284757 Assessing the Robustness of a High-Shear Granulation Process to Meaningful Variability in Raw Material Physical Attributes

Thursday, November 1, 2012: 8:55 AM
Allegheny III (Westin )
Sharvari Borkar, Pharmaceutical Commercialization Technology, Merck & Co., Inc., West Point, PA and Stephen L. Conway, Center for Materials Science and Engineering, Merck & Co., Inc., West Point, PA

While much of a drug product development team's focus is applied to understanding the impact of composition and process parameters on the performance of manufacturing processes, utilizing tools such as risk analysis and multivariate experimental designs, difficulties have arisen when the same tools are inappropriately applied to assessment of process robustness to  raw material variability. The drive for more efficient, flexible supply chains, and the use of both novel and more cost-effective ingredients exacerbates the need for clear assessment of process sensitivity to attributes known to impact performance, as well as variations in attributes previously unrecognized.

By analysis of granulation intermediate properties as surrogates for drug product performance, we illustrate how the sensitivity of a high-shear wet granulation process to meaningful variability in raw material attributes can be assessed. Use of experimental designs intended to rigorously probe process robustness to specified and unspecified raw material attributes, in the presence of inherent process noise, is described.

The results of granulation experiments and associated characterization measurements are presented. We identify performance risks due to shifts in API and excipient attributes that might reasonably be expected in commercial manufacturing. These statistical sensitivities are are related to granulation mechanistic understanding, and potential control strategies are explored. The case study highlights how one aspect of risk can be rigorously explored during development in anticipation of changes in raw material source and manufacturing process that will almost inevitably occur in a drug product lifecycle.

Extended Abstract: File Not Uploaded