Evaluation of Technology for Upscaling Design

Presented by Jincai Li

Dr. Jincai Li is executive director, Biologics Process Development, at WuXi AppTec Biopharmaceuticals Co., Ltd., China.

To move candidate molecules rapidly through the development process, it is necessary to have effective scale-up technology. This means that the platform must be consistent, straightforward, with low variability from sample to sample.
The challenge to speed the move to the clinic was considered by Dr. Jincal Li, executive director of process development at WuXi Biologics (part of WuXi AppTec Ltd). The company touts ist wide range of R&D capabilities, embracing small-molecule and biologic drugs, and gene therapies (including genomic approaches). A section of the company develops medical devices.
Li’s team has extensive experience with the ambr® high-throughput singleuse bioreactor system for parallel fermentation or cell culture using different sized units controlled by an automated liquid-handling platform. The system provides a scaled-down model to explore a wide range of conditions with scalability to larger bioreactors, replicating classical laboratory-scale bioreactors. It is widely used by Major pharmaceutical and biologics companies and academic and research institutes as a reliable microscale model for a range of upstream processes. Li states that the company can fulfill the demands for rapid scale-up, running from gram quantities to 500 kg because of its size and integration.
According to Li, because the ambr® technology allows control of pH and osmolarity, his team was able to obtain consistent results, even with cell lines that generate large amounts of lactate. Such overproduction of lactate can be a confounding factor in assessing cellular growth, causing decreases in extracellular pH and increases in osmolality, thus inhibiting proliferation. Those data allowed them to compare Performance across a range of growth options. Based on the accuracy of these observations, the team was able to scale up from 15 mL to liter levels in experimental cultures. The process was later ramped up to 15 L, 200 L, and eventually 2,000 L scales, allowing an accurate prediction based on the ambr study results.

WuXi Biologics manufacturing facility in operation since 2012

Li continued his description of their findings: The viability and growth Overall matches quite well as the scale of the cultures increased to 2,000 L. He described in detail one case study ramping up from small ambr15 bioreactors, used principally at the cloneselection stage and early phase of process optimization (prescreening of the range of culture conditions). This protocol allows large numbers of bioreactors to be operated simultaneously and efficiently, with substantial savings in labor.
For later process optimization, as multiple product quality samples were needed for each culture, 1 L glass vessels were used. Process was later scaled up to 15 L, 200 L, and eventually 2,000 L scales. Performance of the same process across the various scales was compared. The ambr15 apparatus was sampled every two to three days (compared with daily for other scales).
In these studies, the viable cell Counts were slightly lower in the ambr15 platform than in the sub250 and 2,000 L systems. Li’s team observed that the pH trends, osmolarity, and pCO2 production were very close to the ambr modeling results for the various vessels.
In conclusion, Li found that the ambr system is more representative than the conventional shake flask and spin tube approach, yielding fewer surprises in modeling metabolic and process properties and detailing product quality.