Population pharmacokinetics of treosulfan in children receiving blood or bone marrow transplantation

Abstract

Background: Treosulfan is an alkylating agent administered intravenously to paediatric patients with malignant and non-malignant disease as part of conditioning therapy prior to blood or bone marrow transplantation. Cumulative treosulfan exposure (area under the curve (AUC)) of 4800 mg h/L has previously been identified as a target to maximise transplant engraftment success while minimising mortality in patients with immunodeficiency disorders. This target may not be indicative for patient populations with malignant or genetic disorders. Aim: To develop a population pharmacokinetic model for treosulfan disposition in paediatric patients receiving treosulfan for malignant, immunodeficiency and genetic disorders. Methods: A total of 303 concentration data points were obtained from 46 patients (age = 0.22-17; median = 3.7 years) receiving three daily doses of treosulfan (30-42 g/m2 cumulative dose) and analysed using non-linear mixed effects modelling (NONMEM v7.4) software. Treosulfan clearance (CL) and volume (V) were assessed through different structural, error, covariate, and maturational models. Simulationbased visual predictive checks (VPCs) and non-parametric bootstrapping was performed for evaluation of model predictability and robustness (n = 1000). Results: A one-compartment structural model with inter-individual variability on CL and V best described the data, and proportional and additive error terms described residual error. Aweight-based covariate model, referenced at 70 kg, described both CL and V (15.60 L). CL was further evaluated across each day, with an allometric exponent of 0.75, and a sigmoidalmaturation function using a fixed postmenstrual age of 40 weeks (7.07, 8.18 and 10.66 L/h, respectively). No model misspecification was observed based on goodness of fit plots, and bootstrapping and VPC parameters were acceptable. Comparison with trapezoidalcalculated AUC yielded strong correlation with the model-calculated AUC (R2 = 0.9579). Conclusion: The model described treosulfan pharmacokinetics and variability well. Future directions are to perform a time-to-event analysis, correlating these findings with drug toxicity, engraftment success and event-free survival. Translational Significance: Providing a pharmacokinetic model for treosulfan in children with malignant or non-malignant disease will guide clinicians in providing the best dose to optimise clinical outcomes.L6360744912021-09-30 <br />