Physiologically Based Pharmacokinetic Modeling of Imatinib and Its Main Metabolite for Drug–Drug Interaction Predictions
Helena Leonie Hanae Loer (1), Christina Kovar (1,2), Matthias Schwab (2,3) and Thorsten Lehr (1)
(1) Clinical Pharmacy, Saarland University, Saarbrücken, Germany; (2) Dr. Margarete Fischer-Bosch-Institute of Clinical Pharmacology, Stuttgart, Germany; (3) Departments of Clinical Pharmacology, Pharmacy and Biochemistry, University of Tübingen, Tübingen, Germany
Introduction: Since its approval in 2001, the tyrosine kinase inhibitor imatinib has revolutionized the development of targeted cancer therapy and, despite relatively high resistance rates, remains one of the frontline therapies, e.g., against chronic myeloid leukemia [1,2]. Oral administration is followed by a relative bioavailability of more than 97% and its metabolism occurs primarily via cytochrome P450 (CYP) enzymes 2C8 and 3A4 [3,4]. In vitro studies have additionally identified imatinib as a substrate of numerous influx and efflux transporters, e.g., P-glycoprotein (P-gp) [5]. Imatinib is highly susceptible to drug–drug interactions (DDIs), acting on the one hand as a victim when co-administered with perpetrator drugs, potentially increasing nonresponse to treatment and toxicity [6]. On the other hand, imatinib affects not only its own metabolism but also the exposure of co-medications via inhibition of CYP2C8 and CYP3A4 [7]. To ensure safety and efficacy of imatinib therapy, its pharmacokinetics (PK) and clinically relevant sensitivity to DDIs should be thoroughly investigated. Here, physiologically based pharmacokinetic (PBPK) modeling provides a suitable framework to study not only the PK of individual drugs but also interactive effects during co-medication.
Objectives:
- Development of a whole-body PBPK model for imatinib and its main metabolite N-desmethyl imatinib (NDMI)
- Prediction of DDIs involving different perpetrator drugs
Methods: Model development with the modeling software PK-Sim® (Version 11, www.open-systems-pharmacology.org) was initiated with an extensive literature search for information on the absorption, distribution, metabolism, and excretion of imatinib/NDMI, as well as clinical study data. The model was evaluated by visual comparison of predicted with observed profiles and by quantitative examination of the deviation between predicted and observed PK parameters area under the plasma concentration–time curve (AUC) and maximum plasma concentration (Cmax). A twofold deviation was set as prediction success threshold. DDI modeling was performed by coupling the imatinib PBPK model with previously published models of the CYP2C8/CYP3A4 inhibitor ketoconazole and inducer rifampicin [8,9]. Subsequently, the DDI prediction performance was assessed by comparing predicted with observed DDI AUC and Cmax ratios, i.e., the ratio of the respective PK parameter with versus without perpetrator co-administration, applying the prediction success limits proposed by Guest et al. [10].
Results: A total of 37 and 12 plasma concentration–time profiles of imatinib and NDMI, respectively, were used to build and evaluate the PBPK model. Included studies covered a dose range of 25–750 mg imatinib administered either intravenously or orally in single- and multiple-dose studies. Metabolism was implemented for both imatinib and NDMI via CYP2C8 and CYP3A4, while additional transport via P-gp was included for imatinib, with processes modeled as Michaelis–Menten or first–order kinetics. Furthermore, autoinhibition was considered by incorporating competitive inhibition (CI) of CYP2C8 and mechanism-based inactivation of CYP3A4 by imatinib, as well as CI of CYP2C8 and CYP3A4 by NDMI [7]. The final imatinib PBPK model showed good descriptive and predictive performance regarding imatinib and NDMI plasma profiles, with 92% and 98% of predicted concentration measurements and PK parameters, respectively, within the twofold deviation. Moreover, CYP2C8/CYP3A4 DDI scenarios involving the concomitant use of the strong inhibitor ketoconazole as well as pretreatment with the strong inducer rifampicin were well predicted, as 6/6 of the predicted PK parameter ratios were within the limits proposed by Guest et al. [10].
