Continuous Processing for Viral Vectors

Continuous processing for monoclonal antibodies (mAbs) and recombinant proteins has been shown to enhance the productivity of certain unit operations, such as cell culture and capture chromatography. It is particularly advantageous for labile products sensitive to process conditions. It has the potential to increase process consistency and product quality and yield, reduce process times, eliminate other wasteful activities, and contribute to reduced manufacturing costs. Given these benefits, continuous processing could potentially address the inefficiencies and poor recoveries seen in viral-vector production. The properties of viruses and the nature of transient transfection make moving to continuous operations a challenge. The focus, therefore, is more on process intensification rather than continuous processing, per se.

Batch processing challenging for viral vectors

Batch production of viral vectors is largely achieved via transient transfection of human embryonic kidney (HEK) 293 cells using multiple plasmids. This process typically suffers from low titers. Adherent cell culture is most common, but there is a movement toward suspension-based processing, for which the cell densities are quite low compared to those observed for recombinant proteins and mAbs.

Although adeno-associated viruses (AAVs) are more stable than lentiviruses (LVs), for AAV vectors the transfection step results in the generation of large percentages of “empty” or “partial” viral capsids that do not contain the full gene of interest. LV vectors, meanwhile, are susceptible to degradation by mechanical stress and changes in pH and temperature. These factors contribute to low recoveries during downstream purification.

Learning from continuous processing of antibodies

The production of proteins is less difficult than the manufacture of viral vectors due to the greater complexity of vector products, which involve multiple proteins in a complex three-dimensional structure encapsulating DNA, says Guillaume Freund, scientific support manager at Polyplus. Viral vectors are also more toxic to the cells used to produce them than are proteins.

Tony Hitchcock, a technical director with Charles River, adds that viral vectors are produced at product concentrations several orders of magnitude lower than those for protein products and can be quite variable. In addition, AAV vectors are not secreted like proteins, while LV vectors suffer from low product stability. AAV vectors also require separation of full and empty capsids, which is a more complex separation than that required for proteins and mAbs.

“Even so, the knowledge and expertise built around antibodies can serve as an important starting point for the development of continuous viral-vector processes,” Freund comments. The time-frame will not be short, however, given the numerous hurdles that must be overcome.

Although transfection to generate viral vectors is different from protein expression, the cells used for the transfection step must be grown to the appropriate cell density. “Similar to the so-called N-1 bioreactor in the manufacturing of recombinant proteins or monoclonal antibodies, the cell mass is actually the product of the cell growth step and hence a perfusion solution could be appropriate, using the same solutions as developed for intensifying the cell culture process for recombinant proteins and monoclonal antibodies,” says Marc Bisschops, VP of manufacturing science and technology at Pall Corporation.

Downstream purification steps for viral vectors are similar to those for proteins and mAbs, and therefore, continuous solutions developed for the latter could in principle be applied to viral-vector processing as well, according to Bisschops. He adds that the critical process parameters would not be radically different and the expectations around critical aspects such as bioburden control would remain exactly the same, but the scale may not always be appropriate.

In general, while most technologies applied to viral-vector processing have been inherited from other modalities, Tania Pereira Chilima, CTO, Univercells Technologies, points out that some technologies have been purposely designed to enable integrated and continuous processing in virus production. She highlights Univercells Technologies’ upstream-midstream biomanufacturing platform, which integrates and automates cell culture, transfection/infection or induction, and virus production in a fixed-bed bioreactor run in batch recirculation or perfusion mode with subsequent unit operations to deliver a clarified and concentrated harvest prior to downstream processing.

Many issues to address before continuous processing possible

Most development work for continuous processing focuses on traditional biologics that are produced using platform processes where there is deep process understanding and greater certainty of process behavior. In the field of viral vectors, which are still considered new product modalities, that knowledge base is lacking, according to Joe Makowiecki, enterprise solutions director of business development at Cytiva.

“The susceptibility of viruses to environmental and processing conditions and the fact that viruses are larger in size than traditional proteins and mAbs, along with limited industry experience and expertise, make implementation of process intensification solutions more difficult for viral-vector processes,” Makowiecki says.

The transfection step in particular, as it is currently performed, will be a challenge for transforming the upstream process of viral-vector production into a fully integrated continuous manufacturing platform, observes Bisschops.

