Freeman, J.; Sanders, I. M. J. G.; Harmanus, C.; Clark, E. V.; Berry, A. M.; Smits, W. K.; ESCMID Study Group on C. difficile (ESGCD),

Objectives Increasing resistance to antimicrobials used for the treatment of Clostridioides difficile infections necessitates reproducible antimicrobial susceptibility testing. Current guidelines take a one-size-fits-all approach and/or offer limited guidance. We investigated how the choice of medium affects measured MIC values across two sites. Methods We determined MIC values for the antimicrobials fidaxomicin, metronidazole and vancomycin for a representative collection of European C. difficile strains (n=235) using agar dilution on three different media: Brucella Blood Agar (BBA), Fastidious Anaerobe Agar supplemented with horseblood (FAA-HB) and Wilkins-Chalgren (WC) agar. The study was conducted at two sites to compare reproducibility. Useability (ease of preparation of the media as well as read-out of the assay) was assessed through a survey. Results We found that all media result in highly consistent aggregated MIC data for all antibiotics, with MIC50 and MIC90 within 2-fold of each other across sites. For fidaxomin, MIC values on WC were lower than on the other media. Metronidazole showed the lowest MIC on BBA, and the highest on WC. For vancomycin, there was little difference between media. Though absolute values for individual isolates differed between sites, identified resistant isolates were similar. Results obtained on FAA-HB were most consistent between sites and results obtained on WC showed the most divergence. FAA-HB was positively evaluated in the usability survey. Conclusions This study shows medium-dependent differences in C. difficile MICs for at least two antimicrobials across two sites. We suggest the use of FAA-HB to align with general EUCAST recommendations for susceptibility testing of anaerobic bacteria and deposited reference strains for standard susceptibility testing of C. difficile to increase interlaboratory reproducibility.