Testing for Adventitious Agents in Viral Vector Production

Viral vectors must be free of contaminants, such as adventitious viruses introduced via raw materials and helper viruses introduced purposefully to stimulate expression of the desired viral product. Viral clearance used with incoming raw materials and, where possible, viral-vector products combined with testing is essential. Because conventional viral clearance methods used for protein biologics may degrade or destroy viral vectors, assessment of the potential risk for contamination and the implementation of prevention and control strategies are essential.

The most common source of viral contamination is usually traced back to animal-derived raw materials, such as certain cell-culture media (e.g., serum) and trypsin, according to Artur Padzik, adeno-associated virus (AAV) production manager at Biovian, a contract development and manufacturing organization (CDMO). Non-animal-derived raw materials can be an issue as well, particularly glucose if a chemically defined medium is used. A risk-based approach to process development combined with extensive testing of raw materials (and often the application of virus removal steps at raw material entry points), in-process materials, and final products is essential to ensuring the safety of viral vectors.

Similar to other biologics but challenging

The requirements for adventitious agent testing for viral vectors overall are consistent with those performed for other biological therapeutic products. As with other biologics, too, selection of appropriate testing methods is guided by the need to mitigate the risk of viral contamination taking into account the materials and processes used.

“Viral vectors are biologics, and as with all biologics, the concern around adventitious agents must be addressed with a thorough risk assessment that [considers] control and probability of contamination. The principals from the three-pronged complementary core strategy (material selection, testing, and clearance assessment) to prevent and minimize adventitious agents can and should be applied to viral vector manufacturing processes and products,” says Audrey Chang, executive director, Rightsource and Biologics at Charles River.

Direct adoption of strategies used for recombinant proteins and monoclonal antibodies may not be possible, however, Chang adds, due to biological differences between viral vectors and large protein molecules. Indeed, it may be necessary to modify standard assays or use alternative methods because viral vectors are viruses themselves and could interfere with virus detection methods, notes Leyla Diaz, a technical consultant in the Technical and Scientific Solutions unit of MilliporeSigma.

Specifically, adventitious agent detection can be challenging if cell-based assays are being used in combination with live-virus vaccines, replication-competent gene therapy vectors, or vectors exhibiting the cytopathic effect (CPE), Padzik observes. This issue arises largely because human cell lines are used for cell-based assays, and human cells are susceptible to human viruses. Lytic viruses such as adenoviral products, for example, will infect human and primate-based cells and can cause invalid assays. Robert Schrock, director of bioassay services at Lonza Houston, adds that even if adenovirus is only conditionally replicating and does not replicate, cellular toxicity may occur.

In these cases, neutralization of interfering and abundant product viruses is required, says Padzik. There are three approaches that can be taken, according to Schrock. One is to block antibodies specific to the virus product to “soak up” the product as much as possible without interfering with the detection of other adventitious viruses. Alternatively, the sample can be diluted so as to avoid overwhelming the detection cells. Often these two approaches are used in combination. The third option is to pull the test sample from an uninfected parallel satellite culture that is split off the bioreactor just prior to transduction with the virus bank. “For oncolytic vectors, FDA guidance mentions the use of #3 as acceptable. The first two have limitations, such as with test #2, where the value of the test can rapidly decline due to sample dilution,” Schrock comments.

In general, the key is to select the appropriate test method for a given product. For example, Sophia Nguyen, director and head of quality control at Lonza Houston, points to the selection of cell lines for infectivity assays as being important. “Viruses differ in their ability to infect host cell types.” For the endpoint assay used to determine whether a host cell is infected with a virus, a monolayer culture is preferable to a semi-adherent or suspension culture. With the former, CPE and hemadsorption (HAd) can be visualized by microscopy, making the analysis relevant both to viral detection and subsequent viral identification.

Common test methods

The plethora of potentially contaminating sources arising from the complexity of the viral-vector production process favors a holistic scheme for adventitious agent testing, according to Padzik. In this approach, both general and species-specific assays are used.

