A Pharmacometrics Approach to Understanding Drug Disposition and Drug Interactions in Healthy Individuals and Patients Undergoing Roux-en-Y Gastric Bypass

Abstract

In this dissertation, pharmacometric approaches were used to better understand drug disposition and predict drug interactions in morbidly obese individuals undergoing Roux-en-Y gastric bypass surgery (RYGBS) and to determine the kinetic characteristics of theoretical endogenous biomarkers that would permit the detection of potential drug interactions. Chapter 2 examined how RYGBS affected drug absorption and intestinal and hepatic metabolism in vivo by ascertaining the pharmacokinetics of acetaminophen in patients at three timepoints (e.g., pre-RYGBS and 3- and 12-months post-RYGBS). Following RYGBS, peak APAP concentrations at the 3-month and 12-month visits increased by 2-fold compared to baseline (p < 0.01) and the median time to peak concentration decreased from 35 min to 10 min. The apparent oral clearance of APAP decreased 34% after RYGBS (p < 0.01). The decrease in metabolite-to-parent AUC ratios of all four metabolites at 3-months and 12-months is possibly indicative of a decline in the activities of CYP2E1, UGT1A1, UGT1A9, and SULT1A1 following RYGBS. As drug-drug interaction (DDI) studies have not been conducted in patients undergoing RYGBS, we used an in silico approach to predict DDIs in this patient population (Chapter 3). Using physiologically-based pharmacokinetic (PBPK) modeling, we simulated the impact of RYGBS on the absorption and metabolism of midazolam, acetaminophen, digoxin, and their major metabolites. Secondly, we built PBPK models for verapamil, a highly soluble inhibitor, and posaconazole, a poorly soluble inhibitor, to evaluate CYP3A- and P-gp-mediated DDIs pre- and post-RYGBS. For verapamil inhibition, RYGBS did not affect the fold-change of the inhibited AUC ratio or inhibited peak concentration ratio for either midazolam or digoxin. For posaconazole inhibition, the inhibited midazolam AUC increased by 2.0-fold pre-RYGBS, but only increased by 1.6-fold post-RYGBS due to decreased absorption. For DDI assessment, the use of endogenous biomarkers is a relatively non-invasive approach to provide early detection of potential DDIs during first-in-human clinical trials. Chapter 4 describes a simulation study to investigate how the sensitivity of a theoretical biomarker is affected by metabolite half-life, fraction of the endogenous parent that is metabolized to the metabolite of interest (fm,metabolite), and fraction of the endogenous parent that is metabolized by CYP3A4 (fm,CYP3A4). Several sensitivity indices, including the metabolite concentration (Cm), the metabolite-to-parent concentration ratio (Cm/Cp), the metabolite AUC (AUCm), and the metabolite-to-parent AUC ratio (AUCm/AUCp), were explored in different DDI scenarios. A 20% change in the sensitivity indices was set as the detection threshold for single dose inhibition, mechanism-based inhibition, and induction. Our simulation results demonstrated that the sensitivity of the hypothetical endogenous biomarker was reduced by increasing metabolite half-life. As fm,metabolite was decreased, the change in Cm and AUCm increased, but the change in Cm/Cp and AUCm/AUCp decreased. As fm,CYP3A4 was decreased, the change in all indices decreased. The predicted magnitude of change based on our simulations was comparable to the clinically observed magnitude of change for putative CYP3A4 endogenous biomarkers. In summary, pharmacometrics was used in this dissertation to understand the impacts of RYGBS on drug disposition, predict CYP3A4 and P-gp DDIs in RYGBS patients, and investigate the kinetic characteristics of endogenous CYP3A4 biomarkers.