Cell and Gene Therapies Face Manufacturing Challenges

Many new cell and gene therapies have shown great promise in early clinical studies. While there are a few commercially licensed products, getting these therapies into late-stage trials and approved for use by patients is the next big hurdle. There are numerous manufacturing issues that must be addressed. Fully automated, closed systems that ensure sterility have yet to be introduced. Production processes for patient-specific treatments need to be scaled out, while those for off-the-shelf therapies need to be scaled up. Next-generation characterization techniques must also be developed. Equipment suppliers and cell and gene therapy manufacturers are tackling these issues. Some progress has been made, but the needs of this sector remain significant.

Different therapies, different challenges

First it is important to recognize that there are different types of cell and gene therapies that face very different manufacturing challenges, according to Brian Hampson, vice-president of global manufacturing sciences and technology with PCT, a Caladrius Company. Allogeneic therapies are manufactured from donor cells, while autologous therapies are produced from cells obtained from the patient.

Many companies are developing off-the-shelf (OTS) allogeneic products based on cells obtained from healthy patients that are modified with gene editing to impart new features and knock out undesirable functionality (such as those that cause immune rejection responses). The modified cells are then expanded to produce large quantities of product that can be frozen and stored for use when needed by many patients.

Patient-specific cell therapies (PSCTs) are most often autologous, but can also be allogeneic (from a donor matched to the patient to avoid immune rejection). The key difference from OTS products is the need to use a separate manufacturing batch for each patient. Gene editing is also often used for PSCTs, such as in adoptive cell transfer (ACT) with chimeric antigen receptor (CAR) T cells.

Common challenges

Whether producing a smaller batch from patient cells for an autologous therapy or a large batch from healthy donor cells, major challenges include sterility and avoidance of adventitious agents, which is difficult to maintain using the open systems that are widely employed today, according to Christopher Mach, director and head of bioprocess and director of commercial operations for pharmaceutical technologies at Corning Life Sciences.

Live viral vectors used directly in vivo or in the modification of cells, and the modified cell therapies themselves also require handling and processes that are gentle enough to maintain potency, according to Richard O. Snyder, chief scientific officer with Brammer Bio. “Hold steps and freezing steps should be evaluated carefully when designing these processes,” he notes.

Raw materials can also pose issues. “Cytokines and growth factors that are required are often not available, are very expensive, or cannot be obtained under the appropriate conditions. In addition, continuous manual media feeding/changing for growth factor replenishment and waste removal is laborious and opens up opportunities for mistakes,” Mach says.

The cell expansion environment can be challenging as well, not only generating large numbers of cells, but also growing those cells consistently. Mach points to the use of microcarriers/beads for cell growth as being problematic in this respect. There is also a lack of monitoring technologies for use during the cell expansion phase. Furthermore, the cells can be sensitive to the chemical transfection reagents typically used to introduce the material that programs the cells, according to Mach. Similarly, he notes that consistent and reliable cell isolation that doesn’t introduce foreign particles is an issue. “The same tools and platform technologies used for traditional biologics cannot be used for cell recovery. As a result, currently, cell harvesting can be a very laborious process that takes a lot of steps and has a lot of chances for things to go wrong,” he observes.

Finally, freezing/preserving the cells is a major downstream bottleneck and impacts cell viability and quality, manufacturing locations, and logistics, according to Nina Bauer, senior manager of autologous cell therapy commercial development at Lonza. Mach adds that cryopreservation is a cumbersome process, and there are limited types of vessels that can be used in a cryopreservation environment. Corning will soon be launching new tools that will help customers with containment of cryopreserved materials. The company is also developing new technologies for simple harvesting of live cells that can replace the use of microcarriers and thus avoid particle generation.

Role of gene editing

Gene editing is an enabling technology for the development of novel cell therapies. “Not only are advanced gene-editing techniques allowing the addition and/or removal of cell functionality, they are also influencing the manufacturability of novel cell therapies,” states David Sourdive, executive vice-president of corporate development at Cellectis.

