Risk Mitigation Strategies for Viral Contamination of Biotechnology Products: Consideration of Best Practices
=============================================================================================================

* Amy S. Rosenberg
* Barry Cherney
* Kurt Brorson
* Kathleen Clouse
* Steven Kozlowski
* Patricia Hughes
* Rick Friedman

## Abstract

**CONFERENCE PROCEEDING** Proceedings of the PDA/FDA Adventitious Viruses in Biologics: Detection and Mitigation Strategies Workshop in Bethesda, MD, USA; December 1–3, 2010

Guest Editors: Arifa Khan (Bethesda, MD), Patricia Hughes (Bethesda, MD) and Michael Wiebe (San Francisco, CA)

Viral contamination of biotech product facilities is a potentially devastating manufacturing risk and, unfortunately, is more common than is generally reported or previously appreciated. Although viral contaminants of biotech products are thought to originate principally from biological raw materials, all potential process risks merit evaluation. Limitations to existing methods for virus detection are becoming evident as emerging viruses have contaminated facilities and disrupted supplies of critical products. New technologies, such as broad-based polymerase chain reaction screens for multiple virus types, are increasingly becoming available to detect adventitious viral contamination and thus, mitigate risks to biotech products and processes. Further, the industry embrace of quality risk management that promotes improvements in testing stratagems, enhanced viral inactivation methods for raw materials, implementation and standardization of robust viral clearance procedures, and efforts to learn from both epidemiologic screening of raw material sources and from the experience of other manufacturers with regard to this problem will serve to enhance the safety of biotech products available to patients. Based on this evolving landscape, we propose a set of principles for manufacturers of biotech products: Pillars of Risk Mitigation for Viral Contamination of Biotech Products.

Viral contamination of biotech products is more common than is generally reported or previously appreciated. As recent events demonstrate, virus contamination can also be more devastating, from a number of perspectives, than previously acknowledged. Not only is such contamination, if undetected, potentially a direct risk for infection to patients, but it also poses a risk to the availability of life-saving products for patients with life-threatening illness, a risk to other products in the facility or to products in multiple facilities that share common reagents and/or to equipment, and, finally, a risk to product quality as cells producing therapeutic proteins under the stressed conditions of infection may produce products with alterations in the types and levels of post-translational modifications, potentially altering product stability and in vivo performance.

The majority of viral contaminants of biotech products are thought to originate from raw materials derived from biological sources. However, this is not always the case, as some viral contamination events (e.g., those associated with mouse minute virus (MMV)) may have originated from contamination of non-biological, chemically derived materials that occurred at the supplier's facilities. As an example, parvoviruses are so hardy and chemically resistant (1, 2) that they can survive the hostile conditions used for chemical production and be a potential downstream contamination source. Thus, few raw materials appear to be without some risk for introducing adventitious viruses into a biotech process. Perhaps the most daunting aspect with respect to implementing a control strategy for prevention of adventitious virus contamination is the fact that the viral load in contaminated raw materials that is sufficient to compromise an entire manufacturing process is usually very low. Thus, contamination of incoming raw materials may not be detected during routine testing of these materials, even with highly sensitive polymerase chain reaction (PCR) methods. Indeed, because of this, the source of most viral contamination events is seldom definitively identified. This fact further reinforces the point that the use of testing strategies alone to prevent contamination of a process is inadequate. This point was made strongly by Garnick in 1998, following a series of contamination events of a biotech production facility by MMV. He stated that “the difficulties inherent in detecting low levels of viral contaminants and thus, the potential to unknowingly contaminate a whole manufacturing facility, lie at the center of the concerns about viral contamination” (3). Based on this experience at Genentech, Garnick then proposed a viral control strategy that employed multiple steps: decreased use of raw materials of animal origin; inactivation of potential viruses in raw materials by viral inactivation methods; and robust viral inactivation and removal steps during purification, in addition to testing stratagems including adventitious agent testing of incoming raw materials, cell culture materials, and end-of-harvest materials (3).

