Hyojin Cho 1, Woojin Jung 2, Seonghwa Song 1, Jae Hyun Kim 3, Seol Ju Moon 4,5, So-Jung Park 6, Hwi-yeol Yun 1,7,8, Soyoung Lee 1, Jungwoo CHAE 1,7,8
1 College of Pharmacy, Chungnam National University,South Korea (, Republic of Korea), 2 Graduate School of Clinical Pharmacy, CHA University (, Republic of Korea), 3 School of Pharmacy and Institute of New Drug Development, Jeonbuk National Universit (, Republic of Korea), 4 Department of Pharmacology, Jeonbuk National University College of Medicine (, Republic of Korea), 5 Centre for Clinical Pharmacology and Biomedical Research Institute, Jeonbuk National University Hospital (, Republic of Korea), 6 Department of Korean Internal Medicine, School of Korean Medicine & Korean Medicine Hospital, Pusan National University (, Republic of Korea), 7 Bio-AI Convergence Research Center, Chungnam National University (, Republic of Korea), 8 Senior health Convergence Research Center, Chungnam National University (, Republic of Korea)
Objectives: Gastroesophageal reflux disease (GERD) is highly prevalent and frequently managed with both conventional pharmacotherapy and herbal formulations, highlighting the need to evaluate potential drug–herbal interactions. However, clinical evidence on their co-administration remains limited and largely qualitative1), underscoring the need for quantitative assessment of interaction magnitude and clinical relevance2).
This study developed a PBPK-based approach to evaluate the effects of herbal formulations on the pharmacokinetics of conventional drugs and applied it to commonly used GERD therapies. Through this approach, we aimed to quantitatively characterize the likelihood and magnitude of drug–herbal interactions. The modeling approach was also implemented in an interactive tool to support estimation of interaction risk during concurrent use of herbal and conventional treatments.
Methods: Banhasasim-tang, a representative herbal formulation frequently used for gastrointestinal disorders, was selected as the study formulation. Marker compounds were identified for each herb based on pharmacopeial standards, considering their reported abundance and pharmacokinetic relevance. To develop PBPK models for each marker compound, physicochemical properties were collected from literature sources and in silico prediction tools. In addition, the effects of each marker compound on hepatic enzyme systems were identified through searches of the DIDB and PubMed databases.
Conventional drugs commonly used for GERD treatment, esomeprazole and rabeprazole, were selected. PBPK models for both drugs were developed and validated using published clinical pharmacokinetic datasets, including time–concentration profiles following intravenous administration as well as single and multiple oral doses across different formulations3-9). Model performance was assessed through visual evaluation using overlay and goodness-of-fit plots, as well as quantitative assessment based on the geometric mean fold error (GMFE) for AUC and Cmax.
To assess the influence of formulation variability on drug–herbal interactions, variability in marker compound content due to differences in cultivation conditions and herbal sources was incorporated into the drug interaction model. A 1–3-fold variability range in formulation dose was applied to evaluate its impact on DDI risk.
Given the multi-component nature of herbal formulations, uncertainty in CYP modulation by individual marker compounds was systematically parameterized and propagated in the PBPK simulations. Enzyme inhibition constants (Ki) were sampled from log-normal distributions within a ten-fold range around literature-derived values, and repeated simulations were performed to evaluate the impact of enzyme-effect uncertainty on predicted drug interaction outcomes.
PBPK modeling and simulations were performed using the Open Systems Pharmacology (OSP) Suite (v12.1), with scenario execution and uncertainty analyses conducted in RStudio (v4.4).
Results: The PBPK models demonstrated acceptable predictive performance, with GMFE values for AUC and Cmax of 1.36 and 0.97 for rabeprazole, and 1.57 and 1.24 for esomeprazole. Visual predictive checks showed that more than 80% of observed AUC and Cmax values fell within the 2-fold acceptance range.
