Ji Eun Park (1,2), Wooin Lee (3), Yoshihisa Shitara (1), Shigemichi Morita (1), Jasminder Sahi (1), Kota Toshimoto (2), and Yuichi Sugiyama (2)
(1) Sanofi, Japan and China, (2) RIKEN, Japan, (3) Seoul National University, South Korea
Introduction: Pemafibrate (PMF) is a substrate drug of transporters including OATP1B1 and OATP1B3 as well as multiple isoforms of cytochrome P450 including CYP3A4. The systemic exposure of PMF was increased by a single-dose coadministration of cyclosporine (CsA) or rifampicin (RIF) while it was decreased by repeated-dose coadministration of RIF. We established a physiologically based pharmacokinetic (PBPK) model of PMF to retrospectively investigate these drug-drug interactions (DDIs).
Objectives:
To quantitatively explain the DDIs by using a PBPK model incorporating transporter-mediated hepatic uptake.
To determine appropriate pharmacokinetic (PK) parameters, especially inhibition constants (Ki), for capturing the PK profile of PMF with coadministration of CsA and RIF.
Methods: PK parameters for PMF were obtained and/or calculated based on the information of the common technical documents from Kowa Ltd. PK simulation and fitting analyses were performed by using a computer software, Napp ver 2.31 (University of Tokyo Hospital). Reported PBPK models of CsA and RIF were used with the reported in vivo Ki values of CsA and RIF for OATP1Bs as well as the induction parameters of RIF for OATP1Bs and CYPs [1,2].
Results: The PK profile of a single dose of PMF alone was well captured by using a PBPK model incorporating transporter-mediated hepatic uptake and enterohepatic circulation after optimizing the PK parameters for overall intrinsic hepatic clearance, biliary excretion, and absorption by fitting analysis. Simulation analysis predicted an increase in PMF exposure by a coadministration of single-dose CsA and RIF using reported Ki values for OATP1Bs (0.012 and 0.23 µM for CsA and RIF, respectively [1,2]), but to a lesser extent than the observed data (AUC & Cmax; Observed, CsA 14- & 9-fold, Rif 11- & 9-fold; Simulated, CsA 8- & 5-fold, Rif 3- & 3-fold, respectively). Sensitivity analysis indicated that lower Ki values for OATP1Bs (3~5-fold) than the reported values were necessary to capture the observed data of PMF. Even in this case, the absorption phase was not well captured, and further optimization of absorption rate constants (ka) was required. Finally, 2~8-fold lower Ki value for OATP1Bs and 4-fold increase in ka value well captured the PK profile of PMF when coadministered with CsA and RIF. For multiple-dose RIF, simulation results successfully captured the observed data (AUC & Cmax; Observed, 0.22- & 0.38-fold; Simulated, 0.25- & 0.29-fold, respectively) using reported induction parameters for CYPs and OATP1Bs [2].
Conclusions: Overall, simulation results were in agreement with the observed DDI data (coadministered PMF with CsA and RIF) in the present analysis with two consideration: i) Lower Ki values for OATP than the reported Ki values, albeit obtaining from in vivo DDI studies, suggesting the susbstrate-dependent Ki values for OATP1Bs as previously reported [3]. ii) Higher Ka value, suggesting the altered absorption rate by the inhibitory effect of CsA and Rif on efflux transporters in intestine.
References:
[1] Yoshikado T et al. Clin Pharmacol Ther (2016) 100, 513-523.
[2] Asaumi R et al. CPT Pharmacometrics Syst Pharmacol (2018) 7, 186-196.
[3] Izumi S et al. Drug Metab Dispos (2013) 41, 1859-1866.
Reference: PAGE () Abstr 9347 [www.page-meeting.org/?abstract=9347]
Poster: Drug/Disease Modelling - Absorption & PBPK