4β-Hydroxycholesterol as an endogenous biomarker for CYP3A induction: Scientific rationale, clinical utility, and future perspectives.

Entrectinib is a selective inhibitor of ROS1/TRK/ALK kinases, recently approved for oncology indications. Entrectinib is predominantly cleared by cytochrome P450 (CYP)3A4, and modulation of CYP3A enzyme activity profoundly alters the pharmacokinetics of both entrectinib and its active metabolite M5. We describe development of a combined physiologicallybased pharmacokinetic (PBPK) model for entrectinib and M5 to support dosing recommendations when entrectinib is co-administered with CYP3A4 inhibitors or inducers. A PBPK model was established in Simcyp® Simulator. The initial model based on in vitro-in vivo extrapolation was refined using sensitivity analysis and non-linear mixed effects modeling to optimize parameter estimates and to improve model fit to data from a clinical drug-drug interaction study with the strong CYP3A4 inhibitor, itraconazole. The model was subsequently qualified against clinical data, and the final qualified model used to simulate the effects of moderate to strong CYP3A4 inhibitors and inducers on entrectinib and M5 pharmacokinetics. The final model showed good predictive performance for entrectinib and M5, meeting commonly used predictive performance acceptance criteria in each case. The model predicted that co-administration of various moderate CYP3A4 inhibitors (verapamil, erythromycin, clarithromycin, fluconazole, and diltiazem) would result in an average increase in entrectinib exposure between 2.2- and 3.1-fold, with corresponding average increases for M5 of approximately 2-fold. Co-administration of moderate CYP3A4 inducers (efavirenz, carbamazepine, phenytoin) was predicted to result in an average decrease in entrectinib exposure between 45 and 79%, with corresponding average decreases for M5 of approximately 50%. The model simulations were used to derive dosing recommendations for co-administering entrectinib with CYP3A4 inhibitors or inducers. PBPK modeling has been used in lieu of clinical studies to enable regulatory decision-making.

Maribavir, an orally available antiviral agent, has been approved in multiple countries for the treatment of patients with refractory post-transplant cytomegalovirus (CMV) infection and/or disease. Maribavir is primarily metabolized by CYP3A4; coadministration with CYP3A4 inducers and inhibitors may significantly alter maribavir exposure, thereby affecting its efficacy and safety. The effect of CYP3A4 inducers and inhibitors on maribavir exposure was evaluated based on a drug-drug interaction (DDI) study and physiologically-based pharmacokinetic (PBPK) modeling. The effect of rifampin (a strong inducer of CYP3A4 and moderate inducer of CYP1A2), administered at a 600mg dose once daily, on maribavir pharmacokinetics was assessed in a clinical phase1 DDI study in healthy participants. A full PBPK model for maribavir was developed and verified using invitro and clinical pharmacokinetic data from phase1 studies. The verified PBPK model was then used to simulate maribavir DDI interactions with various CYP3A4 inducers and inhibitors. The DDI study results showed that coadministration with rifampin decreased the maribavir maximum plasma concentration (Cmax), area under the plasma concentration-time curve (AUC), and trough concentration (Ctrough) by 39%, 60%, and 82%, respectively. Based on the results from the clinical DDI study, the coadministration of maribavir with rifampin is not recommended. The PBPK model did not predict a clinically significant effect of CYP3A4 inhibitors on maribavir exposure; however, it predicted that strong or moderate CYP3A4 inducers, including carbamazepine, efavirenz, phenobarbital, and phenytoin, may reduce maribavir exposure to a clinically significant extent, and may prompt the consideration of a maribavir dosing increase, in accordance with local approved labels and/or regulations.

