Minseo Kang 1, Yerin Lee 1, Na Yun Kim 2, Jae Soo Shin 2, Min Soo Park 3, Eunjin Hong 4, Jangik Lee 1
1 College of Pharmacy and Research Institute of Pharmaceutical Sciences, Seoul National University (Seoul, Republic of Korea), 2 IL-YANG Pharmaceutical Co., Ltd, (Yongin, Republic of Korea), 3 Department of Clinical Pharmacology, Severance Hospital, Yonsei University College of Medicine (Seoul, Republic of Korea), 4 College of Pharmacy, Dongguk University (Goyang, Republic of Korea)
Objectives: Radotinib, a weakly basic and lipophilic second-generation tyrosine kinase inhibitor, was approved for the treatment of chronic myeloid leukemia (CML) [1]. Food intake has been reported to affect the pharmacokinetic characteristics of radotinib [2]. However, a mechanistic framework capable of predicting the effects have not been established. This study aimed to develop and verify a physiologically based pharmacokinetic (PBPK) model to quantitatively predict the food effects on radotinib pharmacokinetics.
Methods: A PBPK model for radotinib was developed by incorporating the Advanced Dissolution, Absorption and Metabolism (ADAM) model and the Full PBPK distribution framework [3]. The absorption of radotinib was modeled using gastrointestinal permeability (Peff) estimated from Caco-2 cell permeability tests with inter-laboratory variability corrected using mannitol permeability. To mechanistically capture food-induced changes in radotinib absorption, pH-dependent solubility measured across pH 1.2 to 12 and logarithmic micelle-to-water partition coefficient (log Km:w) were incorporated into the model. The values of log Km:w, a key determinant of bile micelle-mediated solubility enhancement, were optimized against clinical pharmacokinetic data from a dose escalation study (in house data). A lag time was incorporated to reflect food-induced delays in gastric disintegration and subsequent drug transit to the small intestine, based on literature-reported values for capsule formulations [4]. The distribution of radotinib in humans was predicted by scaling tissue-to-plasma partition coefficient (Kp) scalars derived from nonclinical in vivo pharmacokinetic studies. The elimination of radotinib in humans was characterized using recombinant CYP enzyme kinetics in vitro, with intersystem extrapolation factors (ISEF). The model was verified by comparing simulated plasma concentration(Cp)-time profiles following a single oral dose of radotinib 400 mg under fasted and fed conditions with observed Cp data from 24 male healthy volunteers. The model was further verified by comparing simulated geometric mean ratios (GMRs) of maximum Cp (Cmax) and area under the Cp-time curve from time zero to infinity (AUCinf) between fasted and fed conditions with observed GMRs from the healthy volunteers, in which the difference within a 2-fold range was chosen as an acceptance criteria.
Results: Under fasted conditions, the model effectively captured the observed mean Cp-time profile of radotinib, with the ratios of predicted to observed values for time to Cmax (Tmax), Cmax and AUCinf of 1.03, 0.84 and 1.25, respectively. All the ratios satisfied the acceptance criteria within 2-fold range, with Cmax and AUCinf also meeting the more stringent criteria (0.80 – 1.25), which demonstrates successful model performance. Similarly, under fed conditions, the model adequately captured the observed mean Cp-time profile, with the ratios of predicted to observed values for Tmax, Cmax and AUCinf of 1.03, 0.94 and 1.10, respectively, which further demonstrates the robust predictive performance of the model. Considering that food intake substantially increased radotinib exposure with observed GMR increases of 2.65-fold in Cmax and 3.49-fold in AUCinf in previous study [2], this PBPK model reliably predicted the food effects that yielded the simulated GMRs of 2.99-fold for Cmax and 3.68-fold for AUCinf.
Conclusions: A PBPK model for radotinib was developed and verified against clinical data for the first time. The model demonstrated adequate predictive performance for food effects on radotinib pharmacokinetics, with all predicted-to-observed ratios falling within acceptable ranges. This mechanistic framework would provide a scientific basis for optimizing dosing strategies under various meal conditions to ensure safe and effective use of radotinib in clinical practice.
References:
[1] Kwak JY, et al. Phase III Clinical Trial (RERISE study) Results of Efficacy and Safety of Radotinib Compared with Imatinib in Newly Diagnosed Chronic Phase Chronic Myeloid Leukemia. Clin Cancer Res 2017;23(23):7180-8.
[2] Lee Y, et al. Clinical Pharmacokinetic Characteristics and the Effect of Food on Radotinib in Healthy Volunteers. Presented at: EHA 2025 Congress; 2025; Milan, Italy.
[3] Ezuruike U, et al. Guide to development of compound files for PBPK modeling in the Simcyp population-based simulator. CPT Pharmacometrics Syst Pharmacol 2022;11(7):805-21.
[4] Kambayashi A, Sako K, Kondo H. Effects of Diurnal Variation and Food on Gastrointestinal Transit of 111In-Labeled Hydrogel Matrix Extended-Release Tablets and 99mTc-Labeled Pellets in Humans. J Pharm Sci 2020;109(2):1020-5.
Reference: PAGE 34 (2026) Abstr 12185 [www.page-meeting.org/?abstract=12185]
Poster: Drug/Disease Modelling - Absorption & PBPK