The Complex Calculus of Viral-Vector Facility Design: Considerations for Commercial Gene-Therapy Manufacturing

Guided by standardized production and purification platforms, developers of monoclonal antibodies (mAbs) and other recombinant-protein therapeutics can take a relatively straightforward path when designing and establishing facilities for commercial-scale operations. By contrast, designing gene-therapy (GT) facilities involves a more complex calculus. Multiple approaches are available for producing transgene-bearing viral vectors at commercial scales. GT developers must choose carefully, knowing that each manufacturing strategy raises distinctive advantages and limitations for process scalability, equipment implementation, storage, material and personnel flow, and ultimately, facility output. The plurality of manufacturing strategies — and the possibility of further platform advancements — requires GT companies and their facility-design teams to be forward-thinking in their planning. All the while, GT facilities must be engineered for good manufacturing practice (GMP) operations and work with viral vectors, including measures for segregating virus-positive areas and preventing cross-contamination.

Allan Bream (senior fellow in biopharmaceutical processes) and Brita Hobmann (process engineer) of facility-engineering firm CRB have over 40 combined years of experience with biopharmaceutical-facility design and construction, with a particular focus on sites for GTs and gene-modified cell therapies. This past summer, I corresponded with Bream and Hobmann to enumerate some of the many variables involved in establishing commercial GT capabilities. After describing implementation trends for production platforms and related equipment and operational requirements, the duo explore strategies for engineering contamination controls into a facility plan and for optimizing facility use and output. Our discussion also highlights considerations for GT companies that are deciding between establishing internal manufacturing capabilities or outsourcing vector production to contract service providers.

Laying a Foundation for Gene-Therapy Manufacturing

How would you describe the landscape of GT manufacturing? What production approaches currently prevail?

Bream: Adenoassociated viruses (AAVs) remain the most common gene-delivery vehicles for in vivo gene therapy. Lentiviral vectors (LVVs), adenovirus vectors (AdVs), and retrovirus vectors (RVVs) are other available delivery systems. A key consideration is that AAVs do not integrate into a patient-cell’s genome, which makes them effective for in vivo applications but can limit treatment effectiveness over time. LVVs and RVVs do integrate into host-cell genomes and thus are ideal systems for ex vivo manipulation of dividing cells.

Triple transfection still leads the way for gene transfer into cells used to produce recombinant AAV. Three DNA plasmids are incubated in sequence with a production cell line, usually human embryonic kidney 293 (HEK293) cells. One plasmid contains a pair of essential AAV genes, the second has a transgene specific to the GT, and the third is a helper plasmid.

Hobmann: Common themes in GT manufacturing today include scalability of suspension cell culture, the need for single-use (SU) equipment, and the desire to balance in-house and outsourced activities. Typically, GTs are capping at 500-L production volumes in SU bioreactors. However, some companies are operating two 200-L bioreactors because production at that scale often results in more effective transduction than what is observed in 500-L systems.

SU equipment continues to dominate the GT market. Such products require relatively small process volumes compared with those for traditional mAb production. SU enables quick cleaning of biosafety-level (BSL) materials through disposal of spent process-contacting materials. And SU technologies are conducive to speed to market, which is advantageous for contract development and manufacturing organizations (CDMOs) that are trying to attract clients or for start-ups that are producing clinical material.

Discussions continue around the “make or buy” question. GT developers must decide what to invest in now and what to outsource to save capital. Allan and I have found that many CDMOs in the GT market handle the entire process for their clients. An outsourced approach provides biotechnology companies with an option to make their viral vectors off site — e.g., to support production of genetically modified cell therapies, such as those based on chimeric antigen receptor (CAR) T cells.

Alternatively, particular aspects of GT production may be beneficial to outsource initially and then bring back in house. Solution preparation is among the most common activities to be outsourced initially. By purchasing preformulated medias and buffers from a qualified third-party supplier, GT developers can save space and commodities such as water for injection (WFI). When building their own manufacturing facilities, GT companies often have conversations about insourcing and outsourcing to customize investments for current and future process needs.

How do different production approaches influence GT facility design?

