The benefits of single-use technologies for upstream viral-vector processes clearly outweigh their disadvantages.
Cynthia A. Challener
Upstream viral-vector processes can have significant impacts on downstream purification requirements. The presence of serum, culture harvest viability, turbidity, cell lysis reagents, contaminant DNA levels, and residual transfection agents are all upstream factors. Overall process scalability and robustness, says Todd Sanderson, senior R&D manager for Pall Biotech, are dependent on the performance of the bioreactor whether it is adherent- or suspension-based. “The ability to provide a reproducible and uniform ideal microenvironment is a prerequisite for reproducible and uniform vector production in a bioreactor,” he states.
Increasingly, single-use technologies—such as shake flasks, rocking-bioreactor bags, and larger bioreactor bags—have been used for laboratory- to commercial-scale viral vector production, including for upstream inoculum expansion and production steps. These technologies provide several advantages over stainless-steel systems, including reduced capital investment, elimination of cleaning in place (CIP), and cleaning validation. These advantages, according to Atul Mohindra, senior director of R&D for Lonza, drive down overall contamination rates as well as enhance the efficiency of a production facility while reducing manufacturing costs (both operational and maintenance).
While there may be some drawbacks to single-use systems, Tony Hitchcock, technical director at Cobra Biologics, a Charles River company, thinks the benefits far outweigh the problems. “The simple reality is that the COVID-19 pandemic has shown the benefits of single-use systems in that it would not have been possible to deploy the viral vaccines in the time and the scale that has been achieved,” he asserts.
Notably for the manufacture of vectors based on adeno-associated virus (AAV), one of the most common vector types for gene therapy delivery and viral-vector vaccines, recent data have shown that stainless steel can have a negative impact on the infectivity of AAV, which, reinforces the strategy of using single-use technologies, says Emmanuelle Cameau, technical manager for viral vectors and gene therapy at Pall Biotech.
Single-use options from lab to commercial scale
A variety of different single-use systems are used for viral vector production depending on the phase of development, required production scale, the specific vector being manufactured, the type of cells used, and whether the virus is harvested from the culture medium, the cells, or both. The single-use systems most used, according to Chor Sing Tan, strategy manager at Cytiva, include T-flasks, roller bottles, cell factories or cell stacks, rocking bioreactors, fixed-bed culture systems, and stirred-tank bioreactors.
“Traditional laboratory-scale systems, such as roller bottles or cell factories designed for use with adherent cells, are generally used for small-scale production of viral vectors, while fixed-bed culture systems and stirred-tank bioreactors are the predominantly leveraged for scaling adherent and suspension processes, respectively,” Tan says. Use of microcarriers that support adherent cell lines is also being explored as a means for scaling adherent processes in suspension in stirred-tank bioreactors.
Flatware is often used in the seed train for larger-scale processes as well as for small-scale production, according to Tan. T-flasks, for instance, are generally used for subculture, while cell factories are mainly dedicated to the culture of adherently growing cells. Roller bottles are simpler than T-flasks but require a mechanical device for rotation. Cell factories have their own limitations, such as variable volumes and a lack of a sufficient gas exchange for optimum cell growth. For all of these systems, extensive manual procedures and open manipulation preclude their practical use at commercial scale.
Rocking bioreactors, fixed-bed bioreactors, and stirred bioreactors do enable aseptic, closed operations and are amenable to automation. They are also designed to increase cell-culture density while maintaining adequate gas exchange and supply of nutrients and preventing accumulation of cell culture byproducts, according to Tan.
For single-use systems in particular, Sanderson adds that cell-culture compatibility is critical because many of the plastics and polymers used in biologics manufacturing are not inert to biological processes. “In addition to leachable and extractables, there is also risk for nutrient loss by adsorption onto the plastic surface. These concerns can be addressed by screening these materials for compatibility with the materials in each process,” he comments.
Need for higher commercial capacity
Many viral-vector manufacturing processes for gene therapies were initially developed in academic laboratories using traditional plasticware systems such as multi-layer flasks. These processes were often retained when these candidates moved into the clinic. For some low-demand gene therapy indications, these single-use systems remain a technology of choice, offering sufficient capacity and removing the need for process change, says Marie Jourdan, vice-president of marketing and product management at Univercells Technologies.
