(531a) Evaluation of the Impact of Viable Cell Density on kLa for Oxygen Transfer in Perfusion Bioreactors | AIChE

(531a) Evaluation of the Impact of Viable Cell Density on kLa for Oxygen Transfer in Perfusion Bioreactors


Chen, C. - Presenter, Amgen Inc.
Frank, G., Amgen Inc.
The volumetric mass transfer coefficient (kLa) for oxygen transfer is an essential parameter for the design, control and scale up bioreactors used for mammalian cell culture. kLa is dependant on many factors, including bioreactor geometry, gas sparger design, mixing, agitator design, working volume, the composition of media, pH, temperature, as well as the presence of cells in the culture. Due to these dependencies, as well as the time and cost, measurement of kLa in cell culture has been difficult. For operational convenience, and for meaningful comparison of oxygen supply capacities between bioreactors of different scales, engineers in industry typically use a standard, simplified test media which does not include cells or growth promoting media for kLa measurement. However, kLa measured in test media may not be representative of the actual kLa in cell culture due to the dynamic interactions between cells and media, especially in perfusion bioreactors operating at a very high viable cell density (VCD). An in-process measurement of kLa in cell culture usually requires perturbation of dissolved oxygen (DO) from its set point. A non-perturbation measurement, on the other hand, requires precise measurement of oxygen flow into and out of the bioreactor, which is very challenging.

In this study, we implemented an indirect but convenient strategy to evaluate the impact of cells on kLa for oxygen transfer. As part of this study the effect of operating with and without Alternating Tangential Flow (ATF), a widely adopted perfusion technology used in industry, was evaluated. CHO cells expressing a recombinant protein were cultured in two bench-scale perfusion bioreactors at a fixed DO set point. During exponential growth phase (when cell specific oxygen consumption rate is assumed constant), DO controller output required to maintain the DO set point were tracked for a series of conditions of increasing VCDs. After the cell culture finished, kLa was measured at these DO outputs using a standard test media with or without ATF. The predicted oxygen transfer rates at different VCD levels were calculated using kLa measured in test media. By plotting the predicted oxygen transfer rates against VCD, the impact of VCD on actual kLa for oxygen transfer can be inferred by the shape of the curve. Our results indicate that increasing VCD of CHO cells from 10e5 to 350e5 cells/mL results minimal impact on kLa for oxygen transfer. The results also showed that the impact of ATF on kLa is also not significant. In summary, our study presented a useful strategy to understand the role of cells and ATF on kLa for oxygen transfer. The data generated provides justification for using kLa measured in test media without ATF for the study of bioreactors at different scales in industrial settings.