Perfusion process for CHO cell producing monoclonal antibody: Comparison of methods to determine optimum cell specific perfusion rate

Over twenty-five years ago, perfusion was applied to mammalian cells to produce therapeutic proteins, mainly sensitive ones. Nowadays, other issues such as process intensification make perfusion attractive again. However, it is a complex process to set up compared to the classical fed-batch approach. In this paper, we compared three culture methods to determine the perfusion process parameters for a CHO cell line producing a monoclonal antibody. The push-to-low (PTL) and perfusion-without-bleeding (PWB) methods were performed in bioreactors equipped with a cell retention device, while the semi-perfusion (SP) method was performed in spin tubes. In our culture conditions, the PTL reference method, performed at steady state, indicated a minimum cell specific perfusion rate (CSPRmin) estimated at 100 pL.cell−1.day−1. The PWB and SP methods led to lower CSPRmin estimations related to lower cell metabolic fluxes. We therefore propose a three-step approach to determine the perfusion parameters: (i) SP screening method to estimate the optimum medium and temperature, then (ii) PWB run to determine both the growth rate and the maximum cell concentration and finally, (iii) the perfusion process can be operated according to the previously determined parameters, with a final adjustment of the CSPRmin.