Hepatic OATP1B zonal distribution: Implications for rifampicin‐mediated drug–drug interactions explored within a PBPK framework

Abstract

Abstract OATP1B facilitates the uptake of xenobiotics into hepatocytes and is a prominent target for drug–drug interactions (DDIs). Reduced systemic exposure of OATP1B substrates has been reported following multiple‐dose rifampicin; one explanation for this observation is OATP1B induction. Non‐uniform hepatic distribution of OATP1B may impact local rifampicin tissue concentrations and rifampicin‐mediated protein induction, which may affect the accuracy of transporter‐ and/or metabolizing enzyme‐mediated DDI predictions. We incorporated quantitative zonal OATP1B distribution data from immunofluorescence imaging into a PBPK modeling framework to explore rifampicin interactions with OATP1B and CYP substrates. PBPK models were developed for rifampicin, two OATP1B substrates, pravastatin and repaglinide (also metabolized by CYP2C8/CYP3A4), and the CYP3A probe, midazolam. Simulated hepatic uptake of pravastatin and repaglinide increased from the periportal to the pericentral region (approximately 2.1‐fold), consistent with OATP1B distribution data. Simulated rifampicin unbound intracellular concentrations increased in the pericentral region (1.64‐fold) compared to simulations with uniformly distributed OATP1B. The absolute average fold error of the rifampicin PBPK model for predicting substrate maximal concentration (Cmax) and area under the plasma concentration–time curve (AUC) ratios was 1.41 and 1.54, respectively (nine studies). In conclusion, hepatic OATP1B distribution has a considerable impact on simulated zonal substrate uptake clearance values and simulated intracellular perpetrator concentrations, which regulate transporter and metabolic DDIs. Additionally, accounting for rifampicin‐mediated OATP1B induction in parallel with inhibition improved model predictions. This study provides novel insight into the effect of hepatic OATP1B distribution on site‐specific DDI predictions and the impact of accounting for zonal transporter distributions within PBPK models.