INVICTUS II: comparison of disc diffusion and broth microdilution methods for susceptibility testing of ceftolozane/tazobactam and five comparator agents against a subset of INVICTUS I Pseudomonas aeruginosa clinical isolates.

Abstract

Objectives: A subset of clinical Pseudomonas aeruginosa isolates from the INVICTUS I study was selected to compare antimicrobial susceptibility performance for disc diffusion and broth microdilution of ceftolozane/tazobactam and five comparator antipseudomonal agents.

Methods: Overall, 159 non-cystic fibrosis isolates were chosen from INVICTUS I that had the smallest zone diameters from disc testing to ceftolozane/tazobactam, including isolates that were close to the breakpoints. The isolates were tested across 14 centres in the UK. Susceptibility results were interpreted using EUCAST 2026 criteria.

Results: Discrepancy analysis showed combined error rates (very major and major errors) highest for ceftolozane/tazobactam, ceftazidime and piperacillin/tazobactam (15.7%, 14.5% and 12.6%, respectively); when adjusted to remove isolates close to the breakpoint for ceftolozane/tazobactam and piperacillin/tazobactam there were still remarkably high error rates (10.6% and 7.2%, respectively). Essential agreement and bias analysis were not performed as MICs were being compared with zone diameters.

Discussion: Our findings highlight substantial challenges in performing and interpreting susceptibility testing for these two important antipseudomonal β-lactam/β-lactamase inhibitors, which are commonly employed as either first-line empirical therapy (piperacillin/tazobactam) or for serious and MDR infection (ceftolozane/tazobactam). A review of disc testing for these agents may be warranted, and diagnostic laboratories may wish to review the technical limitations and put in place mitigating measures.