Immune cellular therapies are poised to bring about a dramatic change in healthcare. They promise to cure dreadful diseases such as cancer. Yes, cure. Imagine what it would be like to move past symptom management or treatments that are unsatisfactory because they reduce quality of life or leave patients in fear of disease recurrence.
Now that immune cellular therapies are advancing from blood diseases to solid tumors and autoimmune diseases, the overarching question is how these therapies can be manufactured in a timely, cost-effective way so they can actually reach patients. With these therapies, as with most novel therapies, there are challenges in scalability. Moreover, with immune cellular therapies, some of these challenges pertain to autologous approaches, and others pertain to allogeneic approaches. In either case, sufficient supply is critical to extending the reach of cures.
The approval of Kymriah in 2017 marked the beginning of T-cell therapy commercialization. Since then, only a few additional therapies have been approved. However, the FDA projects that this number will grow to 10–20 therapies annually starting in 2025.
It is not surprising that manufacturing issues remain to be addressed. Conventional T-cell therapy manufacturing entails expensive, lengthy, and error-prone processes that span multiple unit operations. Typically, these unit operations include cell enrichment; cell selection; cell activation; gene transfer; cell expansion; and washing, fill-finishing, and formulation.
Of course, conventional T-cell manufacturing involves many more details. For example, the production of a single batch often requires that highly trained teams spend weeks in clean rooms executing over 50 manual processing steps with a plethora of benchtop instruments. Fortunately, conventional manufacturing approaches are being superseded by more innovative approaches.
“If you have 50 manual steps and are manufacturing 10,000 patient doses, you have 500,000 opportunities for operator error,” says Fabian Gerlinghaus, co-founder and CEO, Cellares. “In addition, contamination risk exists at every open manual transfer step. The current unsustainable process failure rates of 18–20% are sky high compared with those in the biologics space.
“We can bring down the process failure rate by more than a factor of three, reduce manufacturing costs up to 70%, and address scalability. We need hundreds of thousands of patient doses today. However, in 2021, the entire industry produced just 4,000 doses. This paints a very stark picture of supply and demand.”
To boost production, Cellares has developed the Cell Shuttle, a high-throughput cell therapy manufacturing platform. It fully automates and closes the manufacturing process from start to finish while allowing 16 independent batches to be manufactured simultaneously.
The heart of the system, a suitcase-sized consumable cartridge, contains modules for all of the unit operations, alleviating the need to transfer cells. After the cell manufacturing process is designed, the starting materials are loaded into the cartridge, which is then transferred into the Cell Shuttle, a truck-sized, self-contained factory. A robotic work cell moves the cartridge from one bioprocessing instrument to the next in accordance with the workflow. At the end, the cartridge contains the final cell therapy product ready for release and infusion.
Gerlinghaus notes that there are many cell therapy modalities. They include T cells modified to express chimeric antigen receptors (CARs) or engineered T-cell receptors (TCRs); natural killer cells; and hematopoietic stem cells. “Every cell therapy process looks a little different,” he continues. “A key requirement was that our platform was flexible enough to address 90% of cell therapy modalities. Without this level of flexibility, it would be a nonviable solution.”
To provide flexibility, the Cell Shuttle cartridge supports different processes out of the box, and reagents are customizable. Workflow design is accomplished through a user-facing software suite, the Process Design Studio. The software empowers process development scientists to digitally design high-level operations, the sequence of unit steps, and the underlying process parameters.
Closed, automated systems
Manufacturing may require fewer manual interventions if closed, automated systems are adopted. These systems reduce the opportunity for error, expedite processing, and increase efficiency.
Historically, cell isolation and bead removal have been completed manually within the cell therapy manufacturing process. Even some FDA-approved therapies use open processes for their isolation and activation steps. “These phases of cell therapy manufacturing can be automated through the use of tools like the Gibco CTS DynaCellect Magnetic Separation System,” says Evan Zynda, PhD, senior scientist, Thermo Fisher Scientific. “Such tools make it possible to carry automated processing through the entire workflow.”
Ultimately, full scalability is needed—scale up, scale out, and scale down. For example, according to Zynda, with autologous workflows, most development is done at 1 L volumes. Developers will need to scale down and out, expanding into decentralized models working with volumes at or below 1 L. As allogeneic therapies move forward, developers will need to scale up to 20, 50, and even 200 L volumes.
Whatever scale applies at the outset of a workflow will need to be sustained throughout the workflow, regardless of whether the workflow is for a low-volume autologous therapy or a high-volume allogeneic therapy. Accordingly, both autologous workflows and also allogeneic workflows can benefit from tools that support continuous processing.
Workflows are evolving to incorporate inline monitoring and remote sensing technologies that make use of feedback to control manufacturing processes and ensure quality without the need for manual manipulation. Using predictive analytics and checkpoint monitoring, manufacturers can eliminate failure modes to achieve a higher manufacturing success rate.
To support the scale necessary for future T-cell therapy manufacturing, manufacturers should look for closed and automated instrumentation that is modular and easy to assimilate with new or existing platforms. “Consider using flexible instrumentation that can integrate and connect with other equipment,” Zynda suggests. The idea, he stresses, is to facilitate communicate between devices.
“Scalability, automation, and decentralized production are focus areas for improvement,” Zynda adds. “Automation will allow for increased use of other technologies in the laboratory, such as artificial intelligence and analytics, to optimize the manufacturing process.”