Adventitious virus detection assays, explains Nguyen, mostly fall into three testing categories: cell-culture-based infectivity assays, targeted nucleic acid-based amplification, and antibody production tests. Both in vitro cell-culture detection methods and in vivo assays are employed. No universal method is available to detect all adventitious agents, and therefore various complementary methods must be used. “There is a clear tendency, however, to move away from assays that require animal models toward assays with higher sensitivity, specificity, and rapid turnover, such as polymerase chain reaction (PCR),” Padzik remarks.

In vitro cell-based detection methods are typically broad-spectrum and intended to detect viruses that are able to replicate in cell lines—a human diploid cell line, a non-human primate cell line, and a cell line of the same species as the one used to produce the vector, Diaz observes. In vivo, animal-based detection methods are also broad-spectrum and intended to detect viruses that replicate in animal models—typically mice and embryonated eggs, but also other animal species based on specific risks associated with the product being tested.

With these non-specific 14- and 28-day adventitious agent tests, CPE or haemagglutination and HAd values serve as the readouts. The obvious limitation, notes Padzik, is the difficulty in detecting adventitious viruses that do not cause CPE. The limits of detection for these assays can also be an issue. Padzik says a partial remedy, but with limited sensitivity, is the use of transmission electron microscopy (TEM) to detect, for example, retroviruses. XC plaque, S+L focus assays, co-cultivation assays, or fluorescent product enhanced reverse transcriptase (FPERT) testing are also commonly used to detect retroviruses.

Molecular methods such as PCR tests are used to detect specific viruses that are either not detectable by other methods or that are emerging as particular biosafety concerns, according to Diaz. Known contaminating entities, including porcine and bovine viruses, minute virus of mice, vesivirus, rhabdovirus, polyomavirus SV40, and arboviruses for various animal and insect cells are addressed with PCR-based specific assays, Padzik says. PCR-based methods that can detect virus families rather than specific viruses are also being employed. They are also used to detect mycoplasma, spiroplasma, mycobacterium species, and mice DNA sequences, according to Nguyen.

Nguyen stresses that each method exhibits a different level of sensitivity, and the testing requirements and sample preparation protocols differ as well. It is imperative, therefore, to select the appropriate test for each vector product.

The importance of raw material testing

Adventitious agent testing of raw materials should be proportional to the risk posed by the raw materials, and the testing methods chosen based on such an assessment, states Horst Ruppach, executive director, scientific and portfolio global biologics for Charles River.

The most common methods of adventitious agent testing for raw materials are cell-culture-based assays using specific indicator cell lines, with Vero cells preferable because they can detect viruses from all sources, according to Nguyen. The added risk for introducing animal viruses posed by animal-derived components requires the use of sensitive cell-based assays that can detect replicating infectious viruses, Ruppach adds.

PCR-based methods are also used depending on the raw material and turnaround requirements. It is important to recognize, though, Ruppach stresses, that while using molecular methods can be advantageous due to their speed, they do not discern between a nucleic acid positive signal and an actively replicating virus.

Viral vectors produced to modify cells for the production of ex vivo cell therapies have some additional raw material concerns with respect to adventitious agents. Human serum albumin (HSA), a commonly used excipient for cell therapies, is a high-risk material with respect to virus and prion transmission, but there is no practical assay for direct prion testing, according to Ruppach. “Here, special emphasis must be placed on HSA donor screening and the manufacturing process in order to mitigate the risk of virus and prion transmission,” he says.

In addition, Ruppach notes that for gene-modified cell therapies, the viral vector is considered a critical raw material, and the same three-pronged complementary core strategy involving material selection, testing, and clearance assessment should be applied to mitigate the introduction of adventitious agents into the final cell therapy product.

