Evaluating the validity of using plasma samples from blood donation archive for influenza sero-epidemiology

Abstract

BACKGROUND: Serology is often regarded as the most reliable method for inferring previous exposure to and estimating population immunity against influenza. Serum is the standard specimen in serological testing, whereas plasma is increasingly preferred as the type of blood sample in human research projects. To ensure blood product safety, blood donation services in developed countries typically retain one plasma sample from each donation for at least 12 months. Upon their expiry, these archived plasma samples can potentially be used for influenza seroepidemiology. However, plasma samples might not necessarily be suitable for influenza serological testing because they contain anticoagulants (citrate, heparin, EDTA), which are known to interfere with the reaction between antigen and antibody. In this study, we compare the hemagglutination inhibition (HAI) and microneutralization (MN) titers of matched serum and EDTA-plasma samples that were obtained from the same blood donation. METHOD: Serum samples were collected from blood donors in the Hong Kong Red Cross Blood Transfusion Services (HKRCBTS) in our previous influenza seroepidemiologic study during the 2009 pandemic in Hong Kong. EDTA-plasma samples that were archived by the HKRCBTS during 2009-2011 were harvested for a separate influenza seroepidemiologic study. Samples from 609 blood donations were found in both seroepidemiologic studies. We perform HAI and MN assays on these matched serum and EDTA-plasma samples against A/California/7/2009 (pandemic H1N1) and A/Victoria/208/2009 (seasonal H3N2). Following conventional practice (to allow for inter-experiment and inter-laboratory variations), we regard titers that differ by one dilution factor or less as identical. RESULTS: The Pearson correlation of titers between matched serum and EDTA-plasma was 0.95 (95% CI: 0.94-0.96) for MN against pandemic H1N1, 0.82 (95%CI: 0.79-0.84) for HAI against pandemic H1N1, and 0.86 (95% CI: 0.84-0.88) for HAI seasonal H3N2. CONCLUSION: We conclude that EDTA-plasma samples could be used for influenza seroepidemiologic study of pandemic H1N1 using MN assay while HAI titers of EDTA-plasma samples should be interpreted with caution because a significant proportion of samples that were HAI seronegative for serum (titer < 1:10) had positive HAI titer for matched EDTA-plasma. Further comparisons would need to be performed for other strains of influenza to establish the general validity of using plasma samples from blood donation archives for influenza sero-epidemiology