Abstract
BACKGROUND: In vitro-in vivo discordance in β-lactams' activities against metallo-ß-lactamase (MBL)-producing Enterobacterales has been described. We aimed to assess whether this discordance is attributed to the supra-physiologic zinc concentration in in vitro testing media. METHODS: A clinical and microbiological observational study of patients with bloodstream infections due to New Delhi metallo-ß-lactamase-producing Klebsiella pneumoniae was performed. Outcomes of patients treated empirically with non-MBL-active β-lactam therapy (carbapenems and ceftazidime/avibactam) and MBL-active β-lactam therapy (ceftazidime/avibactam + aztreonam) were documented. The patients' isolates were used to induce septicemia in mice, and survival upon meropenem treatment was recorded. Meropenem minimum inhibitory concentrations (MICs) were determined in standard media and in the presence of physiological zinc concentrations. RESULTS: Twenty-nine patients receiving empiric non-MBL-active β-lactams (median duration, 4 days) were compared with 29 receiving MBL-active β-lactams. The 14-day mortality rates were 21% and 14%, respectively. In the murine septicemia model, meropenem treatment resulted in protection from mortality (P < .0001). Meropenem MICs in the physiologic zinc concentration broth were 1- to >16-fold lower vs MICs in zinc-unadjusted broth (≥64 mg/L). CONCLUSIONS: Our data provide foundational support to establish pharmacokinetic/pharmacodynamic relationships using MICs derived in physiologic zinc concentration, which may better predict β-lactam therapy outcome.