Gene editing is making it possible to take cells from healthy donors and knock out genes that cause immune responses to foreign tissues. Gene editing can also make the cells refractory to chemotherapeutic drugs or antibodies that patients are treated with, and to checkpoint inhibitors produced by tumors that the body’s immune system fights. Insertion of receptors, meanwhile, allows for targeted tumor attack. Gene editing even makes it possible to produce T cells that are designed to attack other T cells, but not to attack themselves during the expansion phase of the manufacturing process. “As a result, gene editing is enabling an important transition from personal medicine to OTS products that are compatible with the current standard of care,” says Sourdive.

The most common method for genetic modification of cells uses viral vectors, which means that in addition to the cells as starting material, these therapies require a second highly complex raw material-the virus particles-the manufacturing of which brings with it its own upstream bottlenecks, according to Bauer. “The quality (defined by both the design of the viral vectors and their manufacturing) has an immediate impact on the transduction efficiency,” she says.

For instance, manufacturing platforms that rely on transient transfection to generate vectors can be limited by scale and cell density, as well as the need for sufficient raw materials, such as plasmid DNA, according to Snyder. He adds that achieving yield and purity while maintaining potency can also be an issue, as some viral vector classes are fragile.

As a result, there is interest in non-viral alternatives. They can also be complex, however. Cells relevant for autologous cell therapeutic approaches, usually blood or stem cells, are notoriously hard to transfect, according to Andrea Toell, senior product manager at Lonza. Many of these approaches also require quite complex transfection scenarios. “For example, for genome editing using CRISPR (or zinc-finger nucleases [ZFNs] and transcription activator-like effector nucleases [TALENs]), you want to transfer various cargos in parallel, maybe even including different molecule types (plasmid DNA, mRNA, protein). The same is true for generating CAR-T cells using transposon/transposase-based systems or for iPSC [iInduced pluripotent stem cells] reprogramming using episomal vectors,” she notes.

Improved electroporation technologies have been shown to be potent non-viral alternatives for overcoming the complexity of the non-viral approach, according to Toell. Lonza has developed a proprietary system that has been proven to effectively transfect these cell types, even in complex scenarios. The Nucleofector Technology has, to date, been limited to a medium-scale format (of up to 1x10⁷ cells). Recently, however, the company added a functional unit (the LV Unit) to the existing system that enables gene editing with up to 1x10⁹ cells.

“To facilitate scaling of the process, the system was designed so that transfection protocols established on the existing smaller-scale unit can be transferred to the new large-scale unit without the need for re-optimization, thus saving time and material,” notes Toell. She adds that the LV Unit is a closed system, reducing the risk of contamination, and comes with 21 Code of Federal Regulations Part 11-compliant software to fulfill documentation needs in a GMP lab.

Cellectis has developed a proprietary system to deliver TALEN in mRNA format into cells. The Pulse Agile system is a gene-editing engine central to Cellectis’ off-the-shelf CAR T-cell process that can transfect 5x10⁸ cells in one single pulse, yielding more than 97% efficiency and high viability, according to Sourdive.

Scaling up OTS therapies

“The crucial upstream challenge is generation of the master cell bank (MCB), ensuring the quality (stability and long-term viability) of the cells, using freezing technologies,” Bauer asserts. In addition, storage and banking in secure facilities are key. She adds that there is still concern about prion disease contamination, so donor selection/eligibility is another challenge.

Another big challenge for large-scale manufacturing of cell therapies is generating the numbers of cells per batch that are needed, according to Mach. “Typically you need 100-200 billion cells per batch, since patients often require hundreds of millions of cells per treatment,” he says. That means the number of cell population doublings is high, but cells have limited expansion capability, according to Thomas Heathman, business leader for technology development, manufacturing development and good tissue practices services with PCT.

Establishing optimum control parameters at scale is also an issue, which makes it difficult to maintain nutrient, oxygen, and carbon dioxide levels as the surface area-to-volume ratio becomes unfavorable. “This issue is complicated by the fact that the reagents tend to be very expensive and there have been difficulties in maintaining sustainability and reasonable cost of goods at scale with the switch to serum-free culture media,” says Heathman. He adds that there is also a lack of effective potency/quality assays that can be used for the development and operation of scalable up/downstream processes.

There is also need for improved technologies for harvesting live cells, particularly adherent cell therapies, at large scale, according to Heathman. Extended downstream processing times can impact cell quality as the scale increases, which is different from traditional bioprocessing. Similarly, the time that the cells are exposed to toxic cryoprotective agents during cryopreservation increases as the scale increases, and therefore formulation to maintain quality becomes crucial.

