A Physiologically Based Pharmacokinetic Model of Atorvastatin Acid Predicting CYP3A4 and OATP1B Drug–Drug Interactions
Christina Kovar (1), Tobias Kanacher (1), Fenglei Huang (2), Jing Wu (2), Reinhard Sailer (3), David Busse (4), Jose David Gómez-Mantilla (4), and Ibrahim Ince (4)
(1) Pharmetheus AB, Uppsala, Sweden (2) Boehringer Ingelheim Pharma Inc, USA (3) Boehringer Ingelheim Pharma, Biberach, Germany (4) Boehringer Ingelheim Pharma, Ingelheim, Germany
Introduction: The 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase inhibitor atorvastatin is a widely prescribed statin for managing dyslipidemia, a well-established risk factor for cardiovascular disease [1,2]. Its recommended dosing regimen ranges 10–80 mg once daily for adults [2]. Upon oral administration, atorvastatin acid is rapidly absorbed from the gastrointestinal tract achieving maximum plasma concentrations within one to two hours [2]. However, its overall bioavailability remains notably low (~14%) due to extensive first pass metabolism mediated by cytochrome P450 (CYP) 3A4 [2]. Moreover, atorvastatin acid is also a substrate of various transporters, such as the influx transporters, organic anion transporting polypeptides (OATP) 1B1 and OATP1B3, [3] as well as the efflux transporters P-glycoprotein (P-gp), breast cancer resistant protein (BCRP), and multidrug resistance-associated protein 3 (MRP3) [4], making it susceptible to drug–drug interactions (DDIs) mediated by both enzymes and transporters. In this context, physiologically based pharmacokinetic (PBPK) modelling emerges as a powerful tool to quantitively describe and predict different DDI scenarios. The overall aim of this work was to provide a qualified atorvastatin acid PBPK model that might be leveraged for future DDI investigations.
Objectives:
- To mechanistically describe the pharmacokinetics of atorvastatin acid using whole-body PBPK modelling
- To predict enzyme- and transporter-mediated DDIs involving clarithromycin, erythromycin, gemfibrozil, itraconazole, and rifampicin as perpetrator drugs
Methods: Development of a PBPK model of atorvastatin acid using the Open Systems Pharmacology Software Suite (PK-Sim® and MoBi®, version 11.2) [5] was initiated with a comprehensive literature search to gather information about physicochemical properties and absorption, distribution, metabolism, and excretion processes as well as clinical study data. Plasma profiles were divided into a training and a test dataset for model building (7 profiles) and qualification (27 profiles), respectively. To inform CYP3A4- and OATP1B1/1B3-mediated pathways, the DDI studies with itraconazole and rifampicin were allocated to the training dataset. Our established model was linked with published PBPK models of clarithromycin [6], erythromycin [7], gemfibrozil [8], itraconazole [6], and rifampicin [6] to investigate the impact of different perpetrator drugs on the exposure of atorvastatin acid. Model qualification was graphically performed by comparing observed to predicted plasma profiles as well as their respective area under the plasma concentration–time curve from the first to the last time point of measurement (AUClast) and maximum plasma concentration (Cmax) values. For DDI prediction performance, predicted and observed AUClast and Cmax ratios were calculated, and the prediction success limits proposed by Guest et al. were applied [9]. Here, a stricter criterium for DDI predictions than the traditional two-fold deviation is utilized [9].
Results: The developed PBPK model for atorvastatin acid includes a total of 34 plasma profiles after single and multiple oral doses across the range of 1 to 80 mg in healthy subjects. The final PBPK model comprises metabolism via CYP3A4 as well as transport processes via BCRP, MRP3, OATP1B1/1B3, and P-gp, implemented via Michaelis-Menten kinetic processes. While Michaelis–Menten constant values could be utilized from the literature, the respective catalytic rate constants were estimated bridging the gap between in vitro and in vivo settings. The established PBPK model showed a good predictive and descriptive performance, with 85% of AUClast and Cmax values, respectively, falling within the two-fold deviation of observed values. Moreover, a good DDI performance was demonstrated by 11/14 AUClast and Cmax ratios, respectively, lying within the limits proposed by Guest et al. [9].
Conclusions: A whole-body PBPK model for atorvastatin acid has been successfully developed, including relevant active disposition processes, and applied to describe and predict the impact of different DDI scenarios on the exposure of atorvastatin acid. Moreover, our established PBPK model enriches the openly accessible PBPK model library by a new CYP3A4 and OATP1B1/1B3 substrate.
References:
[1] Ascunce RR et al. Curr. Atheroscler. Rep. (2012) 14, 167–174.
[2] https://labeling.pfizer.com/ShowLabeling.aspx?id=587
[3] Vildhede A et al. Drug Metab. Dispos. (2014) 42(7), 1210–8.
[4] Deng F et al. Drug Metab. Dispos. (2021), 49(9), 750–759.
[5] Lippert J et al. CPT pharmacometrics Syst. Pharmacol. (2019) 8, 878–882.
[6] Hanke N et al. CPT pharmacometrics Syst. Pharmacol. (2018) 7, 647–659.
[7] https://github.com/Open-Systems-Pharmacology/OSP-PBPK-Model-Library/tree/master/Erythromycin
[8] Türk D et al. Clin Pharmacokinet. (2019) 58, 1595–1607.
[9] Guest E J et al. Drug Metab. Dispos. (2011) 39, 170–173.