A second limitation of current viral testing strategies is that non-specific, in vitro screening tests, which take weeks to perform, are generally employed to assess in-process materials, while PCR tests that provide timely feedback on process status, if employed, are biased towards detection only of adventitious agents previously known to infect a company's specific biotech production system. Under this control strategy, identification of a contamination event may occur only after downstream processing. More critically, some biotech manufacturers do not adequately leverage the experience of their contemporaries in designing testing programs more broadly for agents that have infected similar manufacturing processes. For example, MMV has contaminated multiple biotech processes, yet not all manufacturers routinely evaluate this specific risk to product quality and employ a risk reduction strategy that utilizes a specific PCR test for early detection of MMV. Typical adventitious agent screens use a panel of indicator cell lines that do not support the growth of all potential viral agents (and should not be presumed to do so) or show a cytopathic effect (4) and thus contribute to the potential for breakthrough viral contamination (5). Moreover, the potential for interference by components of the cell culture matrix, especially those containing fetal bovine serum (FBS), may further confound the ability to detect viruses using indicator cell lines. For example, antibodies to vesivirus found in FBS, either present in the original test article or added as a component of the enriched growth medium for the indicator cell lines used, can interfere with detection of the adventitious viruses for which they bear specificity. Thus, current testing paradigms have often been insufficient to reduce the risk of contamination of entire production facilities by novel or unanticipated viral agents, even when their potential prevalence in raw materials was known, thereby echoing the famous military dictum: The generals are always fighting the last war.

Based on these observations, we consider the following, adapted from Garnick with updates for the current decade, to be “Pillars of Risk Mitigation for Viral Contamination of Biotech Products”: 

1.  Remove or replace, when possible, naturally or biologically sourced raw materials from cell culture. This is feasible for some but not all production systems (particularly non-adherent cell culture processes). This directive is also consistent with current European efforts to eliminate such materials from biotech production.

2.  Routinely employ and improve robust viral inactivation/removal steps to eliminate adventitious agents in raw materials, where possible. Consideration should be given to implementation of several orthogonal methodologies, each with advantages and disadvantages including irradiation (gamma or UV), heat inactivation, high temperature–short time pasteurization protocols, and robust viral filters for non-viscous materials. These techniques should target those agents identified during risk assessments and should include those agents known to be prevalent in source animals (see item 7 below).

3.  Use advanced techniques as tools for understanding the extent of viral contamination of source materials commonly used in bioprocessing. While impractical for routine screening, if multiple batches of raw materials are used per year or many cell banks of the same type (e.g., Chinese hamster ovary) exist in a facility, the technologies elucidated below can provide a risk profile by detecting with sensitive methods and identifying viruses which pose a potential threat to the manufacturing process. Those technologies which can screen for a wide variety of viruses and thus guide comprehensive risk mitigation strategies, include massively parallel sequencing, triangulation methods (PCR, electrospray ionization mass spectrometry, mass analysis), and microarrays. Thus, they could be employed to understand potential risks posed by viral contamination of certain raw materials and parental cell lines, particularly in the setting when other strategies, for example, inactivation steps, are not an option. The identified viral agents can then be targeted for assessment on a lot-by-lot basis directly in production culture cells by more rapid, specific PCR tests and/or by infectivity assays using sensitive indicator cell lines that show a cytopathic effect when exposed to the infectious virus. Furthermore, the use of these techniques for investigation of viral contamination of production processes will allow a more timely and focused response in the event of a viral contamination.

