Delivering operational optimisation of cell therapy manufacturing through appropriate process characterisation and modelling

Cell therapies (CT) use cells to treat health conditions, and pluripotent stem cells (PSCs) have been identified as a promising cell candidate given their therapeutic potential. However, manufacturing these cells poses challenges because of the insufficient understanding of how cellular behaviour responds to process operations. A commonly used methodology to overcome this problem is characterising how the variability in process parameters affects cell-product quality. Despite the availability of tools to characterise this operational region, CT manufacturing is challenging to characterise due to the lack of understanding of their mechanisms of action and the inherent complexity of live cell products.

This study presents a framework to characterise the interactions between a PSC and its manufacturing process, understanding better how its behaviour determines manufacturing outcomes and establishing strategies for its manufacturing control. Areas relevant to manufacturing operations were characterised as novel case studies: cell expansion, feeding operations and cell type selection. The first developed a computational model to represent cell dynamics and identified growth inhibitors affecting cell expansion. The second explored cells-medium interactions as drivers of growth inhibition, as well as of feeding operations. Then a framework was developed to supplement medium components on an as-needed basis. Thus, reducing the medium volume required for cell expansion. The third used the computational model to identify manufacturing control strategies for genetically and non-genetically variant PSCs. By exploring the prevalence of both populations in culture, the model predicted how variability of process parameters affects population prevalence, leading to the identification of manufacturing control strategies.

Overall, this thesis shows that process characterisation and modelling help us to better understand how PSC behaviour interacts with manufacturing outcomes, thereby broadening knowledge in CT process operations compared to previous work. The developed framework is transferable across various cell lines and can be adapted to support different objectives such as population selection and feed optimisation. This understanding can be used to develop more efficient manufacturing processes, as well as predict the impact of process changes, providing a valuable tool in CT development.