Integrating high-throughput experimentation with advanced decision-support tools for chromatography process development

The development and commercialisation of a new therapeutic drug is a lengthy and expensive process hindered with uncertainties and high attrition rates. Monoclonal antibodies are a major contributor to the continuous growth of the global biopharmaceutical industry. Chromatography remains the workhorse in antibody purification despite its complex process development and the high operating cost. The research here presents the establishment of an integrated and data-driven decision-support framework in early-stage protein chromatography process development. The key focus of the research is the development of a systematic and rational methodology to automate and accelerate data analysis and decision-making. A novel workflow was developed that combined high-throughput experimentation (HTE) at micro-scale with design of experiments (DoE), multi-variate data analysis, multi-attribute decision-making and a robustness analysis technique to screen and optimise chromatography resins. DoE was linked with an advanced chromatogram analysis method to cope with the large datasets resulting from HTE by automating raw data manipulation. Additionally, the approach offers the ability to correlate the trade-offs between purity and yield with process parameters through a regression analysis. High-throughput purification data were further leveraged using a decision-support tool for the chromatographic purification train linked with a bioprocess economics spreadsheet model. The bioprocess economics model was also used to provide insights regarding the cost-effectiveness of pre-packed chromatography columns as an alternative to conventional self-packed columns for clinical and commercial manufacture. The implementation of the framework demonstrated the synergy of different decision-support tools and allowed for the rapid evaluation of multiple chromatographic purification trains in order to determine the most cost-effective resin sequence and column type considering the whole manufacturing process. Additionally, it is demonstrated that chromatography process development activities could be accelerated by defining platform purification processes and identifying manufacturing bottlenecks fast and with limited feedstock material