Mavacamten is a first-in-class, oral, selective, allosteric, reversible cardiac myosin inhibitor approved by the US Food and Drug Administration for the treatment of adults with symptomatic New York Heart Association functional class II-III obstructive hypertrophic cardiomyopathy. Mavacamten is metabolized in the liver, predominantly via cytochrome P450 (CYP) enzymes CYP2C19 (74%), CYP3A4 (18%), and CYP2C9 (8%). A physiologically-based pharmacokinetic (PBPK) model was developed using Simcyp version 19 (Certara, Princeton, NJ). Following model verification, the PBPK model was used to explore the effects of strong CYP3A4 and CYP2C19 inducers, and strong, moderate, and weak CYP2C19 and CYP3A4 inhibitors on mavacamten pharmacokinetics (PK) in a healthy population, with the effect of CYP2C19 phenotype predicted for poor, intermediate, normal, and ultrarapid metabolizers. The PBPK model met the acceptance criteria for all verification simulations (> 80% of model-predicted PK parameters within 2-fold of those observed clinically). A weak induction effect was predicted when mavacamten was administered with a strong CYP3A4 inducer in poor metabolizers. Moderate reductions in mavacamten exposure were predicted with a strong CYP2C19/CYP3A4 inducer in all CYP2C19 phenotypes. Except for the effect of strong CYP2C19 inhibitors on ultrarapid metabolizers, steady-state area under plasma concentration-time curve and maximum plasma concentration values were weakly affected (< 2-fold) or not affected (< 1.25-fold), regardless of CYP2C19 phenotype. In conclusion, a fit-for-purpose PBPK model was developed and verified, which accurately predicted the available clinical data and was used to simulate the potential impact of CYP induction and inhibition on mavacamten PKs, stratified by CYP2C19 phenotype.

ABSTRACTA physiologically based pharmacokinetic (PBPK) model was developed and verified for dordaviprone, a small molecule with antitumor effects in glioma patients. The model was applied to assess the drug–drug interaction (DDI) potential of dordaviprone as a victim of CYP3A4 inhibitors and inducers, and as a perpetrator of CYP3A4, CYP2C8, CYP2D6 inhibition. A combination of in vitro and clinical data was used to develop a minimal distribution PBPK model with a single adjusting compartment and mechanistic absorption using the Simcyp Population‐Based Simulator (V21). Simulated maximum concentration (Cmax) and area under the concentration time curve (AUC) of the 3 clinical studies used to verify the PBPK model were within 1.4‐fold of observed exposures. The simulated increase in dordaviprone AUC and Cmax (4.6‐ and 1.7‐fold) following administration of multiple doses of itraconazole was consistent with the observed values (4.4‐ and 1.9‐fold). All PBPK‐simulated changes in dordaviprone plasma exposure when administered with CYP3A4 moderate (erythromycin, fluconazole) and weak (cimetidine) inhibitors, and moderate (efavirenz) and strong (rifampicin) inducers were consistent with their CYP3A4 potency classification (AUC ratio = 2.68, 2.48, 1.42, 0.35, and 0.17, respectively). The simulated AUC and Cmax of probe substrates for CYP3A4 (midazolam), CYP2C8 (repaglinide) and CYP2D6 (desipramine) after coadministration with 625 mg dordaviprone were the same as those in the absence of dordaviprone (ratio = 1.0) and remained unchanged after a sensitivity analysis using 10‐fold more potent inhibition constants. Due to changes in dordaviprone plasma exposure when co‐administered with CYP3A4 inhibitors, dordaviprone dose adjustments may be necessary; CYP3A4 inducers should be avoided.

PurposeIpatasertib, a potent and highly selective small-molecule inhibitor of AKT, is currently under investigation for treatment of cancer. Ipatasertib is a substrate and a time-dependent inhibitor of CYP3A4. It exhibits non-linear pharmacokinetics at subclinical doses in the clinical dose escalation study. To assess the DDI risk of ipatasertib at the intended clinical dose of 400 mg with CYP3A4 inhibitors, inducers, and substrates, a fit-for-purpose physiologically based pharmacokinetic (PBPK) model of ipatasertib was developed.MethodsThe PBPK model was constructed in Simcyp using in silico, in vitro, and clinical data and was optimized and verified using clinical data.ResultsThe PBPK model described non-linear pharmacokinetics of ipatasertib and captured the magnitude of the observed clinical DDIs. Following repeated doses of 400 mg ipatasertib once daily (QD), the PBPK model predicted a 3.3-fold increase of ipatasertib exposure with itraconazole; a 2–2.5-fold increase with moderate CYP3A4 inhibitors, erythromycin and diltiazem; and no change with a weak CYP3A4 inhibitor, fluvoxamine. Additionally, in the presence of strong or moderate CYP3A4 inducers, rifampicin and efavirenz, ipatasertib exposures were predicted to decrease by 86% and 74%, respectively. As a perpetrator, the model predicted that ipatasertib (400 mg) caused a 1.7-fold increase in midazolam exposure.ConclusionThis study demonstrates the value of using a fit-for-purpose PBPK model to assess the clinical DDIs for ipatasertib and to provide dosing strategies for the concurrent use of other CYP3A4 perpetrators or victims.

