Christina Schräpel (1,2), Lukas Kovar (1), Simeon Rüdesheim (1,2), Boian Ganchev (2), Patrick Kröner (2), Svitlana Igel (2), Reinhold Kerb (2), Thomas E. Mürdter (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 and University Tübingen, Stuttgart, Germany, (3) Departments of Clinical Pharmacology, and of Pharmacy and Biochemistry, University Tübingen, Tübingen, Germany
Introduction: Polycystic ovary syndrome (PCOS) is one of the major causes of female infertility, which affects about 15-20% of couples worldwide. Clomiphene, a selective estrogen receptor modulator (SERM), is the first line therapy for the treatment of anovulation for about 50 years. However, since clomiphene is metabolized primarily through CYP2D6 and CYP3A4, altered exposure of clomiphene and its active metabolites can occur due to CYP2D6 polymorphisms (drug-gene interactions (DGI)) and concomitant use of clomiphene with CYP2D6 and CYP3A4 inhibitors (drug-drug interactions (DDI) and drug-drug-gene interactions (DDGIs)) [1]. This may be one of the reasons for the high rate of nonresponding patients (8-30%) [2]. To overcome this issue, physiologically based pharmacokinetic (PBPK) modeling can be applied as a valuable tool for quantifying DGI, DDI and DDGI scenarios [3].
Objectives:
- Development and evaluation of a PBPK model of the enantiomer (E)-clomiphene (enclomiphene) including three metabolites after oral administration
- Subsequently, prediction of DGI, DDI and DDGI effects
Methods: A PBPK model of enclomiphene with its three metabolites (E)-4-hydroxyclomiphene, (E)-desethylclomiphene and (E)-4-hydroxydesethylclomiphene was built in PK-Sim® (Version 9.0) as part of the Open Systems Pharmacology Suite [4]. Data for model development were extracted from literature including physicochemical parameters. Moreover, data from an internal multi-center pharmacogenetic panel study were used with plasma concentration-time profiles and urinary data from 20 healthy females, including 6 poor metabolizers (PM), 6 intermediate metabolizers (IM), 5 extensive metabolizers (EM) and 3 ultra-rapid metabolizers (UM). For PBPK model development, 40 plasma profiles were split into an internal training (8 plasma profiles and 8 urinary profiles of enclomiphene and its metabolites from EM and PM patients) and an external test (32 plasma profiles, 32 urinary profiles) dataset. When necessary, parameters were estimated based on the internal dataset. Model performance was evaluated by comparing observed plasma profiles with predicted profiles and by comparison of predicted to observed area under the plasma concentration-time curve (AUC) values and maximum plasma concentration (Cmax) values. Moreover, as quantitative measures of the descriptive and predictive model performance the geometric mean fold errors (GMFEs) of predicted and observed AUC and Cmax, as well as the mean relative deviations (MRDs) of all predicted compared to observed plasma concentrations were calculated.
Results: The PBPK model includes metabolism of enclomiphene and its metabolites primarily through CYP2D6 and CYP3A4 as reported in literature, while other CYP enzymes play minor roles [1]. The values for the corresponding Michaelis-Menten constants KM were obtained from literature. Separate kcat values were assigned to the different CYP2D6 phenotypes and informed through the internal dataset as well as in vitro data. The final model was capable to describe and predict all plasma concentration-time profiles of the internal and external dataset with GMFE values for AUC of 1.42 and for Cmax of 1.45 for all plasma concentration-time profiles as well as an overall MRD value of 1.92. CYP2D6 PM, IM, EM and UM profiles could be predicted successfully based on CYP2D6 activity scores with GMFE values for the DGI ratios (predicted AUC ratio PM, IM or UM to EM versus observed AUC ratio PM, IM or UM to EM) of 1.43 for AUC and 1.46 for Cmax. Furthermore, an effective coupling with an existing clarithromycin model [5] enabled the prediction of DDI and DDGI effects of concomitant use of clarithromycin with enclomiphene. The GMFE values for the DDI and DDGI ratios (predicted AUC ratio enclomiphene + clarithromycin to enclomiphene versus observed AUC ratio enclomiphene + clarithromycin to enclomiphene) were 1.45 for AUC and 1.47 for Cmax.
Conclusion: A whole-body PBPK model of enclomiphene including three important metabolites (E)-4-hydroxyclomiphene, (E)-desethylclomiphene and (E)-4-hydroxydesethylclomiphene was successfully developed. The model was able to predict plasma profiles within CYP2D6 and CYP3A4 DGI, DDI and DDGI scenarios.
Funding: This work was funded by the Robert Bosch Stiftung (Stuttgart, Germany), the European Commission Horizon 2020 UPGx grant 668353, and a grant from the German Federal Ministry of Education and Research (BMBF 031L0188D, “GUIDE-IBD”).
References:
[1] Mürdter T et al.: Genetic polymorphism of cytochrome P450 2D6 determines oestrogen receptor activity of the major infertility drug clomiphene via its active metabolites. Hum Mol Genet (2012) 21(5):1145-1154.
[2] Kim MJ et al.: Effect of the CYP2D6*10 allele on the pharmacokinetics of clomiphene and its active metabolites. Arch Pharm Res (2018) 41:347-353.
[3] U.S. Food and Drug Administration. Clinical Drug Interaction Studies – Study Design, Data Analysis, Implications for Dosing, and Labeling Recommendations. Draft Guidance for Industry (2017).
[4] Eissing T et al.: A computational systems biology software platform for multiscale modeling and simulation: integrating whole-body physiology, disease biology, and molecular reaction networks. Front Physiol (2011) 2:4.
[5] Hanke N et al.: 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(10): 647-659.
Reference: PAGE 29 (2021) Abstr 9819 [www.page-meeting.org/?abstract=9819]
Poster: Drug/Disease Modelling - Absorption & PBPK