Improving antimicrobial susceptibility testing of arcobacter butzleri: comparative methods and epidemiological cut-off value (ECOFF) determination

Arcobacter butzleri is an emerging zoonotic pathogen associated with gastrointestinal illnesses in humans, yet standardized antimicrobial susceptibility testing (AST) methods and interpretive breakpoints for this microorganism remain unavailable. Broth microdilution, despite being endorsed by the CLSI-VAST subcommittee, is often impractical in several laboratories. Therefore, this study aimed to evaluate agar dilution as a more feasible alternative to broth microdilution. A total of 415 A. butzleri isolates from various meat sources in Bangkok, Thailand were confirmed by PCR and tested against ciprofloxacin, erythromycin, gentamicin, and tetracycline. AST was conducted using broth microdilution under aerobic conditions for 24 hours as the reference method. This was compared to agar dilution methods performed under aerobic conditions at 37 °C for 24 and 48 hours, and under microaerobic conditions at 37 °C for 48 hours, mimicking one of the Campylobacter AST standards. Tentative epidemiological cut-off (ECOFF) values were determined using eligible MIC distributions and the ECOFFinder. Percentage agreement, Cohen’s kappa and Gwet’s AC1 were used to assess agreement levels between methods by using proposed tentative ECOFF values, as well as the commonly applied Campylobacter breakpoints. Tentative ECOFFs proposed in this study were as follows: ciprofloxacin 0.5 µg/ml, erythromycin 16 µg/ml, gentamicin 2 µg/ml, and tetracycline 16 µg/ml. The agar dilution method under aerobic conditions for 24 hours showed the highest agreement with the reference method when the tentative ECOFF values were applied, with very high percentage agreement of more than 99% for all antibiotics. Kappa values showed very good levels of agreement for all antibiotics (>0.90) except for erythromycin that could not be calculated due to imbalanced data. Paradoxically low kappa values were addressed by Gwet’s AC1 values, which also showed agreement levels ranging from 0.999 – 1.000 for all antibiotics. The application of Campylobacter breakpoints yielded the same values for percentage agreement, kappa, and Gwet’s AC1 for ciprofloxacin and gentamicin due to the similar threshold values. Lower percentage agreement was observed for erythromycin (97.58) and especially tetracycline (76.14%). The values obtained for Cohen’s kappa and Gwet’s AC1 were used synergistically to suggest that the method with the best agreement was aerobic conditions for 24 hours for all antibiotics, except tetracycline, for which aerobic conditions for 48 hours was the best method. The lowest agreement was observed in the microaerobic conditions for 48 hours for all antibiotics, especially with the application of Campylobacter breakpoints. These findings support the reliability of the agar dilution method under aerobic conditions for 24 hours as an alternative AST method for A. butzleri and highlight the need for species-specific interpretive criteria to strengthen AMR surveillance.