The well-known limitations of flatware in terms of scalability, reliability, and process optimization have caught the attention of end-users, however. “Scaling-out to achieve large quantities of viral vectors presents several drawbacks that lead to footprint, capital, and operational expense, and reproducibility issues. Developers are consequently turning to more scalable technologies with higher throughputs and economies of scale while reducing footprint, manual operations, and cost of goods compared to 2-D technologies,” Jourdan observes.
Large-scale single-use bioreactors, when compared to flatware systems, also require much less plastic to produce the same quantity of viral vector product, according to Cameau. That means the quantity of plastic that must be disposed of afterwards is also reduced.
Suspension cell culture in stirred-tank bioreactors and adherent culture in fixed-bed bioreactors have become the predominant choices. Both, according to Cameau, offer more process control than flatware systems because bioreactors include sensors for monitoring of pH, dissolved oxygen (DO), carbon dioxide percentage, and other parameters.
Automated bioreactors, Cameau adds, provide opportunities for optimization with regard to the cell-culture media/surface ratio in adherent processes. “Compared to flatware, the ratios in fixed-bed bioreactors are generally significantly lower, resulting in lower overall volumes to handle upstream, and lower volumes to be processed downstream,” she explains. Cell-culture media utilization also tends to be more efficient, and information gained from media utilization studies can be used to balance cost and performance at all scales and formats, according to Sanderson.
Adherent versus suspension cell culture
As mentioned, viral-vector producers have an initial choice between suspension and adherent production platforms. A key difference between adherent and suspension bioreactors is whether there is a physical surface provided to support cell growth. This physical surface, according to Sanderson, complicates the engineering involved to provide a uniform controlled environment through the entirety of the bioreactor.
He also notes that adherent expression systems generally rely on serum-containing media formulations, and animal-derived components pose an additional risk for the introduction of adventitious agents and biological raw material consistency, which can result in more biological variation than chemically defined media systems.
Adherent cultures require the use of trypsin to detach cells. “The presence of active trypsin changes cellular metabolism and intracellular morphology. Serum is used to inactivate the trypsin, but lot-to-lot variability and the instability of the serum may fluctuate the degree of trypsin inactivation, and thereby result in less control over the host-cell state. Residual activated trypsin can also inactivate many viral vectors,” says to Natraj Ram, vice-president of innovation in Thermo Fisher Scientific’s BioProduction business.
Adherent processes, Sanderson continues, also involve more complex seed trains with larger numbers of flatware vessels or another adherent bioreactor system due to the large number of cells required to inoculate adherent bioreactors.
For these reasons—and also due to the limited number of options for large-scale production in adherent systems—many viral-vector and gene-therapy developers are now focusing on suspension platforms, according to Hitchcock.
The wide availability of scale-down models for suspension processes, such as the ambr platform from Sartorius, also allow for process intensification to increase vector productivity to comparable levels to the adherent platforms, Hitchcock says. “These systems also provide a route for scaling to commercial production, which is especially true for adeno-associated virus (AAV)-based therapies where an increasing number of producers are looking to operate at scales up to 2000 L,” he says.
Cobra Biologics is also seeing an increasing number of companies developing producer and stable cell lines, particularly for the production of lentiviral (LV) vectors, which are more readily transferred into suspension platforms and can potentially be run at scales greater than 2000 L.
Even so, Sanderson observes that adherent bioreactors have a process advantage over suspension bioreactors for intracellular products such as those based on the AAV2 serotype, in which the attachment of the cells allows for cell lysis in relatively low volumes, which reduces the required downstream purification steps.
In addition, both adherent and suspension single-use bioreactors integrate well with other single-use technologies, according to Ram. Furthermore, while it is necessary to ensure uniform distribution of cells throughout the reactor matrix as well as the uniform distribution of the transfection complex for reactions in fixed-bed bioreactors, he notes that the volumes transferred during both the inoculation and transfection steps can be decreased compared with those required for suspension processes in traditional stirred tanks.
Indeed, stirred-tank reactors have been demonstrated to offer a scalable solution for biologics products, notes Jourdan. However, she says, with this system, capacity increases linearly with operating volumes, which for specific very high demand-low productivity applications, multiple units in parallel may be required, which has footprint and capital expenditure implications.
Some additional decrease in volumes is also possible during the harvest step, Ram adds, if using the fixed-bed bioreactor system with a process where the vector remains in the cells. In this way all of the reactor medium can be disposed prior to harvest and a low-volume harvest step can be accomplished. “For suspension processes, a separate harvest step must be developed, which adds some complexity but is not a showstopper,” he comments.
