The influence of individual gastrointestinal physiology parameters on prediction of regional drug absorption.

Physiologically based biopharmaceutics modeling (PBBM) is crucial for drug absorption prediction due to its ability to integrate drug-specific, physiological, and biopharmaceutic properties. Recent reports highlight the need to improve prediction accuracy for regional, colonic absorption, and extended-release (ER) formulations, often using average physiological data due to rare concurrent individual measurements (e.g., luminal pH and transit time). This investigation evaluated if in-vivo measured individual pH and transit time data could enhance absorption predictions. Metoprolol (BCS I), with low colon absorption risk, and AZ3 (BCS III), with high risk, were chosen based on IntelliCap study data. Metoprolol had 6-hour and 14-hour release profiles, and AZ3 had a 12-hour profile. Individual PBBMs used GastroPlus and GI-Sim, comparing default and individual gastrointestinal models. Model accuracy was evaluated using Cmax, AUC0-t, and relative bioavailability (Frel) with absolute average fold error (AAFE). For metoprolol's 6-hour profile, individual models' AAFE ranged 1.18-1.23 (GastroPlus) and 1.23-1.45 (GI-Sim), similar to defaults (1.15-1.27, 1.22-1.32). For the 14-hour profile, AAFE was 1.56-1.68 (GastroPlus), 1.35-1.37 (GI-Sim) against defaults (1.61-1.66, 1.37-1.40). AZ3's individual AAFE was 1.23-1.56 and 1.35-1.72 versus defaults (1.29-1.58, 1.29-1.56). The results indicate that using individual data did not significantly improve prediction accuracy compared to default models, highlighting the effectiveness of current small intestine models but also emphasizing the need for better models in predicting colonic absorption.