Advancing biopharmaceutics modelling of extended-release formulations with sequential TIM-1/TIM-2 biorelevant upper...

Predicting the in vivo performance of extended-release (ER) products is difficult because drug release is excipient-controlled and may continue into the colon, a region poorly represented by conventional dissolution methods. Here, during ER formulation development, a biorelevant in vitro-in silico workflow was established to support formulation development of rocavorexant, a weakly water-soluble, high-permeability BCS II compound formulated as a HPMC-matrix ER tablet. Formulations were screened for food effect risk using the dynamic in vitro TIM-1 gastrointestinal model, before dissolution profiles measured in sequential TIM-1/TIM-2 upper-GI and colonic transit were used as dissolution input for PBBM. TIM-derived dissolution profiles were constructed from luminal samples, distinguishing apparently dissolved from dispersed drug, and from filtrate/dialysate fractions from the jejunal, ileal and colonic compartments representing dissolved drug removed from the systems. PBBM using conventional USP-I dissolution, either directly or after level A IVIVC transformation, failed to reproduce clinical pharmacokinetics, underpredicting early exposure and missing a late 20-24 h plasma plateau. Sequential TIM-1/TIM-2 input predicted clinical exposure well (predicted/observed C_max 1.03; AUC 0.91-0.94), recovered the plateau and linked regional release, solubilisation, absorption and systemic pharmacokinetics.