For instance, Maxime Dumont, a product manager with Polyplus, points out that the transfection mix generated from plasmid DNA and transfection reagent has a limited stability, and it is critical to determine the optimal timeframe of addition. “Transfection reagents that yield much more stable transfection mixes are required to perform transient transfection in a truly continuous operation,” he says. Better stability of complexes combined with biodegradable transfection agents will be needed to extend the time that transfection processes can proceed, adds Freund.

The issues posed by transient transfection could potentially be solved with the development of producer cell lines that can produce vectors without the need of transient transfection. Generating high-titer stable producer cell lines is quite difficult, however, according to ViroCell Biologics’ process development director, Carlo Scala.

In addition, some viruses, notably AAV and adenovirus, are produced intracellularly, creating the need after the transfection step for chemical lysis of the cells followed by treatment with nucleases to degrade cellular DNA. Performing such a step may not be possible in a continuous fashion, notes Hitchcock. He adds that to separate empty and full capsids, a number of producers still use ultra-centrifugation, and while there are continuous ultra-centrifuges available, achieving the required purities may be challenging.

Another challenge highlighted by Bisschops is the current scale of viral-vector manufacturing. Applications for these products typically target somewhat smaller patient populations, and hence batch sizes are much smaller than for most mAb therapeutics and require smaller manufacturing systems that are still current good manufacturing practice (CGMP)-compliant. “With process intensification and/or continuous bioprocessing, the challenge to realizing such systems may become more significant. Single-use components such as connectors and sensors are not always available in the right size to support the appropriate scales,” he explains.

Potential of perfusion processes for transient transfection

Upstream viral-vector production, says Scala, could vastly benefit from continuous manufacturing if solutions can be developed. For secreted, extracellular labile products (e.g., LV vectors), continuous processing would enable continuous virus collection, clarification, concentration, and cooling, agrees Chilima.

While inoculation and virus production don’t always lend themselves to translation into continuous processes, there are technologies being investigated as enablers of process intensification. The use of fixed-bed bioreactors, for instance, allows growth of adherent cells in very compact, fully controlled bioreactors rather flatware. “The transfection process and cell lysis can both take place in the bioreactor, minimizing operator interventions and maintaining a well-controlled (and functionally closed) state throughout the viral-vector expression process,” observes Bisschops.

Other viral-vector manufacturers are moving toward continuous processing by switching from batch to fed-batch mode, according to Dumont. “The use of perfusion processes as a way to increase biomass is a promising means for intensifying the transfection process,” he notes. The key is exchanging out old media, which affords more optimal transfection conditions. While the use of perfusion will depend on the route of production, Hitchcock explains that the use of perfusion-based processes can increase cell densities ahead of viral infection or plasmid transfection to increase productivities or prior to induction of production in stable cell lines, both of which lead to higher productivity.

One challenge to this approach that must be tackled is the fact that HEK cells tend to clump or aggregate at higher cell densities. Cell aggregation can significantly limit the efficiency of the transient transfection step. “Improvements and optimization of culture media along with bioreactor parameters will be necessary to reduce or prevent aggregation at higher cell densities,” Makowiecki states.

Opportunities for intensifying downstream operations

Given the sensitivity of many vectors, particularly enveloped ones such as LV vectors, to degradation by shear forces, high salt concentrations, and temperature, the timing of downstream processes is crucial to avoid product loss. Consequently, Scala emphasizes that continuous upstream vector production and harvest must be combined with rapid and efficient downstream processing.

There are, in fact, vector purification steps that, as is the case for recombinant proteins and mAbs, may be well-suited for process intensification, enabling increased production and recovery, according to Makowiecki.

Even for intracellular serotypes, which might benefit less from upstream continuous processing because the cells must be lysed to collect the product, which is performed at the end of the process in batch mode, continuous processing in downstream operations would be beneficial, notes Chilima. “The use of continuous processing techniques such as simulated moving bed chromatography (SMB) may lead to higher efficiency of capacity utilization, faster processing times, and lower reagent usage, among other benefits,” she says.

Chromatography can also be intensified, Makowiecki observes, by using large-pore separation media such as those found in membrane and fiber-based devices that can more efficiently process larger virus molecules and thus allow rapid cycling. “The key to application of this type of solution is to ensure that the holistic system (the pump, the flow path, and the device) does not create disruption due to higher flow rates,” he comments.

