Predictive Dissolution Models for Real-Time Release Testing: Development and Implementation - Workshop Summary Report

Abstract

To date, few examples of dissolution models for real-time release testing (RTRT) have been approved for commercial drug products or published in literature. Thus, a structured approach has not been established by which a novice to the field could design, develop, validate, and implement an RTRT dissolution model. Moreover, with scant examples available, there has not been a body of work by which to learn of general regulatory expectations for such models.To address these gaps and to encourage conversation between regulatory and industrial experts on these topics, a virtual (web-based) workshop entitled "Predictive Dissolution Models for Real-Time Release Testing: Development and Implementation" was held November 11-12, 2021. This article summarizes key points from the podium presentations, panel discussions, and breakout sessions focusing on (1) the current best practices to establish predictive model specifications; (2) designing models to predict the "safe space" of a release test and creating models utilizing process analytical technology (PAT); and (3) exploring the strategy of compliant regulatory submissions, including model validation and post-approval lifecycle management. Industrial case studies were presented showcasing attempted approaches to and successful implementations of RTRT of dissolution for drug product manufacturing.