LB-026 - PHYSIOLOGICALLY-BASED PHARMACOKINETIC MODELING FOR THE PREDICTION OF BCRP-MEDIATED DRUG INTERACTIONS USING R STATISTICAL SOFTWARE

M. Hartauer¹, B. Thompson², S. Hate², S. Hullett², M. Posada²; ¹The University of North Carolina - Chapel Hill, , , ²Eli Lilly and Company, , .

Background: The Breast Cancer Resistant Protein (BCRP) is expressed widely throughout the body, including in the intestine and liver, and impacts the absorption and disposition of many drugs that are BCRP substrates. Thus, co-administration of a BCRP inhibitor and BCRP substrate can lead to clinical drug-drug interactions (DDIs). Rosuvastatin is a widely prescribed statin and is often used as a clinical index substrate to assess BCRP inhibition, as BCRP limits its intestinal absorption and mediates its biliary efflux. Pirtobrutinib is a BCRP inhibitor in the clinic and was shown to increase the AUC of orally administered rosuvastatin by 2.18-fold. Physiologically-based pharmacokinetic models (PBPK) can be leveraged to predict such BCRP-mediated DDIs using in vitro data.

Methods: To this end, a PBPK modeling platform was developed using the mrgsolve package in R statistical software. Rosuvastatin PBPK input parameters were obtained from literature, and PK predictions were compared with observed data. Population variability in rosuvastatin pharmacokinetics was modeled by introducing variability on transporter expression, renal clearance, and body weight. Furthermore, the model was used to predict the impact of single dose Pirtobrutinib on rosuvastatin pharmacokinetics. The measured in vitro BCRP Ki of 1.3 µM was incorporated into a Pirtobrutinib PBPK model, and the predicted extent of interaction was compared with observed clinical data.

Results: As summarized in Table 1, the rosuvastatin model showed excellent concordance with observed data, and the interaction between Pirtobrutinib and rosuvastatin was well captured. Additionally, the model closely captured observed rosuvastatin plasma-concentration time profiles following a 20 mg oral dose and predicts extensive enterohepatic recirculation, as the apparent fraction absorbed (0.48) was over 50% greater than the fraction absorbed (0.30).

Conclusion: A fully customizable modeling platform combining mechanistic absorption, disposition, and population variability was developed using the mrgsolve package in R statistical software. The model reproduced the PK of rosuvastatin and its DDI with Pirtobrutinib. This flexible PBPK platform can be utilized for modeling in drug discovery and development.