A physiologically-based pharmacokinetic (PBPK) model network for the prediction of CYP1A2 and CYP2C19 drug–drug–gene interactions with fluvoxamine, omeprazole, S-mephenytoin, moclobemide, tizanidine, mexiletine, ethinylestradiol and caffeine
Tobias Kanacher (1*), Ingrid Michon (1**), Andreas Lindauer (1***), Enrica Mezzalana (1*), Chantaratsamon Dansirikul (2), Celine Veau (2), Jose David Gómez Mantilla (2), Valerie Nock (2) and Angèle Fleury (2)
(1) SGS-Exprimo, (2) Boehringer Ingelheim Pharma, (*) currently employed by Pharmetheus, (**) currently employed by Certara, (***) currently employed by Calvagone
Objectives: Physiologically-based pharmacokinetic (PBPK) modeling is a well-recognized method for quantitatively predicting the effect of intrinsic/extrinsic factors on drug exposure. However, there are few verified, freely accessible and modifiable, comprehensive drug-drug interaction (DDI) PBPK models. This study aimed to develop a qualified PBPK DDI network for cytochrome P450 CYP2C19 and CYP1A2 substrates and strong or moderate CYP2C19 inhibitors to expand the library of available models for DDI prediction with qualified whole-body PBPK models. Additionally, this study aimed to describe and predict drug–drug–gene interactions (DDGIs) between substrate and inhibitor drugs for patients with genetic polymorphisms affecting CYP2C19 metabolism.
Methods: Reference index substrates and inhibitors recommended by the FDA were selected to include of a range of strong and moderate CYP1A2 and CYP2C19 inhibitors. As a selection, template PBPK models were developed for interactions between fluvoxamine, S-mephenytoin, moclobemide, omeprazole, mexiletine, tizanidine and ethinylestradiol as the inhibitors and/or substrates. Each substrate or inhibitor was validated by at least two independent clinical drug inhibition combination studies if data were available. Concentration-time data from relevant publications were digitized in addition to unpublished clinical data sponsored by Boehringer Ingelheim. PBPK model development and simulations were performed using the open source software Open Systems Pharmacology (OSP) suite. Key endpoints were the predicted inhibitor/substrate drug concentration-time profiles, area under the plasma concentration time curve (AUC) and maximum plasma concentration (Cmax) for each evaluated DDI pair, as well as the AUC and Cmax ratios for predicted vs. observed.
Results: DDI simulations with fluvoxamine as a strong CYP1A2 inhibitor, and caffeine, tizanidine or mexiletine as victims, demonstrated an excellent prediction of the substrate concentrations, with mean predicted/observed Cmax and AUC ratios of around 1. Excellent predictions were also obtained in DDI simulations with fluvoxamine as a strong CYP2C19 inhibitor and omeprazole or S-mephenytoin as substrate. Predictions of the inhibitory potential of fluvoxamine on CYP2C19 were excellent for both CYP219 extensive (EM) and poor metabolizer (PM) subjects. The PBPK model for omeprazole, developed as racemate for the esomeprazole and R-omeprazole enantiomers, predicted the observed concentration-time profiles after single and multiple doses esomeprazole/R-omeprazole in both CYP2C19 EM and PM well. In addition, DDI simulations with omeprazole as a moderate CYP2C19 inhibitor and moclobemide as substrate demonstrated a good prediction of moclobemide levels. DDI simulations with omeprazole as substrate and moclobemide as a moderate CYP2C19 inhibitor also demonstrated good prediction of both omeprazole and moclobemide levels. DDI simulations with mexiletine as moderate CYP1A2 inhibitor and caffeine or tizanidine as substrates demonstrated an underprediction of levels for both substrates, but within 2-fold. DDI simulations with ethinylestradiol as moderate CYP1A2 inhibitor and caffeine or tizanidine as substrates demonstrated a good prediction of both substrates.
Conclusions: The presented PBPK models qualified as a CYP2C19 and CYP1A2 PBPK DDI-network. In addition, the models were able to account for impact of CYP2C19 polymorphism within this network. The models are freely provided on the Open Systems Pharmacology Suite platform, expanding the library of publicly available models for DDI predictions with a qualified CYP2C19 – CYP1A2 PBPK network.
 Drug Development and Drug Interactions: Table of Substrates, Inhibitors and Inducers, https://www.fda.gov/drugs/drug-interactions-labeling/drug-development-and-drug-interactions-table-substrates-inhibitors-and-inducers.
 Open Systems Pharmacology Suite: PK-SIM® AND MOBI® FOR PBPK AND QUANTITATIVE SYSTEMS PHARMACOLOGY, http://www.open-systems-pharmacology.org/