External Evaluation of a Population Pharmacokinetic Model for Dosing Busulfan in Children – Body Surface Area better than Body Weight
Trame MN(1), Bartelink IH(2), Boos J(3), Boelens JJ(2), Hempel G(1,3)
(1)Department of Pharmaceutical and Medical Chemistry – Clinical Pharmacy -, University of Muenster, Germany; (2)Departments of Haematology, Immunology and Clinical Pharmacy, University Medical Center Utrecht, The Netherlands; (3)Department of Paediatric Haematology and Oncology, University Children’s Hospital Muenster, Germany
Objectives: A previously developed population pharmacokinetic model to evaluate the best method for dosing busulfan in children was tested by external evaluation using pharmacokinetic data from children receiving a different schedule of administration.
Methods: The development dataset included plasma concentrations of 94 paediatric patients, aged 0.4 to 18.8 years (median age 9.2), receiving either oral or i.v. busulfan for four consecutive days before bone marrow transplantation. Out of the 94 children, 48 children received oral busulfan every 6 h. The dosing varied between 13 and 20 mg/kg with seven patients receiving a dose of 600 mg/m2. The other 46 children received IV busulfan as 2 h infusions every 6 h in 15 doses of 0.7 to 1.0 mg/kg. The first infusion was given as a double dose over 4 h followed by the second infusion 12 h thereafter. The evaluation dataset consisted of 24 children who received IV busulfan as 3 h infusion once-daily for 4 consecutive days. By means of population pharmacokinetic modelling using nonlinear mixed-effects modelling (NONMEM) plasma concentration-time data of the development dataset were analysed. Several covariates such as age, body weight and body surface area (BSA) were tested on their effects on the pharmacokinetic parameters. The next step was to evaluate the developed one-compartment model by external evaluation.
Results: A one-compartment model with BSA as a covariate for clearance (Cl/F) and volume of distribution (V/F) described the busulfan kinetics of the development dataset sufficiently. The final population estimates of the development dataset were: Cl/F 4.23 l/h per m2 ± 27%, V/F 19 l/m2 ± 35% and ka 0.963 h-1 ± 91%. Interoccasion variability (IOV) for Cl/F (10%) and V/F (22%) was lower than interindividual variability. Prediction of the population parameters of the evaluation dataset on the basis of the developed one-compartment model resulted in very similar population values compared to the developed dataset. Furthermore, the precision and robustness of the model could be confirmed by comparison of the goodness-of-fit plots of the development and the evaluation dataset.
Conclusions: In the paediatric population, BSA, not body weight, is the best predictor for Cl/F and should be considered for dose adjustment. By external model evaluation we were able to confirm the findings and show robustness of the model with data from different dosing and schedule of administration.