LB-003 - A NOVEL MECHANISTIC PHYSIOLOGICALLY BASED PHARMACOKINETIC-PHARMACODYNAMIC (PBPK-PD) MODEL TO ASSESS DRUG-DRUG INTERACTION (DDI) RISKS FOR ACID REDUCING AGENTS (ARA).

U. Ezuruike, Y. Yeap, O. Hatley, S. Neuhoff; Certara UK (Simcyp Division), Sheffield, United Kingdom.

Background: Co-administration of ARAs with drugs exhibiting pH-dependent solubility can lead to DDIs mediated by gastric pH changes. PBPK modelling, recognized by the FDA as an alternative approach to clinical trials for evaluating DDIs, may be particularly useful in delineating the gastric pH-mediated and drug metabolizing enzyme and transporter (DMET)-mediated components of a DDI. Using a new ‘Gastric pH Interaction’ functionality (Simcyp Simulator, Version 23) that enables dynamic modelling of gastric pH changes, which, in turn, are fed back into the PK and consequently DDI estimations, we assess and categorize the DDI risks of the H2-receptor antagonist ranitidine.

Methods: A full PBPK model including permeability-limited models for the gut (ADAM), the liver (PerL), and the kidneys (MechKiM) was developed for ranitidine, which also accounted for its weak CYP3A4 inhibition (Ki of 14.09 µM). Employing the new ‘Gastric pH Interaction’ functionality, PD effect parameters were built into the model, describing its impact on baseline gastric pH levels using a Sigmoid Emax model (Emax=1, EC50=0.45 µM, N=1), with an effect compartment (ke0=2 h-1). Independent verification of the ranitidine PBPK-PD model was performed against single and multiple doses of ranitidine following IV and oral dosing. Then, the model was used to simulate clinical DDI studies of ranitidine as a perpetrator (n=15) with midazolam, triazolam, nifedipine, metoprolol, rivaroxaban, and ketoconazole to categorize them as either gastric pH-mediated and/or DMET-mediated DDIs. Mechanistic vs. empirical (adjusting the pH of the stomach to 6) approaches of assessing pH-dependent DDIs were also compared.

Results: After verifying the PK of ranitidine with the developed PBPK model, the simulated clinical DDIs were well recovered with the model with AFEs of 1.09 and 1.38 for Cmax and AUC ratios, respectively. The DDIs were categorized as either predominantly gastric pH-mediated (e.g., ketoconazole), or predominantly DMET-mediated (e.g., midazolam). The new mechanistic model also improved the quantitative assessment of pH-dependent DDIs.

Conclusion: Using ranitidine as a model drug ARA, the utility of the mechanistic PBPK-PD model with feedback to the PK in assessing and categorizing ARA DDI risks have been demonstrated.