December 1, 2017 (Vol. 37, No. 21)
Viral Clearance Calls for Bioprocess Engineers to Implement Control Strategies
If dealing with viral contamination seems like opening a can of worms, try not dealing with it. Yes, viral contamination occurs infrequently, but when it does occur, the consequences can be severe. Viral contamination incidents can shut down production facilities and interrupt drug supplies.
Even before production commences, viral contamination demands attention. For example, failure to demonstrate virus removal during clinical programs can lead to costly delays in product approvals.
Unfortunately, the viral safety of a drug cannot be adequately assured by simply testing the drug itself. As the FDA indicated in its 2012 Guidance, Q11 Development and Manufacture of Drug Substances, “with sterilized chemical entities or biotechnological/biological drug substances, there is an inherent limitation in the ability to detect low levels of bacterial or viral contamination.”1
Viral safety is not a discrete test at the end of drug manufacture or even at any point upstream during production. Viral safety is more of a process, one that is intertwined with the drug manufacturing bioprocess itself, and one that demands a control strategy. Such a strategy may incorporate upstream assessments of critical quality attributes or in-process controls.
Control strategies and related topics were discussed at the 2017 BioProcess International conference that was held recently in Boston. At this event, speakers addressed diverse virus filtration issues—upstream filtration, filtration of non-protein products, and filtration in continuous processing—as well as more general viral safety issues.
Viral Safety and Process Bottlenecks
The difficulty of maintaining viral safety while expediting process development was addressed at the meeting by Min Zhu, Ph.D., director of protein science, Boehringer-Ingelheim. To design and demonstrate viral clearance takes considerable resources, said Dr. Zhu. No fewer than three of the six major downstream unit operations are devoted exclusively or primarily to remove viruses. These are low pH virus inactivation, anion-exchange chromatography, and virus filtration. Measurable clearance also occurs through Protein A capture, cation exchange, and ultrafiltration/diafiltration.
Viral clearance imposes bottlenecks throughout downstream purification, thus slowing down overall process speed. At the same time, virus filtration is considered one of the most expensive downstream steps. It represents a significant portion of the cost of goods (COG).
To overcome process bottlenecks and contain filter costs, Boehringer-Ingelheim is assembling a knowledge database consisting of data from the literature, cross-industrial collaboration, vendors, and in-house research. The database will be utilized to support development of process design, risk assessment for process characterization studies, and control strategy.
“This is an ongoing project relying on open-source information,” Dr. Zhu declared. At this stage, the project is proceeding without any formal collaborations with vendors.
Given the cost and time required to exploit virus filtration and other clearance strategies for COG improvement, bioprocessing engineers often face the tradeoffs between throughput, flux, log removal, and cost. Navigating these variables becomes trickier for continuous processes.
“Viral inactivation and filtration are required for any manufacturing process from mammalian cell culture, as two independent, orthogonal viral clearance steps,” Dr Zhu tells GEN. “The challenge is how to incorporate them into continuous processing.”
Viral inactivation is typically performed at batch mode. To incorporate it into continuous liquid flow, Boehringer-Ingelheim has designed an incubation chamber to allow continuous flow and maintain minimal product contact time and target pH.
“The challenge of incorporating viral filtration into continuous processing is demonstrating robust viral removal when the product concentration in feed is variable,” Dr. Zhu stated. “A small-scale viral clearance validation strategy needs to be developed further.”
Harsh Conditions
Brian Buesing, senior research associate at Asahi Kasei Bioprocess America, discussed work on his company’s PlanovaTM BioEX virus filter, which contains a bundle of straw-shaped hollow fibers and relies on a filtration mechanism that is based on size exclusion. The membrane structure of Planova allows proteins to easily pass through the hollow fiber walls while viruses are efficiently captured in the membrane pores.
An advantage of the Planova platform is parvovirus clearance with or without buffer flush under a wide range of modest to harsh pH levels, salt concentrations, pressures, and protein concentrations.
