Human cell culture process capability: a comparison of manual and automated production



Cell culture is one of the critical bioprocessing steps required to generate sufficient human-derived cellular material for most cell-based therapeutic applications in regenerative medicine. Automated cell expansion is fundamental to the development of scaled, robust and cost effective commercial production processes for cell-based therapeutic products. This paper describes the first application of process capability analysis to establish and compare the short-term process capability of manual and automated processes for the in vitro expansion of a selected anchorage-dependent cell line. Estimates of the process capability indices (Cp, Cpk) have been used to assess the ability of both processes to consistently meet the requirements for a selected productivity output and to direct process improvement activities. Point estimates of Cp and Cpk show that the manual process has poor capability (Cp = 0.55, Cpk = 0.26) compared to the automated process (Cp = 1.32, Cpk = 0.25), resulting from excess variability. Comparison of point estimates, which shows that Cpk < Cp, indicates that the automated process mean was off-centre and that intervention is required to adjust the location of the process mean. A process improvement strategy involving an adjustment to the automated process settings has demonstrated in principle that the process mean can be shifted closer to the centre of the specification to achieve an estimated seven-fold improvement in process performance. In practice, the 90% confidence bound estimate of Cp (Cp = 0.90) indicates that that once the process is centred within the specification, a further reduction of process variation is required to attain an automated process with the desired minimum capability requirement. Copyright © 2009 John Wiley & Sons, Ltd.