4.  Know the spectrum of infectious viruses potentially present in your source animals or plants. This should include an assessment of the current prevalent and emerging infections of the relevant source animals or those naturally associated with plants (e.g., in the soil, etc.), even if they have not yet been known to pose a problem for biotech manufacture. This proactive strategy is vital for anticipating viral contamination events. This point becomes ever more important as the supply chain for materials becomes increasingly global, since recent reports indicate that control of animal diseases, particularly in the developing world, is lacking (6). Indeed, it has been alleged that a new infectious disease emerges every four months and that 75% originate in animals (7). Failure to take note of prevalent animal infections, and the repercussions thereof, are illustrated by the emergence of Vesivirus 2117 as a major contaminant in biotech production. This virus is a pathogen of cattle, which contaminated the production processes of several life-saving products used for treatment of orphan diseases (8). Vesivirus 2117, related to the Calicivirus family of viruses, was first identified during an investigation in 2003 of Chinese hamster ovary (CHO) cells demonstrating cytopathology (9). At that time, a Reverse Transcriptase PCR assay was established to detect this virus in biological raw materials and, in particular, in FBS. Although it was not known at that time whether Vesivirus 2117 would pose a major threat to the biotech supply chain, nonetheless, by 2006, Vesivirus 2117 was found to be prevalent in serum samples obtained from dairy and beef cattle in the United States (10). In a 2006 report, a clear warning was given to those in the biotech industry whose production systems required FBS: “The laboratory-based study reported here provides evidence of widespread vesivirus infections in cattle across a large area of the United States. The clinical, zoonotic, and other implications of this finding in a major food animal species warrant further investigation” (10). Despite the published widespread prevalence of this virus in source animals, its documented ability to grow in CHO cells, and the description of a method for detecting its presence, this virus was not routinely screened for in FBS or other bovine-derived materials, nor in production cultures, by biotech product producers. Moreover, the U.S. Food and Drug Administration (FDA) was likewise not aware of the potential for this organism to contaminate biotech products and thus did not request testing for this virus in production processes that were at risk. In fact, it took a major contamination event in 2009 that limited the supply of life-saving therapeutics and necessitated a complete shutdown and comprehensive decontamination of a biological manufacturing facility to bring the problem to the forefront (8), whereby this virus is now routinely tested for, at least by some manufacturers. Scientific expectations have been modified by this event and Vesivirus 2117 should be tested for in certain raw materials used for production and during cell culture for all “at risk” production processes. Of course, risk-based scientific concerns should always be considered in a regulatory dossier review and/or facility inspection. However, testing for a single virus known to be a risk, even employing the most highly sensitive PCR assay, is an example of perpetuating the practice of “fighting the last war.” This example emphasizes the need to implement strategies to prevent contamination with the next novel and emerging viruses. To identify which specific viruses may pose a threat via contamination of FBS and other raw materials, a collaborative and comprehensive approach is required. Industry and FDA, either separately or together, could form a task force to ensure continuous monitoring of source animal/plant health reports, or commission such studies where lacking. This way the emergence of viruses that pose a threat to the safety of biotech-derived products can be anticipated and mitigated. A joint task force to review the literature and initiate studies, where appropriate, would be one option. A related task force, the “Consortium on Adventitious Agent Contamination in Biomanufacturing,” has recently been initiated by the MIT Center for Biomedical Innovation.

5.  Use the most sensitive tests to screen relevant raw materials and early production cultures for known potential viral contaminants, as described in items 3 and 4 above. Several viruses, such as reovirus, MMV, Cache Valley virus, and vesivirus, have been responsible for the majority of contamination events of biotech products reported to the Center for Drug Evaluation and Research (CDER) and FDA. Given the heightened risk associated with these viruses, an effective risk-reduction strategy warrants routine testing of relevant raw materials with sensitive and specific tests for these viruses, and, in some cases, assessment for the presence of antibodies to viruses in these materials may be warranted. Furthermore, for evaluating viral contamination of FBS using indicator cell lines, removal or inactivation of maternally derived antibodies may allow for improved detection of viruses. However, any raw material can be contaminated with a virus and, as mentioned above, even the most sensitive detection methods will not necessarily detect a low viral load in raw material that can nonetheless compromise the production process and facilities. In such cases, the use of rapid detection methods for known contaminants at multiple points during cell culture provides timely information that can mitigate the extent of the contamination event and prevent a complete shutdown of a manufacturing facility and loss of vital product.

