December 1, 2016 (Vol. 36, No. 21)
Would-Be Viral Intruders Are Always Waiting for You to Lower Your Guard
Since its inception, biotechnology has been preoccupied with virus filtration, the clearance of invasive and adventitious viruses from biopharmaceutical products. And now, far from waning, virus filtration is set to become even more important. Virus filtration, reports Markets and Markets, constitutes a global market that is growing 12% each year. At present, this market is valued at $1.8 billion. By 2020, it will reach $3.3 billion.
These figures encompass diverse products and applications, which is fitting because virus filtration is often identified with viral risk mitigation, a comprehensive strategy that incorporates both clearance and inactivation practices, as well as procedures related to procurement and/or treatment of raw materials. Viral risk mitigation, so broadly conceived, was a key topic at CHI’s Bioprocessing Summit, which was held recently in Boston.
Faux Viruses
Viral clearance validation is demonstrated by conducting costly “spiking studies” that introduce live model viruses into scaled-down manufacturing processes, typically in BSL-2 facilities. “A noninfectious viral surrogate would be useful for scientists developing or characterizing downstream purification process steps,” said David Cetlin, founder and CEO of MockV Solutions, which produces recombinant, noninfectious mock virus particles (MVPs). Cetlin expects that one day biomanufacturers may routinely use stock panels of MVPs, or similar agents, for viral clearance studies.
MockV’s lead candidate is a parvovirus surrogate designed to mimic minute viruses of mice (MVM) particles. “We designed the surrogate to have a size, surface charge, and hydrophobicity similar to that of MVM particles, typical model viruses,” Cetlin explained.
MVPs are currently undergoing proof-of-concept studies at partners’ commercial installations to determine the comparability of their clearance to that of actual MVM particles. The MVP idea should be broadly applicable to the creation of surrogates for other common model viruses, for example, retrovirus, reovirus, and rabies virus.
Are We Overdoing It?
Xinfang Li, Ph.D., an associate director at ImmunoGen, notes that bioprocessors expend considerable resources on viral clearance studies based on widely used downstream unit operations with inherent clearance capabilities. The BioPhorum Development Group Viral Clearance Working Team, of which Dr. Li is a member, considers low-pH virus inactivation and virus filtration to be the most effective means of clearing viruses from biopharmaceutical preparations.
Low-pH virus inactivation normally precedes affinity capture, during which the antibody is maintained in a low-pH environment. However, acidic inactivation following capture is also a convenient way to inactivate enveloped viruses because minimum sample manipulation is required during low pH treatment purification.
Common wisdom holds that filtration and pH inactivation are orthogonal, but that is not the entire story. “Due to the difference of mechanism, low-pH inactivation works only for enveloped viruses,” noted Dr. Li, “whereas filtration is suitable for both enveloped and nonenveloped viruses.”
Bioprocessors use both methods, but separately. Low-pH inactivation occurs after affinity capture, while viral filtration is implemented after the final polish step.
“Currently, by using platform process, especially for early-phase work, the industry is engaging in modular filing by using historic data to save resources,” informed Dr. Li. “For later-stage work, tremendous resources are invested in viral clearance, including new and old resin evaluation. But new resins provide only minimal improvements, so the hope is that to save resources, industry will eventually eliminate the viral clearance study for used resins.”
Efficiency vs. Throughput
Virus filtration is an expensive unit operation for which bioprocessors must often trade removal capability with throughput. Pall has commercialized a line of virus filters, Pegasus Prime filters, that provide enhanced throughput while maintaining removal and providing value.
During manufacture, the hydrophilic polyethersulfone filter membranes are layered onto templates that peel away before assembly and packaging. This novel casting technology imparts a unique surface geometry that increases available surface area relative to the nominal frontal surface area. Pall refers to this as a “surface-enhanced substrate.”
“Pegasus incorporates micron-dimensioned channels that increase virus retention,” said Aernot Martens, Ph.D., global product manager for virus filtration. This relatively inexpensive fabrication technology does not substantially affect the cost of Pegasus relative to conventional virus filters.
All bioprocess membranes eventually plug, but because of their enhanced product contact area, Pegasus Prime membranes foul significantly later than conventional filters, asserts Pall.
“Most plugging occurs upstream where the big pieces get stuck, beyond the region where the antibody products sieve through on the downstream side,” Dr. Martens explained. Since the Pall manufacturing process does not affect the virus-retentive region, it retains large particles while allowing more product to pass through. “You can do whatever you like on the upstream side,” added Dr. Martens, “without affecting retention.”
What You Don’t Know: Emerging Viruses
Threats from never-before-encountered viruses in bioprocesses keep regulators and manufacturers on their toes. Porcine circovirus appeared on bioprocessors’ radar screen in 2009, when scientists found through deep sequencing that stocks of commercial rotavirus vaccines had come into contact with the potential pathogen.
“Viruses are a major risk in viral vaccines because their production does not incorporate viral clearance steps,” explained Barbara Potts, Ph.D., senior consultant with Potts and Nelson Consulting. “Something comes up seemingly every 10 years. The biopharmaceutical industry employs very few virology specialists, but once a contamination occurs, everyone becomes an expert in viruses.”
Circovirus have been known since 1978, and detection by means of the polymerase chain reaction (PCR) became possible in 1999. The virus comes in two flavors: type 1, which is ubiquitous in pigs but not pathogenic and type 2, which is rare but can be devastating to farmed animals. DNA for both virus types were found in pig-derived vaccines, but the implications were unclear.
The contamination was traced to the use of porcine trypsin, used to help cultured cells thrive in suspension. Vaccine makers were flummoxed originally since every batch of product contained viral DNA. That was true because the virus existed in the master cell bank. “Whenever you make a new lot, you go back to the master bank,” Dr. Potts noted, “and every time you pick up fresh contamination.”
