For the last two decades, the application of physiologically based pharmacokinetic (PBPK) models has grown exponentially in the field of oral absorption and in a regulatory context. Although these models are widely used, their predictive power should be validated and optimized in order to rely on these models and to know exactly what is going on "under the hood". In this study, an automated sensitivity analysis (ASA) was performed for 11 gastrointestinal (GI) variables that are integrated into the PBPK software program Simcyp®. The model of interest was a previously validated workspace that was able to predict the intraluminal and systemic behavior of two different suspensions of posaconazole in the Simcyp® Simulator. The sensitivity of the following GI parameters was evaluated in this model: gastric and duodenal pH, gastric and duodenal bicarbonate concentrations (reflecting buffer capacity), duodenal bile salts concentration, gastric emptying, the interdigestive migrating motor complex (IMMC), small intestinal transit time (SITT), gastric and jejunal volumes, and permeability. The most sensitive parameters were gastric/duodenal pH and gastric emptying, for both suspensions. The outcome of the sensitivity analyses highlights the important GI variables that must be integrated into an in vivo predictive dissolution test to help and create a rational and scientific framework/design for product development of novel and generic drug products.
Keywords: bioequivalence; dissolution; modeling; oral absorption; pharmacometrics; physiologically based pharmacokinetics; precipitation; simulations; supersaturation.