Cefepime model suggests plasma targets may overestimate lung exposure during renal replacement

Researchers developed a population pharmacokinetic model to inform cefepime dosing for critically ill adults with hospital-acquired pneumonia requiring renal replacement therapy (RRT). Eight patients undergoing continuous RRT contributed pre- and post-filter plasma and effluent samples; another 43 patients requiring RRT contributed plasma and pulmonary epithelial lining fluid (ELF) samples. The non-parametric model used an 80/20 subject-level development–validation split, with Monte Carlo simulations assessing plasma and ELF target attainment across dosing regimens, body weights, and minimum inhibitory concentrations (MICs).

The final model integrated systemic disposition, native clearance, extracorporeal removal, and ELF distribution. Simulations showed adequate plasma target attainment, but ELF attainment varied with MIC, pharmacodynamic target, regimen, and body weight. Continuous infusion at 4 g/day provided adequate simulated ELF target attainment at MICs of 4 mg/L or lower; the abstract did not report numerical attainment probabilities.

These findings suggest that plasma-based cefepime targets may overestimate pulmonary exposure during RRT, especially at higher MICs, body weights, and targets for time above MIC. The results concern modeled drug exposure rather than demonstrated clinical benefit and do not establish a dosing recommendation. The dedicated continuous RRT cohort was small, and external validation and prospective studies are needed to determine whether plasma-guided Bayesian dosing reliably ensures exposure at the infection site.