Thesis (DR. P.H.)--University of Michigan.

Thesis (DR. P.H.)--University of Michigan.

Background Swine influenza virus (SIV) was first isolated in the United States in 1930 and was thereafter widely reported in many countries. Most SIVs that have been identified are influenza A viruses. There was no report of influenza B viruses isolated in swine. Seroepidemiological study in UK has shown a low seroprevalence of influenza B antibody in pigs. The primary serological test used to detect influenza antibody is the hemagglutination inhibition (HI)test. Enzyme-linked immunosorbent assay (ELISA) are also available commercially for detection of antibodies against influenza A viruses but not for the detection of influenza B antibodies. Objectives 1) To examine the prevalence of influenza B antibodies in pig sera sampled at the abattoir in Hong Kong. 2) To develop the cell-based ELISA assay for the detection of antibodies against influenza A and B viruses. 3) To compare the cell-based ELISA assays with three commercial ELISA kits, namely the IDVet ID Screen influenza A antibody competition ELISA, the IDEXX Influenza A Ab test and the IDEXX AI MultiS-Screen Ab test using swine sera. 4) To test swine sera using the influenza B cell-based ELISA assay to complement data on swine seroprevalence obtained with HI tests. Methods The first part of this study involved HI screening of 4643 pig sera from 2009 to 2012. These sera were tested for the presence of antibodies against B/Brisbane/60/2008 and B/Wisconsin/1/2010whichrepresent the B/Victoria and B/Yamagata lineages respectively. The second part of this study involved the development and performance evaluation of the cell-based ELISA assays. The cell-based ELISA assays were developed using influenza virus infected cells as the capture antigens and fluorescence-labelled anti-IgG antibody as the detection antibody. The viruses that were used to prepare the assays were A/California/04/2009, B/Brisbane/60/2008 and B/Wisconsin/1/2010. All three cell-based ELISA assays were tested with WHO reference sera and swine sera and the results were analyzed using paired t-test and receiver operating characteristic analysis. In addition, the results of the influenza A cell-based ELISA assay were compared with the commercial ELISA assay using Fisher’s exact two-tailed test, Pearson’s correlation analysis and Bland-Altman plot. Results A low prevalence (0.28%; 95%CI: 0.16%-0.47%) of influenza B antibody was observed inthe swine sera samples. The seroprevalence for B/Victoria was higher than that of B/Yamagatain 2010to2012. Co-existence of B/Victoria and B/Yamagata antibodies were found in the swine population during 2010 and 2011. The influenza A cell-based ELISA was found to have low sensitivity (64.1%;95%CI: 52.4%-74.4%) and high specificity (94.7%; 95%CI:80.9%-99.1%) when compared with the commercial ELISA assays. In contrast, using HI as the reference test influenza B cell-based ELISA prepared using B/Wisconsin/1/2010 infected cells were shown to have high sensitivity (92.31%; 95%CI:64.0%-99.8%) but low specificity (63.16%;95%CI:38.4%-83.7%) in detection of influenza B antibodies in swine sera. Conclusion Sporadic transmission of influenza B virus may occur in swine but there is no evidence for efficient and sustained transmission of the virus between them. Cell-based ELISA assay prepared with B/Wisconsin/1/2010 may be considered as an alternative screening testprior to HI subtyping. / published_or_final_version / Public Health / Master / Master of Public Health

 The emergence of the pandemic H1N1 2009 virus of swine-origin and its transmission back to swine highlighted the need for global surveillance of swine influenza. Serology can help to address the epidemiological situation of influenza infection. Since typical serology tests such as hemagglutination inhibition or microneutralization assays are subtype and partially virus-lineage specific, it is important to select appropriate viral antigens for such studies. A poorly chosen panel of antigens will lead to underestimation of the seroprevalence. The choice of well-matched antigen is difficult if there is no prior virological surveillance in that area and even if there was virological surveillance data, transient infections may go undetected. Hence an universal influenza A type reactive serological test is needed. While such tests are available for poultry, there is little published data on the performance of these commercial influenza ELISA assays for serology on swine sera. In this study we evaluated 3 commercially available competitive ELISA assays, IDEXX? Influenza A Ab test, IDEXX? AI MultiS-Screen Ab Test and IDVet ID Screen? Influenza A Antibody Competition ELISA kit for detecting influenza type A reactive antibodies in swine. The virus antigens and the serum samples were obtained from a 14-year systematic abattoir-based virological and serological surveillance for swine influenza in southern China. The performance was evaluated by ROC curve and scatter plot, together with other statistical parameters including the Youden index to optimize the cut-off levels. Using the optimized cut-off levels, sensitivity and specificity of the IDEXX? Influenza A Ab test was 86% and 89% respectively; for IDEXX? AI MultiS-Screen Ab Test was 91% and 87% and for IDVet ID Screen? Influenza A was 95% and 79%, respectively. These findings help to provide different cut-off levels to maximize the sensitivity or specificity to suit different purposes. We found that the ELISA assay was useful in detecting serum samples that may be positive for influenza antibody but missed in the serology screening tests due to limitations in the chosen antigen panel. The ELISA assay maybe helpful in global swine influenza surveillance programs. / published_or_final_version / Microbiology / Master / Master of Medical Sciences