Techniques for Virus Removal
Nanofiltration is perhaps the most robust technique for virus removal and is achieved by size exclusion. Many different nanofilters
are offered by key filter manufacturers including Pall, Millipore, Sartorius, and Asahi. Biomanufacturers will typically select
the filter that they are most familiar with; usually the brand they use in their platform process. The "generic" filter may
not always be the most appropriate filter for a given product, hence it is important to evaluate filter performance and optimize
it if necessary during the process develop-ment activities. Filters for mAbs and other products have pore sizes of 15–20 nm.
The aim is to use a nanofilter with the smallest filter pore size that will allow the product to pass through, while still
retaining viral material.
These filters effectively remove small or difficult to inactivate viral contaminants, such as minute virus of mice (MVM) that
cannot be inactivated at low pH. Regulators are keen to see a step within the process that will give, ideally, at least a
four-log reduction of these non-specific model viruses in validation runs. However, the filter pore sizes are small, and filter
performance is influenced by the quality of the feed stream, as well as the quality of the virus spike during validation.
It may not, therefore, be possible to achieve the specified filter capacity in a validation study without appropriate optimization
of the virus spike ratio and the use of a highly purified virus stock.
Biomanufacturers must minimize the filter area they use because of the cost of the filter cartridges. To take advantage of
the true in-use capacity of the nanofilter, test runs need to be carried out carefully. Without careful design of the virus
spiking study, the biomanufacturer may need to increase the filter membrane area from the size they predicted. It is, therefore,
important that the quality of the virus spikes being used in these steps does not compromise the ability to validate the required
capacity for nanofiltration steps.
While all of the available virus filters are able to achieve comparable reductions for model viruses, performance in terms
of flow decay and filter capacity in process use is more variable; for example, the "generic" filter used in a platform process
may not be optimal for the purification of a new product. If that is the case, filters from different manufacturers should
be compared to identify which brand is most compatible with each product.
Another technique, chromatography, also removes viruses through separation, but it is generally not considered to be sufficiently
robust because operational parameters vary greatly in terms of flow rate, operating capacities, buffer pH, and conductivities.
It would be difficult and expensive to demonstrate definitively that any changes in these parameters had no impact on the
effectiveness of viral clearance. Instead, chromatography steps tend to be used for additional viral reduction, on top of
inactivation and filtration, instead of as the main method for removal.
Depending on the conditions specified for the chromatographic step and the nature of the chromatography, the virus may bind
to the column while the product flows through. The virus is then removed from the column by a regeneration procedure. Other
chromatographic procedures separate the virus and product by differences in the binding affinity to the chromatographic matrix.
Removal of the virus can also be implemented through a combination of removal and inactivation, as is the case for the Protein-A
affinity step. In such cases, it is important to be able to discriminate between the contributions of the two mechanisms to
the removal of the virus. This can be achieved by implementing a PCR-based assay detecting the viral genome, which does not
discriminate between infectious and inactivated virus.
The viral safety of biological products is carefully controlled and regulated because of the potential impact contaminants
can have on patients. Although there have been several examples of viral contamination in continuous cell culture, no biopharmaceutical
product manufactured in this way has been implicated in the transmission of infectious virus to humans thus far. Continued
control and vigilance is still important none-theless as is the implementation of new analytical tools that detect previously
unknown viral contaminants.
Kate Smith is principal scientist, development services, at BioReliance, firstname.lastname@example.org
1. S. Smith, "Process Management Applications in Biopharmaceutical Drug Production," thesis MBA and MSc in Engineering Systems
(Massachusetts Institute of Technology, June 2011) http://ow.ly/n53L0, accessed July 15, 2013.