Modified-release (MR) drug products are designed to provide controlled drug delivery over time, offering therapeutic and compliance advantages. However, ensuring consistent in vivo performance requires a thorough understanding of the relation between in vitro dissolution behavior and in vivo drug absorption. In vitro-in vivo correlation (IVIVC) serves as a critical tool in this context, enabling formulation optimization, supporting regulatory decision-making, and streamlining product development. This study aimed to use tofacitinib as a model compound to evaluate three deconvolution methodologies within GPX™ — numerical, compartmental, and mechanistic. Prototype formulations with varying release rates were assessed in a randomized crossover study in healthy volunteers. In vivo fraction absorbed profiles were derived and convoluted to simulate plasma concentration–time profiles, which were then compared to observed clinical data. Prediction errors for key pharmacokinetic parameters (i.e., plasma C_max and AUC) were determined, and 90% confidence intervals for both parameters were calculated to assess bioequivalence between the simulated (convoluted) and observed plasma profiles. The results demonstrate the utility of deconvolution-based IVIVC models for MR product development in a physiologically-based pharmacokinetic (PBPK) framework and offer a strategy for assessing dissolution variability in support of regulatory flexibility and robust formulation lifecycle management.

By Bart Hens