“We recently collaborated with Janssen Pharmaceuticals to better define the design space of operating conditions that would work for three of their antibody products,” Buesing tells GEN. Results from that collaboration appeared June 2017, in Biotechnology Progress.2 “We found that both filters under all combinations of operating conditions were capable of achieving a virus removal factor of at least 4 log.”
Additionally, hollow-fiber Planova filters achieve high flows and mass throughput even in the presence of 500 mM salt, and are compatible with a variety of buffers, according to Buesing. He added that the hollow-fiber membrane material, hydrophilic modified polyvinylidene fluoride, also has the strength and integrity to withstand steam-in-place operations.
An advantage of hollow-fiber membranes is the capability to increase filtration capacity without requiring conformational changes in the filter cartridge. “Each of our filter types, in sizes spanning laboratory and manufacturing scales, is available in the same basic form factor,” Buesing noted. While connection types differ among sizes, the effective surface area of the filter membrane is increased by increasing the length or number of the hollow fibers in the bundle rather than by changing the filter conformation.
While Planova filters were designed for batch processing, nothing precludes their use in a continuous process. “Certainly, the operational aspects of total loading volume, variation in load solution parameters, and representing this manufacturing strategy at validation scale need to be considered, but these questions would need to be addressed for any candidate nanofilter prior to implementation in a continuous process,” Buesing asserted. “Asahi is currently working with regulators to determine the best path forward to demonstrate the suitability of our filters for continuous processing.”
Asahi Kasei Bioprocess America provides custom equipment supporting the automated use of Planova™ filters.
Virus- and Cell-Based Products
The production of vaccines, viral vectors, and cell-based therapies presents unique challenges for virus safety. If the product itself is a virus or a living cell that is capable of harboring viral infections, size exclusion and chemical inactivation are inappropriate.
Michael Cunningham, Ph.D., associate director, applications, mechanical system analysis/design tool (MSAT, MilliporeSigma, noted that contaminating viruses are as small as 20 nm, which is close to the limits of virus filtration, which makes them difficult to separate. “Just one viral particle per liter of process fluid is enough to contaminate an entire process,” he remarked.
Given available technology, ensuring the total absence of contaminating viruses in starting materials is impossible, so the focus during biomanufacturing is on virus detection and removal using standard methods. Well-characterized sourcing of raw materials can also reduce virus risk.
Safety testing of cell and virus stocks can include detection of virus nucleic acid by the polymerase chain reaction, or detection/quantitation of viruses using cell-based assays. In vivo testing can also be utilized if the virus is a known entity in the testing system.
Another strategy involves the judicious use of helper viruses—tools that facilitate the propagation of recombinant virus preparations that will, in the absence of helper viruses, reflect how they have been engineered to lack the ability to proliferate. In nature, the hepatitis B virus “helps” the replication-defective hepatitis D virus replicate by providing envelope proteins.
“Many helper viruses—a good example is adeno-associated virus—are larger than their counterpart therapeutic viruses, and can often be separated by filtration employing appropriately-sized membranes,” stated Dr. Cunningham.
In instances where helper viruses are more susceptible to chemical or pH inactivation than their corresponding therapeutic viruses, virus inactivation may facilitate their removal. In instances where helper viruses are of similar size to the therapeutic virus, chromatographic separation of these two virus populations may be a consideration, assuming that the virus preparations are well characterized.
MilliporeSigma’s Mobius FlexReady Solution for Virus Filtration features an optimized single-use flowpath to support parvovirus filtration needs. It accommodates Viresolve® Pro Modus devices, which are designed to combine high virus-retention assurance and fast and efficient processing.
Viral Safety in Continuous Processing
Continuous bioprocessing, which is slowly gaining on batch processing, raises new issues in viral safety. Morven McAlister, Ph.D., senior director, Pall Life Sciences, made the case in her talk for better understanding of the “virus filter design space” during continuous processing.
She explained how with first-generation virus filters, virus retention was affected by factors such as low-pressure conditions and process interruptions. “The newer generation filters have been designed to overcome these problems, such that they show consistent virus retention at both upper and lower operating ranges.”
