521b Data Exploration Analysis In Process Development: Application to API Characterization

Jose E. Tabora, Jessica Defreese, Prashant Deshpande, and Jean Tom. Bristol-Myers Squibb, One Squibb Drive, New Brunswick, NJ 08901

The characterization of the physical properties of the API is an enduring challenge in pharmaceutical process development. Multiple attributes such as particle size distribution, specific surface area, bulk and tapped density are routinely measured in efforts to define the acceptability criteria for subsequent processing. In addition, due to the complexity of the processes involved, a phenomenological relationship between the process parameters from crystallization and drying and the resulting physical attributes of the isolated material is generally intractable. In this work we present the application of data exploration analysis and chemometrics to facilitate the characterization of the API powder properties in relationship to process parameters. Material isolated from multiple lots of Pilot plant scale processing was characterized by up to 12 measured properties. Multivariate analysis (PCA and multiway-PCA) was applied to the descriptors and to the batch processing profiles to aide visualization of a large data set. The trends identified from data mining were evaluated in the context of known physical processes.