Drug–herbal interaction assessment using PBPK models incorporating marker compounds of Banhasasim-tang showed that, compared with drug alone, predicted AUC ratios for esomeprazole and rabeprazole were 1.15 and 1.04, respectively, both below the interaction threshold (≥1.25). When variability in herbal composition and uncertainty in CYP inhibition were incorporated, esomeprazole exposure increased in a dose-dependent manner, resulting in a higher probability that its AUC would exceed the interaction threshold (≥1.25), whereas rabeprazole exposure remained largely unchanged with negligible risk across all scenarios.
Based on these results, an interactive prediction platform was developed to support risk assessment for drug–herbal co-administration. The application allows users to select the PPI (esomeprazole or rabeprazole) and input the administration day to generate the probability of pharmacokinetic interaction. In addition, detailed information on dosing regimens for the conventional drug and herbal formulation, simulation methodology, Ki distributions of marker compounds, and predicted interaction outcomes can be explored within the platform.
The Shiny app is available at: https://hyojin.shinyapps.io/Herb_drug_interaction/.
Conclusions: In this study, we established a PBPK modeling framework for multi-constituent herbal formulations, demonstrating its utility for mechanistic evaluation of drug–herbal interactions. Model simulations suggested minimal interaction under typical exposure conditions, while probabilistic analyses highlighted the importance of accounting for variability inherent to herbal products when assessing interaction risk. Application of the framework to esomeprazole, rabeprazole, and Banhasasim-tang enabled quantitative characterization of interaction potential across dosing scenarios. Future work will focus on clinical validation and expansion to additional drug–herbal combinations.
References:
Reference:
1) Han, G., Lim, J. T., Kim, J. S., & Lee, J. H. (2016). Analysis of domestic guidelines for developing clinical trial guidelines of herbal medicinal products for gastroesophageal reflux disease and comparison with previous herbal clinical trials. Journal of Internal Korean Medicine, 37(1).
2) Jung, Y. J., Kim, Y. J., & Hong, S. H. (2020). Research trends over the past 10 years on the effects of Banhasasim-tang for gastroesophageal reflux disease in Chinese and Korean databases. Journal of Internal Korean Medicine, 41(3).
3) Hassan-Alin, M.; Andersson, T.; Niazi, M.; Liljeblad, M.; Persson, B.A.; Röhss, K. Studies on drug interactions between esomeprazole, amoxicillin and clarithromycin in healthy subjects. Int. J. Clin. Pharmacol. Ther. 2006, 44(3), 119–127.doi:10.5414/cpp44119
4) Wilder-Smith, C.H.; Röhss, K.; Nilsson-Pieschl, C.; Junghard, O.; Nyman, L. Esomeprazole 40 mg provides improved intragastric acid control as compared with lansoprazole 30 mg and rabeprazole 20 mg inhealthy volunteers. Digestion. 2003, 68(4), 184–188.doi:10.1159/000075697
5) Xu, Y.; Tian, X.; Wang, W.; Tian, W.; Zhang, T.; Sun, J.; Zhou, Q.; Shao, C.Pharmacokinetics of esomeprazole in critically ill patients. Front. Med. 2021, 8, 621406.doi:10.3389/fmed.2021.621406
6) U.S. Food and Drug Administration, Center for Drug Evaluation and Research (CDER). Clinical Pharmacology and Biopharmaceutics Review for NDA 204655: Nexium 24HR (Esomeprazole Magnesium) Delayed-Release Capsules. 2014. (Reference ID: 3457453)
7) Wang Y, Zhang D, Li Y, et al. Pharmacokinetics and absolute bioavailability of rabeprazole in healthy Chinese volunteers. Xenobiotica. 2011;41(3):206–212.
8) Swan SK. The pharmacokinetics of rabeprazole in health and disease. Alimentary Pharmacology & Therapeutics. 1999;13(Suppl 3):11–17.
9) Miura M, Tateishi T, Suzuki T. Stereoselective metabolism of rabeprazole-thioether to rabeprazole by human liver microsomes. European Journal of Clinical Pharmacology. 2006;62(2):113–117. doi:10.1007/s00228-005-0077-8
Reference: PAGE 34 (2026) Abstr 11998 [www.page-meeting.org/?abstract=11998]
Poster: Drug/Disease Modelling - Absorption & PBPK