Subsequent testing on the blood of vaccinated children confirmed the presence of viral DNA but indicated that the virus had not actually replicated in patients. The issue died down, but the incident underscored the inherent risk of viral contamination originating from raw materials.
“You can do a lot of mitigation for viruses at the raw material stage, but not with actual viral vaccine products,” Dr. Potts remarked. Biomanufacturers have sourced acid-inactivated trypsin since the 1980s, but evidently a batch or several batches that had not been properly treated slipped through. Alternative sources are usually unsuitable for vaccine production, but the industry-wide push for animal component–free ingredients eventually led to a corn-based trypsin, which works well.
Regulatory Perspective
The FDA’s Center for Drug Evaluation and Research (CDER) reviews and evaluates virus clearance data supplied by firms developing clinical phase or marketed biotech products. “We look for the validity of the small-scale models and the overall virus-removal capacity of the purification process,” said Kurt Brorson, Ph.D., a staff scientist at the CDER’s Office of Biotechnology Products.
The CDER also conducts “critical path” laboratory research aimed at improving strategies for measuring viral clearance of biotech production. This research has focused on such issues as the efficiencies of chromatography columns in clearing viruses over extended use.
Adventitious viruses, which appear ubiquitously in some species, are capable of contaminating mammalian bioreactor cultures. This has happened only occasionally, but when it does, the costs are steep. “It could cause drug shortage situations if the facility shutdown is prolonged,” Dr. Brorson pointed out. While several types of virus are capable of growing in CHO cells, one type of virus in particular, MVM, seems to be the most common and problematic.
Certain new technologies have provided incremental improvements for manufacturing over existing bioprocessing steps and are also able to clear viruses reasonably well. Recently, multimodal chromatography resins have provided improvements over traditional anion-exchange resins because they can run at higher salt concentrations and still remove viruses. Additionally, disposable systems such as membrane adsorbers have been introduced as replacements for bead-based resins.
“These can have cost advantages over reusable bead chromatography for low-volume products, and seem to clear viruses efficiently,” Dr. Brorson noted. Also, companies are introducing barrier technology such as high-heat/short-time media treatment steps to prevent adventitious agents from getting into bioreactors in the first place.
“There has been increasing acceptance of new unit operations and such concepts as modular clearance, whereby unit operations shared between two or more similar products are validated only once,” elaborated Dr. Brorson. “Regulatory authorities are also generally accepting the concept of evaluating virus filter performance using a small virus like parvovirus, because this is a worst-case scenario, and the claims from this data can be applied toward clearance of larger viruses such as retroviruses.”
Finally, industry and regulators are moving toward standardization in viral clearance. In 2012, the first American Society for Testing Materials (ASTM) standard for viral clearance was published in a document titled “E2888-12: Standard Practice for Process for Inactivation of Rodent Retrovirus by pH.” This standard specifies the low-pH incubation conditions for inactivating murine retroviruses by 5 log10. The ASTM is working on additional standards with active CDER participation.
Evolving Standards
For years, the biotech industry has relied on International Conference on Harmonization (ICH) Q5A, the most highly cited and followed guidance for biopharmaceutical viral clearance, and the United States Pharmacopeia (USP) general chapter 1050, which is a word-by-word replica of ICH Q5A. In August 2016, the USP issued a new general chapter, 1050.1, which provides greater clarity on the design, conduct, and interpretation of viral clearance studies.
“Chapter 1050.1 expands the list of potential model viruses that should be selected, with an eye on new and emerging viral agents that industry needs to be aware of,” commented Mark Plavsic, Ph.D., head of process development and manufacturing at Torque Therapeutics. The new chapter also provides guidance on preliminary “matrix studies,” such as interference and toxicity studies, to assist in procedure harmonization across industry and within organizations. Importantly, it also provides examples of experimental design for virus removal and inactivation to further assist harmonization of clearance study design and execution.
Virus stock quality attributes is an issue that previous guidances glossed over or ignored. The new chapter provides recommendations on virus identity, purity, stability, and infectious titer that allow more accurate calculation of log removal. “It also covers virus stock traceability, documentation, passage number, and reduction of aggregation, all important elements of viral quality that give users more understanding on the virus and viral stock,” added Dr. Plavsic.
Chapter 1050.1 encourages biomanufacturers to incorporate inactivation and/or physical removal in their processes. According to Dr. Plavsic, inactivation is sometimes insufficient, which requires adding a removal step. Conversely, removal may be sufficient because as a unit operation it is more robust than inactivation and clears both small and live viruses.
“In the literature, the terms robustness and effectiveness are often confused, but they don’t mean the same thing. A process (or a single step) can be effective without being robust, or vice versa,” Dr. Plavsic explained. “The older ICH and USP guidances did not fully address this issue, and they did not define these common terms.” The new chapter stipulates that “effectiveness” applies to steps that clear greater than four-log of virus, a regulatory first. It also states that steps that clear between one- and three-log, although not technically effective, still contribute to overall clearance.
Dr. Plavsic observed that over the years more manufacturers have been using nanofiltration, an expensive but highly effective way to remove viruses: “Nanofiltration has almost become the norm, and its acceptance can only grow as not every process lends itself to inactivation. But even when inactivation works, processes are increasingly adding nanofiltration as well.”
He also indicated that he sees more innovator companies becoming involved in viral clearance studies rather than leaving them to contract testing firms. This reflects in part the seriousness with which regulators take viral contamination as well as the perception that contractors may lack the appropriate understanding of the sponsor’s purification process details.