(T0930-02-11) Delineating the Role of Transporters in the Absorption and Disposition of Digoxin Using the Physiologically Based Pharmacokinetic (PBPK) Modeling

Purpose: Digoxin (DIG) is one of the cardiac glycosides that inhibits sodium-potassium ATPase, an enzyme that regulates the intracellular concentration of sodium and potassium. DIG falls under Class IV of the Biopharmaceutics Classification System (BCS), i.e., a low solubility and low permeability and is the most commonly recommended probe substrate to investigate the drug-drug interaction (DDI) potential of investigational drugs that are P-gp inhibitors. The absorption and disposition of DIG is governed by multiple transporters: P-gp, MDR3, OATP4C1, and Na⁺/K⁺-ATPase. The purpose of this project was to 1) develop a mechanistic PBPK model delineating the role of transporters in the absorption and disposition of DIG, and 2) verify the contributions of transporters by simulating clinical DDIs studies with rifampicin (RIF), itraconazole (ITZ), and clarithromycin (CLM).

Methods: The GastroPlus® v.9.8.3 was used to build and validate DIG PBPK model. The Advanced Compartmental Absorption and Transit (ACAT^TM) model was used to describe the intestinal dissolution and absorption of DIG after PO administration, and PBPKPlus™ module was used to describe DIG systemic distribution and elimination. Human physiologies were generated using the Population Estimates for Age-Related Physiology (PEAR Physiology^TM) module within the GastroPlus. The systemic distribution of DIG was described using a whole body PBPK model with a permeability-limited model for kidney and muscle and a perfusion-limited model for the remaining tissues. Tissue/plasma partition coefficients were calculated from tissue composition and compound physiochemical properties using the default methods (Lukacova for perfusion-limited and Poulin and Theil -extracellular for permeability-limited tissues).  The permeability-surface area products (PStc) for tissues described by a permeability-limited model were calculated from Specific PStc (Spec.PStc) value (PStc per mL of tissue cell volume) and the individual tissue cell volumes. The Spec.PStc (0.35 mL/s/mL tissue) was fitted against data the clinical pharmacokinetic (PK) data following IV infusion doses^1-3 ranging from 0.5 mg to 1.5 mg to correctly capture the tissue distribution. To match the observed plasma exposure after IV and PO administration, blood to plasma concentration ratio (Rbp) was fitted (0.55) across various doses. DIG is primarily eliminated by renal excretion (~50-70% of the dose excreted in the urine as unchanged drug) via passive glomerular filtration and transporter-mediated active secretion. And ~25-30% of the drug is eliminated through biliary secretion and the remaining ~25 to 28% is excreted through non renal routes. OATP4C1-mediated influx at the basolateral side and P-gp efflux at the apical side of kidney proximal tubules are responsible for the DIG transporter-mediated active secretion into urine. The impact of these transporters on DIG renal secretion was modeled using Michaelis-Menten kinetics with in vitro K_m values^{4, 5} and the V_max optimized to match the observed DIG cumulative urinary excretion (FDA 2016). The basolateral uptake of DIG into muscle tissue through Na⁺/K⁺-ATPase⁶ was added, K_m value was assumed to be closer value as that of OATP4C1 and the V_max was optimized accordingly to capture the initial distribution phase in DIG PK profile after IV administration. DIG biliary secretion was modeled by simple flux across the apical liver membrane by fitting the PStc of 0.5 mL/s such that the biliary secretion represents ~25% of the total drug elimination. The DDI module in GastroPlus was used to predict the effect of RIF, ITZ, and CLM on DIG PK for varying study designs^3,8-15.  Table 1 summarizes the induction and inhibition parameters for perpetrator drugs used in the DDI simulations.

Results: The model accurately captures the DIG PK (Figure 1) after single dose IV and PO administrations of doses ranging from 0.25 mg to 1 mg in fasted or fed state, with more than 70% of the predicted Cmax and AUC values within the bioequivalence (BE) limit of 0.8-1.25, 95% of predictions within 50% of the observed data, and all the predictions were within 2-fold of the observed data ^{1-3, 8-11,16-21} . DIG C_P‑time profiles before and during co-administration with interacting drugs were reasonably predicted over the full range of administered doses. The predicted Cmax and AUC ratios were mostly within the Guest limits²² as shown in Figure 2. It is worthwhile to notice the RIF impact, which depending on the timing of DIG and RIF administration may show a net induction effect (ratio < 1) or net inhibition effect (ratio > 1) and model accurately captured these scenarios (orange points in in Figure 2).

Conclusion: The work here aimed to develop the DIG PBPK model and validate it as a sensitive substrate model for use in predicting the potential DDI interactions between DIG and P-gp perpetrators. The PBPK approach incorporates all the relevant processes in drug ADME, and all the perpetrator mechanisms. The overall results presented in Figures 1 and 2 show that the model accurately captures DIG PK and predicts the effect of perpetrators on IV and  PO doses of DIG. The validated DIG model can be used to evaluate potential DDI interactions with other P-gp perpetrators.

References: 1. Hager. Eng J Med., 300(22):1238 (1979)
2. Ochs. Amer Heart J., 96(4):507 (1978)
3. Greiner. J Clin Invest., 52(7):147 (1999)
4. Mikkaichi. Proc Natl Acad Sci USA., 101(10):3569 (2004)
5. Troutman. Pharm Res., 20(8):1200 (2003)
6. Lee CA. CPT., 96(3):298 (2000)
7. Smith. The J Biol Chem., 275(31):23530 (2000)
8. Wiebe. Clin Pharmacokinet., 59:1627 (2020)
9. Kirby. Drug Metab Dispos.,40(3):610 (2012)
10. Gurley. Mol Nutr Food Res., 52(7):772 (2008)
11. Jalava. Therap Drug Monit., 19(6):609 (1997)
12. Gurley. Drug Metab Dispos., 35(2) (2007)
13. Gurley. Drug Metab Dispos., 34(1): 69 (2006)
14. Rengelhausen. Br J Clin Pharmacol., 56:32 (2003)
15. Tsutsumi. J Clin Pharmacol., 42:1159 (2002)
16. Erjrfalt. Mol Pharm., 12(11):4166 (2015)
17. Westphal. Clin Pharmacol Ther., 68(1):6 (2000)
18. Tayrouz. Clin Pharmacol Ther., 73(5):397 (2003)
19. Johnson. Clin Pharmacol Ther., 23(3):315 (1978)
20. Greenblatt. Clin Pharmacol Ther., 16(3):448 (1974)
21. Jounela. Eur J Clin Pharmacol., 8 :365 (1975)
22. Guest. Drug Metab Dispos., 39(2) :170 (2011)


Table1 Induction and Inhibition parameters of Transporters for Rifampicin, Itraconazole, and Clarithromycin


Figure 1 Baseline Model Predictions (Cmax and AUC)


Figure 2 Goodness of Fit Plots for DDI Predictions with P-gp Inducers and Inhibitors