Providing a definition for virus safety is nebulous at best. Industry and regulators recognize that the holy grail of viral safety — absolute freedom from extraneous agents or residual pathogenicity — is a myth. ‘You find what you are looking for’ is a common axiom and nowhere is it more applicable than in the case of virus detection to ensure the safety of biopharmaceuticals. Recent reports of adventitious virus contamination of products and production environments have accentuated the necessity of a holistic approach to virus safety. This paper reviews the approaches applied to ensure viral safety and discusses some of the recent contamination incidents.

The biopharmaceutical industry has experienced considerable technology maturation in the last several decades. The biopharma market has grown for a variety of reasons, including blockbuster drugs going off patent, the need for a more extensive and robust drug pipeline, and the impetus to pursue biosimilars/biobetters. According to a report published in January 2011 by BCC Research (www.bccresearch.com/report/BIO079A.html), the global biologics market was valued at an estimated $149 billion in 2010 and is expected to reach $239 billion by 2015. The monoclonal antibodies segment is expected to increase at the fastest rate within the total biologics market to $86 billion by 2015. The therapeutic protein segment, valued at $72 billion in 2010, is expected to reach $107 billion in 2015.

Biopharmaceuticals pose several unique manufacturing and regulatory challenges due to their intrinsic complex profile. They have a high level of structural complexity and heterogeneity, are produced in living systems or supplemented with reagents derived from living systems, and consequently have a complex purity/impurity profile that poses unique analytical challenges. Additionally, production involves 10 or more manufacturing stages encompassing 18-30 unit operations with several hundred process parameters. A single change could have a cascade effect; the impact on the quality, safety, and efficacy of the biological is not predictable. Therefore, determining what constitutes a critical process parameter (CPP) and the specific critical control points (CCPs) in the manufacturing process poses significant challenges. Nevertheless, quality and safety requirements similar to those applied to small molecules (chemical entities) are applied to biopharmaceuticals as well.

Ensuring virological safety of biologicals is even more challenging. Iatrogenic accidents in the past have occurred due to contamination of the production system (e.g., avian retrovirus type C in yellow fever vaccine, SV40 in inactivated poliovirus vaccine), manufacturing process-related concerns (e.g., incomplete inactivation of live virus vaccines such as polio and rabies vaccines) and the use of contaminated excipients (hepatitis B virus (HBV) was transmitted via the human serum used in the yellow fever vaccine).

There are several lessons to be learned from the abovementioned incidents. Adventitious agents can potentially contaminate the production system and go undetected; in many cases, particular raw materials have been implicated. Changes in critical process parameters can impact the safety profile, and extrapolation of inactivation data must be done with caution; in the case of the polio vaccine (Cutter incident), the presence of cations contributed to the thermostability of the virus and resultant infection transmission. Also, virus detection systems are not always sensitive enough to detect low levels of pathogenic virus — the human serum used as an excipient in the yellow fever vaccine was implicated in HBV transmission; however, this was not proven until the availability of molecular techniques such as polymerase chain reaction (PCR).

The safety profile of blood- and plasma-derived products has improved considerably during the last three decades, largely due to advances in detection methods, donor selection measures, and virus inactivation and removal methodologies that continue to evolve. Nevertheless, there is a constant threat of new and emerging agents. Globalization is not limited to the flow of money, goods and people across geographic boundaries, but also includes distribution of pathogens into locations not considered indigenous to the pathogen.

Recent reports of contamination of bulk harvests, adventitious virus contamination of manufacturing environments and even a marketed vaccine product have shone the spotlight on the vulnerability of all pharmaceutical/biopharmaceutical operations.^1-5 Fortunately, to date, biopharmaceuticals produced in recombinant cell lines have had an excellent safety record; there has been no report of iatrogenic virus transmission of pathogenic virus through administration of these products.

Virus contamination has several serious consequences, paramount being the potential to impact patient safety. In addition to a direct impact, lack of drug availability for patients with life-threatening illnesses has serious consequences, as many products are still under patent and may be single sourced. Table 1 summarizes the consequences of contamination. Several of the issues listed in Table 1 have been highlighted in recent contamination events with Vesivirus 2119 that occurred at a U.S. biopharmaceutical facility.⁶