Bream: Suspension culture facilitates processing better than anchorage-dependent operations do. Operating suspension cell culture in a rocker or agitated bioreactor system also offers better gas and nutrient exchange, therefore providing cells with optimal growth conditions. But scale-up is limited by current SU bioreactor equipment. Some bioreactors work at volumes up to 4000 L. However, most commercial processes are capped at 1000 L.

Although suspension cell culture is relatively straightforward, having an effective transfection method is key. With higher cell densities and volumes, transfection efficiency decreases. As Brita explained, that factor often compels manufacturers to use two small production bioreactors instead of one large unit.

Hobmann: Regarding anchorage-dependent platforms, starting with an adherent process can bring some clinical advantages, but unfortunately those benefits often are overcome with disadvantages during larger, commercial-scale operations. Compared with the many available options for suspension bioreactors, choices for large-scale adherent-culture equipment remain limited. Thus, commercial-scale processes based on adherent platforms max out around 500 m2, and increasing capacity requires adding equipment to scale out. Harvest steps also may require dislodging cells from the culture surface between media exchanges. In that regard, adherent bioreactors act like perfusion processes, which require larger media volumes than do suspension processes operated in batch mode. Large media volumes entail significant solution-preparation demand and require ample space to make, store, and transfer those solutions.

Suspension- and adherent-cell platforms have their own pros and cons. Sometimes GT developers will have a business driver to develop flexibility for using both approaches. Doing that is achievable with strategic facility planning and foresight.

What considerations do packaging/helper and producer cell lines present for materials use, equipment configuration, and facility layout?

Hobmann: Historically, packaging/helper cell lines have dominated the viral-vector space, and stable producer cells have been an aspiration for GT companies. In a survey for CRB’s 2020 Horizons: Cell and Gene Therapy report, 65% of respondents said that their companies were developing or intended to develop stable producer cells because of the potential for lowering costs and improving scalability (1). In processes based on producer cell lines, transgene conversion occurs earlier in the process, at the start of the growth cycle. That reduces the need for costly plasmids compared with quantities needed for processes based on packaging cell lines and triple transfection at large production volumes. Producer cell lines also could improve vector yield and batch consistency because they are pretransfected with the needed genetic elements, effectively eliminating the need for triple transfection. The challenge will be to scale up these “holy grail” cell lines to be commercially viable. Advancements in producer cells will be exciting to watch develop over the next few years.

Establishing Facility Controls

Which biosafety requirements come with viral-vector manufacturing?

Bream: Cleanroom classification for viral-vector manufacturing depends on whether a process is open or closed. Typically, inoculum preparation occurs in a biosafety cabinet within a grade-C room background (see opening photo). Closed upstream operations (e.g., cell-culture expansion and viral-vector propagation) are carried out in a grade-D environment. When transitioning to downstream operations, process volumes decrease, and the need for truly closed, small-scale operations drives the cleanroom classification to grade C.

Bulk viral-vector filling occurs within a grade-C space, then is transferred to drug-product filling within the same facility. Such areas are subject to EU GMP Annex 1 (2), and they often are segregated from drug-substance manufacturing areas. Small-batch, manual filling of vials requires a grade-A environment with a grade-B background. That, of course, drives the need for airlocks and gowning protocols that are compatible with aseptic operations. For large-scale operations, regulators prefer automated vial filling within a high-efficiency particulate air (HEPA)–filtered isolator system. Being closed systems, isolators allow for a grade-C room background.

The 2023 update to EU GMP Annex 1 calls for manufacturers to implement contamination control strategies (CCSs) along with preuse, poststerilization integrity testing (PUPSIT) while following quality risk management (QRM) principles (2).

Hobmann: Viral vectors typically are classified as BSL-2 materials because they are “agents associated with human disease and pose moderate hazards to personnel” (3). BSL-2 materials entail special handling considerations, such as safety-exhausting bioreactors, isolators, and cleanrooms. A GT manufacturer should perform a risk assessment to evaluate the likelihood and impact of viral-vector escape from a functionally closed system. For instance, what is the probability of a bioprocess bag leaking? What would be the effect of recirculating air from vector-positive to vector-free areas? Such questions should be discussed with the goal of establishing engineering controls to minimize risk.

