Over the past two decades, the pharmaceutical industry has begun a shift toward a new paradigm of “Quality by Design” (QbD); quality is ensured by developing a detailed understanding of the impact of variability in process parameters on product quality attributes and using this understanding to set process parameter operating ranges such that the process is always run in a parameter space demonstrated to assure quality (the “design space”). Because it is not possible to test product quality at every point in the design space, the QbD approach requires mathematical models that predict product quality attributes at points where no data exist. Moreover, the variability in process parameters must be modeled to predict the expected variability in quality attributes that will result from a given design space definition.<br/><br/>We have developed a methodology to predict the variability in process performance and product quality attributes that will result from a given downstream process design space definition. Multivariate studies were conducted to characterize the process for purification of an enzyme product. Regression analysis was then performed to model the multivariate experiment results and construct mathematical models which predict the outputs of each column operation from the column operating parameters. Finally, a putative process design space was selected, Monte Carlo simulation was used to simulate the variability in process parameters expected to result from the choice of design space, and the resulting variability in recovery and product quality attributes was predicted from the regression models. The putative design space was iteratively modified until the expected distributions of recovery and all product quality attributes met acceptance criteria. Our approach is easily generalizable, and application to bioprocesses for production of other molecules will be straightforward.

Monte Carlo tool for prediction of variability in product quality attributes and definition of downstream process design space

253rd National Meeting (2017)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.