Mimicking physiology with hollow-fiber perfusion technology: an interview with Mindy Miller

The challenges with manual cell expansion are multifaceted. First, manual culture requires skilled technicians or scientists to perform the task without error or contamination. Cell manipulation and sampling are open events and must occur in a biosafety cabinet within a dedicated clean space like a GMP lab, which is expensive to operate. Moreover, even with highly skilled technicians and a clean room, contamination still poses a major risk.

Once the cells are loaded into the bioreactor, an automated bioreactor system can eliminate all the open events, thereby reducing the amount of skilled labor required and eliminating the risk of contamination during these events. Furthermore, with an automated bioreactor platform such as the Quantum Flex Cell Expansion System, the device can be monitored through a web-based software called Cell Processing Application, or CPA. This means that anyone with access can check on the status of the run and monitor any device inputs remotely without paying a technician to suit up and enter the clean room environment.

As an immunologist, when I first saw a hollow-fiber perfusion-based bioreactor in action, I immediately noticed how it resembled a lymph node. The inside of the capillary fibers, where the cells reside, is a relatively small space — only 200µm in diameter — and using a unique fluidics technique the cells are kept in close proximity to one another. Throughout a culture we can achieve cell densities close to that of a lymph node. Unlike other bioreactors, though, perfusion feeding allows cells at this high density to thrive because they are provided with a constant supply of nutrients and oxygen. These nutrients and oxygen are perfused through the space outside the capillary fibers. This means that base media is pumped through a gas transfer module and delivered to our cells, much like the oxygenation of blood in the alveolar capillaries of our lungs, which is then carried to our cells through our circulatory system. Even more compelling for me was that the perfused media that is delivered to the hollow fibers actually diffuses across the fiber membrane in distances that are equivalent to the distances at which diffusion occurs at distal sites in the human body. I’ve never seen a device that captures the benefits of physiology so well.

Hollow-fiber bioreactors can support both adherent and suspension cell expansions. With adherent cells, we utilize coating reagents to help the cells adhere to the walls of the hollow fibers and release agents to help release the cells at the end of the run. Suspension cells don’t require these reagents, but we do employ different fluidic strategies to hold the cells in the bioreactor. For example, we can flow media through a hollow fiber in one direction with adherent cells because the cells are stuck to the fibers. If we did this with suspension cells, the cells would easily flow out of the reactor. So instead, we use bidirectional feeding for suspension cells. This means that we flow complete media into the bioreactor from both sides of the fibers toward the center. Since the fibers are semi-permeable, fluid can cross the membrane and exit the bioreactor from the extracapillary side, thus holding the cells in the center of the bioreactor.

The small bioreactor is great for process development work because you can utilize fewer reagents while optimizing a process. For T cells specifically, we have shown that the small bioreactor works great at generating cell numbers that can support autologous applications. The standard bioreactor can support the generation of cells for multi-dose or allogeneic applications or when the time to dose is critical given higher starting material populations. However, the standard bioreactor can be used for additional applications, such as generating viral vectors or collecting exosomes.

The primary goal of automation is to increase patient accessibility and safety. Automation is reducing the cost of therapies by reducing the amount of highly skilled labor required, reducing reagent costs through more efficient usage, especially when paired with advanced technology and reducing the time to dose through more efficient expansion. Many cell therapies are still unreachable to patients and we hope that by providing automated devices and workflows, we will see the costs reduced enough to reach substantially more patients.

Cell therapy is such a fun field to be in because our fundamental understanding of how these therapies work and the new iterations being developed by academics, clinicians and biopharma are occurring in parallel to the development of manufacturing tools used for their implementation. Device manufacturers aren’t just sitting around waiting for a therapy to be developed — they are working alongside therapy developers and getting a head start to ensure that the technology already exists to support and even enhance the final product. And right now, we are looking toward manufacturing cells that are safer, more effective and hopefully allogeneic in the future.