Development of a Limited Sampling Strategy for the Estimation of Exposure to High-Dose Etoposide After Intravenous Infusion in Pediatric Patients.

Abstract

Etoposide (VP-16), a podophyllotoxin derivative, is used in conditioning regimens before allogeneic hematopoietic stem cell transplantation in children with acute lymphoblastic leukemia. The aim of this study was to develop a limited sampling strategy (LSS) suitable for the prediction of exposure to VP-16 defined as area under time-concentration curve (AUC). The study included 28 pediatric patients with acute lymphoblastic leukemia, who were administered a 4-hour infusion of 60 mg/kg VP-16. VP-16 concentrations were determined in samples collected 4-124 hours after the beginning of infusion. On obtaining the pharmacokinetic (PK) profiles, a population PK model was developed in NONMEM (ICON Development Solutions, Hanover, MD) with first-order conditional estimation with interaction algorithm. LSSs were chosen by means of a multivariate regression analysis and cross-validated with a leave-one-out approach. Predictive performance of LSSs was assessed by calculating relative prediction error (PE), mean PE, mean absolute PE, and root mean squared PE for model-predicted and observed AUC. VP-16 PKs was best described by a 2-compartment first-order model, and a large variability in the PK parameters was observed. A 3-sample strategy allowed the estimation of VP-16 with highest accuracy and precision (mean relative PE = 0.18%, 95% confidence interval, 1.73%-2.09%; mean absolute relative PE = 3.47%, 95% confidence interval, 2.28%-4.66%; root mean squared PE = 4.43%). The final equation was AUC = 6.85 × C6 h + 3.88 × C12 h + 46.11 × C28 h + 282.0 (adjusted R = 0.9540). In conclusion, developed LSS allows accurate and precise estimation of VP-16 AUC and might be useful for therapeutic drug monitoring.