Conclusions: In the presented study, a parent–metabolite whole-body PBPK model for imatinib and its main metabolite NDMI was successfully developed and applied to predict DDI scenarios involving imatinib as a victim drug. The DDI model could be extended to cover additional DDI scenarios as appropriate observed data and perpetrator models become available. However, the PBPK model could also be used to predict the effects of perpetrator drugs not yet studied in clinical DDI trials on the PK of imatinib/NDMI. Moreover, the model could support model-informed precision dosing by providing dose adjustments of imatinib to compensate for altered exposure due to perpetrator co-administration.
References:
[1] Kantarjian, H.M.; Jain, N.; Garcia-Manero, G.; Welch, M.A.; Ravandi, F.; Wierda, W.G.; Jabbour, E.J. The Cure of Leukemia through the Optimist’s Prism. Cancer 2022, 128, 240–259, doi:10.1002/cncr.33933.
[2] Roskoski, R.J. Properties of FDA-Approved Small Molecule Protein Kinase Inhibitors: A 2023 Update. Pharmacol. Res. 2023, 187, 106552, doi:10.1016/j.phrs.2022.106552.
[3] Peng, B.; Dutreix, C.; Mehring, G.; Hayes, M.J.; Ben-Am, M.; Seiberling, M.; Pokorny, R.; Cupdeville, R.; Lloyd, P. Absolute Bioavailability of Imatinib (Glivec®) Orally versus Intravenous Infusion. J. Clin. Pharmacol. 2004, 44, 158–162, doi:10.1177/0091270003262101.
[4] Filppula, A.M.; Neuvonen, M.; Laitila, J.; Neuvonen, P.J.; Backman, J.T. Autoinhibition of CYP3A4 Leads to Important Role of CYP2C8 in Imatinib Metabolism: Variability in CYP2C8 Activity May Alter Plasma Concentrations and Response. Drug Metab. Dispos. 2013, 41, 50–59, doi:10.1124/dmd.112.048017.
[5] Hamada, A.; Miyano, H.; Watanabe, H.; Saito, H. Interaction of Imatinib Mesilate with Human P-Glycoprotein. J. Pharmacol. Exp. Ther. 2003, 307, 824–828, doi:10.1124/jpet.103.055574.
[6] Bowlin, S.J.; Xia, F.; Wang, W.; Robinson, K.D.; Stanek, E.J. Twelve-Month Frequency of Drug-Metabolizing Enzyme and Transporter-Based Drug-Drug Interaction Potential in Patients Receiving Oral Enzyme-Targeted Kinase Inhibitor Antineoplastic Agents. Mayo Clin. Proc. 2013, 88, 139–148, doi:10.1016/j.mayocp.2012.10.020.
[7] Filppula, A.M.; Laitila, J.; Neuvonen, P.J.; Backman, J.T. Potent Mechanism-Based Inhibition of CYP3A4 by Imatinib Explains Its Liability to Interact with CYP3A4 Substrates. Br. J. Pharmacol. 2012, 165, 2787–2798, doi:10.1111/j.1476-5381.2011.01732.x.
[8] Hanke, N.; Frechen, S.; Moj, D.; Britz, H.; Eissing, T.; Wendl, T.; Lehr, T. PBPK Models for CYP3A4 and P-Gp DDI Prediction: A Modeling Network of Rifampicin, Itraconazole, Clarithromycin, Midazolam, Alfentanil, and Digoxin. CPT Pharmacometrics Syst. Pharmacol. 2018, 7, 647–659, doi:10.1002/psp4.12343.
[9] Marok, F.Z.; Wojtyniak, J.-G.; Fuhr, L.M.; Selzer, D.; Schwab, M.; Weiss, J.; Haefeli, W.E.; Lehr, T. A Physiologically Based Pharmacokinetic Model of Ketoconazole and Its Metabolites as Drug–Drug Interaction Perpetrators. Pharmaceutics 2023, 12, 679.
[10] Guest, E.J.; Aarons, L.; Houston, J.B.; Rostami-Hodjegan, A.; Galetin, A. Critique of the Two-Fold Measure of Prediction Success for Ratios: Application for the Assessment of Drug-Drug Interactions. Drug Metab. Dispos. 2011, 39, 170–173, doi:10.1124/dmd.110.036103.