Comparative evaluation of the Vitek 2 yeast susceptibility test and CLSI broth microdilution reference method for testing antifungal susceptibility of invasive fungal isolates in Italy: the GISIA3 study.

The newly available AST-YS01 Vitek 2 cards were evaluated, and the results were compared with those obtained by the CLSI M27-A2 microdilution reference method. Clinical fungal isolates, including 614 isolates of Candida spp., 10 Cryptococcus neoformans isolates, 1 Geotrichum capitatum isolate, and 2 quality control strains, were tested for their susceptibilities to amphotericin B, fluconazole, and voriconazole using both methods. The majority of fungal isolates were susceptible to all antifungal agents tested: the MIC90 values determined by the Vitek 2 and CLSI methods were 0.5 and 1 mu g/ml, respectively, for amphotericin B; 8 and 16 mu g/ml, respectively, for fluconazole; and <0.12 and 0.25 mu g/ml, respectively, for voriconazole. Overall there was excellent categorical agreement (CA) between the methods (99.5% for amphotericin B, 92% for fluconazole, 98.2% for voriconazole), but discrepancies were observed within species. The CAs for fluconazole were low for Candida glabrata and Candida krusei when the results of the CLSI method at 48 h were considered. Moreover, the fully automated commercial system did not detect the susceptibility of Cryptococcus neoformans to voriconazole. The Vitek 2 system can be considered a valid support for antifungal susceptibility testing of fungi, but testing of susceptibility to agents not included in the system (e.g., echinocandins and posaconazole) should be performed with other methods.