Abstract
Objectives: Monte Carlo simulation (MCS) was used to evaluate optimal dosage for cefepime (FEP), moxalactam (MOX), and cefperazone/sulbactam (CFZ/SBT) against extended-spectrum β-lactamase (ESBL) producers isolated from the Blood Bacterial Resistant Investigation Collaborative System. Methods: Minimum inhibitory concentration (MIC) was tested by agar dilution, and ESBL producers were identified by modified Clinical and Laboratory Standards Institute tests. Pharmacokinetic parameters were derived from data on healthy individuals, and probability of target attainment (PTA) and cumulative fraction of response (CFR) %fT >MIC values were estimated by MCS. Results: A total of 2032 Escherichia coli (875 ESBL-producing) and Klebsiella pneumoniae (157 ESBL-producing) strains, and 371 other Enterobacteriaceae strains, were isolated from patients with bloodstream infections (BSIs). MIC(90) values for FEP, MOX, and CFZ/SBT against ESBL-producing E. coli and K. pneumoniae were 64/64 mg/L, 2/32 mg/L, and 64/128 mg/L, respectively. Conventional MOX and CFZ/SBT doses failed to reach 90% PTA against isolates with MICs ≥8 mg/L and ≥4 mg/L, respectively. Against ESBL producers, neither FEP nor CFZ/SBT achieved ≥90% CFR, while CFRs for MOX (1 g iv q6h, 2 g iv q12h, and 2 g iv q8h) exceeded 90% against ESBL-producing E. coli. Simulated CFRs for FEP and MOX were similar (>90%) against non-ESBL-producing Enterobacteriaceae, and higher than CFRs for CFZ/SBT. Conclusion: ESBL producers from BSIs were highly susceptible to MOX, and PTA values were generally higher for MOX than FEP or CFZ/SBT for conventional dosing regimens. This large MCS analysis shows that MOX but not FEP or CFZ/SBT can be used empirically to treat BSIs caused by ESBL-producing E. coli strains.