(27e) Dynamic Feeding Via Automated Sampling for High Productivity Fed Batch Cell Culture Process
During cell culture process development, significant productivity improvement can often be realized using advanced feeding strategies based on at-line measurements. Recent advances in bioreactor auto-sampling technology have enabled automation of complex feeding algorithms. This work demonstrates the application of fully integrated cell culture control, monitoring, and data processing to achieve significant productivity improvement via automated dynamic feeding strategies.
Chinese hamster ovary cells were cultured in 2L bioreactors for an antibody production process. Automated sampling was implemented to deliver samples from the reactors, at regularly scheduled intervals, to a multi function analyzer that measures cell counts, pH, gases, electrolytes, nutrients and metabolites. The analyzer then communicated data through OPC to a custom run manager within our bioreactor control system that calculated appropriate feed volumes based on a) viable cell density (VCD), or b) glucose concentration. The run manager then activated feed pumps and dynamically fed two distinct stoichiometrically balanced media based on real-time in process measurements. These dynamic feeding algorithms and optimized feed formulations resulted in significant productivity improvement of up to 50% over a bolus feed process.
While automation dramatically reduced the time required to operate the cultures, new challenges of maintaining culture osmolality and adjusting to differences in stoichiometric consumption ratios over time were still difficult to overcome. Nevertheless, fully automated, highly productive processes were achieved using a commercially available bioreactor control system with in-house enhancements and an automated sampling system, with feeding based on industry standard off-line measurements (VCD and glucose). This capability not only increases efficiency of process development studies, but potentially increases the relevance of fully automated processing to large scale manufacturing by demonstrating the ease and flexibility of complex feed strategies at small scale.