Résumé
Background: Optimal antibiotic exposure is a vital but challenging prerequisite for achieving clinical success in ICU patients.
Objectives: To develop and externally validate a population pharmacokinetic model for continuous-infusion meropenem in critically ill patients and to establish a nomogram based on a routinely available marker of renal function.
Methods: A population pharmacokinetic model was developed in NONMEM (R) 7.3 based on steady-state meropenem concentrations (C-ss) collected during therapeutic drug monitoring. Different serum creatinine-based markers of renal function were compared for their influence on meropenem clearance (the Cockcroft-Gault creatinine clearance CLCRcG, the CLCR bedside estimate according to Jelliffe, the Chronic Kidney Disease Epidemiology Collaboration equation and the four-variable Modification of Diet in Renal Disease equation). After validation of the pharmacokinetic model with independent data, a dosing nomogram was developed, relating renal function to the daily doses required to achieve selected target concentrations (4/8/16 mg/L) in 90% of the patients. Probability of target attainment was determined for efficacy (C-ss >= 8 mg/L) and potentially increased likelihood of adverse drug reactions (C-ss >32 mg/L).
Results: In total, 433 plasma concentrations (3.20-48.0 mg/L) from 195 patients (median/P-0.05 - P-0.95 at baseline: weight 77.0/55.0-114 kg, CLCRCG 63.0/19.6-168 mL/min) were used for model building. We found that CLCRCG best described meropenem clearance (CL = 7.71 L/h, CLCRCG = 80 mL/min). The developed model was successfully validated with external data (n = 171, 73 patients). According to the nomogram, daily doses of 910/1480/2050/2800/ 3940 mg were required to reach a target C-ss = 8 mg/L in 90% of patients with CLCRCG = 20/50/80/120/180 mL/min, respectively. A low probability of adverse drug reactions (<0.5%) was associated with these doses.
Conclusions: A dosing nomogram was developed for continuous-infusion meropenem based on renal function in a critically ill population.