Imipenem pharmacokinetics/pharmacodynamics in preclinical hollow fiber model, dose-finding in virtual patients, and clinical evidence of efficacy for Mycobacterium abscessus lung disease

Background: Guideline-based therapy (GBT) for Mycobacterium abscessus (Mab) lung disease achieves sputum culture conversion rates (SCC) of 35%. This poor GBT efficacy is mirrored in the hollow fiber system model of Mab (HFS-Mab). While imipenem is part of GBT, biological effect with or without β-lactamase inhibitors, is unproven.

Methods: We performed imipenem/relebactam minimum inhibitory concentration (MIC) in 122 Mab isolates, and an exposure-response study in the HFS-Mab using human intrapulmonary pharmacokinetics. The % time that concentration persisted above MIC (TMIC) mediating maximal effect in the HFS-Mab was used as the exposure target in 10,000 virtual subjects Monte Carlo experiment (MCE)-based dose-finding. For real-world evidence, we performed a patient, intervention (imipenem), comparison (no β-lactam), and outcome (SCC) (PICO) analysis.

Results: Imipenem killed 1.32 log10 CFU/mL below day 0 in HFS-Mab. Imipenem target exposure was TMIC=47.9±9.77%. 1g infusion, every 6h, achieved the target in >90% of virtual patients in MCEs. The pharmacokinetics/pharmacodynamics MIC breakpoint was 1mg/L. In PICO analyses, median days-to-SCC were 470 in comparators, 311 for imipenem added on failing regimen, and 37 in newly treated (p=0.049). The odds ratio for SCC when imipenem was part of the initial regimen versus comparators was 12.5 (95% confidence interval: 1.47 to 84.55). Patients on imipenem experienced no treatment-limiting adverse event, while 2/7 comparators did (p=0.0457). Middlebrook 7H9 broth MIC distribution, read at 24 hours, correlated better with patient responses than cation-adjusted Mueller Hinton broth.

Conclusion: Imipenem demonstrated biologic effect in the HFS-Mab and in patients. Imipenem/relebactam doses of 1g every 6h are recommended.