Currently most OTS therapy manufacturers are looking to leverage bioreactor technologies from traditional bioprocesses due to the large quantities of available information on these systems, says Heathman. “In general, there has not been a fast pace of development with respect to large-scale bioreactors for cell or gene therapy. We do know, however, that suppliers are working on new technologies, but by and large our customers are just getting around the issue by using what they can and then adapting it to the environment they need to work in,” Mach adds. He notes, though, that disposable biocontainer/bag films are cleaner, microcarriers are pre-sterilized, and high-performing media formulations, reagents, and supplements have been introduced.

Sourdive agrees that there is plenty of room for improvement with respect to cell separation. “The issues of throughput and volume have not yet been addressed to the extent required today. We now have a need to produce lots of cells at one time with high levels of purity,” he observes. Sourdive is hopeful about new technologies under development, though. “We expect to see quantum leaps in advances in the near future, such as in-line process characterization and microfluidics. In fact, there are many developments underway that pertain to increasing throughput, such as multiplexing by scaling out. The multiplicity of small developments will be combined to form integrated systems that offer high throughput with lots of inline control,” he states.

For cell separation, Bauer notes that there is interest in magnetic bead selection for highly selective harvesting and bubble-based antibody selection. For harvesting, she notes that two technologies-a GE harvester based on tangential flow filtration (TFF) and a counter-flow centrifugation system being developed by PCT in collaboration with Invetech that is potentially superior to TFF-are attracting attention.

Scaling out PSCTs

Not surprisingly, there are also several challenges to cell processing for patient-specific cell therapies. “Manufacturing PSCTs is very different from producing OTS therapies,” asserts Hampson. “It means manufacturing an individual therapy for one patient at a time and requires the isolation of the cells of interest from primary human material. PSCTs are not homogeneous cell lines; they are primary human cells, and cells from each patient or matched donor behave differently. As a result, isolating the cells can be challenging,” he explains.

Specifically, collection of adequate starting material from the patient or donor can be an issue. “Although collection is performed typically at the clinical site and some may not consider it part of the manufacturing process, we believe it should be; because the source material comes from individual patients, it can be highly variable. Standardization of collection is a challenge because it typically needs to be performed at many different sites with multiple variables in play,” says Hampson.

Because the quality of the cells can vary greatly depending on the severity of the condition or potential previous treatments the patient has received, defining clear quality criteria and having rapid tests available to assess the material are crucial, according to Bauer. “To date, however, the quality criteria for many therapies are still fairly rudimentary, and testing takes a long time. In addition, for cases in which the cells do not pass the tests, the question remains: are these patients automatically not eligible for treatment? For ethical reasons, efforts should be undertaken to develop technologies that can either improve the starting quality (an upstream bottleneck) or the process, so that the final product can ultimately conform to release criteria (methodology bottleneck),” she comments.

The second challenge is access to reliable and cost-effective methodologies for isolation of the cell type of interest. The number of available technologies has been growing for a number of years now, but there are still issues. Hampson notes that for instance, droplet-based flow sorting is still a very manual, open, and technique-sensitive process that is not suitable for large-scale commercial production of many thousands of doses per year. It is labor-intensive, time consuming, and subject to errors and problems of various sorts.

For therapies that require cell culture, the challenge is gaining access to closed, automated, bioreactors. “As with cell isolation, available technologies have been growing, but there still is a lack of an elegant, closed, automated solution for many of these processes,” says Hampson. Formulating the final product has its own challenges, too, particularly given the very low volumes and precise formulation and compositional requirements, and again the need for closed, automated, systems, according to Hampson.

To produce large quantities of products, the issue for PCSTs is not the need for larger bioreactors. “The need is for small-scale bioreactors that are suitable for making one patient lot at a time, and scale out is needed, not scale up,” Hampson observes. As examples of existing systems, he points to the Wave/Xuri, G Rex, and Quantum bioreactors offered respectively by GE, Wilson Wolf, and Terumo.

For washing/volume reduction/final product formulation, Hampson notes that there have been a couple of new platforms introduced in recent years. For cell washing, Fresenius offers the Lovo platform and Biosafe--which was recently acquired by GE--the Sepax. Acoustic focusing, meanwhile, is a novel approach under development. As mentioned previously, PCT is also developing its own closed, automated, cell processing system to address the unmet need in this area.