Allogeneic, off-the-shelf cell therapies in which the cells used to produce ex vivo-modified cell products are derived from multiple human donors are worth a special mention, according to Diaz. They require a carefully derived risk-mitigation strategy, because each donor contributes a separate, individual virus risk to the product. “Upfront donor screening plays a large role in the development of a strategy that can mitigate the risk of virus contamination of these cell-based therapeutic products,” she states. In addition to normally performed adventitious agent product testing, it may be necessary to also perform additional virus testing to address potential risks associated with a specific set of donors or even a specific geography.

A few limitations of existing methods

While traditional in vivo and in vitro methods for adventitious agent testing of viral vectors are sufficient to ensure the production of safe and effective products, they do suffer from some limitations, including sensitivity, the time required to perform the assays, throughput, and the detection of unknown contaminants, according to Padzik.

“Delivery times of adventitious virus test results have always been a serious challenge regardless of the method used, with some of them taking four to five weeks, and up to a few months with full analysis and documentation,” Padzik observes. For some gene therapy products that have a short shelf life, Diaz notes that these lengthy periods can be prohibitive. This concern is particularly relevant for autologous gene-modified cell therapies that must be administered to patients shortly after production. Molecular methods are being used to fill this gap, in particular PCR-based methods are being fine-tuned or developed to fill this need.

“The growing scope of required assays is creating continuous pressure, especially on early-stage clinical projects where results from the first human clinical studies are pivotal in securing the funding for full clinical medicine development,” Padzik adds.

“A partial solution could be a new pan-microbial microarray comprising 29,455 sixty-mer oligonucleotide probes for screening on a single chip vertebrate viruses, bacteria, fungi, and parasites,” Padzik comments. He cautions, though, that PCR-based methods are not able to help with the recognition of undefined adventitious agents and they suffer from noticeable instances of false positive cases.

Sample volume is another concern, particularly for viral vectors used in gene-therapy products. “Depending on the scale of vector production, adventitious virus testing can take up a significant amount of volume that could be otherwise used for patient dosing,” Diaz explains. This issue is of particular concern in early stages of product development before a production scale-up strategy has been implemented.

Nguyen reiterates that the different methods used for adventitious agent testing have different sensitivities for the detection of different contaminants and limitations. “These methods are intended to complement each other to provide a greater understanding of the detection of adventitious agents. In addition, the methods also have specific pharmacopeial and regulatory requirements/guidelines, such as sampling size. As always, these requirements must be considered when establishing a testing strategy,” she contends.

Need for new analytical methods

Tried-and-true test methods such as in vivo and in vitro virus assays and mycoplasma culture and microbiology culture assays have been used to demonstrate the biosafety of biologics for decades and can be applied in viral-vector manufacturing for certain steps such as cell-line and viral-seed-stock characterization, Ruppach emphasizes. During the past few years, adds Nguyen, understanding and methodologies have evolved significantly for adventitious agent testing, and Lonza believes this trend will continue to accelerate.

Given the limitations of existing methods and the advances being made in molecular-based methods, Ruppach expects the eventually the industry will switch to molecular test methods and rapid alternatives such mycoplasma PCR and rapid microbiological testing.

NGS-based assays are attractive for a number of reasons. The biopharma industry is focusing on using the full potential of NGS for adventitious testing requirements due to its low turnaround time in particular, according to Nguyen. The other main reason NGS is attractive is its ability to detect unknown contaminants, says Padzik. NGS can also be used to detect low-level known virus DNA/RNA sequences, according to Schrock. Identification will, of course, still require a comprehensive library of known and potential viral genomic sequences, and verification will be necessary by comparing results to those obtained using established methods.

In addition to these technical benefits, rapid NGS development has the potential to reduce costs and increase accessibility. “These advantages are driving significant interest and the urgency within the industry to validate this technique so that it can be swiftly implemented and applied for routine use,” Padzik remarks. The Advanced Virus Detection Technology Interest Group (AVDTIG) is an industry-based organization facilitating the exchange of experiences with the development of NGS-based assays to accelerate the introduction of practical solutions.

