Abstract Objectives Optimal antimicrobial exposure in epithelial lining fluid (ELF) is critical for meropenem efficacy in pneumonia, yet ELF pharmacokinetic data remain scarce, particularly in children. To address this, we aimed to develop a model capable of predicting meropenem concentrations in both plasma and ELF for evaluating pharmacodynamic target attainment under clinical dosing strategies. Methods A physiologically based pharmacokinetic (PBPK) model was developed to simulate unbound meropenem concentrations in plasma and ELF. An empirical penetration coefficient (ρ) was incorporated to link lung intracellular concentrations to ELF concentrations, modelled as a function of clinical and inflammatory covariates. Following validation, the percentage of time over a dosing interval that the free drug concentration remains above the MIC(%ƒT MIC), of meropenem plasma and ELF were related to in-hospital mortality. Monte Carlo simulations were conducted to assess the PTA for 40%ƒT MIC under varying regimens, MIC ranges (0.25–16 mg/L) and penetration scenarios. Results The PBPK model accurately predicted meropenem exposures in both plasma and ELF. ELF penetration was significantly influenced by physiological and pathological factors. ELF %ƒT MIC showed higher interindividual variability compared with that of plasma and was more strongly correlated with survival in both adult (P = 0.073) and paediatric patients (P = 0.013). Although prolonging the infusion improved ELF target attainment for susceptible pathogens (MIC ≤4 mg/L) with adequate penetration, it failed against high-MIC strains or with poor lung penetration. Conclusions These findings underscore the importance of targeting infection-site pharmacokinetics over plasma exposure for better therapeutic efficacy in pneumonia. The model can be used to optimize dosing strategies.
Liu et al. (2026) studied this question.