Gene therapy developer AskBio has created a model for recombinant adeno-associated virus (rAAV) manufacturing, which they hope will streamline product development and improve large-scale manufacturing.
The model was made by reverse engineering the company’s manufacturing process. The aim is to identify, at an early stage, the drug therapy candidates that will scale up successfully.
“We’ve built a highly-functional model to get our team to innovate,” explains Jacob Smith, senior director of process development at AskBio. “The aim is to encourage a fail-early mindset to reduce manufacturing process changes and risk in our therapeutic programs.”
Speaking to GEN in advance of last month’s Bioprocessing Summit, Smith explains how his team scaled down their upstream manufacturing process into Ambr® high-throughput bioreactors. They also scaled down their purification process using Tecan automated liquid handlers and microscale chromatography columns.
Using their existing products, they studied vector productivity and encapsidation in AskBio’s Pro10™ production cell line for the smaller-scale process. They were then able, Smith says, to develop an empirical model to examine how other products might behave.
“I think it’s assumed that if a candidate [molecule] produces [well in the R&D phase], you can optimize it later on,” Smith says. “But with gene therapy you don’t always have that extra time.”
The company says they’ve successfully evaluated the model several times in process development and plan to incorporate it into their next generation of products. According to Smith, they have also evaluated their three clinical-stage programs with at least three iterations of each therapeutic candidate.
“We haven’t used the new model yet, but it’s something we’ve been building,” Smith says. “And we already incorporate manufacturability into our candidate selection criteria.”
As one department in a full-service gene therapy company, they are also testing and tweaking the model while manufacturing vectors for AskBio’s R&D and translational research teams. According to Smith, they typically run 8–10 preparations a week in smaller bioreactors, and a 50–250 L bioreactor every other month.