6.  Employ and improve viral clearance steps in the product purification scheme. FDA guidance (Points to Consider in the Manufacture and Testing of Monoclonal Antibody Products for Human Use, 1997) has recommended that each purification process include at least two orthogonal robust viral clearance steps. Furthermore, International Conference on Harmonisation guideline ICH Q5A indicates that manufacturers should provide “some level of assurance that adventitious viruses which could not be detected or might gain access to the production process would be cleared.” Both documents suggest that purification processes should be designed to have a robust capacity to clear adventitious viruses. While ICH Q5A also indicates that “for viruses that are not known or expected to be present, … a specific clearance value need not be achieved,” this statement should not be interpreted to indicate that the inability to clear certain viruses is acceptable without a full risk-versus-benefit analysis and additional process testing. Thus, in circumstances where certain viruses may be present but (1) at levels below the limit of detection and (2) not expected to be cleared by the manufacturing process, manufactures should consider the need to implement all feasible and valuable measures to assure the safety of their biotech products and include additional clearance steps in the manufacturing process. Consideration of such additional steps must take into account not only the potential viruses but also the sensitivity of the product to modifications that may result from these steps, such as aggregate formation (heat, irradiation), deamidation (increased pH), etc. Areas for technological improvement include gaining a true understanding of critical unit operation process parameters for robust and reliable clearance, as extensively discussed at the 2009 Indianapolis Viral Clearance Symposium (11), as well as standardization of unit operations. Recent efforts by the Parenteral Drug Association (PDA) and FDA to set standards for virus filter nomenclature (PDA TR41) and preparation of virus spikes for clearance studies (PDA TR47) are steps in the right direction.

7.  The final key to improving our ability to prevent or limit the impact of viral contamination of biotech products focuses on the use of a robust quality risk management system as discussed in ICH Q9. This system should be used to assess and document risks associated with viral contamination to reduce such risk to an acceptable level and to facilitate continual improvement in the control of viruses as new knowledge and technologies become available. This approach can be an effective strategy in circumventing the conservative aspect of human nature—the reluctance to deviate from old and accustomed practices—and instead take action to improve our ability to detect and eliminate adventitious agents from biotech manufacturing processes. Thus, as an ongoing component of a well functioning quality system, a firm should undertake periodic reassessments of the adequacy of risk mitigation afforded by existing viral contamination strategies. This assessment should include evaluation of the production history, viral controls, new analytical and production technologies, relevant literature, standards as they evolve, and current surveys on infectious agents potentially prevalent in raw material sources.

It is clear that both FDA and industry have serious unmet challenges and obligations to better ensure the freedom from adventitious agent contamination of biotech products. These challenges have been discussed above and require both short- and long-term planning to best assure product quality and safety. Regulators must become more active in encouraging industry to mitigate adventitious agent risk and not just be in a position of reacting following catastrophic events. This should involve a regular schedule of reevaluating and updating current standards with regard to manufacturing practices to further mitigate the risk of viral and other adventitious agent contamination, or performing such reevaluations as more sensitive and specific methods evolve. Moreover, strategies for control of adventitious agents should be updated routinely and enforced to reflect the current risk landscape, including these changing paradigms and technologies to better assess or mitigate risk, in accordance with current good manufacturing practice (cGMP). Indeed, as stated in ICH Q5A, “since the most appropriate techniques may change with scientific progress, proposals for alternative techniques, when accompanied by adequate supporting data, may be acceptable.” In conclusion, all firms must provide assurance that a drug or biologic meets the regulatory requirements of the U.S. Federal Food, Drug, and Cosmetic Act and the Public Health Service Act as to safety and has the quality and purity characteristics that it purports to possess. But as technology and knowledge evolve and a clearer picture of the risk landscape emerges, manufactures should reevaluate and adapt control strategies to address their evolving understanding of risks to process and to product. Appropriate risk review, risk control, CAPA (corrective and preventative action), and continual improvement programs should be in place. Failure to address such risks to product quality brings into question compliance with cGMP, and regulatory inspections should continue to evaluate whether firms meet contemporary standards for viral control.

*   © PDA, Inc. 2011

## References

1.  1. Boschetti N., Wyss K., Mischler A., Hostettler T., Kempf C. Stability of minute virus of mice against temperature and sodium hydroxide. Biologicals 2003, 31 (3), 181–185.
    
    [CrossRef](http://journal.pda.org/lookup/external-ref?access_num=10.1016/S1045-1056(03)00037-X&link_type=DOI) 
    
    [PubMed](http://journal.pda.org/lookup/external-ref?access_num=12935806&link_type=MED&atom=%2Fpdajpst%2F65%2F6%2F563.atom) 
    
    [Web of Science](http://journal.pda.org/lookup/external-ref?access_num=000185322900005&link_type=ISI) 

2.  2. Boschetti N., Niederhauser I., Kempf C., Stühler A., Löwer J., Blümel J. Different susceptibility of B19 virus and mice minute virus to low pH treatment. Transfusion 2004, 44 (7), 1079–1086.
    