Evofosfamide is a cytotoxic small‐molecule prodrug preferentially activated under hypoxic conditions. The cytotoxicity of evofosfamide impacted the generation of in vitro drug‐drug interaction (DDI) data, especially in vitro induction results. Therefore, a novel physiologically based pharmacokinetic (PBPK) approach was used, which involved available in vitro and clinical data of evofosfamide and combined it with induction data from the prototypical cytochrome P450 (CYP)3A inducer rifampicin. The area under the concentration‐time curve (AUC) ratios of midazolam were above 0.80, indicating that induction of CYP3A by evofosfamide administered weekly is unlikely to occur in humans. Moreover, static and PBPK modeling showed no clinically relevant inhibition via CYP2B6, CYP2D6, and CYP3A4. In conclusion, PBPK models were used to supplement in vitro information of a cytotoxic compound. This approach may set a precedent for future studies of cytotoxic drugs, potentially reducing the need for clinical DDI studies and providing more confidence in the clinical use of approved cytotoxic compounds for which DDI information is sparse.

In the past, rifampicin was well-established as strong index CYP3A inducer in clinical drug-drug interaction (DDI) studies. However, due to identified potentially genotoxic nitrosamine impurities, it should not any longer be used in healthy volunteer studies. Available clinical data suggestcarbamazepine as an alternative to rifampicin as strong index CYP3A4 inducer in clinical DDI studies. Further, physiologically-based pharmacokinetic (PBPK) modeling is a tool with increasing importance to support the DDI risk assessment of drugs during drug development. CYP3A4 induction properties and the safety profile of carbamazepine were investigated in two open-label, fixed sequence, crossover clinical pharmacology studies in healthy volunteers using midazolam as a sensitive index CYP3A4 substrate. Carbamazepine was up-titrated from 100 mg twice daily (b.i.d.) to 200 mg b.i.d., and to a final dose of 300 mg b.i.d. for 10 consecutive days. Mean area under plasma concentration-time curve from zero to infinity (AUC(0-∞)) of midazolam consistently decreased by 71.8% (ratio: 0.282, 90% confidence interval (CI): 0.235-0.340) and 67.7% (ratio: 0.323, 90% CI: 0.256-0.407) in study 1 and study 2, respectively. The effect was adequately described by an internally developed PBPK model for carbamazepine which has been made freely available to the scientific community. Further, carbamazepine was safe and well-tolerated in the investigated dosing regimen in healthy participants. The results demonstrated that the presented design is appropriate for the use of carbamazepine as alternative inducer to rifampicin in DDI studies acknowledging its CYP3A4 inductive potency and safety profile.

Predicting bioequivalence and developing dissolution bioequivalence safe space in vitro for warfarin using a Physiologically-Based pharmacokinetic absorption model

Background: Physiologically based pharmacokinetic (PBPK) models utilize computer-based simulations to predict the pharmacokinetics of drugs. By using mathematical modeling techniques consisting of differential equations to simulate blood flow, tissue compositions, and organ properties, the pharmacokinetic properties of drugs can be better understood. Specifically, PBPK models can provide predictive information about drug absorption, distribution, metabolism, and excretion (ADME). The information gained from PBPK models can be useful in both drug discovery, development, and regulatory science. PBPK models can help to address some of the ethical dilemmas that arise during the drug development process, particularly when examining patient populations where testing a new drug may have significant ethical concerns. Patient populations where significant physiological change (i.e., pregnancy, pediatrics, geriatrics, organ impairment populations, etc.) and pathophysiological influences resulting in PK changes can also benefit from PBPK modeling. Additionally, PBPK models can be utilized to predict variations in drug metabolism resulting from genetic polymorphisms, age, and disease states. Methods: In this mini-review, we examine the various applications of PBPK models in drug metabolism. Current research articles related to drug metabolism in genetics, life-stages, and disease states were reviewed. Results: Several key factors in genetics, life-stage, and disease states that affect metabolism in PBPK models are identified. In genetics, the role of CYP enzymes, genetic polymorphisms, and ethnicity may influence metabolism. Metabolism generally changes over time from neonate, pediatric, adult, geriatric, and perinatal populations. Disease states such as renal and hepatic impairment, weight and other acute and chronic diseases also can also alter metabolism. Several examples of PBPK models applying these physiological changes have been published. Conclusions: The utilization and recognition of these specific areas in PBPK modeling can aid in personalized dosing strategy, clinical trial optimization, and regulatory submission.