Detection: You Find Only What You Are Looking For

Industry professionals and regulators indeed recognize that there is no all-encompassing virus detection method and no ‘one-(virus cultivation) medium-for-all’ to be able to detect all viral contaminants. Our current infectivity and molecular detection methods have several limitations. Endogenous and adventitious viruses may escape detection for a variety of reasons: limited assay sensitivity (small sample volumes assayed), limitations in detection methods as, for example, necessity for virus-specific assays, need for permissive cell lines to detect viral variants — to name a few. However, sophistication in computational power and advances in bioinformatics have proven that the concept of ‘virus free’ is only as good as our detection method. Methodologies such as massive parallel sequencing, degenerative PCR and panmicrobial microarrays have resulted in detection of viral sequences in cell substrates and virus seed stocks.^{7, 8} Victoria et al.⁸ reported detection of endogenous retroviral sequences and one adventitious virus in eight live attenuated human viral vaccines. Testing of veterinary vaccines has revealed presence of endogenous feline retrovirus RD114 in live feline vaccines and Torque tenovirus (TTV) nucleic acid in five poultry and 10 mammalian vaccines.⁹

Are Biopharmaceuticals Safe From a Virus Safety Standpoint?

The viral risk profile of any biological is contingent on a variety of factors including: source of the biological, raw materials used, production systems, purification reagents and excipients. Table 2 provides a brief summary of the factors influencing viral risk profile.

Evaluating the safety of any product begins with a risk assessment. Viral contaminants may gain access via the intrinsic/endogenous viral load associated with the cell line, raw materials used in production, purification and formulation reagents, or manufacturing environments (equipment, facilities, personnel). As cell banks are extensively characterized, any viral contaminant associated with them will not be cytolytic; however, chronic or latent viruses could potentially be present. Recombinant products produced in human/humanized (human/rodent) cell lines are preferred from an immunological standpoint, however, the absence of a species barrier raises additional viral safety concerns. The widespread use of murine cell lines in the manufacture of monoclonal antibodies is a potential source of introduction of rodent zoonotic agents. Chinese Hamster Ovary (CHO) cell lines are frequently used in monoclonal antibody production, and several viruses — Cache Valley Virus (CVV), Reovirus, Sindbis, Vesivirus 2117, Encephalomyocarditis Virus, Mouse Minute Virus (MMV) — can replicate in CHO cell lines[10]. According to the CDC National Center for Emerging and Zoonotic Infectious Diseases, approximately 75% of recently emerging infectious diseases affecting humans are diseases of animal origin and approximately 60% of all human pathogens are zoonotic (http://www.cdc.gov/ncezid/).

Raw materials are a significant source of vulnerability to virus contamination, as large volumes of complex process media and gases are used during production. Reovirus type 2, CVV and Epizootic Hemorrhagic Disease Virus contamination have been potentially linked to the use of non-irradiated bovine serum, while contamination of rotavirus vaccine with porcine circovirus-1 likely originated from the porcine trypsin used in the establishment of master cell banks several decades ago. Many processes have moved away from the use of animal-derived materials, but where this is not possible the risk of introduction of adventitious agents must be evaluated and mitigated.

There have been some instances of bioreactor contamination in the past; this has been summarized by Kerr and Nims.³ Depending on the contaminant (e.g., Reovirus), bioreactor virus contamination may be silent, i.e., little/no change in cell viability and other parameters, in which case the production may go to completion, and will likely only be detected during adventitious virus testing of the bulk harvest. The concern with this scenario is that it has the potential to contaminate entire manufacturing environments, which would then necessitate decontamination of the entire manufacturing facility. Other viruses such as CVV could cause the bioreactor to crash (i.e., changes in cell viability, biochemical parameters), resulting in premature aborting of the production run. A few instances of MMV contamination have also been reported.^{2, 4} Batch contamination may either be overt and detected by poor cell viability or growth; in some cases, it may be less apparent and detected only at the bulk harvest stage.³

Are Upstream Virus Barriers Necessary?

Several considerations deserve mention here: rodent-derived cell lines (e.g., CHO) are often used in mammalian cell culture bioreactors, and these are viable hosts for zoonotic viruses and other rodent viruses (MMV). Standard raw materials lot testing will not detect low-level viral contaminants, especially considering that contamination is distributed non-homogenously, but detection is based on probability. Routinely, heat sterilization and sterile filtration of heat-stable and heat-labile raw materials, respectively, is employed. However, these methods cannot be relied upon to result in clearance of the adventitious viral agent due to the relative physicochemical resistance and small size of the contaminant viruses.

In view of recent reports of virus contamination of bioreactors and the concomitant financial losses, regulatory implications, and facility decontamination costs, some companies have started to incorporate upstream barrier technologies such as y–irradiation,¹¹ UV-C inactivation and HTST (high-temperature-short-time) treatment of raw materials.¹² It is important to remember that these methods must not, in any way, impact target protein yield, downstream process performance or product quality. In the final analysis, while incorporation of upstream virus barriers is not mandated, proactive implementation of adequate virus barriers does provide some business insurance against virus contamination events.

