Busulfan is administered intravenously in conditioning regimens prior to allogeneic hematopoietic stem cell transplantation (HSCT), with model-informed precision dosing (MIPD) targeting a 4-day cumulative AUC of 80-100 mg*hour/L. Three pharmacokinetic (PK) models-Bognar, Langenhorst, and McCune-were evaluated using data with a minimum of 1 and a median of 2 PK sampling days from pediatric and adult patients at two tertiary hospitals in the Netherlands. Simulations with the best-performing model evaluated dose optimization and the role of therapeutic drug monitoring (TDM). The Bognar and Langenhorst models, with overlapping datasets (25% between models, 21% with the current), accurately described busulfan concentration-time curves in all 535 patients (0.18-72 years), especially in those > 60 years (bias and precision in clearance; -5.6%, 6.48% and 16.55%, 16.95%; in observed concentrations -1.86%, -6.52% and 9.12%, 3.58% for both models > 60 years). The McCune model underestimated clearance by 36% in children < 2 years of age. Despite accounting for age-dependent glutathione depletion, the Langenhorst model did not outperform the simpler Bognar model. Model-based initial dosing did not improve target attainment over SmPC-based dosing. However, TDM-based MIPD with a single TDM on days 2-4 significantly improved target attainment (49% vs. 87%), with a number needed to treat for TDM of 2.5. This study demonstrates the clinical efficacy of simple PK models such as the Bognar model, providing that simplicity does not compromise performance. It highlights the inadequacy of nomogram-based dosing alone and reinforces the critical role of TDM in optimizing busulfan exposure for pediatric and adult myeloablative HSCT patients.
© 2025 The Author(s). Clinical Pharmacology & Therapeutics published by Wiley Periodicals LLC on behalf of American Society for Clinical Pharmacology and Therapeutics.