Physiologically based pharmacokinetic (PBPK) modeling was used to predict the human pharmacokinetics and drug-drug interaction (DDI) of GDC-2394. PBPK models were developed using in vitro and in vivo data to reflect the oral and intravenous PK profiles of mouse, rat, dog, and monkey. The learnings from preclinical PBPK models were applied to a human PBPK model for prospective human PK predictions. The prospective human PK predictions were within 3-fold of the clinical data from the first-in-human study, which was used to optimize and validate the PBPK model and subsequently used for DDI prediction. Based on the majority of PBPK modeling scenarios using the in vitro CYP3A induction data (mRNA and activity), GDC-2394 was predicted to have no-to-weak induction potential at 900 mg twice daily (BID). Calibration of the induction mRNA and activity data allowed for the convergence of DDI predictions to a narrower range. The plasma concentrations of the 4β-hydroxycholesterol (4β-HC) were measured in the multiple ascending dose study to assess the hepatic CYP3A induction risk. There was no change in plasma 4β-HC concentrations after 7 days of GDC-2394 at 900 mg BID. A dedicated DDI study found that GDC-2394 has no induction effect on midazolam in humans, which was reflected by the totality of predicted DDI scenarios. This work demonstrates the prospective utilization of PBPK for human PK and DDI prediction in early drug development of GDC-2394. PBPK modeling accompanied with CYP3A biomarkers can serve as a strategy to support clinical pharmacology development plans. SIGNIFICANCE STATEMENT: This work presents the application of physiologically based pharmacokinetic modeling for prospective human pharmacokinetic (PK) and drug-drug interaction (DDI) prediction in early drug development. The strategy taken in this report represents a framework to incorporate various approaches including calibration of in vitro induction data and consideration of CYP3A biomarkers to inform on the overall CYP3A-related DDI risk of GDC-2394.

Monitoring endogenous biomarkers is increasingly used to evaluate transporter-mediated drug-drug interactions (DDIs) in early drug development and may be applied to elucidate changes in transporter activity in disease. 4-pyridoxic acid (PDA) has been identified as the most sensitive plasma endogenous biomarker of renal organic anion transporters (OAT1/3). Increase in PDA baseline concentrations was observed after administration of probenecid, a strong clinical inhibitor of OAT1/3 and also in patients with chronic kidney disease (CKD). The aim of this study was to develop and verify a physiologically-based pharmacokinetic (PBPK) model of PDA, to predict the magnitude of probenecid DDI and predict the CKD-related changes in PDA baseline. The PBPK model for PDA was first developed in healthy population, building on from previous population pharmacokinetic modeling, and incorporating a mechanistic kidney model to consider OAT1/3-mediated renal secretion. Probenecid PBPK model was adapted from the Simcyp database and re-verified to capture its dose-dependent pharmacokinetics (n = 9 studies). The PBPK model successfully predicted the PDA plasma concentrations, area under the curve, and renal clearance in healthy subjects at baseline and after single/multiple probenecid doses. Prospective simulations in severe CKD predicted successfully the increase in PDA plasma concentration relative to healthy (within 2-fold of observed data) after accounting for 60% increase in fraction unbound in plasma and additional 50% decline in OAT1/3 activity beyond the decrease in glomerular filtration rate. The verified PDA PBPK model supports future robust evaluation of OAT1/3 DDI in drug development and increases our confidence in predicting exposure and renal secretion in patients with CKD.

Physiological changes during pregnancy can affect antiretroviral drug processes and further influence drug efficacy and safety. Physiologically-based pharmacokinetic (PBPK) modeling offers a unique modality to predict PK in pregnant women. The objective of this study was to establish a PBPK modeling of tenofovir disoproxil fumarate (TDF) in pregnant women, to provide a reference for the clinical use of TDF. A full PBPK modeling of tenofovir (TFV) and TDF following i.v. and p.o. administration was developed using the simulation software PK-Sim®. The modeling was then extrapolated to pregnant women based on pregnancy- related physiological parameters in Mobi® Simulator. The mean fold error (MFE) and geometric mean fold error (GMFE) methods were used to compare the differences between predicted and observed values of PK parameters (Cmax, tmax, AUC0-∞) to evaluate the accuracy of PBPK modeling. The developed PBPK modeling successfully predicted the TDF disposition in the non-pregnant population, wherein the MFE average and GMFE of all predicted PK parameters were within a 1.5-fold error range, and more than 96.30% of the predicted drug concentration values were within a 2-fold error range of the measured values. After the extrapolation of these models to the third trimester of pregnancy, the scaling anatomy/physiology and hepatic intrinsic clearance made the pregnant population PBPK modeling meet the standard requirement of 0.5 < MFE and GMFE value < 2. It was more appropriate to simulate the in vivo process of low-dose TDF in pregnant women. The non-pregnant population PBPK modeling of TDF established in our study can be extrapolated to pregnant women. Our study provides a reference for realizing clinical personalized medication for pregnant women.