    [PubMed](http://journal.pda.org/lookup/external-ref?access_num=15225251&link_type=MED&atom=%2Fpdajpst%2F65%2F6%2F563.atom) 

3.  3. Garnick R. L. Raw materials as a source of contamination in large-scale cell culture. Dev. Biol. Stand. 1998, 93, 21–29.
    
    [PubMed](http://journal.pda.org/lookup/external-ref?access_num=9737373&link_type=MED&atom=%2Fpdajpst%2F65%2F6%2F563.atom) 
    
    [Web of Science](http://journal.pda.org/lookup/external-ref?access_num=000075761500003&link_type=ISI) 

4.  4. 1.  Brown F., 
    2.  Lubiniecki A. S.
    
    Lee C. K. Issues of Biological Assays for Viral Vaccines. In Viral Safety and Evaluation of Viral Clearance from Biopharmaceutical Products; Brown F., Lubiniecki A. S., Eds.; Karger: Basel, 1996.
    
    

5.  5. Willkommen H. Presentation at PDA/FDA Adventitious Viruses in Biologics, December, 2010. Hannelore.wilkommen{at}gmx.de.
    
    

6.  6. Gilbert N. Livestock plagues are spreading. Nature News February 11, 2011, doi:[10.1038/news.2011.87](http://journal.pda.org/lookup/doi/10.1038/news.2011.87).
    
    

7.  7. ILRI Clippings. Animal surveillance needed to stop bird flu and other human epidemics—World Bank. March 11, 2011 [http://ilriclippings.wordpress.com/2011/03/11/animal-surveillance-needed-to-stop-bird-flu-and-other-human-epidemics-world-bank/](http://ilriclippings.wordpress.com/2011/03/11/animal-surveillance-needed-to-stop-bird-flu-and-other-human-epidemics-world-bank/)
    
    

8.  8. Genzyme Press Release, [http://www.genzyme.com/corp/media/GENZ%20PR-062509.asp](http://www.genzyme.com/corp/media/GENZ%20PR-062509.asp)
    
    

9.  9. Oehmig A., Büttner M., Weiland F., Werz W., Bergemann K., Pfaff E. Identification of a calicivirus isolate of unknown origin. J. Gen. Virol. 2003, 84 (Pt. 10), 2837–2845.
    
    [Abstract/FREE Full Text](http://journal.pda.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MzoidmlyIjtzOjU6InJlc2lkIjtzOjEwOiI4NC8xMC8yODM3IjtzOjQ6ImF0b20iO3M6MjI6Ii9wZGFqcHN0LzY1LzYvNTYzLmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

10. 10. Smith A. W., Iversen P. L., Skilling D. E., Stein D. A., Bok K., Matson D. O. Vesivirus viremia and seroprevalence in humans. J. Med. Virol. 2006, 78 (5), 693–701.
    
    [CrossRef](http://journal.pda.org/lookup/external-ref?access_num=10.1002/jmv.20594&link_type=DOI) 
    
    [PubMed](http://journal.pda.org/lookup/external-ref?access_num=16555277&link_type=MED&atom=%2Fpdajpst%2F65%2F6%2F563.atom) 
    
    [Web of Science](http://journal.pda.org/lookup/external-ref?access_num=000236657000023&link_type=ISI) 

11. 11. 1.  Blumel J., 
    2.  Brorson K.
    
    Miesegaes G., Bailey M., Willkommen H., Chen Q., Rousch D. Viral Clearance Symposium Participants, Blumel J., Brorson K. Proceedings of the 2009 Viral Clearance Symposium. Dev. Biol. (Basel) 2010, 133, 3–101.
    
    [PubMed](http://journal.pda.org/lookup/external-ref?access_num=21516942&link_type=MED&atom=%2Fpdajpst%2F65%2F6%2F563.atom)