Detecting Echinocandin Resistance in C. glabrata Using Commercial Methods: Are CLSI or EUCAST Breakpoints Suitable for Categorical Classification?

Abstract

Previous studies correlated Sensititre YeastOne and gradient diffusion plastic strips with standard procedures for the detection of echinocandin-resistant C. glabrata isolates. However, these studies were limited by the low number of resistant isolates studied; the inclusion of sufficient numbers of mutant isolates is essential to test the procedures' capacity to detect resistance. We assessed the performance of Sensititre YeastOne and Etest strips to detect echinocandin resistance in susceptible or resistant C. glabrata isolates (n = 80) in which the FKS genes were sequenced, and MICs interpreted using EUCAST and CLSI breakpoints. Isolates were echinocandin-susceptible (n = 50) or echinocandin-resistant according to EUCAST 7.4 methodology. Echinocandin susceptibility using Sensititre YeastOne and Etest strips and categorical agreement were assessed. All except one anidulafungin-resistant isolate had an anidulafungin Sensititre YeastOne MIC ≥ 0.25 mg/L, while most micafungin-resistant isolates had a MIC ≥ 0.25 mg/L. Likewise, all anidulafungin-resistant isolates had an anidulafungin Etest strip-obtained MIC ≥ 0.03 mg/L, whereas micafungin-resistant isolates were ≥ 0.125 mg/L. Overall, these commercial methods correctly classified > 90% of isolates by using any breakpoint. Despite the low number of errors detected, these were mostly false resistance (major errors) with EUCAST breakpoints and false susceptibility (very major errors) with CLSI breakpoints. Sensititre YeastOne and Etest strips were suitable procedures to detect echinocandin resistance in C. glabrata. The high number of FKS mutants included reinforces our study and opens the door for multicentre validations.