Behaviour of GS-CHO Cell Lines in a Selection Strategy

Abstract

Selection of a cell line with desirable growth and productivity kinetics is a critical path activity for first-in-human studies with new therapeutic proteins. Typically, programmes to develop a desirable cell line have multiple evaluation stages, where each stage has different growth conditions. At each stage, the results are used to reduce the number of cell lines taken further, by removing the lowest producers and those with unacceptable growth characteristics. Analysis of Lonza’s data for 9 cell line development programmes revealed a highly heterogeneous response of cell lines to their environment, which did not appear to be well predicted by earlier screens. This raised the question: is it possible that we are discarding the ‘best’ cell lines leaving ‘good’ cell lines? The work reported is part of a programme to determine how cell line selection strategies should be modified to increase the probability of identifying the ‘best’ cell lines.

One hundred and seventy five randomly selected cell lines were progressed from transfection, through four assessment stages. Overall, the ranking positions of cell lines changed between each stage. Subsequent analysis showed that it was much easier to identify the poorest producing cell lines at each stage than the cell lines with high ranking positions. However, prediction of cell line behaviour upon transfer between stages is not obvious with these data.

These data therefore suggest that although very good cell lines can be isolated using the current cell line development strategy, there is an opportunity to introduce improvements to the process.