Evaluation of piperacillin-tazobactam disks using contemporary Enterobacterales isolates suggests the need for disk potency optimization

Piperacillin-tazobactam (TZP) is a commonly used broad-spectrum agent. OXA-1 β-lactamases drive global Enterobacterales TZP resistance and raise MICs to the clinical breakpoints (8/4-16/4 µg/mL), making susceptibility testing challenging. Two TZP disks are used globally. The first with 100 µg piperacillin and 10 µg tazobactam, established in 1992, is used by laboratories following U.S. FDA and Clinical and Laboratory Standards Institute (CLSI) standards. The second, with 30 µg piperacillin and 6 µg tazobactam, was developed by the European Committee on Antimicrobial Susceptibility Testing (EUCAST). When CLSI updated Enterobacterales TZP MIC breakpoints in 2022, it became apparent that the 100/10 µg disk may not accurately predict TZP resistance or susceptibility. In this study, we performed a disk mass titration study using 100 contemporary Enterobacterales isolates, including 40 harboring bla_OXA-1. Relative to reference broth microdilution with CLSI breakpoints, categorical agreement (CA) for the 100/10 µg disk was 68% with 0.6% major errors (MEs) and 32% minor errors. The CA for the 30/6 µg disk against EUCAST breakpoints was 88% with 15% very major errors and 10% ME. A third disk developed in this study, 20/5 µg, yielded 81% CA with CLSI MIC breakpoints and disk breakpoints generated using the error-rate-bounded method. CA was 62.4%, 63.2%, and 70.9% for isolates with bla_OXA-1 using the 100/10, 30/6, and 20/5 µg disks, respectively, whereas it was 71.6%, 86.9%, and 83.1% for isolates without bla_OXA-1. Decreasing TZP disk potency improved the overall performance of disk diffusion and improved separation between susceptible and non-susceptible isolates, particularly those harboring OXA-1, but no disk yielded optimal results.

Importance: In this article, we address major gaps in contemporary data for piperacillin-tazobactam (TZP) susceptibility testing and evaluate the performance of disk diffusion. TZP is the most common empiric broad-spectrum agent against Gram-negative pathogens and is used as a carbapenem-sparing regimen. OXA-1 β-lactamases drive global Enterobacterales TZP resistance and raise MICs to the clinical breakpoints, making susceptibility testing challenging. In 2022, CLSI revised the Enterobacterales TZP MIC breakpoints. Due to the lack of contemporary correlates, disk diffusion breakpoints were revised using outdated historical data from 1991 and 2003 and yielded unacceptable error rates. Additionally, there is a lack of global consensus on disk potency. The EUCAST TZP disks contain 30 μg piperacillin and 6 μg tazobactam. The CLSI disks, established after TZP approval in 1979 and prior to the widespread prevalence of ESBLs, contain 100 μg piperacillin and 10 μg tazobactam. Here, we evaluate disk diffusion using the 100/10 and 30/6 μg TZP disks with 100 contemporary Enterobacterales isolates, including 40 harboring bla_OXA-1. We conducted a disk development study to determine if an alternative potency might address accuracy issues with TZP susceptibility testing. We demonstrate that decreasing TZP potency improves the performance of disk diffusion and improves separation between susceptible and non-susceptible isolates, particularly those harboring OXA-1, but no disk yielded optimal results. The alternative 20/5 μg disk yielded the lowest errors using CLSI MIC breakpoints and the best separation between susceptible isolates and isolates harboring bla_OXA-1. Our study addresses an unmet need, shows that further optimization of the TZP disk potency is possible, and provides clinical laboratories with a better understanding of the performance of TZP disks using contemporary, challenging isolates. A larger, multicenter study is needed for further optimization but has been limited by a lack of funding for an off-patent antimicrobial. Our struggles in accessing funding underscore the frequent challenge with AST for older but heavily used antimicrobials.