An improved membrane-filtration enzyme immunoassay for the rapid serological diagnosis of viral infections

Removal of small nonenveloped viruses by antibody-enhanced nanofiltration during the manufacture of plasma derivatives

BACKGROUND

Filters with nominal pore sizes in the nanometer range are well-established tools for enhancing the virus safety margins of plasma-derived products, yet intrinsically less successful for smaller viruses such as hepatitis A virus (HAV) and human parvovirus B19 (B19V). The formation of virus-antibody complexes increases the effective size of these smaller viruses and would thus improve their removal by nanofiltration. While the principle of virus removal by antibody-dependent nanofiltration has been demonstrated with animal antisera and viruses spiked into human plasma product intermediates, the significance of these results remains unclear due to the potential contributions of xenoanti-bodies and/or heteroagglutination in such heterologous systems.

STUDY DESIGN AND METHODS

The current study investigated antibody-dependent virus removal by nanofiltration in a heterologous animal parvovirus system to establish the concentration dependence of the effect. In addition, the phenomenon was investigated in a homologous system with custom-made HAV and B19V antibody-free and -containing human immunoglobulin intermediates. Viruses were analyzed with infectivity assays and fully validated polymerase chain reaction assays that also circumvent the obscuring effects of neutralizing antibodies with infectivity assays.

RESULTS

By use of the heterologous mice minute virus and the homologous HAV and B19V systems, viruses passed the 35-nm (Planova 35N) filter in the absence of specific antibodies. Beyond a threshold virus antibody concentration, nanofiltration resulted in effective virus removal of viruses smaller than the nominal pore size of the filter used.

CONCLUSION

HAV and B19V are effectively removed by antibody-dependent 35N nanofiltration, already at intermediate antibody concentrations well below those comparable to human plasma pools for fractionation.

Measurement of IgM responses to a subunit influenza A vaccine by sucrose-gradient centrifugation and membrane-filtration enzyme immunoassays

IgM responses to a deoxycholate-split influenza vaccine containing the surface antigens of the H3N2 virus A/Philippines/2/82 were studied in five volunteers, three of whom were seronegative by haemagglutination inhibition (HI) tests. Responses were measured by a sucrose-gradient centrifugation technique, in which IgM-specific HI activity was computed as a proportion of total IgM and IgG-specific HI activity, and by a membrane filtration-enzyme immunoassay (MF-EIA). Responses could be detected in all volunteers when measured by sucrose-gradient centrifugation within 1-2 weeks, and the IgM induced was 5-40% of total HI-specific activity. The response was also observed in the presence of low levels of pre-existing antibody. Levels of IgM antibody, when measured by MF-EIA, could be easily detected in two of the volunteers, while those of two others were very low and there was no response in a fifth. No biphasic virus-specific response to vaccination, involving first IgM and then IgG, could be measured by either technique. From these studies, the sucrose-gradient fractionation technique appears to be the more sensitive procedure.