Mengxu Zhang1, Thomas I.F.H. Cremers2,3, Vivi Rottschäfer4,5, Elizabeth C.M. de Lange1
1Division of Systems Pharmacology and Pharmacy, Leiden Academic Centre of Drug Research,, 2Quantall Holding BV, 3Department of Analytical Biochemistry, University of Groningen, 4Mathematical Institute, Leiden University, 5Korteweg-de Vries Institute for Mathematics, University of Amsterdam
Objectives: Clozapine is the only antipsychotic officially approved for treatment-resistant schizophrenia (TRS) [1], yet its clinical use is complicated by a narrow therapeutic window, high interindividual variability, and severe side effects [2, 3]. Effective dose optimization is challenging, as direct measurement of unbound clozapine concentrations in the brain extracellular fluid (brainECF) is not feasible in humans. Traditional plasma therapeutic drug monitoring (TDM) [4] does not reliably reflect target-site exposure, necessitating alternative approaches to predict brain pharmacokinetics (PK) and pharmacodynamics (PD). This study aimed to develop a physiologically-based pharmacokinetic (PBPK) and PD model to characterize clozapine disposition in the brainECF and its dopamine D2 receptor occupancy (RO). The model integrates active blood-brain barrier (BBB) influx transport, brain metabolism, and drug accumulation in neutral lipids, enabling translation from rats to humans for improved dose optimization strategies. Methods: A previously established rat PBPK model (LeiCNS-PK3.5) [5], which accounts for brain metabolism, was extended to incorporate active BBB influx transport and neutral lipid binding (LeiCNS-PK3.6). Transporter and enzyme kinetic parameters were obtained from in-vitro or in-vivo studies [6, 7] and extrapolated using in vitro-in vivo extrapolation (IVIVE) techniques. The model was calibrated using rat microdialysis-derived brainECF clozapine PK data [8] and validated against experimental observations via visual predictive checks (VPC) and symmetric mean absolute prediction error (SMAPE). For human translation, species-specific enzyme and transporter kinetic parameters were incorporated to estimate brain metabolism and BBB influx transport [6, 9]. Plasma PK parameters were derived from population PK models and literature-reported clinical data [10-13]. The PK-PD relationship between brainECF clozapine concentrations and D2 RO was modelled using a Langmuir equation derived receptor binding equation, and predictions were validated against PET imaging data from schizophrenia patients treated with clozapine [14]. Results: The updated LeiCNS-PK3.6 model significantly enhanced the prediction of clozapine brainECF pharmacokinetics in rats. Notably, the SMAPE was reduced from 93 % to 41% compared to LeiCNS-PK3.5. Although variability data were unavailable, the VPC based solely on the median predicted profile demonstrated a markedly improved fit to the observed rat PK data. In particular, the model captured the delay in brainECF exposure, as reflected by an increase in Tmax from 42 minutes in LeiCNS-PK3.5 to 127 minutes in LeiCNS-PK3.6. In human simulations, the model predicted steady-state clozapine average concentrations (Cav) in brain homogenate of 10,956 ng/ml, aligning with therapeutic exposure ranges inferred from plasma data in previous reviews. Furthermore, the integrated PK-PD model accurately predicted dopamine D2 RO in a subset of patients. However, in those with lower than predicted D2 RO, incorporating interindividual variability factors—such as CYP1A2 polymorphisms, sex, age, and smoking, which have been documented to influence clozapine response [15] — aligns the model predictions more closely with the observed data. Conclusions: This integrated PBPK-PD modelling approach successfully characterized clozapine brain disposition across species and provides a potential tool for exploring interindividual variability (e.g., CYP1A2 polymorphisms, transporter activity, drug-drug interactions) in support of personalized therapy. Future work will focus on refining model predictions by incorporating pharmacogenetic factors, such as CYP1A2 polymorphisms, and environmental influences (e.g., smoking and alcohol consumption) to improve dose individualization. Furthermore, the model will be expanded to assess drug-drug interactions and other factors, enhancing its utility in supporting model-informed drug development (MIDD) strategies aimed at optimizing treatment in TRS patients within the framework of personalized medicine.
[1] Peitl, V., et al., Clozapine in Treatment-Resistant Schizophrenia and Its Augmentation with Electroconvulsive Therapy in Ultra-Treatment-Resistant Schizophrenia. Biomedicines, 2023. 11(4). [2] Haidary HA, P.R., Clozapine. [Updated 2023 Nov 10]. . 2024, In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing. [3] Wills, K.H., et al., Combining Therapeutic Drug Monitoring and Pharmacokinetic Modelling Deconvolutes Physiological and Environmental Sources of Variability in Clozapine Exposure. Pharmaceutics, 2021. 14(1). [4] Olesen, O.V., Therapeutic drug monitoring of clozapine treatment. Therapeutic threshold value for serum clozapine concentrations. Clin Pharmacokinet, 1998. 34(6): p. 497-502. [5] Zhang, M., et al., Exploring K(p,uu,BBB) values smaller than unity in remoxipride: A physiologically-based CNS model approach highlighting brain metabolism in drugs with passive blood-brain barrier transport. Eur J Pharm Sci, 2024. 203: p. 106883. [6] Dickens, D., et al., Cellular Uptake of the Atypical Antipsychotic Clozapine Is a Carrier-Mediated Process. Mol Pharm, 2018. 15(8): p. 3557-3572. [7] Fang, J., Metabolism of clozapine by rat brain: the role of flavin-containing monooxygenase (FMO) and cytochrome P450 enzymes. Eur J Drug Metab Pharmacokinet, 2000. 25(2): p. 109-14. [8] Cremers, T.I., et al., Microdialysis evaluation of clozapine and N-desmethylclozapine pharmacokinetics in rat brain. Drug Metab Dispos, 2012. 40(10): p. 1909-16. [9] Fang, J., et al., Elucidation of individual cytochrome P450 enzymes involved in the metabolism of clozapine. Naunyn Schmiedebergs Arch Pharmacol, 1998. 358(5): p. 592-9. [10] Reeves, S., et al., A population pharmacokinetic model to guide clozapine dose selection, based on age, sex, ethnicity, body weight and smoking status. Br J Clin Pharmacol, 2024. 90(1): p. 135-145. [11] Woo, Y.S., et al., Bioequivalence of Generic and Brand Name Clozapine in Korean Schizophrenic Patients: A Randomized, Two-Period, Crossover Study. Psychiatry Investig, 2015. 12(3): p. 356-60. [12] Margareth, d.O.T.o.t., et al., Steady-state bioequivalence study of clozapine psychotic patients in Brazil. African Journal of Pharmacy and Pharmacology, 2016. 10(14): p. 289-297. [13] Golden, G. and G. Honigfeld, Bioequivalence of clozapine orally disintegrating 100-mg tablets compared with clozapine solid oral 100-mg tablets after multiple doses in patients with schizophrenia. Clin Drug Investig, 2008. 28(4): p. 231-9. [14] Takano, A., et al., Time course of dopamine D2 receptor occupancy by clozapine with medium and high plasma concentrations. Prog Neuropsychopharmacol Biol Psychiatry, 2006. 30(1): p. 75-81. [15] Albitar, O., et al., Population Pharmacokinetics of Clozapine: A Systematic Review. Biomed Res Int, 2020. 2020: p. 9872936.
Reference: PAGE 33 (2025) Abstr 11765 [www.page-meeting.org/?abstract=11765]
Poster: Drug/Disease Modelling - Other Topics