A biotech company working on cancer immunotherapies has adopted a novel tool for designing experiments, which it hopes will support future data-driven decision making.
Wilhad H. Reuter, lead engineer for upstream process development at Mural Oncology, and his team adopted a Design of Experiment (DoE) approach for their studies into worst-case scenarios for upstream processes. According to Reuter, taking a cue from DoE to worst-case studies is unique for smaller companies.
“DoE is a tool that I’ve primarily only seen in drug substance and drug product development at larger companies,” he told to GEN. “Not everyone uses a DoE design for performing worst-case studies, so I think we’re using a powerful tool for a problem that can be approached in different ways.”
Worst-case studies aim to understand how variability in manufacturing processes might affect the quality of the output. For example, Reuter explains, what impact a one-degree difference in temperature in a bioreactor might have on the resulting drug molecule. DoE, meanwhile, is a statistical tool designed to gather as much information as possible with the minimum resources or experiments required. Using this approach, allowed Reuter to narrow down the factors that affect product quality.
By relying on JMP statistical software, Reuter’s team identified three parameters (pH, temperature, and initial cell seeding density) that predicted cell culture performance.
“This work shows the power of using a statistical approach, not only for analysing data, but also designing experiments to improve overall efficiency,” he continued.
The team plans to go onto carry out more statistical studies for Mural Oncology, and the researchers encourage biomanufacturers elsewhere in the industry to try them out.
“Historically, I think people shy away from statistics because they can seem confusing and convoluted,” notes Reuter. “Showing the power of modelling in gathering valuable information will encourage more people to use these tools for their own work in the future.”
