PII-023 - PBPK MODEL DEVELOPMENT AND VALIDATION FOR ROSUVASTATIN: INTEGRATING IN VITRO, CLINICAL PHARMACOKINETIC, AND LIVER PET IMAGING DATA TO IMPROVE PREDICTIONS OF OATP1B- AND BCRP-MEDIATED DRUG-DRUG INTERACTIONS

M. Ladumor¹, M. Jiang², J. Guo², X. Chu²; ¹Merck & Co., Inc., West Point, PA - 19486, USA, ²Merck & Co., Inc., West Point, PA, USA.

Background: Rosuvastatin, a widely prescribed statin for dyslipidemia, primarily targets the liver. It is also a clinical index substrate for OATP1B and BCRP transporters. Altered hepatic or systemic exposure of rosuvastatin due to transporter-mediated drug-drug interactions (DDIs) or pharmacogenomic (PGx) variations can result in severe muscle toxicity or treatment failure. Accurate prediction of these interactions is crucial for its efficacy and safety. However, existing physiologically based pharmacokinetic (PBPK) models for rosuvastatin cannot predict liver tissue exposure, raising concerns about the accuracy of transporter kinetic parameters. This study aimed to develop a reliable PBPK model for rosuvastatin to improve predictions of transporter-mediated DDIs.

Methods: The PBPK model was developed using Simcyp V23, incorporating full PBPK distribution (method 2), permeability-limited liver (PerL) model, and advanced dissolution, absorption, and metabolism (ADAM) model. The PerL model includes passive diffusion, sinusoidal influx (OATP1B1/3, OATP2B1, NTCP), sinusoidal efflux (MRP4), canalicular efflux (BCRP) transporters, and metabolism (CYP2C9, UGT1A1/3). Hepatic transporter kinetic parameters were derived from in vitro and human PET imaging liver and gallbladder data. The ADAM model incorporates passive permeability, intestinal apical influx (OATP2B1), and efflux (BCRP) transporters. Intestinal transporter kinetics were derived from in vitro and clinical pharmacokinetic data. Model validation involved comparing predicted PK parameters (Cmax and AUC) with clinical data.

Results: The PBPK model accurately simulated systemic and PET imaging liver and gallbladder concentrations and predicted within 2-fold of observed clinical intravenous and oral rosuvastatin PK data (single and multiple doses). The contribution of transporters was confirmed with clinical PGx (OATP1B1 & BCRP) and clinical inhibitors (OATP1B: rifampin & cyclosporine; BCRP: fenebrutinib & febuxostat), with PK predictions within 2-fold of observed data.

Conclusion: The developed rosuvastatin PBPK model reliably predicted clinical PK, PGx, and transporter-mediated DDIs using multiple clinical data. This model can be used to assess OATP1B- and BCRP-mediated DDIs for investigational drugs, supporting clinical study design and drug labeling.