Risk Mitigation Strategies

The current risk minimization strategy to guard against inadvertent virus exposure of patients is a combination of three efforts:

1. prevention of access of virus by screening of starting materials (cell banks, tissues, or biological fluids) and raw materials/supplements used in production processes (culture media, serum supplements, transferrin, etc),
2. incorporation of robust virus clearance steps, into the manufacturing process, and
3. monitoring production using a relevant screening assay.¹³

Table 3 summarizes risk mitigation strategies for virus contamination control.

Source materials evaluation is a critical component of a risk minimization strategy. Cell bank source, history and exposure to ‘complex’ raw materials must be assessed. Cell banks for many current marketed products were established decades ago and documentation related to use of animal-derived materials may not be available. Detection methods will test for ‘known-knowns’ (viruses we know and suspect to be present); the potential for presence of ‘unknown-unknowns’ (viruses that are not in our current lexicon and cannot be detected) always exists. In terms of raw materials, it is essential to look beyond the obvious. Just because a raw material is labeled recombinant or synthetic does not imply that it has not been exposed to animal-derived materials. In one report, an antibiotic with a claim of ‘no exposure to animal-derived materials’ was, in fact, produced by bacteria grown in a medium containing bovine peptones, bile, and several other bovine constituents.¹⁴

It is necessary to understand the supply system and recognize the complexity and interlinking nature of the components. Replacement of animal-derived materials with materials sourced from plants poses different challenges and risks. You must perform a comprehensive risk assessment of critical components and make risk assessment a priority with suppliers. Additionally, there should be a program for ongoing monitoring and application of risk mitigation as needed.
Virus Clearance Methods and Clearance Evaluation (Validation) Studies

Several unit operations employed in biomanufacturing provide virus clearance, while others — virus filtration, for example — are deliberately included in the manufacturing processes to provide an adequate level of viral safety. Some unit operations such as affinity steps (Protein A) and chromatographic separations may not be optimized for virus clearance. Nevertheless, they afford a certain measure of virus clearance. For example, anion exchange chromatography in the flow through mode binds impurities such as host cell protein, DNA and endotoxin along with providing viral clearance. Low pH inactivation, effective against enveloped viruses, has been demonstrated to be a robust step for virus inactivation (critical process parameters are pH, time and temperature). Robust virus removal of large viruses (> 50 nm) can be demonstrated with both large- and small-pore-size filters; removal of small viruses (~ 25 nm) may be process dependent even with the small-pore-size virus filters.

Demonstration of viral clearance by unit operations in the downstream purification process is a critical component of ensuring the overall safety of biopharmaceuticals. The objective of virus clearance evaluation studies is not only to evaluate the ability of the manufacturing process to clear (inactivate/remove) known viral contaminants, but also to estimate the robustness of the manufacturing process to clear known and unknown viruses. The choice of virus for use in virus validation studies is governed by the type of product (e.g., plasma-derived, cell line-derived), unit operation, phase of study, and regulatory and industry expectations. While regulations and guidance documents require that virus validations be conducted with relevant viruses as well as specific and non-specific model viruses, a typical panel for biotech/recombinant products would include X-MuLV (xenotropic Murine Leukemia Virus; retrovirus), HSV (Herpes Simplex Virus)/PRV (Pseudorabies virus), small non-enveloped virus (MMV) and Reovirus.

ICH Q8¹⁵ has established the concept of design space. Bacteriophages such as PR 772 and PP7 may be used as surrogates for mammalian viruses during process evaluation for viral clearance by size-exclusion filtration and defining design space. Design space conditions established with a phage will be predictive of filter clearance of the corresponding mammalian virus. However, current regulatory expectations for filter validation studies for submission to regulatory authorities require the use of mammalian viruses.¹⁶

The extent of virus clearance required (number of logs clearance) is not provided in any prescriptive guideline but a framework approach is applied based on the evaluation of the potential for presence of any baseline viral load and incorporation of an adequate safety factor.

Parameters critical to the manufacturing process need to be explored and the robustness of the unit operation demonstrated. Multivariate analysis and other statistical techniques are applied including a Design of Experiments (DoE) Approach. DoE reduces testing burden and goes beyond conventional ‘worst-case scenario’ approach. Manufacturers that develop platform technologies can also apply their considerable knowledge base related to their production system — cell line, raw materials and manufacturing production and purification conditions — and leverage that information to reduce product development timelines and streamline process validation that may facilitate regulatory approval. However, the onus of demonstration of applicability of this strategy across multiple products lies with the manufacturer.

In-process Surveillance

In-process surveillance can limit the spread of contamination. In addition to testing of bulk harvests, enhanced analytical methods such as quantitative PCR for known contaminants and newer analytical methods for broader detection of adventitious agents would facilitate early detection and can reduce the requirements for decontamination, protect downstream equipment and mitigate the need for facility decontamination.