Finally, Ram points out that the use of perfusion in fixed-bed bioreactors is simpler than in traditional stirred tanks because the attachment matrix essentially acts as the cell retention device, while in suspension culture an external cell retention device is required. “Even in the case where the AAV serotype remains predominately in the cell, there is still usually 20–30% present in the spent media, which most drug manufacturers opt to retain. If the research team wants to keep the supernatant, detergent lysis should occur in the presence of the spent media in order to use this detergent treatment as a viral inactivation step as well.”
Despite the advantages of adherent cell culture noted by Ram, he believes that the maturity of single-use technologies in the traditional stirred-tank space combined with the the ability to further scale-up processes should be considered in terms of adopting suspension versus adherent modalities. “Advances in suspension-based viral-vector manufacturing should be emphasized, and newer technologies that incorporate suspension-based cultures and provide the same advantages as adherent modalities should be developed,” he asserts. “As with many new technologies,” Ram continues, “adoption will remain a challenge and needs to be addressed through collaboration across the industry.”
Single-use, fixed-bed bioreactors
Adherent processes, which initially were notoriously difficult to scale-up in regard to operational space and manpower required, have become scalable in recent years due to the emergence of 3D fixed-bed bioreactors.
These single-use solutions, according to Ram, have alleviated the operational challenges for adherent processes at the terminal reactor stage, but are known to have issues with even distribution of fluid, cells, and reagents, thereby making process control challenging. “With each new 3D fixed-bed bioreactor that emerges on the market, however, these concerns are being addressed,” he says. Regardless, Ram also notes that adherent platforms still require more manipulations and manpower for the seed train preceding the terminal bioreactor when compared to suspension platforms.
Recently introduced fixed-bed bioreactors offer scalable solutions that deliver high capacity for adherent processes in a more industrial manner, says Jourdan. “These single-use bioreactors deliver high cell densities in a reduced footprint while offering in-line process monitoring capabilities, facilitating operations. In addition, their reduced operational volumes enable optimization of the use of expensive media and other materials, such as transfection reagents,” she states.
The Univercells Technologies single-use scale-X fixed-bed bioreactor portfolio consists of automated, intensified bioreactors utilizing a structured spiral wound fixed-bed available in different scale configurations to facilitate the transition of processes from R&D to commercial-scale good manufacturing practice (GMP) manufacture, Jourdan says. She also notes that the design of the scale-X bioreactor has been shown to provide significantly higher yields than alternative adherent-cell based systems for virus production, while the inline product concentration offers a decrease in operational volumes and delivers a concentrated, clarified bulk harvest.
In the Pall iCELLis automated, single-use, fixed-bed bioreactors, the compact three-dimensional fixed beds are filled with proprietary macrocarriers composed of polyethylene terephthalate microfibers that provide a large surface area for cell growth in a small footprint. Even media flow and a unique waterfall oxygenation approach combined with gentle agitation enable high cell densities equaling the productivity of much larger stirred-tank units. The platform has been used for the development of scalable commercial processes, including production of the Novartis’ gene therapy Zolgensma (1) and Johnson & Johnson’s COVID-19 viral-vector vaccine (2).
Single-use suspension bioreactors
Single-use suspension bioreactors are increasing in popularity for viral vector production. New media formulations have been developed that allow direct adaptation of HEK293 adherent cell lines to serum-free suspension growth conditions, according to Sanderson. Suspension-adapted cell lines are also commercially available.
In addition to these approaches, viral vectors are occasionally produced in suspension bioreactors using adherent growth conditions on solid support microcarriers. Other expression systems, such as engineered producer lines and insect cell/baculovirus systems, are typically used in suspension bioreactors as they are suspension-based cell types (typically Sf9, HEK293, or HELA cells).
Challenges around decreased transfection efficiencies for suspension processes and difficulty removing transfection reagents prior to harvest, which results in the need for careful design of downstream processing steps and related analytics, are being addressed through the development of new transfection agents with improved properties and performance, according to Ram. He also notes that other process intensification efforts, such as high cell density cultures, have also helped alleviate these issues. Furthermore, Ram believes that as the field progresses to using packaging and producer cell lines for viral vector production challenges associated with suspension-transfection platforms could become less significant.
In the meantime, in-house work at vendors and collaborations between vendors and end users are enabling viral-vector manufacturing to transition from adherent to suspension culture, according to Tan,. For instance, Cytiva has introduced subtle modifications to existing bioreactor platforms such as improved impeller and baffle designs and the use of different gas sparging sizes to increase transfection efficiencies. The use of computation fluid dynamics tools is also helping with optimization of flow rates and the location for plasmid addition to improve productivity.