Brigatinib is an oral anaplastic lymphoma kinase (ALK) inhibitor approved for the treatment of ALK-positive metastatic non-small cell lung cancer. Invitro studies indicated that brigatinib is primarily metabolized by CYP2C8 and CYP3A4 and inhibits P-gp, BCRP, OCT1, MATE1, and MATE2K. Clinical drug-drug interaction (DDI) studies with the strong CYP3A inhibitor itraconazole or the strong CYP3A inducer rifampin demonstrated that CYP3A-mediated metabolism was the primary contributor to overall brigatinib clearance in humans. A physiologically-based pharmacokinetic (PBPK) model for brigatinib was developed to predict potential DDIs, including the effect of moderate CYP3A inhibitors or inducers on brigatinib pharmacokinetics (PK) and the effect of brigatinib on the PK of transporter substrates. The developed model was able to predict clinical DDIs with itraconazole (area under the plasma concentration-time curve from time 0 to infinity [AUC∞] ratio [with/without itraconazole]: predicted 1.86; observed 2.01) and rifampin (AUC∞ ratio [with/without rifampin]: predicted 0.16; observed 0.20). Simulations using the developed model predicted that moderate CYP3A inhibitors (e.g., verapamil and diltiazem) may increase brigatinib AUC∞ by ~40%, whereas moderate CYP3A inducers (e.g., efavirenz) may decrease brigatinib AUC∞ by ~50%. Simulations of potential transporter-mediated DDIs predicted that brigatinib may increase systemic exposures (AUC∞) of P-gp substrates (e.g., digoxin and dabigatran) by 15%-43% and MATE1 substrates (e.g., metformin) by up to 29%; however, negligible effects were predicted on BCRP-mediated efflux and OCT1-mediated uptake. The PBPK analysis results informed dosing recommendations for patients receiving moderate CYP3A inhibitors (40% brigatinib dose reduction) or inducers (up to 100% increase in brigatinib dose) during treatment, as reflected in the brigatinib prescribing information.

Drug-Drug Interaction Risk Assessment of Esaxerenone as a Perpetrator by In Vitro Studies and Static and Physiologically Based Pharmacokinetic Models

Cobimetinib is eliminated mainly through cytochrome P450 (CYP) 3A4-mediated hepatic metabolism in humans. A clinical drug-drug interaction (DDI) study with the potent CYP3A4 inhibitor itraconazole resulted in an approximately sevenfold increase in cobimetinib exposure. The DDI risk for cobimetinib with other CYP3A4 inhibitors and inducers needs to be assessed in order to provide dosing instructions. A physiologically based pharmacokinetic (PBPK) model was developed for cobimetinib using in vitro data. It was then optimized and verified using clinical pharmacokinetic data and itraconazole-cobimetinib DDI data. The contribution of CYP3A4 to the clearance of cobimetinib in humans was confirmed using sensitivity analysis in a retrospective simulation of itraconazole-cobimetinib DDI data. The verified PBPK model was then used to predict the effect of other CYP3A4 inhibitors and inducers on cobimetinib pharmacokinetics. The PBPK model described cobimetinib pharmacokinetic profiles after both intravenous and oral administration of cobimetinib well and accurately simulated the itraconazole-cobimetinib DDI. Sensitivity analysis suggested that CYP3A4 contributes ~78% of the total clearance of cobimetinib. The PBPK model predicted no change in cobimetinib exposure (area under the plasma concentration-time curve, AUC) with the weak CYP3A inhibitor fluvoxamine and a three to fourfold increase with the moderate CYP3A inhibitors, erythromycin and diltiazem. Similarly, cobimetinib exposure in the presence of strong (rifampicin) and moderate (efavirenz) CYP3A inducers was predicted to decrease by 83 and 72%, respectively. This study demonstrates the value of using PBPK simulation to assess the clinical DDI risk inorder to provide dosing instructions with other CYP3A4 perpetrators.