Send to printer »

Feature Articles : Mar 1, 2006 (Vol. 26, No. 5)

QA/QC for Viral Clearance

Methodologies Are Evolving, But No Clear Trends Are Emerging
  • Gail Dutton

Viral clearance processes and the issues surrounding them are becoming more sophisticated but, as yet, few clear trends are emerging. However, industry experts are noting some currents and eddies that are beginning to swirl around the issue.

For example, the biotech industry is giving greater attention to raw materials. Companies, where possible, are moving away from animal- or tissue-derived raw materials to minimize the risk of adventitious virus contamination, notes Mani Krishnan, program manager, Millipore (www. millipore.com).

There also is a tendency to use set, defined studies for Phase I trials, which then are expanded in Phase III for global submissions, according to Joe Hughes, Ph.D., executive vp and general manager, Apptec (www.apptec-usa.com).

The amount of viral clearance needed at early phases of clinical development can be less than at late clinical development. The amount of clearance from individual steps depends on the input titer and volume of sample that you titrate, notes Martin Wisher, Ph.D., scientific director, BioReliance, Invitrogen bioservices (www. bioreliance.com).

Missteps

There also is some indication that companiesparticularly young firms with little or no experience in viral spiking studiestend to conduct more viral clearance studies than actually are needed, notes Barry Rosenblatt, Ph.D., director of technical services, Charles River Laboratories (www.crl.com).

Conversely, some new companies underestimate what they need to do for submissions to the U.S. and the E.U., including the time and cost to perform these studies. They dont realize that they need to plan well ahead.

To minimize such mistakes, companies need to evaluate viral clearance methods to ensure theyre applicable, Dr. Rosenblatt says, and thereby avoid situations in which viral clearance methods are ineffective for the virus in question. New regulatory people tend to be stricter than their more experienced colleagues when interpreting guidances, thus causing firms to be more cautious.

Dr. Rosenblatt warns against the temptation to invest in unproven new viral reduction technologies. If there are no calls for it, it wont go to market, so vendors will try to build an audience of early adopters to enhance their new products marketability. If its five years away from market, dont bother trying it.

Q-PCR, however, is proven, popular, and gaining acceptance. Many industry leaders are beginning to favor Q-PCR to either replace or complement infectivity assays.

Validation Emphasized

According to Dr. Wisher, Before licensing applications, regulatory bodies are asking that robustness studies for downstream processing be performed, including chromatographic resin re-use and sanitization studies. The EMEA guidance document on viral clearance studies for early-phase clinical trials is eagerly awaited, and the EU and Japan are wanting validation of plasma protein purification steps for TSE removal.

With the increased emphasis on validationparticularly for Phase III studies and product licensure applicationsmanufacturers are increasing their use of new and used resin studies, Dr. Hughes says. The object is to prove that well-used resins from chromatography columns are as effective as new resins at viral clearance.

The bracketing matrix approach championed by Genentech (www.gene.com) and Kurt Brorson, staff scientist, the Office of Biotechnology Products, CDER, FDA, uses data gathered during past studies with similar process steps and product to justify not having to perform new viral clearance studies.

As companies gain experience with products such as Mabs and r-proteins derived from well-defined cell lines, it is likely that they will be able to use their in-house databases to support virus reduction claims of their processes, at least for early development and human trials, Krishnan suggests.

There also is increased awareness of the quality of the virus preparations used in the validation studies and their impact on the results, he continues. The purity of the virus preparations can have a significant impact on the filter throughput achieved with some virus filters. In some cases, the quality of the virus preparations also can affect the log reduction value obtained with the virus filters. This has both economic and regulatory implications.

To complicate the matter, Jerold Martin, senior vp, scientific affairs, Pall Life Sciences, biopharmaceuticals (www.pall.com), notes that regulators are seeing different results from what appear to be identical studies by different manufacturers. The differences in outcome appear to be caused by differences in the way the virus solutions are prepared. Theres no standard regarding the preparation of viral spike material, and those differences may affect the materials sensitivity to pH, for example, thus altering the kinetics, says Martin.

There also are small differences in the running parameters that can change viral clearance results, notes Dr. Wisher, co-chair of the Parenteral Drug Associations Virus Preparation Standardization Task Force, which is just beginning to address the preparation of viral spiking material.

Martin notes that the more we look at the issue, the bigger it gets. The worst case for viral spiking for a membrane filter is for the virus not to aggregate, because filters appear to perform better when the particles aggregate. But, the unaggregated particles are more sensitive to heat or other clearance methods. The worst case for one viral clearance method isnt the worst case for others. Therefore, he says, it is likely that viral spiking preparation standards will be developed for each different clearance process.

How much Clearance?

The question of how much viral clearance is enough is the million dollar question, according to Luke A. Pallansch, Ph.D., director, pharmaceutical research services at Cell Trends (www.celltrends.com).

According to Dr. Rosenblatt, The answer is readily available if you know the parameters, noting that it depends upon the number of virus particles per unit volume detected in the feed stock, the anticipated dose, process productivity, and cell line yield. Consequently, Dr. Pallansch believes, fixed log reduction values probably would not be appropriate for all manufacturer biopharmaceuticals. Instead, he advocates understanding the critical parameters of individual unit operations within a manufacturing process in order to assure the robustness of viral clearance associated with those unit operations.

The fairest answer, says James Gilbert, Ph.D., senior director, MDS Pharma Services Biopharmaceuticals (www. mdsinc.com), is that there should be a demonstration of sufficient viral clearance by the manufacturing process to provide a significant margin of safety in potential viral load based upon the viral content of the starting materialgenerally below the limits of assay detection. Of course, what is a significant margin of safety?

