Abstract
OBJECTIVES: Extended-spectrum β-lactamases (ESBLs) present a serious challenge in the treatment of Gram-negative bacterial infections. ESBLs mediate resistance to most β-lactams, which may be reversed with the addition of an active β-lactamase inhibitor (such as tazobactam, relebactam and avibactam). However, various ESBLs may exhibit different susceptibilities to these inhibitors, which could impact efficacy. We proposed a framework for comparing the efficacy of these inhibitors when combined with the same β-lactam. METHODS: Three clinical isolates of Klebsiella pneumoniae harbouring CTX-M-15 and one Escherichia coli isolate with SHV-12 were examined. Piperacillin MICs were determined by broth dilution using escalating concentrations of tazobactam, relebactam and avibactam. The resulting MICs were characterized as response to inhibitor concentrations using an inhibitory sigmoid E(max) model. Using the best-fit parameter values, the model was conditioned with fluctuating inhibitor concentrations to simulate instantaneous MIC(i) profiles for each isolate-inhibitor pair. Using a simulated exposure of 4 g piperacillin every 8 h, %fT > MIC(i) was estimated for each piperacillin/inhibitor combination. A hollowfibre infection model was subsequently used to validate the predicted effectiveness of selected combinations. RESULTS: In all scenarios, piperacillin MIC reductions were well characterized with increasing inhibitor concentrations. As predicted by %fT > MIC(i), combining piperacillin with avibactam (61.4%-73.6%) was found to be superior to tazobactam (13.5%-44.5%) for suppressing bacterial growth over time. CONCLUSION: We illustrated a practical approach to compare the performance of different inhibitors. This platform may be used clinically to identify the optimal pairing of various β-lactams and β-lactamase inhibitors for individual isolates producing ESBLs.