Population Pharmacokinetic Modeling of Thiotepa in Pediatric and Young Adult HCT Recipients

Background Thiotepa (TT) is an alkylating agent with immunosuppressive properties often included in high-dose chemotherapy prior to hematopoietic cell transplantation (HCT) to support stem cell engraftment. However, the pharmacokinetics (PK) of TT and its active metabolite, TEPA, remain largely uncharacterized in children and young adults undergoing HCT. Using prospective multicenter data, this study aimed to characterize TT and TEPA PK, and to identify key patient covariate-PK relationships to inform future model-informed precision dosing (MIPD) approaches. Methods TT PK data were collected from 2017 to 2023 across 5 pediatric bone marrow transplant sites as a part of a prospective study (NCT03609827). TT was administered intravenously and all patients underwent PK sampling during a single dosing interval. Samples were analyzed using a previously validated LC/MS/MS assay. Covariates evaluated included age, weight, height, sex, ancestry (patient identify ethnicity and race), liver function tests, and renal function. A nonlinear mixed effects modeling approach was used to characterize TT and TEPA PK in which R (v4.3) was utilized for data visualization, NONMEM (v7.6) for model estimation. Results The dataset included a total of 210 and 235 quantifiable concentrations for TT and TEPA from 91 patients. Among the study subjects, the median age of patients was 4.5 years (0.23-22.5) years with a median body weight of 16.7 kg (4.6-110). Final model parameters are shown in Table 1. The best fit model for these data was a 2-compartment structure for TT, a nonlinear conversion from TT to TEPA, and a 1-compartment structure for TEPA. Covariate relationships included in the final model included age and allometrically-scaled fat-free mass (FFM). For age, separate maturation function terms were estimated for PK parameters of clearance (CLTT, CLTEPA, Vmax), and Vc,TT. For FFM allometric scaling, exponents 0.75 for CLTT and CLTEPA, and 1.0 for Vc,TT and Vp,TT were used. Given the nature of modeling parent and metabolite data, the VTEPA term is unidentifiable and fixed to 1. The visual predictive checks (VPC) shows the distribution of the data is well-captured by the model across the majority of the PK-time curve, with slight misspecification at TT peak concentrations, likely driven by the limited information on peak data due to the sampling strategy for this study (Figure 1). Conclusions The developed population PK model characterizes TT and TEPA in children and young adults undergoing HCT and identifies FFM and age as impactful covariates and supports future MIPD efforts of thiotepa. Given limited existing PK data for TT and TEPA, internal and external validations were not available; future prospective trials with expanded data will be vital to improve the model for personalized dosing.