Predictive Modeling of End-of-Run Product Titer in CHO Cell Fed-Batch Processes Using Multivariate Analysis

Fed-batch cell culture processes in the biopharmaceutical industry faces the challenge of accurately predicting end-of-run titer concentrations, which is the mass per volume of monoclonal antibodies produced by Chinese Hamster Ovary (CHO) cells. This prediction is mainly beneficial for process efficiency, while also benefitting yield and product quality. Nutrient feed strategies, metabolic states, and environmental conditions within the bioreactor can influence these titer outcomes. Variables such as viable cell concentration, dissolved oxygen, glucose, and glutamine levels play pivotal roles in dictating productivity and quality of the CHO Cell antibody product. In this report, Principal Component Analysis (PCA) and Partial Least Squares (PLS) regression are leveraged for enhanced prediction accuracy in titer outcomes utilizing the initial CHO Cell antibody product titer, cell growth/health, and analyte measurements in a CHO Cell fed-batch production process.