An example of a newer generation filter is Pall’s Pegasus Prime. It can be used for batch processes at higher system pressures, but it shows equivalence at low flow or system pressures, which is more representative of a continuous process.
“A filter that shows equivalence in performance over upper and lower process parameters is what we define as having a robust virus filter design space,” Dr. McAlister explained. “The idea behind continuous processing is that processing occurs in real time, so the filter must be capable of operating under very low flow conditions, which is a very different filter design space than batch filtration.”
One could, theoretically, achieve some level of process continuity by expressing, harvesting, and holding process fluid for batch virus filtration, but that scenario does not fully exploit the value of continuous processing, Dr. McAlister continued.
“One of the main advantages of continuous processing is the ability to reduce the size of the manufacturing environment reduced by eliminating the need for large hold tanks,” she noted. “By fully understanding the design space for each process step, there is constant control over production which can reduce the time for a monoclonal antibody to be completely processed.”
There are various models that could be followed and still allow a batch to be defined. For example, one option may be to use multiple filters in a parallel configuration. When a defined throughput is met, that filter can be taken offline and the flow diverted to a second filter. In this way, the fluid that is filtered with the first virus filter could be viewed as a batch. However, with this model, consideration has to be given to process interruptions and flushes.
Raw Material Qualification
Raw materials qualification is a critical part of overall process and product safety that intersects significantly with virus safety. Given the discrete inputs and outputs for raw materials, a risk-based strategy for qualification makes sense.
According to Aurora Henry, raw materials specialist at CMC Biologics, the riskiest materials are those that are non-compendial and are used late in the process.
A compendial material will often be manufactured to GMP standards, with suppliers incorporating controls for residual solvents, elemental impurities, and other contaminants. “Compendial materials are produced with pharmaceutical manufacturing in mind and therefore avoid animal-derived components, genetically modified organisms, and allergens as much as possible,” Henry said. “They are tested to standards based on typical manufacturing conditions in the manufacturing region, and so they are well characterized.”
Non-compendial materials do not incorporate the same controls. Suppliers will not provide as much information about their process and may not include enough testing to fully characterize their materials. “This means that they are inherently less controlled than compendial materials, which makes it more challenging to demonstrate suitability for use through testing,” Henry added. “Lack of characterization by the supplier typically requires additional characterization internally.”
Given those realities, eliminating all risky materials is impossible. Media, for example, are often not well characterized, so non-compendial materials will always be part of processes. As a consequence, the strategy often is to eliminate or mitigate the risk rather than revise the process to remove the risky material.
Raw Material Parameters
At CMC Biologics, risk assessment examines a variety of raw material parameters that might affect the process such as sourcing, quality, and the potential to affect the process. Assessments are completed for each material/supplier used in the process with the support of a multifunction team consisting of process science, manufacturing, quality, supply chain, and others. The initial risk assessment is completed internally, but mitigation often employs outsourced services such as contract testing or third-party storage.
“As for supplier verification or qualification, I expect information on material sourcing, including their supplier, if they are a repackager,” Henry explained. “I will also ask about the process itself to ensure that potential impurities are captured in the specification. I expect that their assays are validated and that they have stability data for the container and closures, and will request test methods and elemental impurity profiles.”
Qualification is more than a process to minimize the end user’s test burden. “It’s an ongoing process,” Henry advised. “As materials are used long term in the commercial product, new issues will be discovered and additional controls may be required. As a consequence, risk assessments should be revisited periodically to ensure that the qualification of the materials included in the commercial process are still appropriate and the initial reasoning is still valid.”
References
1. International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). Q11 Development and Manufacture of Drug Substances (Chemical Entities and Biotechnological/Biological Entities), (November 2012), accessed November 10, 2017.
2. D. Strauss et al., “Characterizing the Impact of Pressure on Virus Filtration Processes and Establishing Design Spaces to Ensure Effective Parvovirus Removal,” Biotechnol. Prog. 33(5), 1294–1302 (September 2017).