Chromatography steps that involve a gradient elution are notoriously more difficult to translate into continuous processes, however, warns Bisschops. Even so, he stresses there still are many opportunities for process intensification in downstream viral-vector manufacturing. A specific example would be the polishing steps in which membrane adsorbers can be used to effectively separate full and empty capsids. Here again, the larger pore sizes of membrane adsorbers work effectively for larger products such as viral vectors because they mitigate slow diffusion behavior observed with traditional resin/bead-based chromatography media.

Other technologies such as single-pass tangential flow filtration (TFF) to reduce process volumes and buffer exchange ahead of chromatography operations may well be suitable for downstream viral vector purification processes, according to Hitchcock. “For the recovery and purification steps, there is a key drive to reduce process times and reduce product losses, and these techniques offer the potential to do that,” he explains.

Bisschops does caution, though, that many of the filtration steps in downstream viral-vector processing are operated in flow-through mode anyway, and sizing these filters to accommodate an entire batch rather than configuring these steps in a cyclic continuous mode may make a lot of sense.

Stable producer cell lines needed

The transient transfection step by its nature does not make conversion to continuous processing easy. The need for an unstable transfection-reagent plasmid mix and the fact that the viral products are toxic to the HEK cells used for vector manufacturing are just two features of the process that make continuous operation difficult.

To make continuous processing possible, therefore, there is a need for stable producer cell lines that can produce viral vectors at high titers over long periods of time, according to Scala. Stable cell lines, notes Hitchcock, offer the potential for increased production levels but also more consistent productivity levels.

“Having a large enough bioreactor with stable producer cell lines that provides enough material for subsequent downstream purification in batches would allow running of an effective continuous [process],” Scala says. In turn, continuous vector manufacturing by stable vector-producing cells could substantially improve yields and reduce costs and manufacturing times, adds Farzin Farzaneh, Chief Scientific Officer, of ViroCell Biologics.

A number of groups, including the innovations and process development teams at ViroCell, are working on the generation of stable, high-titer vector producing suspension cells, according to Farzaneh. Indeed, a few stable cell lines have been developed for LV vectors and many researchers are working on AAV cell lines as well.

“It’s a natural progression within the industry to get to stable producing cell lines for viral vectors; this pathway was already followed for mammalian cell culture, leading to the industry’s most frequently used mammalian cell line, Chinese hamster ovary. Viral vectors are heading in the same direction,” remarks Makowiecki. The best strategy, Scala believes, is to keep investing in the development of stable producer cell lines and simultaneously design/develop vectors that can be manufactured to high titer and with higher physical stability.

Advanced purification and filtration technologies

The rapid growth in demand for viral vectors and other new modalities has led to growing interest in the development of fit-for-purpose downstream purification solutions for these challenging molecules.

“In terms of filters, there is scope to look to formats to reduce product losses with these large vector products, especially for the processing of enveloped LVs, where significant product losses can be encountered,” Hitchcock comments. He reiterates that single-pass TFF technology from Pall has been shown to be applicable for AAV vectors, while Repligen’s tangential-flow depth-filtration systems give high recoveries for LV products.

There has also, according to Makowiecki, been a renewed focus on fiber-based technologies. “Hollow-fiber technologies are experiencing a rebirth, with applications expanding beyond the traditional vaccine space to use with new product modalities for TFF, for which this process is a focal point,” he says.

Newer, nanofiber technologies are also being introduced to the market that have been designed for use with viral vectors and other large molecules such as plasmid DNA and messenger RNA. “These fiber technologies can be easily employed to intensify chromatographic processes,” Makowiecki observes.

Novel fiber-based chromatography formats such as those from Cytiva and Astrea are very well-suited to large biological complexes and offer significant reduction processing times in terms of both flow rates and improved binding and elution kinetics, which are needed for high-throughput processing platforms, Hitchcock observes.

Makowiecki anticipates such technologies being first used for batch processing of new modalities, with the learnings gained in these operations then applied, in combination with the knowledge developed around continuous processing for traditional biologics, to the development of continuous solutions in the new-modality space.

In addition to the new fiber technology itself, there is also significant work being done to develop new ligands for attachment to the fibers, as well as for use with more traditional chromatography resin-based beads and membranes. “We are already beginning to see many different types of ligands coupled with devices based on not only larger-pore media, but more traditional chromatography media as well. These affinity ligands are specific for particular vectors and vector serotypes. Eventually, though, more broad-based ligands that are specific for vectors and provide good performance for many different serotypes will eventually emerge, as happened with protein A for mAbs,” says Makowiecki.