For Ruppach, one of the main challenges to effective viral-vector adventitious agent testing today is the resistance to new analytical methods. “There are several barriers to the adoption of new techniques such as NGS, including fear of regulatory delays, the risk associated with changing platforms, the resources (financial, personnel, time) needed, the tendency within the biopharma industry to be more focused on the short-term and less on the long-term, and the hesitancy to share information and desire to protect company trade secrets,” he explains.

“While all are valid reasons, it is essential to develop and share data that will support progress in the NGS field,” Ruppach says. He points specifically to the need for reference standards that will allow sharing of data obtained during development, comparability studies, validation and re-validation, technical transfer and assay trending, and ultimately the design of appropriate bridging studies that will enable translation to commercial application.

“Switching to rapid methods for adventitious agent testing must be done carefully, taking into consideration the fact that viral-vector manufacturing processes offer limited viral clearance steps,” states Ruppach. “Certainly, new alternatives such as NGS have applicability for viral assessment,” he adds. Padzik is hopeful as well. “The requests for using unbiased NGS are significantly increasing, especially in vaccine development. The current COVID-19 pandemic accelerated this transition even further,” he comments.

Ruppach also observes that a revision to International Council for Harmonisation (ICH) Q5A Quality of Biotechnological Products: Viral Safety Evaluation of Biotechnology Products Derived from Cell Lines of Human or Animal Origin is currently in the works and expected to include updated guidance to reflect updated scientific knowledge and advances in analytical technologies for new modalities, including viral vectors. For Padzik, the changing regulatory landscape and pressure to reduce animal testing are important drivers for NGS adoption, along with the apparent increasing acceptance of NGS-based assays in vaccines and other products where contaminating virus detection by the classical test is difficult and vaccine release is time-sensitive. “It is possible to envision gradual supplementation and, ultimately, the replacement of in vivo and other nucleic-acid technologies by NGS,” he concludes.

Taking a holistic view is the best strategy

“For most virus-based vaccines and viral vectors, the removal of contaminating viruses is not feasible, so the prevention and detection of virus contamination become more critical in a virus risk mitigation strategy,” emphasizes Diaz.

Although the principles used for adventitious agent testing of monoclonal antibodies (mAbs) apply to viral vectors, it is certainly not a “plug-and-play situation”, according to Chang. She stresses that a holistic view of the product, manufacturing, and ultimate use of the viral vector must be taken into account when performing risk assessments and establishing risk-control strategies.

For instance, the batch size for viral vectors is different from that for mAbs, with only a small number of doses produced per batch. In addition, some aspects of viral-vector manufacturing, such as the level of empty capsid production and consequent impact to the patient, are not fully understood. “Furthermore,” says Chang, “it is not possible to introduce robust viral removal/inactivation steps that exist with antibody manufacturing. Therefore, closer characterization of the starting raw materials, including biological characterization that assesses adventitious agents thoroughly, is prudent.”

Diaz agrees that an effective adventitious agent testing strategy starts with an appropriate risk assessment that evaluates all possible sources of virus contamination, including cell banks, virus seeds, plasmids, media, and other substances used in production. The risk assessment should also include identification of any steps in the manufacturing process (upstream and downstream) where virus contamination could be introduced.

“Evaluating the manufacturing components and processes for their contribution to the virus contamination risk allows for a multi-pronged approach that includes appropriate sourcing of materials, the selection of relevant virus detection methods for the risks identified, and the potential removal of contaminants associated with specific viruses in downstream processes,” Diaz summarizes.

There are several key aspects of any strategy for adventitious agent testing, according to Nguyen. Appropriate testing regimens must be included that monitor the possible introduction of adventitious agents. These methods must be suitably sensitive or use indicator cell lines or animal models in which a specific viral entity is known to replicate. The potential for matrix interference must be avoided; spiking of the test sample matrix with a positive control virus enables assessment of potential interference and the limit of detection of the method. “As a general rule of thumb,” she says, “the assays should be sufficiently developed with the appropriate set of predetermined system suitability and acceptance criteria, including relevant positive and negative controls, and fully established before validation.”

About the author

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