Instrumentation decisions should support long-term efficiencies. For instance, GMP-compliant instrumentation ensures that any work done during process development and optimization can transfer over to commercial production.
According to Zynda, more and more users are moving from rockers to stirred tank bioreactors. Stirred tank bioreactors offer more environmental control, boast a smaller footprint relative to volume, are extremely scalable and flexible, and are best suited to support real-time monitoring and feedback-controlled processes.
“Interestingly, this technology is not new,” Zynda points out. “Although it has been in use commercially for almost 100 years, it has not been applied in cell therapy. The barrier to adoption should, therefore, be lower. Newer versions are available that are fit for purpose.”
A capacity crunch
Contract development and manufacturing organizations (CDMOs) and contract manufacturing organizations (CMOs) are bedeviled by capacity limitations that complicate the manufacture of large quantities of T cells. “The general capacity crunch is likely to increase,” says Matthew Hewitt, PhD, executive director, Scientific Solutions, Cell and Gene Therapy, Charles River Laboratories. “Most CDMO/CMOs are focused on clinical manufacturing rather than commercial manufacturing.”
Manufacturing is also challenged by product heterogeneity. “Although there are methods to reduce heterogeneity between lots, lot-to-lot variability will be difficult to completely eliminate since we are dealing with living therapies,” Hewitt notes. “Performing proper process and analytical development is key to process stability and consistency. The process is the product—do not neglect it.”
In general, automation improves product consistency and lowers costs, but implementation timing is key. Many companies enter the clinic with open and manual processes to compress development timelines. After seeing positive clinical data, they circle back to discuss closing and automating their processes. However, companies that follow this route need to conduct expensive and time-consuming clinical product comparability studies.
More money may be spent up front if automation is implemented prior to entering the clinic. The advantage is that comparability studies are then unnecessary. By developing a closed and automated platform process, therapeutic developers have the opportunity to reduce development for subsequent programs. Fully closed and automated manufacturing processes generally yield about 30–40% savings over open, manual processes. Automating quality control analytics testing can further decrease costs.
Hewitt sees several emerging trends. One is process and analytical automation, which will lead to changes in the clean room classification required for therapeutic manufacturing, changes that will further reduce costs. “We are also likely to see manufacturers tested by rising demand for commercial allogeneic cell therapies,” Hewett predicts. To meet demand, manufacturers will need to scale up properly.
If autologous therapies continue to be a significant portion of commercial therapies, then alternative hub-and-spoke decentralized manufacturing methods may emerge to help stratify workforces in a field where talent is tight. Such methods may also help to put therapies closer to patients.
Timelines remain critical. Companies must move efficiently from discovery to clinic, and there are still areas where regulators can assist. According to Hewitt, several groups have discussed parent-child regulatory filings, where a developer establishes a standard process and analytical backbone to advance therapeutic programs to clinic.
After they progress through the backbone once, the developer can use the same framework to advance subsequent programs, provided process and analytical parameters remain unchanged. Regulators have indicated they are open to these structures but have provided very little guidance on the data needed to implement them. About all the regulators have indicated is that child programs must be “comparable” to the parent program.
“The field needs more clarity,” Hewitt insists. “Otherwise, we will have difficulty building a fast-test, fast-fail environment that can deliver promising therapies to patients sooner.”
Exothera, a Belgium-based CDMO, delivers customized process development and GMP manufacturing services for viral vector–based therapeutics such as gene therapies, cell therapies, oncolytic viruses, and vaccines. “Our technology selection can meet clients’ needs all the way from early development until the commercial scale, in a one-stop-shop fashion,” says Hanna Lesch, PhD, the company’s chief technology officer. “We are an experienced CDMO using scale-X technology and other adherent and suspension bioreactors to provide a flexible and high-quality approach to accommodate projects at all stages of development.”
At Exothera, process development begins with technology transfer at the flask scale. Then, leveraging design of experiment tools, the company proceeds directly to small- and mid-scale bioreactors with minimal process changes. “We have accomplished successful transfer to both adherent and suspension bioreactors,” Lesch asserts. “And we have done so with both transiently transfected and stably transfected cell lines.”
According to Lesch, meeting GMP and CMC standards in the early stage of process development is very important as well as defining an adequate analytical strategy early on to meet the critical quality attributes. “If you can meet high GMP and CMC requirements at a small scale, then process scalability depends only on the technologies you select,” Lesch points out. “This is possible through the quality by design approach, which meets regulatory requirements and creates a robust, scalable, and reproducible cGMP manufacturing process.”
The personalized nature of cell therapies and rare diseases and small patient populations has made the scaling of cell therapy production challenging and expensive. Lesch notes that whether cell therapy manufacturing follows the point-of-care (autologous) model or the centralized (allogeneic) model, improving automation can help reduce development and manufacturing times and costs.
CDMOs have shown that ready-to-use viral vector platforms can make viral vectors more readily available. Accessibility to viral vectors for ex vivo therapy will be easier in the future. Moreover, new manufacturing technologies that reduce times and costs and analytical technologies that improve process understanding are opening the way to automation.
“Smaller developers usually reach out to us to fast-track their new drug to clinical trials, whereas bigger companies ask us to scale up their process to commercial scale,” Lesch remarks. “In both cases, we can contribute to making life-changing therapies available and affordable for patients.”