(M1030-10-69) Elucidation of Interaction Mechanisms between Irinotecan and Oroxylin A in Rats through Physiologically Based Pharmacokinetic Modeling

Purpose: Irinotecan, as a topoisomerase I inhibitor, is mainly hydrolysed to its active metabolite SN-38 by carboxylesterase 2 (CES2) and subsequently deactivated by glucuronidation (UTG1A1) in rats[1]. Elimination of irinotecan and its metabolites are also mediated by transporters such as P-glycoprotein (P-gp) and breast cancer resistant protein (BCRP) on the bile canalicular membrane and intestinal epithelium membrane[2]. Our previous study[3] found that oroxylin A (OXA) from Radix Scutellariae could significantly inhibit intestinal CES2 activities of rats. The current study was designed to further evaluate the roles of metabolic enzymes and transporters during the drug-drug interaction process between irinotecan and OXA in rats through physiologically based pharmacokinetic (PBPK) modeling.

Methods: Plasma and tissue (intestine, liver, kidney, lung at 0.5, 2 and 5 hr post dosing) pharmacokinetics (PK) profiles of irinotecan in absence (Group I: intravenous (IV) 35 mmol/kg) and presence of OXA (Group I+O: IV 35 mmol/kg irinotecan + Oral 2800 mmol/kg OXA) were obtained from our previous study[3] and current study with five/six rats at each time point. PK data of OXA after oral administration of 40, 120 and 360 mg/kg OXA were digitized from reference[4] using WebPlotDigitizer. The modeling and simulations were carried out with PK-SIM® version 11.1. Physicochemical properties (LogP, pKa, f_up and solubility) of irinotecan, SN-38, SN-38 glucuronide, OXA and oroxylin A 7-O-glucuronide (OAG) were taken from Drugbank or SciFinder. Unbound fraction (f_up) values of each analyte were predicted with fu,p Predictor[5] if they were not available from the literature. Enzyme expression profiles of CES2, UTG1A1, P-gp and BCRP were estimated from the reported relative enzyme activities[2, 3] or by assuming the same levels of expression in major tissues if no supporting data was available. The kinetics of these enzymes and transporters were defined using V_max and K_m in rat liver microsomes or estimated with PK-SIM®’s parameter identification. The best methods for partition coefficients of each compound were also determined through parameter identification. Sensitivity analyses were performed for all the parameters of the PBPK model of each compound to identify critical parameters of the models. Visual predictive check interval was obtained by checking the “Calculate Sensitivity” during parameter identification. The PK-Analysis in PK-SIM® was used to calculate non-compartmental analysis PK parameters for the observed and simulated PK profiles.

Results: Tissue distribution of irinotecan and SN-38 in major organs shown in Figure 1 suggested that i) both irinotecan and SN-38 were highly accumulated in intestine after its IV administration and ii) OXA could significantly reduce such accumulation in intestine, especially for those of SN-38 possibly due to its inhibition on the CES2-mediated hydrolysis of irinotecan. To study the in-vivo inhibition effect of OXA, PBPK of OXA was developed using the reported three doses of oral PK data[4]. OXA’s PBPK model could closely predict the plasma concentration at three dose levels and the C_max and AUC_0-t of the simulated PK profiles were all within 2-fold changes when compared with PK parameters of the observed data (data not shown). Similarly, PBPK of irinotecan was developed and interacted with OXA’s PBPK through Ki of OXA (determined in our lab). The schematic interaction mechanisms for irinotecan and OXA with metabolic enzymes and transporters were summarized in Figure 2. After oral administration, OXA entered the intestinal mucosa and inhibited the CES2-mediated irinotecan hydrolysis, which significantly lowered the systemic exposure of SN-38 (50% reduction in AUC_0-t) as shown in Figure 3. Our established PBPK model could successfully predict not only the PK profiles of irinotecan and SN-38 after IV administration of irinotecan but also their inhibited PK profiles in presence of OXA with the observed data (Group I & Group I+O) within the 5^th-95^th percentile of predicted PK profiles and the fold errors for PK parameters ranging from 0.99 to 1.27 (Figure 3).

Conclusion: Our developed PBPK model successfully integrates the major known metabolic enzymes and transporters to describe the normal and OXA inhibited PK profiles of irinotecan and SN-38 in rats. For the first time, the model indicated that the inhibition of CES2-mediated irinotecan hydrolysis in intestinal mucosa could be the underlying interaction mechanism between irinotecan and OXA.

References: [1] N. Kaneda, T. Yokokura, Nonlinear pharmacokinetics of CPT-11 in rats, Cancer Res, 50 (1990) 1721-1725.
[2] M.M. Parvez, A. Basit, P.B. Jariwala, Z. Gaborik, E. Kis, S. Heyward, M.R. Redinbo, B. Prasad, Quantitative Investigation of Irinotecan Metabolism, Transport, and Gut Microbiome Activation, Drug Metab Dispos, 49 (2021) 683-693.
[3] J. Zhang, M. Xiao, X. Ji, Y.S. Lai, Q. Song, Y. Zhang, C.M. Ip, W.L. Ng, Z. Zuo, Inhibition of Radix Scutellariae flavones on carboxylesterase mediated activations of prodrugs, Life Sci, 305 (2022) 120743.
[4] G. Ren, H. Chen, M. Zhang, N. Yang, H. Yang, C. Xu, J. Li, C. Ning, Z. Song, S. Zhou, Y. Bian, Y. Lu, N. Li, Y. Zhang, X. Chen, D. Zhao, Pharmacokinetics, tissue distribution and excretion study of Oroxylin A, Oroxylin A 7-O-glucuronide and Oroxylin A sodium sulfonate in rats after administration of Oroxylin A, Fitoterapia, 142 (2020) 104480.
[5] R. Watanabe, T. Esaki, H. Kawashima, Y. Natsume-Kitatani, C. Nagao, R. Ohashi, K. Mizuguchi, Predicting Fraction Unbound in Human Plasma from Chemical Structure: Improved Accuracy in the Low Value Ranges, Mol Pharm, 15 (2018) 5302-5311.

Acknowledgements:This work was supported by the General Research Fund of Hong Kong Research Grants Council (project number: 14108119).

Figure 1. Tissue distribution of irinotecan and SN-38 at 0.5, 2 and 5 hours after intravenous administration of irinotecan (A, C) or co-administration with OXA (B, D) in rats.

Figure 2. Metabolism and distribution of irinotecan and oroxylin A (OXA) in major tissues of rats.

Figure 3. The 95% visual predictive check intervals (upper) and observed versus predicted PK parameters (bottom) for irinotecan (A) and SN-38 (B) after intravenous administration of irinotecan alone (Group I) or co-administration with oroxylin A (Group I+O). Lines are the simulated plasma concentration of irinotecan and SN-38. The upper and bottom edges of the shaded areas are the 5th and 95th percentiles of the simulated concentration.