, E test 31.6% and Sensititre YeastOne 53.2%. Among the 16 patients treated with fluconazole, no correlation between clinical outcome and any MIC value obtained with either method was observed. The only variable independently associated with a poor outcome was having a disseminated disease. Conclusions There is a weak correlation between fluconazole MICs against Cryptococcus spp. as determined by CLSI, EUCAST, E test and Sensititre YeastOne. Neither procedure could predict the clinical outcome of patients with cryptococcosis receiving fluconazole-based therapy. With present methods, fluconazole resistance in Cryptococcus may be clinically misleading.

Additional details