Visiting Scientist The University of Manchester Manchester, England, United Kingdom
Background: Coproporphyrin-I (CP-I) is a sensitive endogenous biomarker (EB) of organic anion-transporting polypeptide (OATP) 1B. Early clinical monitoring of CP-I is recommended to assess the risk of OATP1B-mediated DDIs. This study critically evaluated the ability of recently published CP-I models to predict the effect of SLCO1B1 521T>C genotype, ethnicity, and sex on CP-I baseline and CP-I-drug interactions using the largest clinical dataset to date. Methods: Simcyp V23 was used to evaluate two CP-I PBPK models: one from Simcyp database, and the other adapted from our previous work1 (“in-house” model). CP-I plasma baseline of 731 individuals, and plasma concentrations after the administration of rifampicin (150-600 mg), cyclosporine (20-100 mg), and probenecid (500-1000 mg) in diverse populations were used for model velification. Sensitivity analysis was performed to understand factors contributing to differences in estimated in vivo inhibition constant (Ki) for rifampicin using EB PBPK model. Results: Despite differences in CP-I parameters, both models successfully predicted CP-I baseline (94% of predictions within 1.5-fold of the observed data). Interaction between CP-I and rifampicin was better predicted with the Simcyp model (95% of AUCR and CmaxR predicted within Guest criterion2 vs. 77% for “in-house” model) (Figure 1). Primary reason for this difference was the 4-fold smaller contribution of renal elimination (fe) of CP-I in the Simcyp model. Varying CP-I fe from 0-15% led to 7-fold differences in estimated OATP1B1 Ki for rifampicin. In contrast, there was no considerable difference between models in their predictive performance of cyclosporine or probenecid interactions. Conclusion: The robustness of two CP-I PBPK models was confirmed using the largest and most diverse dataset to date. Estimated in vivo Ki for new inhibitor may be sensitive to assumed contributions of alternative elimination pathways in the biomarker model, and this needs to be considered when using EB PBPK models for such purposes.
1. Takita et al. CPT Pharmacometrics Syst Pharmacol. 2021;10(2):137-147 2. Guest et al. Drug Metab Dispos. 2011;39(2):170-3