Single-use bioreactor simplification
What if a single, single-use bioreactor design could be used for both suspension and adherent cell culture? That would help dramatically simplify the equipment needs for viral-vector manufacturers and alleviate the pressure surrounding the adherent versus suspension decision.
Univercells Technologies believes its NevoLine Upstream platform is just such a solution. The technology can accommodate both suspension- and adherent-based processes in a high-capacity, low-footprint manner. “The scale-X structured fixed-bed bioreactor and the NevoLine Upstream platform enable technology selection to be decoupled from process and scale considerations, providing an unprecedented level of flexibility,” Jourdan asserts.
In addition, because the NevoLine Upstream platform integrates cell culture with in-line midstream operations such as virus clarification, concentration, and ultrafiltration/diafiltration, it delivers a clarified concentrated bulk product ready for downstream processing. “The integration of midstream unit operations provides a considerable reduction in operational volumes prior to downstream directly impacting the overall manufacturing footprint. The automated approach improves process reliability via centralized process control and reduced manual operations,” states Jourdan. She notes that the single-use aspect is key to the ability to offer such a high level of flexibility that can accommodate a variety of products, processes, and scales.
Evolution of process development support
As single-use technologies have increasingly been adopted for commercial-scale viral-vector production, single-use bioreactor vendors have experienced growing demand for additional support. “At Pall, we are increasingly seeing customers reaching out for help with process scale up using our scalable manufacturing platform,” Cameau comments. The company’s Accelerator Process Development Services group can, she says, provide robust industrial-scale processes, including the development of a process analytical technology (PAT) approach, in six months to one year depending on the scope of the project.
In addition, Jourdan notes that a recent evolution in the industry sees more and more suppliers offering process-development support for handling the configuration of single-use systems based on customer process needs. “As a result,” she says, “technology providers are working to expand their portfolios to offer tailored, comprehensive solutions that help users circumvent challenges in terms of sourcing, validation, supply chain, and storage.”
Managing capacity demand
Sourcing has, in fact, become an issue with single-use components and systems, as well as many other materials used in biopharmaceutical manufacturing. Single-use technology has been a key enabler of the rapid ramp up of production for COVID-19 vaccines, and the sudden increase in manufacturing capacity demand has put more pressure on the supply of polymer resins used for the manufacture of single-use components, resins that are also used to produce clinical devices, personal protective equipment, masks, and packaging materials, according to Tan.
The result has been materials shortages and lead times up to nine months for some single-use items, which are impacting not just bioreactors, but all stages of production processes that now rely on single-use systems and has become a worldwide problem affecting clinical and commercial manufacturing, says Hitchcock. He does add that suppliers have invested heavily in new production facilities and are seeking to provide additional capacity, but it will take some time for any new production capabilities to come online.
With single-use bioreactors subjected to long lead times along with other consumables and single-use biocontainer assemblies, the best mitigation strategy is good planning, according to Wenling Dong, associate director, of process development at Lonza. Mohindra comments that increased levels of transparency around supply versus demand between companies can also help to address some of the areas of particular concern. Building more flexibility into processes to allow for use of materials and equipment from multiple suppliers and maintaining stocks of common materials are also important strategies being employed by Thermo Fisher Scientific, Ram observes.
One benefit of the current market situation, according to Jourdan, is that it has helped to raise awareness about the challenges that might come with relying on turnkey solutions from vendors, which in turn has led to changes in approaches to sourcing such as those noted by Ram. “In many cases,” she explains, “it can really pay off to already have alternatives identified to mitigate the risk of loss of continuity.” However, Jourdan realizes that such an approach does introduce additional complexities that require additional effort in terms of logistics, supplier qualification, and process validation, often making this solution difficult to put in practice.
Challenges for process monitoring
The ability to monitor and control the bioreactor environment has been shown to enable more reproducible processes and improve viral vector product titer and quality/potency including the empty/full ratio, according to Sanderson. Modern single-use adherent and suspension bioreactors generally provide a basic level of process monitoring and control including dissolved oxygen, pH, temperature, etc.
In addition, across the industry, Sanderson notes that vendors are investigating the use of novel PAT tools and sensors such as Raman spectroscopy and carbonate and glucose sensors to enhance process monitoring and control strategies.