Beyond closed process equipment, heating, ventilation, and air conditioning (HVAC) systems play a critical role in GT manufacturing, helping to protect personnel and the environment. Air-handling zoning, HEPA filters, and room pressurization are a few mechanical elements that provide another level of control when used properly. If a viral-vector breach occurs within a room or isolator, chemical decontamination is recommended using vaporized hydrogen peroxide or an equivalent agent. Decontamination approaches should be discussed early in facility design and as part of a risk assessment.

How else can manufacturers ensure segregation of vector-positive areas?

Bream: Several methods provide layers of protection to prevent cross-contamination.

As in other biopharmaceutical facilities, personnel, materials, and waste airlocks serve as transitions into and out of process areas. However, for GT operations, unidirectional flows of personnel, materials, and waste are preferred rather than the bidirectional flows typically used in facilities for mAbs and other biopharmaceutical products.

A two-corridor facility approach is typical, and that aids unidirectional flow. Personnel and materials enter into process areas through “supply” corridors, while waste exits and personnel return to locker rooms through a return corridor.

Host-cell expansion (before infection of a production bioreactor) is segregated from the viral suite.

Validated chemical sanitization/wipe-down and decontamination steps are performed at entry/exit points into/out of process suites.

Generally, viral-vector suite HVAC pressurization is negative relative to that in facility supply and return corridors, helping with containment and preventing risk of cross-contamination.

Hobmann: In addition to the methods that Allan mentioned, closed processing serves as another level of engineering control for vector containment. Luckily, closed-process technology is not new to the industry, and GT manufacturers can choose from a variety of SU offerings to maintain closure. Aseptic connectors, tubing welders, and preassembled manifolds are a few components that aid in maintaining closure for an SU-based platform. Specific steps might need to be functionally open, in which case operations should be handled in higher-classified areas such as grade-A isolators or biosafety cabinets.

For small batch volumes, few options are commercially available for closed, GMP-grade equipment options. Already small GT production volumes are reduced further after downstream processing. Thus, purification steps in a GT manufacturing process might require high room classifications or secondary containment measures to mitigate risks associated with functionally open steps.

Forward-Thinking Facility Design

What kinds of mistakes do GT manufacturers make when establishing production facilities?

Bream: Early stage GT companies feel pressure to produce enough material quickly to support clinical trials. Thus, companies often make process-development (PD) compromises. With not enough time to optimize a given process, companies are forced to into a decision: Aim for a “perfect” process or settle for an “adequate” manufacturing approach. The problem is that modifying a process later can be disruptive and expensive. As a program advances to phase 3 clinical trials, it might reach the point at which process changes cannot be made.

Problems also come from using process methods that cannot be scaled up. PD often begins with open processes in research-scale equipment that does not perform robustly. Mitigating such deficiencies in future processes can be expensive, though. For instance, if an open step cannot be closed, then it will need to be housed in a cleanroom with a high classification. That would increase both facility capital costs and operational expenses.

Effective technology transfer from a partner CDMO can be troublesome. At some point, a GT developer might want to transfer a process from its CDMO back to its own manufacturing facility. During that process, communication and documentation regarding the process and associated methods will be vital, as will details about the quality control (QC) testing approach. Clear goals, intermediate objectives, and regularly scheduled meetings with progress tracking will be key to success.

Hobmann: Commercial GT products are relatively new to the market compared with mAbs and traditional modalities. Thus, there is no “gold standard” for GT manufacturing, and each company is developing its own best practices. Those different approaches can lead to creative innovatives, but companies can run into obstacles along the way if they lack a forward-looking mindset. For instance, that takes us back to the question of whether to use an adherent- or suspension-based platform and to the importance of considering long-term success with each format. GT developers need to balance speed to market with scalability for future demand.

The past five years have witnessed a significant increase in GT companies vying for established commercial facilities. From this surge comes industry best practices and expert guidance. When building a facility, look to such guidance and to experts who have seen these obstacles before.

How can manufacturers ensure that they are optimizing facility use and output?

Bream: There are myriad factors to consider, including flexibility and redundancy, process closure, risk tolerance, and issues related to SU technologies. Companies also need to consider the basic make-or-buy choice carefully.

Regarding flexibility, where possible, specify equipment that can support a breadth of process capabilities and product types. And whether for routine calibration, part replacement, or unexpected failure, process equipment requires periodic maintenance and will be taken out of service. Manufacturers need to establish strategies to ensure that equipment maintenance does not interrupt production significantly. That includes N + 1 redundancy for key equipment.

