S118: Mathematical modeling approaches to complement cell culture process development

Animal cell culture technology has advanced significantly over the last few decades and animal cells are routinely cultured in controlled bioreactors in scales as big as 25,000L for the synthesis of a variety of therapeutic proteins.  While most of the process understanding required for the robust implementation of this technology will continue to be derived from empirical experimentation, mathematical modeling approaches could be used to complement these efforts and ensure robust execution of cell culture processes.  Given the complexity and inherent variability associated with biological processes, careful consideration should be given to the kinds of mathematical models to be used with due recognition to the appropriate context(s) and limitations to their interpretation. Specifically, this presentation will discuss the use of a dynamic model to characterize effect of certain process parameters on the observed level glycation, which is caused by the non-enzymatic addition of glucose to specific sites in the synthesized monoclonal antibody.  In addition to characterizing the effect of culture performance on glycation, applicability of stochastic modeling techniques to characterize variability arising both due to operational and inherent biological considerations will also be discussed.  Simulations using this model were used to successfully guide experimentation to test the effect of various glucose feeding strategies to control glycation.  Through this example we aim to show that, in spite of the limitations of using mathematical models to describe biological processes, carefully selected modeling techniques could aid in practical process development by summarizing knowledge, formalizing hypotheses and elucidating relationship between process parameters.