Antimicrobial susceptibility of Corynebacterium diphtheriae and Corynebacterium ulcerans in Germany 2011-17

Background: Diphtheria is mainly caused by diphtheria-toxin-producing strains of Corynebacterium diphtheriae and Corynebacterium ulcerans. The recommended first-line antibiotic is penicillin or erythromycin, but reliable susceptibility data are scarce.

Objectives: To define WT MIC distributions of 12 antimicrobial agents and provide data for the determination of tentative epidemiological cut-off values (TECOFFs) for potentially toxigenic corynebacteria and to evaluate the potential usefulness of a gradient test (Etest) for susceptibility testing of penicillin, erythromycin and clindamycin.

Methods: For the 421 human or veterinary isolates from the period 2011-17, MICs of 12 antimicrobial agents were determined. Etest performance was evaluated for penicillin, erythromycin and clindamycin.

Results: MIC distributions were characterized and TECOFFs could be set for 11 out of 24 antibiotic/species combinations. The current EUCAST clinical breakpoints, predominantly determined for Corynebacterium species other than C. diphtheriae and C. ulcerans, divide the WT MIC distributions of penicillin and clindamycin, thereby making reproducible susceptibility testing of C. diphtheriae and C. ulcerans difficult. For erythromycin, 4% of C. diphtheriae and 2% of C. ulcerans had MICs higher than those for WT isolates. Phenotypically detectable resistance to other antibiotics was rare. Etest underestimated MICs of penicillin and lower concentrations needed to be included for erythromycin, while for clindamycin the Etest was not a good surrogate method.

Conclusions: MIC distributions based on reference broth microdilution for potentially toxigenic Corynebacterium spp. were developed. For five and six agents, TECOFFs were suggested for C. diphtheriae and C. ulcerans, respectively, but for Corynebacterium pseudotuberculosis the number of isolates was too low.