Need for closed, automated systems

There is a general theme to the challenges discussed herein--the need for closed-system automation solutions that enable commercially viable manufacturing of viral vectors and modified/unmodified cells and ensure consistent quality, reasonable cost of goods, and scale up/out to meet commercial demand, according to Hampson. “The number of available systems is increasing, but there are still too few to cover anywhere near all of the different applications of cell therapies that are out there,” he states. For instance, there remains a need for closed, automated bioreactors that can, in a single-unit operation, take some isolated cells, perhaps at very low starting volume, expand them in culture, perhaps to a much larger volume, while accommodating the variability in starting material from one patient/donor to the next and have a harvested product ready for formulation into a final product.

There are a few early systems on the market worth noting, according to Bauer. The Cocoon, developed by Octane Biotech in collaboration with Lonza, is a GMP-in-a-box approach particularly suited for autologous manufacturing, allowing for out-scaling and potential downgrading of GMP levels. While this system is still in development, Miltenyi’s Prodigy (mainly for non-adherent approaches) and Terumo’s Quantum (predominantly MSCs and related adherent cell types) are technologies of interest that are already on the market. GE offers a modular manufacturing approach with automated unit operations in a flexible system. “Other important developments currently under way include sensors for in-line monitoring of culture conditions, which allows for automated fine tuning and adjustments of nutrients, gasses, etc.,” Bauer remarks.

In addition, current methods to harvest, concentrate, transduce, and wash cells at large scale are labor intensive, result in liquid handling and contamination issues, and require too much time, leading to reduced cell viability, therapeutic potency, and product quality, according to Mach. Automated systems that can handle this multi-step process are also desired. “Filtration using hollow fibers, acoustic separation, and continuous centrifugation have shown promise but it is too early to tell if these technologies will be beneficial for cell therapy manufacturing,” Mach observes.

Snyder adds that single-use systems are important for manufacturing viral gene therapies and modified cell therapies. “The technology facilitates rapid turnaround times between campaigns for changeover, cleaning, and set-up,” he notes. Examples include disposable stirred tank and rocker reactors, as well as HYPERStack and iCELLis Fixed Bed Bioreactor platforms, respectively from Corning and Pall Corporation.

Characterization issues

Characterization of both the starting cells and final cell-based products is important for both autologous and allogeneic therapies. Doing so is challenging, however, because cells are not defined in the same manner as proteins and small molecules, according to Sourdive. “Cells are identified and their phenotypes assigned based on the presence of certain markers on their surfaces. They are then placed into a category based on their size and complexity,” he explains.
Characterization methods are improving, however. Additional markers are regularly being identified, and researchers are learning what features make a cell a specific type (stem, effector, etc.). These advances are both setting the bar higher when characterizing cells and enabling manufacturers to increase the accuracy of cell selection when collecting a specific population for production of cell therapies.

Quality control is also a challenge, according to Sourdive. “Functional assays are getting increasingly accurate but also demanding and may rely on live control or target material that must remain consistent over time,” he notes.

Other implications

Cell and gene therapies are a new class of drug products that has not only direct manufacturing implications, but also implications for regulation of the manufacturing processes involved in their production, according to Sourdive. “We are breaking new grounds, for example with OTS CAR T-cell products, and the regulations are developing as products are introduced. Industry is working closely with regulatory agencies to develop an understanding of what appropriate regulatory controls should look like, bearing in mind that the products are live cells with their own descriptive models and analytical techniques,” he observes.

The introduction of off-the-shelf cell therapies also has the potential to expand the availability of these advanced treatments to a much wider patient population. “To date, cell therapies, especially autologous approaches, have been limited to patients with access to sophisticated points of care with highly skilled experts and advanced cell processing platforms. Allogeneic off-the-shelf therapies will expand that reach to much broader populations and bring profound changes to the industry for the benefit of patients,” Sourdive asserts.

Article Details

BioPharm International
Vol. 30, No. 1
Pages: 20–25

Citation

When referring to this article, please cite it as C. Challener, “Cell and Gene Therapies Face Manufacturing Challenges," BioPharm International 30 (1) 2017.