Value of Virus Filtration as a Method for Improving the Safety of Plasma Products
T. Burnouf
DOI: https://doi.org/10.1111/j.1423-0410.1996.tb01334.x
1996-05-01
Vox Sanguinis
Abstract:Recently, Roberts [I] stressed some of the potential benefits of virus filtration to improve the safety of plasma products. However, some points in his interesting letter deserve deeper consideration when assessing the potential of this technology. Roberts’ paper focuses largely on the issue ofwhether or not 35-nm filters eliminate parvovirus Bl9, depending its aggregation state. Although there is presently a great deal of concern about the residual risk of transmission of small, non-enveloped viruses from plasma derivatives, it might be an error to focus or limit R&D programmes only towards the reduction of parvovirus B19 at the expense of effective reduction treatment of other viruses. Human plasma being a raw material which can be contaminated by new infectious agents, one should, not reject nanofiltration ofhigh-molecular-weight plasma protein concentrates on ‘large-pore’ filters only because it might not eliminate parvovirus B19. It should be kept in mind that viral validation experiments indicate that nanofiltration even on ‘large-pore’ filters, may be the only current technique completely eliminating 5 to 7 logs of infectivity of highly resistant non-enveloped model viruses such as SV40 and reovirus type 3 [2], and possibly the Creutzfeld-Jacob agent [3]. In other words, fractionators only targeting their attention on the reduction of B19 might make wrong strategical choices with regards to potential future pathogenic agents and the overall safety profile of plasma products. Our experimental data on highly-purified FIX and FXI concentrates show that nanofiltration on 35-nm filters may remove large infectious doses of bovine parvovirus. We have obtained similar clearance factors with two other high-purity FVlll and von Willebrand factor concentrates. As we have suggested, parvovirus removal could be due to aggregation [2] either during virus stock preparation, or after spiking it in low-protein-content solutions, or both. Roberts is certainly right when he says that the potential of nanofiltration on ‘large-pore filters’ to remove B19 is likely to be influenced by many factors. The same indeed holds true for all viral reduction treatments because their efficacy is influenced by several parameters: the efficacy of a pasteurisation treatment is directly influenced by the kind of stabilisers used, or the protein content of the solution, that of a severe dry-heat treatment by the residual moisture content of the product after freeze-drying. In other words, it can be said that there are as many heat treatment procedures as there are products that are heat-treated. Each viral reduction process must be validated on a case-by-case basis and experimental data showing elimination of bovine parvovirus on a 35-nm filter for a given product might not be confirmed for another one having different physicochemical characteristics. It must also be stressed that small-scale viral validation experiments must be carried out under conditions mimicking as closely as possible those to be performed at the production scale. Roberts also states that nanofiltration of high-molecular-weight protein concentrates on ‘large-pore filters’ (35 nm) ‘only’ provides additional security against the large enveloped viruses when products have already been treated with a specific virus reduction step. However, the benefit of nanofiltration to further limit the risk of transmission of the major plasma-borne viruses should not be undervalued, considering the episodic transmissions of HIV, HCV, or HBV which have been suspected in the last few years from prothrombin complex concentrate, FVIII, or intravenous immunoglobulins G preparations [ 4 7 ] . Nanofiltration can act as a filter clearing protein solutions of residual viruses that may not have been properly inactivated or eliminated during the earlier phase of the production process or have been accidentally introduced. Implementation of additional viral inactivatiodelimination treatments, as strongly recommended by regulatory agencies, is mandatory but must be made following sound criteria. Indeed, it would be risky to select, for example, a heat tratment as an additional viral-reduction step of a given SD-treated product only because of the potential of such treatment to inactivate HAV or B19 (although B19 is known to be relatively heat-resistant) if this might induce risks of alteration of plasma protein structure, increase immunogenicity and induce inhibitors in the patients. In other words, a delicate balance must be found in choosing a new virucidal treatment between its benefits in terms of viral safety and the risk of neoantigen formation. In this respect, current in vitro [2] and in vivo [8] data indicate that nanofiltration does not induce alteration of plasma proteins, presumably because it is carried out under mild physiological conditions in terms of pressure, pH, osmolarity, and temperature. In conclusion nanofiltration appears to be a mild, highly effective viral-reduction procedure which could be essential in the improvement of the overall safety of plasma as well as other biological products. As with any viral reduction process, its efficacy must be validated on a case-by-case basis and confirmed by clinical experience. In the meantime, R&D programs to find and develop other, new virucidal processes must be encouraged.