Aims: To characterise the pharmacokinetics of busulfan in paediatrics and investigate changes in clearance over a 4-day treatment course of once daily intravenous therapy.
Methods: A population pharmacokinetic model was developed using NONMEM®, based on first-order conditional estimation with interaction, in patients ≤18 years receiving once daily intravenous busulfan for haematopoietic stem cell transplant (HSCT) who had concentration-time measurements taken following each administered dose.
Results: Data for model development were collected from 40 paediatric patients characterising 159 dosing days and 1113 busulfan plasma concentration-time measurements. A two-compartment model with linear elimination best described the data with a typical clearance (CL) of 0.27L/kg/h (IIV 19.5%; IOV 9.1%), typical central volume of distribution of 0.67L/kg (IIV 14.5%), and typical peripheral volume of distribution of 12.24L/kg. Actual patient weight, post-menstrual age, a maturation component and time since start of treatment were found to be significant covariates influencing CL. Actual patient weight significantly influenced both central and peripheral volume of distribution. A reduction in CL of 29% was observed in a standard patient from day 1 to 4 of dosing with a maximal CL reduction of -36.1%. The time to achieve 50% of maximal reduction occurred at approximately 6.06 hours after initiation of the first dose . Model estimates suggested that a 35.0% higher dose of busulfan for dose 1 of therapy than that currently recommended in the product information(PI) (5.95mg/kg vs 4.4mg/kg for a 17kg child) followed by a 24.9% (4.47/kg) dose reduction from dose 2 onwards would increase the likelihood of achieving an AUCcum target of 90mg/L·h and a daily AUC0-24 target of 22.5mg/L·h. With no dosage adjustments the typical patient (17kg) was expected to attain an AUCcum of 82.6mg/L·h (range 66.9-110) with most expected to achieve below 22.5mg/L·h exposure on day 1. Alternatively, adjusting the PI recommended dose following dose 1 based on exposure using this model within a Bayesian forecasting software program may help optimise treatment.
Conclusions: A population PK model has been developed that can be used to target a desired busulfan exposure level on each day of treatment as well as cumulative exposure over the course of therapy in an individual child undergoing a HSCT. Further investigation of the potential influence of busulfan metabolites on the time-dependant CL of busulfan is warranted.