(64d) Process Characterization for Commercial Cell Culture Process

When commercial manufacturing processes are modified to take advantage of the latest cell culture medium and process technologies to improve productivity, they need to be characterized to ensure that they can robustly produce the product.  This presentation will focus on a specific case study in which multivariate experiments were combined with univariate experiments to design an overall process characterization strategy for a new version of the cell culture process for an existing commercial product.  The process parameters to test were selected based on a risk-based approach after extensive analysis of the collected process development data.  Designing multivariate experiments to characterize cell culture processes can be a challenge since establishing a multivariate design space requires that the acceptable range for one or more parameters be narrowed to mitigate failures due to the additive effects of multiple factors.  The test ranges for these experiments needs to be established by balancing the competing priorities that require that the ranges be narrow enough to successfully establish a multivariate design space without excessive experimentation and be broad enough to meaningfully characterize the process.  This can be especially challenging for existing commercial products for which it is desired that all product quality attributes meet previously established specifications.  To simplify the design of the multivariate experiments and to improve the interpretability of the results, parts of the cell culture process with different growth and protein synthesis characteristics were treated as two different unit operations, which were studied in separate multivariate experiments. The ranges for the these parameters were carefully established by using stochastic simulations that utilized data collected during process development to ensure that the test ranges were sufficiently broad to characterize the process without risking excessive number of failures.  The results of these studies will be discussed along with opportunities and pitfalls associated with this approach.