In terms of make-or-buy decisions, early-stage and prerevenue companies have tight budgets for personnel and capital equipment. Such companies should consider outsourcing activities such as media and buffer preparation (which also would help to decrease storage requirements) and QC release testing. Of course, outsourcing will require added planning to ensure that a vendor can support given manufacturing requirements.

GT companies should work with suppliers of SU technologies to ensure that such equipment can support a given manufacturing process and fit into a particular facility design. GT developers also should investigate how effectively equipment can interface with a facility’s process-control system.

Developing a closed process will enable parallel processing without prompting additional regulatory scrutiny.

As far as risk tolerance, companies must explore how comfortable they are with risk concerning GMP process flows, HVAC capabilities, equipment redundancy, open processing, and so on. Managers should get key stakeholder buy-in as early as possible.

Hobmann: The key to optimizing facility efficiency and throughput is planning. That seems simplistic, but it is important that GT companies take time at the start to identify facility drivers and capacity goals, with key stakeholder buy-in. Initial planning will provide a road map throughout facility design, and detours should be addressed as a team.

Simulation modeling can be used to inform early facility-design stages by pressure-testing GMP flows, process schedules, and equipment use. Process simulations also can test ratios of upstream, downstream, and filling functions to prevent schedule bottlenecks and reduce equipment downtime. Other use cases include evaluation of parallel trains to achieve throughput projections and warehouse capacity modeling to support incoming consumables. Planning sets the framework for an optimized facility and heads off challenges associated with significant course-correction.

How can GT manufacturers help to ensure that their facilities will address future needs?

Bream: At early project stages (preferably during feasibility assessment or preconcept activities), defining facility-user requirements for future needs is essential. Process modeling and exploration of different facility approaches are important in that regard. Those activities enable testing of different facility conditions, capabilities, and concepts before serious design work begins. Performing site master planning alongside manufacturing assessment can be valuable as well, considering site amenities such as parking, administrative buildings, and other needed supports. A strategic approach to intra- and extra-building planning will enable streamlined, well-conceived expansions.

Hobmann: Continuous innovation in biomanufacturing technology and science means that change is inevitable in the GT industry. The key is to be adaptable to such advances throughout the life cycle of a manufacturing facility.

Flexibility has become an industry buzzword, and it means something different for each person. If we look at one production suite as an example, a number of approaches could make that suite “flexible” for future operations. Gas utilities, data connections, and electrical outlets could be standardized and spaced around the suite for different equipment configurations. Ceiling heights and room footprints could be designed to facilitate future scale-up/out.

A GT developer should work with its suppliers and facility engineers to define what flexibility means for its own purposes. To prevent overdesign, the team should not deviate from that definition. GTs are the product of cutting-edge science, and they will continue to evolve and change as the science does. Being flexible — within defined goals — will be key to mitigating risk and extending facility longevity.

References

1 Maestre N, et al. Horizons: Cell and Gene Therapy, 2020. CRB: Cary, NC, 2020; https://go.crbgroup.com/horizons-atmp-thanks-0.

2 EU Guidelines for Good Manufacturing Practice for Medicinal Products for Human and Veterinary Use: Annex 1 — Manufacture of Sterile Medicinal Products. European Commission: Brussels, Belgium, 2022; https://health.ec.europa.eu/system/files/2022-08/20220825_gmp-an1_en_0.pdf.

3 Biosafety in Microbiological and Biomedical Laboratories (sixth edition). Meechan PJ, et al., Eds.

US Centers for Disease Control and Prevention: Atlanta, GA, 2020; https://www.cdc.gov/labs/pdf/SF__19_308133-A_BMBL6_00-BOOK-WEB-final-3.pdf. https://www.cdc.gov/labs/pdf/SF__19_308133-A_BMBL6_00-BOOK-WEB-final-3.pdf

Brian Gazaille, PhD, is managing editor of BioProcess International; [email protected]; 1-212-600-3594. Allan Bream, MSChE, is senior fellow in biopharmaceutical processes, and Brita Hobmann, EIT, is a process engineer, both with CRB.