(645e) Multi-Factor Modeling to Predict Significant Factors in Tablet Disintegration and Dissolution Using the FBRM

We have previously demonstrated the application of Focused Beam Reflectance Measurement (FBRM) for analysis during immediate-release tablet disintegration to evaluate changes in particle size and count. This approach could distinguish changes in tablet particle size during disintegration due to varying properties, such as tablet hardness. Here we present a proof-of-concept study demonstrating the generation of a predictive statistical model of a drug product design of experiments (DoE) using parameterization of FBRM measurements taken during tablet disintegration. The FBRM was used to monitor the disintegration of tablets generated in an 18-run DoE evaluating five variables at three levels. The DoE evaluated API properties as well as dry granulation and tableting conditions. Statistical models were generated using the DoE input variables fit to predict FBRM measurement parameters. Robust, multi-factor models identified that several FBRM measures such as total particle count, time of maximum particle count, mean chord (particle) length, or bounded particle size ranges were dependent on several DoE inputs. Furthermore, these factors identified via FBRM disintegration corresponded with the significant factors of dissolution prediction models. In contrast, a USP disintegration method failed to provide a predictive model with multiple significant factors. The FBRM approach to monitoring disintegration allows for a more comprehensive understanding of tablet performance and provides greater insight into the impact of upstream material properties or processing parameters. Material properties and process parameters ultimately impacting dissolution could be predicted earlier and in a material-sparing way. It further offers an orthogonal measurement to gain a better understanding of tablet dissolution explained in a framework of underlying particle behavior.