Appropriate analytical solutions essential too

Continuous processing is not possible without access to real-time, in-process analytical data. The largest hurdle to continuous manufacturing of viral vectors, in Bisschops’ view, could well be the analytical side. “Appropriate in-process controls with rapid turn-around times are a necessity to further drive process intensification,” he states.

Hitchcock agrees that one of the greatest challenges will be process analytics, with regard to both the ability to determine product concentration during operations and putting the required controls in place to ensure process consistency and stability.

One particular issue highlighted by Hitchcock relates to the ability of current process analytical technologies to detect the very low protein concentrations encountered in viral-vector manufacturing. “There is a requirement to improve online monitoring and control systems generally, but this capability will be critical for continuous processing of viral vectors,” he observes.

On the positive side, Bisschops does note that there are developments being made that will help enhance manufacturing process understanding and quality standards while simultaneously reducing variability in manufacturing and improving—as a result of that—product quality consistency. “These developments are a key to success for further improving process efficiency and driving process intensification for viral vector manufacturing,” he concludes.

Potential regulatory benefits

Improved process analytical technologies will also be crucial for continuous viral-vector manufacturing from a regulatory perspective, according to Hitchcock. “The data generated from in-process analytics will be essential for showing product consistency during the operation, both with regards to purity but also potency and comparability to material produced through batch processes,” he explains.

Fortunately, continuous processing by itself is encouraged by regulatory authorities and therefore provides opportunities for viral vector producers to align the development and characterization of manufacturing platforms with the regulatory expectations, according to Bisschops.

“If we take a more holistic perspective on process intensification, saving time seems to be an area where good science and learnings from regular biopharmaceutical product development can help a lot. The use of the principles of quality by design (QbD), for instance, can address regulatory challenges early on in the clinical development of viral vectors. With such approaches, the time to market can probably be reduced while reducing risks and regulatory challenges at the same time,” Bisschops says.

In that vein, Pall published a framework for the use of QbD for AAV vectors to support scientists in addressing regulatory aspects in an early stage of process development (1).

All about improving efficiency

While continuous processing can provide measurable benefits, there clearly are significant hurdles before viral-vector manufacturing can be deployed in fully continuous mode. It is even possible that continuous processing may not be an appropriate solution.

“The driver behind moving to continuous operations is to improve efficiency, productivity, and overall quality. Many times, this can be achieved through a combination of optimization, new product introduction, and intensification. Achieving the goal of improved efficiency, productivity, and quality should be the focus, rather than implementing continuous solutions,” Makowiecki emphasizes.

It is likely, agrees Hitchcock, that continuous processing will be introduced in a stepwise manner, with specific unit operations becoming more intensified as enabling technologies become available. “Movement towards process intensification will take place based on technical achievements that are associated with demonstrable operational benefit in terms of costs yield, and product quality,” he notes.

Developing stable producer cell lines that yield as much product as transient transfection processes and do so over time must take place before continuous upstream processing can become a reality, adds Dumont. He agrees, though, that fully continuous operations may never work for viral vectors, and that other approaches to boosting productivity may be necessary.

Indeed, continuous viral vector manufacturing may not be an immediate focus area as the driving force may not be strong enough to address the hurdles involved, Bisschops adds. Instead, he believes an approach that leans more toward a holistic perspective of process intensification, including reducing timelines for development, technology transfer, and regulatory approval may add more value.

“In my opinion, we should not focus so much on continuous manufacturing as a goal by itself. Continuous manufacturing is a subset of process intensification. This has many angles, but is not limited to reducing manufacturing costs. It also includes smarter ways of working to shorten development timelines,” comments Bisschops. He adds that existing technologies such as fixed-bed bioreactors for adherent cell culture and membrane adsorbers for chromatographic purification can already be used to support process intensification for viral vector manufacturing.

Reference

1. P. Cashen and B. Manser, “Quality by Design (QbD) for Adeno-Associated Virus (AAV): A Framework for a QbD Assessment for AAV Products Within the Chemistry Manufacturing and Controls (CMC) Documentation,” Pall Corporation White Paper, September 2021.

About the author

Cynthia A. Challener, PhD is a contributing editor to BioPharm International.

Article details

BioPharm International
Vol. 35, No. 8
August 2022
Pages: 31–35

Citation

When referring to this article, please cite it as C. Challener, “Continuous Processing for Viral Vectors,” BioPharm International 35 (8) 2022.