Currently, however, Mohindra points out that the need for process monitoring means that external probes (for pH, dissolved oxygen tension) must be autoclaved and then inserted into the single-use vessels in an aseptic manner.
For smaller-scale vessels with less than 50 L of working volume, this operation must often be performed within a biosafety cabinet (BSC). It is also prone to error, according to Mohindra, because of the length of the probes and the limited space available.
At larger scales, disposable aseptic connectors do allow for autoclaved probes to be inserted into single-use vessels in a sterile manner outside of BSCs. These solutions, however, says Mohindra, are not ideal because improper installation or excess pressure can lead to junction compromise and failure of the process.
“These challenges highlight the need for single-use sensors that are integrated into single-use bags,” Mohindra concludes. He notes that some progress has been made, but there is still scope for the development of sensors that measure more process variables and in a more accurate manner.
Ram adds that ultimately real-time PAT might be the factor determining the choice of suspension versus adherent cell culture for viral-vector manufacturers. “Many of these technologies have been designed in the context of monoclonal antibody production, which is based on suspension platforms, so the adoption of existing PAT for viral-vector production will likely be easier for suspension-based viral vector platforms,” he says. The current lack of standardization of analytical methods for titer comparison, however, prevents effective assessment of available upstream technologies, according to Tan.
Challenges of working with large-scale single-use technologies
While the benefits of single-use technologies outweigh their disadvantages, there are some challenges unique to using disposable systems, particularly on a large scale. One of the key issues with implementing single-use systems for commercial-scale viral vector production is the generation of excessive biohazard dry/semi-dry waste, both in regard to the material itself but also its packaging, says Ram. The shelf life with respect to sterility of gamma-irradiated materials is another operational planning concern.
Large-scale, single-use systems also present challenges with respect to loading and unloading of the single-use bags into the equipment, higher risks for leaks, and the need for increased warehouse space to store single-use materials, according to Ram.
Looking ahead
Overall process robustness is the most crucial consideration, asserts Cameau, and work continues across the industry to overcome the challenges to commercialization of advanced therapies.
For instance, Cameau notes that leachables and extractables remain a concern any time single-use technologies are implemented. In addition, more understanding is needed regarding how the upstream manufacturing process impacts the full versus empty ratio of AAV. “There may be parameters there that could be used to better control this ratio, and we work towards those solutions daily,” she observes.
Furthermore, some vectors can be sensitive to shear, to light, to foam, and/or easily degraded by proteases released from the cells when entering apoptosis, according to Cameau. “When scaling up, it can sometimes be challenging to apply the correct scale-up strategy to get reproductive and scalable results, whatever the platform chosen. This difficulty can lead to a lower titer when transitioning from flatware to bioreactors, whether they are stirred-tank for suspension cultures or fixed-bed types. More needs to be learned about appropriate scaling strategies for specific vector types,” she says.
The emphasis on speed to market in the fast-growing gene therapy segment also presents an opportunity for technology providers to deliver valuable, sustainable solutions that help viral-vector manufacturers facilitate process optimization and scalability and reduce the cost of goods, according to Jourdan.
Single-use technologies, for instance, have limitations with respect to pressure and flow rates used to purify certain products, centrifugal forces, temperature, and—due to sparger designs—oxygen and carbon dioxide stripping rates, issues that Mohindra would like to see vendors address. Ram, meanwhile, would like to see improvements in mixing technologies for flatware used in seed trains or entirely new means of building up seed trains to achieve greater uniformity and consistency of cultures.
Adoption of advanced PAT tools not only improves process monitoring and control, but enables process intensification strategies, including continuous manufacturing (perfusion), in both adherent and suspension systems, observes Sanderson. A perfusion strategy, according to Cameau, can help reduce vector titer loss due to external aggressions in the bioreactor.
Finally, Tan notes that there is a general consensus among end users that the next innovation in gene therapy and gene-modified cell therapy will be achieving stable producer cell lines for viral vector production combined with the use of large single-use bioreactors for increased titers and reduced costs.
References
1. R. Legmann, Cytotherapy, 22(5) Supplement S151 (May 2020).
2. Clive Glover, “Rapid Development and Industrialization of a COVID-19 Viral Vector Vaccine Manufacturing Process,” Pall Biotech Blog, Jan. 27, 2021.
About the author
Cynthia A. Challener, PhD, is a contributing editor to BioPharm International.
Editor's Note
This is an expanded version of the article published in the August 2021 issue of BioPharm International.