Population Pharmacokinetics (PK) of Vancomycin in Children on Extra-Corporeal Life Support (ECLS)

Learning Objectives: VAN remains first line therapy for the treatment of life-threatening infections caused by methicillin resistant Staphylococcus aureus (MRSA) and ampicillin resistant Enterococci. Current evidence suggests that VAN trough levels of 15–20 mg/L or an AUC:MIC ratio of 400:1 are needed to maximize outcomes. Many factors can affect VAN PK in the pediatric ICU including sepsis and ECLS. Currently, no VAN PK data exists for children on ECLS with the current ECLS equipment. The purpose of this study was to describe VAN PK in children with a contemporary ECLS operation, including the Quadrox-i pediatric oxygenator

Methods: During the ECLS run, routine blood samples were collected from 26 children who received VAN for prophylaxis or empiric therapy. VAN concentrations were measured by fluorescent polarization and the population PK was determined using PMetrics. Multiple compartmental and covariate models were explored to determine the best fit of the data.

Results: 26 children with a median age of 1.8 yr (3 mo-7 yr) accounted for 33 VAN treatment courses. Each child contributed a median of 2 VAN levels (1–14) to the model. A 2 compartment model with weight as a covariate fit the VAN concentration data the best. Mean (SD) population estimates for clearance (CL), volume of the central compartment (Vc), total volume of distribution (Vd), and intercompartment transfer constants were 2.02(2.79) mL/min/kg, 0.05(0.05) L/kg, 0.12 (0.17) L/kg, 1.63(1.31) h-1, and 2.34(1.55) h-1, respectively. This resulted in a mean (SD) elimination half-life of 2.03(2.47) h. The CL estimate in this population of ECLS patients is significantly higher than previous estimates (1 mL/kg/min) with older oxygenators yet the Vd appears to be smaller with a contemporary ECLS arrangement.

Conclusions: These are the first VAN PK data in critically ill children with a contemporary ECLS operation utilizing the Quadrox-i pediatric oxygenator. These data demonstrate significantly higher CL estimate than previous data which needs to be recognized when selecting a dosing regimen.