Conclusions

Providing a definition for virus safety is nebulous at best. Zero risk is a myth, but virus safety can be enhanced by incorporation of multiple overlapping virus containment and clearance strategies. Recent contamination concerns have demonstrated the necessity for a proactive approach to virus safety that includes incorporation of risk assessment, risk mitigation and management strategies in order to provide maximum possible assurance of a sufficiently low risk of harm that is significantly outweighed by the therapeutic benefits.

While the so-called safety tripod — appropriate sourcing, demonstrating viral clearance capacity through the manufacturing process, and in-process controls — for virus safety assurance has stood the test of time and safely delivered products with an excellent safety profile (from a virological safety standpoint), we will continue to be vulnerable to emerging and evolving viral agents. Contamination events are rare, but they are catastrophic when they occur. Current industry initiatives towards establishment of the consortium on adventitious agent contamination in Biomanufacturing (CAACB), a collaboration between the Massachusetts Institute of Technology’s Center for Biomedical Innovation and biopharmaceutical manufacturers, seeks to leverage industry knowledge and experience and share industry risks and prevention strategies regarding mitigation and management of adventitious agent contamination.

To quote American author and playwright Alfred Sheinwold, “Learn all you can from the mistakes of others. You won’t have time to make them all yourself.”  

References

1. Baylis, S.A., et al., Analysis of porcine circovirus type 1 detected in Rotarix vaccine. Vaccine, 2011. 29(4): p. 690-7.
2. Skrine, J., A biotech production facility contamination case study – mouse minute virus. Paper presented at: PDA/FDA Adventitious Viruses in Biologics Detection and Mitigation Strategies Workshop; 2010. Dec.1–3; Rockville, MD., 2010.
3. Kerr, A. and R. Nims, Adventitious viruses detected in biopharmaceutical bulk harvest samples over a 10 Year Period. PDA J Pharm Sci Technol, 2010. 64(5): p. 481-5.
4. Moody, M., MMV contamination – a case study: detection, root cause determination and corrective actions. Paper presented at: PDA/FDA Adventitious Viruses in Biologics Detection and Mitigation Strategies Workshop; 2010 Dec. 1–3; Rockville, MD., 2010.
5. Pierard, P., Presence of porcine circovirus (PCV1): findings, investigations and learning. Paper presented at: PDA/FDA Adventitious Viruses in Biologics Detection and Mitigation Strategies Workshop; 2010. Dec.1–3; Rockville, MD., 2010.
6. Bethencourt, V., Virus stalls Genzyme plant. Nat Biotech, 2009. 27(8): p. 681-681.
7. Onions, D. and J. Kolman, Massively parallel sequencing, a new method for detecting adventitious agents. Biologicals, 2010. 38(3): p. 377-80.
8. Victoria, J.G., et al., Viral nucleic acids in live-attenuated vaccines: detection of minority variants and an adventitious virus. J Virol, 2010. 84(12): p. 6033-40.
9. Dodet, B., et al., Viral safety and extraneous agents testing for veterinary vaccines. Biologicals, 2010. 38(3): p. 326-31.
10. Berting, A., M.R. Farcet, and T.R. Kreil, Virus susceptibility of Chinese hamster ovary (CHO) cells and detection of viral contaminations by adventitious agent testing. Biotechnol Bioeng, 2010. 106(4): p. 598-607.
11. Gauvin, G. and R. Nims, Gamma-irradiation of serum for the inactivation of adventitious contaminants. PDA J Pharm Sci Technol, 2010. 64(5): p. 432-5.
12. Weaver, B. and S. Rosenthal, Viral risk mitigation for mammalian cell culture media. PDA J Pharm Sci Technol, 2010. 64(5): p. 436-9.
13. ICH.Q5A(R1), Viral Safety Evaluation of Biotechnology Products Derived From Cell Lines of Human or Animal Origin. 1999.
14. Potts, B.J., TSE Case Studies Associated with Japanese and Other Regulatory Authorities–Talk Transcript. PDA J Pharm Sci Technol, 2010. 64(5): p. 442-4.
15. ICH. Q8 (R2), Pharmaceutical Development. International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use; Current Step 4 version, August 2009.
16. Miesegaes G, S.L., H. Aranha, and K Brorson, Virus Retentive Filters. Encyclopedia of Industrial Biotechnology, John Wiley and Sons, 2009.

Hazel Aranha, Ph.D., RAC is manager, Viral Clearance and Safety, at Catalent Pharma Solutions in Morrisville, NC. She can be reached at hazel.aranha@catalent.com or 919-465-8123.