Physiologically Based Pharmacokinetic Modeling to Predict Human Starting Dose and Pharmacokinetics for Rocbrutinib

Rocbrutinib is a fourth-generation Bruton's tyrosine kinase (BTK) inhibitor that covalently binds wild-type BTK and non-covalently engages the C481S mutant. Its pharmacokinetic (PK) characteristics in healthy Chinese subjects remain unclear. This study aimed to support maximum recommended starting dose (MRSD) selection and predict exposure of rocbrutinib in healthy Chinese subjects via an integrated model-informed strategy. We determined key extrapolation parameters and preclinical PK in CD-1 mice and beagle dogs. Three approaches were employed to determine the MRSD: The no observed adverse effect level (NOAEL) dose method based on body surface area, the NOAEL exposure method based on a physiologically based pharmacokinetic (PBPK) model, and the minimum effective dose method based on a PBPK model. The PBPK model was developed and validated using preclinical and clinical PK data. Predicted MRSD values by the NOAEL dose method, the NOAEL exposure method, and the minimum effective dose method were 48.7, 9.8, and 10.5 mg, respectively. Considering the lowest predicted value and available tablet strength, a starting dose of 12.5 mg was selected. Predicted plasma concentration-time profiles were consistent with those observed in animals and humans. The fold error for C_max and AUC fell within the 0.5-2.0 range. The model further predicted brain tissue exposure across species. Moreover, the predicted brain BTK occupancy at 12.5 mg was approximately 37% for wild-type BTK and 6% for C481S mutant BTK. The PBPK model serves as a valuable tool for dose selection, PK prediction, and CNS target engagement evaluation, supporting the clinical development of rocbrutinib.