Multicentre validation of a modified EUCAST MIC testing method and development of associated epidemiologic cut-off (ECOFF) values for rezafungin.

Abstract

Objectives: Rezafungin EUCAST MIC testing has been associated with notable inter-laboratory variation, which prevented ECOFF setting for C. albicans. We assessed in vitro susceptibility and reproducibility for a modified EUCAST methodology and established associated wild-type upper limits (WT-ULs). Methods: MICs against 150 clinical Candida isolates (six species), molecularly characterized fks mutants (n = 13), and QC strains (n = 6) were determined at six laboratories according to E.Def 7.3 but using Tween 20 supplemented medium. WT-ULs were determined using the derivatization method, the ECOFFinder programme and visual inspection. Consensus WT-ULs were determined. Results: The laboratory- and species-specific MIC distributions were Gaussian with >99.5% MICs within four 2-fold dilutions except for C. parapsilosis (92.8%). The following consensus WT-UL were determined: C. albicans 0.008 mg/L; C. dubliniensis and C. glabrata 0.016 mg/L; C. krusei and C. tropicalis 0.03 mg/L; and C. parapsilosis 4 mg/L. Adopting these WT-UL, six clinical isolates were non-wild-type, five of which harboured Fks alterations. For 11/13 mutants, all 670 MICs were categorized as non-wild-type whereas MICs for C. glabrata Fks2 D666Y and C. tropicalis Fks1 R656R/G overlapped with the corresponding wild-type distributions. Repeat testing of six reference strains yielded 98.3%-100% of MICs within three 2-fold dilutions except for C. albicans CNM-CL-F8555 (96%) and C. parapsilosis ATCC 22019 (93.3%). Conclusions: The modified EUCAST method significantly improved inter-laboratory variation, identified wild-type populations and allowed perfect separation of wild-type and fks mutants except for two isolates harbouring weak mutations. These consensus WT-UL have been accepted as ECOFFs and will be used for rezafungin breakpoint setting.