FDA and ICH guidances indicate the margin of safety has been three to five log reduction value higher than the estimated viral particles in one dose.

For Mabs and recombinant proteins produced using well-defined cell lines, such as CHO and NSO, this roughly translates to about 1215 log10 clearance for endogenous retroviruses and about 6 log10 clearance of adventitious viruses, Krishnan elaborates. Typically, x-MuLV is used as a specific model virus for endogenous retroviruses, and MMV is used as a relevant adventitious virus. That said, Dr. Hughes notes that, because of high endogenous retroviral particles or high dose levels, some products have a target of 1622 log10 clearance.

Tissue samples, however, are a little different, and are undergoing some changes, Dr. Hughes says. Historically, viral clearance has been achieved by chemical inactivation, followed by titration to determine how much virus remained. Now regulatory bodies question whether the tissue absorbed and protected the virus, so you must measure the inactivation by taking off the supernatant, disrupting the tissue sample, and taking a second sample to measure any protective effect from the tissue.

Dr. Gilbert suggests that the desired extent of viral clearance be determined by case-by-case evaluation of the data, testing results, nature of the product, nature of the starting material, and a variety of other considerations. The important thing, in my mind, is for drug developers to get guidance from the appropriate regulatory agencies on the expectation of the agency.

That guidance, however, is often open to interpretation. As Martin says, None of the regulatory agenciesin the U.S. or internationallyhave documented how much viral clearance is needed. It is done on a case-by-case method.

The Committee for Proprietary Medicinal Products of the EMEA, for example, says a given viral clearance step with 4 log reduction is robust. The industry widely interprets that standard as requiring a 4 log reduction but, that same document later states that if log reduction is less than 1 log is it not significant. Therefore, viral clearance doesnt have to be 4 logs, but is better if it is, Martin concludes.

Regulatory bodies dont want manufacturers to rely upon such a number analysis, however, Dr. Wisher says. Companies sometimes lack a good sense of their goals for target removal based upon dosage size and, therefore, may not have planned enough clearance steps. Most have a five or six log safety factor, so with a viral load of 1012 particles per liter, and a safety factor of 5 logs, the target viral clearance will be 1017. To achieve this, you have to add three or four steps together, Dr. Hughes says.

Regulators want manufacturers to perform a virus risk assessment and to discuss the viral clearance obtained in the light of this assessment, Dr. Wisher says. In that vein, The European Pharmacopoeia has published a draft monograph that requires all manufacturers of biological products to prepare a virological risk assessment.

Removal vs. Inactivation

The debate between the relative merits of viral inactivation and viral removal is waning as companies take an orthogonal approach that uses both, because none of the viral clearance methods employed to date can be expected to remove all viruses present in a given process, Dr. Pallansch says.

Inactivation and removal methods are limited by the detection system. For removal, the assumption must be that any viruses present at concentrations below the detection limit would be active, Dr. Gilbert says. Conversely, If you inactivate virus below the limit of detection, are there still active viruses lurking?

The method of clearance, clearly, depends upon the threat and a products stage of development. Viral inactivation using low pH or using chaotropes is relatively inexpensive when compared with virus clearance using filtration, but inactivation is generally effective only against enveloped viruses. Krishnan says. In contrast, filtration can remove both non-enveloped and enveloped viruses and is relatively insensitive to small changes in protein concentration, temperature and other process conditions.

European regulators and guidance documents have expressed the view that viral inactivation steps are more robust than removal steps. There are probably fewer variables with inactivation than with a column chromatography step, Dr. Wisher says, but, chromatography steps can give reproducible viral removal and can contribute to overall clearance.

Biotech drug producers tend to have a preference to avoid inactivation other than low pH, which often is intrinsic, because chemical inactivation requires more complex purification to remove the inactivating agent, and physical inactivation can reduce biological activity of the product. Thus, the preference for added clearance steps for biotech tends to lean toward removal by virus membrane filtration, says Martin.

Additionally, Martin says, inactivation for steps beyond intrinsic clearance is perceived to be less expensive. Inactivation appears to be the first choice for plasma derivatives, followed by removal by virus membrane filtration. There are, however, no universal virus inactivation or removal methods, so often inactivation and removal are used complementarily.

Design Early

Viral clearance studies should be designed as early as possible. As Dr. Gilbert advises, A strong collaboration between the process development scientist and the viral clearance investigator is key to good design, because the nature of the method to address each process step is dictated by the nature of the process step itself.

Integration, Dr. Pallansch notes, does not always occur because pharmaceutical and biotech companies frequently in-license biotherapeutics that are developed in academic settings where the research program is frequently focused on bioactivity rather than biosafety.

Good process development for biopharmaceuticals involves designing purification processes for high yield, low cost, and efficiency, as well as for product viral safety from the very beginning, Dr. Gilbert comments. Advice from a viral clearance expert can be particularly important for novel processes or novel steps within a process.

For retrovirus clearance, the FDA requires that clearance be demonstrated before Phase I trials begin. Likewise major regulatory bodies in Europe and North America require that companies demonstrate that their purification process has the ability to clear endogenous and adventitious viruses before the drug product receives marketing authorization, Krishnan notes.

The basis for evaluating viral clearance methodologies will continue to evolve as the risk-based approach of the Pharmaceutical cGMP Initiative encourages the use of existing and emerging science to ensure that limited resources are best targeted to address manufacturing issues and biopharmaceutical safety/quality, Dr. Pallansch concludes.