Performance-Based Quality Specifications: The Relationship Between Process Critical Control Parameters, Critical Quality Attributes, and Clinical Performance

The quality of pharmaceutical products is currently evaluated through a series of tests that do not explicitly communicate the clinical consequences of product variability. A previously published risk simulation platform was used to generate quantitative estimates of inefficacy and toxicity for 288 uniform lots of extended-release theophylline tablets displaying various levels of content uniformity and dissolution variability. These data were used to evaluate the univariate specifications utilized in the United States Pharmacopeia (USP) < 711> and < 905>. Simulation revealed that the specifications are too lenient for content uniformity, both in terms of inefficacy and toxicity, whereas the criteria for dissolution testing are too strict for inefficacy and inaccurate for toxicity. The USP tests also failed to pinpoint the clinical interaction between content uniformity and dissolution variability. Additionally, the simulation platform was used to define the underlying relationship between product quality attributes and clinical performance. Here, content uniformity and Weibull dissolution time constants were used as inputs to the design spaces, which were conditioned on quantitative estimates of inefficacy and toxicity. This methodology enhances the information content of the design space by omitting quality surrogates (e.g., dissolution, moisture content) that are utilized in current design space practices.