Optimizing imatinib sampling strategies through a population approach: a crucial step to model-informed precision dosing validation for point-of-care therapeutic drug monitoring

Background Imatinib therapeutic drug monitoring (TDM) contributes at optimizing exposure, yet its implementation is limited by logistical constraints, including strict sampling time requirements. Model-Informed Precision Dosing (MIPD) offers a promising approach to dosage individualization by leveraging population pharmacokinetic (popPK) models, thereby mitigating the constraints of sample collection time. Integrating MIPD and Point-of-Care (POC) analytical methods in ambulatory settings could further improve TDM feasibility and accessibility. This study serves as a proof of concept for MIPD integration in the workflow of imatinib TDM. To identify optimal sampling time for predicting imatinib steady-state trough concentrations (C min, SS ) using a popPK model for an MIPD application. Methods A popPK model developed from data of 146 patients (244 concentrations) was used to simulate individual concentration-time profiles of 1000 patients under standard dosing (400 mg once or twice daily) from treatment initiation to steady-state (reached after 11 days of treatment). Empirical Bayes estimates were generated from single or paired sampling time points and used to predict C min, SS . Predictive performance was evaluated using mean prediction error, coefficient of determination, root mean square prediction error (RMSPE), and successful prediction rates based on recommended therapeutic ranges. Results Sampling 5–24 h and 1–12 h post-dose for once and twice daily administrations, respectively, yielded an RMSPE < 41% and a successful prediction rate around 70%. Conclusion The proposed time windows provide flexible sampling strategies, confirming the benefits of MIPD for the convenience of imatinib TDM, thus improving the clinical feasibility of POC TDM.