ABSTRACT
This study determined the pharmacokinetics of cefepime in patients requiring extracorporeal membrane oxygenation (ECMO) support to guide dosage selection. Cefepime population pharmacokinetics where characterized in Pmetrics for R for six critically ill patients receiving ECMO. Simulation was employed to determine the fT>MIC and total trough concentration of varying regimens in each patient to evaluate ability to achieve optimal pharmacodynamic exposure and thresholds for cefepime-associated neurotoxicity. Of the six participants, two required continuous veno-venous hemodiafiltration (CVVHDF) while four had a CrCL between 92-199 ml/min. All patients received 2 g q8h as a 3h infusion. A two-compartment model fitted the data best with median (range) parameter estimates as follows: clearance, 5.99 (4.10-10.29) L/h; volume of central compartment, 10.08 (2.45-15.14) L; and intercompartment transfer constants (k12), 3.58 (2.01-4.99) h-1 and k21, 1.70 (1.00-2.88) h-1. The 2g q8h (3h infusion) regimen resulted in >70% fT>MIC in all patients up to an MIC of 16 µg/mL, whereas 2g q12h (0.5h) resulted in 5/6 patients achieving 70% ƒT>MIC at 8 µg/mL but only 1/6 at 16 µg/mL. Aggressive dosing regimens resulted in trough concentrations exceeding conservative neurotoxicity thresholds. No patient demonstrated signs or symptoms of neurotoxicity during treatment. For ECMO patients with normal to augmented renal clearance similar to those presented here, or those receiving CVVHDF, these data support dosing regimens of 2g q8h (3h infusions) to empirically target MICs up to 16 µg/mL. Larger studies are needed to determine how ECMO affects cefepime pharmacokinetics.