Epidemiological Investigation On Infections Of Influenza A Virus In Poultry And Occupational Exposed Human Population

Recently, there have been cumulating reports on events of human infection withavian influenza viruses (AIVs). Since its emergence in human in1997, avianinfluenza (H5N1) virus has caused658laboratory-confirmed human cases in15countries. Of these cases,388have died, with a case fatality rate (CFR) of nearly60%.HPAI H7N7also has been reported to infect human and cause1death in Netherlandsin2003. There also have been reports on human infection with avian influenza (H9N2)and (H7N3) virus in Hong Kong and Canada, respectively. In early2013, avianinfluenza A (H7N9) virus emerged in Southeast China. As of28February2014,375laboratory-confirmed human cases of avian influenza A (H7N9) have been officiallyreported to WHO and115of them have died. Therefore, infection of influenza A virusamong poultry and human has been paid much attention.The detecting method is essential in the field investigation of infection with avianinfluenza A(H7N9) virus. To find out a rapid, efficient method to detect H7N9virus infield studies, we evaluated five commercial influenza A assays which were commonlyused to detect influenza in humans: Life Technologies VetMAXTM-Gold SIVDetection Kit, Quidel Molecular Influenza A+B assay, Remel Xpect Flu A&B assay,Quidel Quickvue influenza assay and Quidel Sofia Influenza A+B assay. Firstly,cloacal swabs were collected from live poultry in markets or farms with active poultryoperations in Nanjing and Wuxi, Jiangsu Province. RNA was extracted and screenedwith the WHO qRT-PCR assays for any influenza A and avian H7N9virus.Of the specimens initially screened,27were found to be H7N9-positive, which werepaired with70random-number selected specimens that were negative by both theWHO influenza A and H7N9assays. Later, these97samples were selected for use inthe second round of testing. This final panel of swab specimens was tested in ablinded fashion employing the above mentioned5commercial influenza A assays andthe WHO qRT-PCR H7N9assay (repeat testing). We ran each antigen detection assaytwice and each molecular assay three times. Sensitivity, specificity, and confidenceintervals around each parameter were calculated employing the WHO H7N9assay asthe gold standard using SAS v9.3(SAS Institute, Inc., Cary, NC, USA). Results:(1)There was excellent concordance between the first and second round of WHO H7N9assays.(2) The Quidel Quickvue influenza assay failed to detect any influenza Aspecimens among the97specimens in the panel. The other four assays had excellentspecificity, ranging from96to100%.(3) The Life Technologies VetMAXTM-Gold SIV Detection Kit had the highest sensitivity at100%(95%CI0.87–1.00), regardlessof a false positive probability of4%(95%CI1–12%). The Quidel MolecularInfluenza A+B assay had a sensitivity of85%(95%CI66–96%) and a false positiveprobability of3%(95%CI0–10%).To understand the status of AIVs infections among poultry in Wuxi, we monthlycollected swabs from poultry and related environment in live poultry markets (LPMs)and confined poultry feeding operations from June to December,2013. Using RNAextracted from the specimens, WHO qRT-PCR influenza A assay was used to screenfor influenza A viruses. The positive samples were further processed with reversetranscription-PCR to amplify the whole genome of the virus, as well as the genecloning and sequencing. The available sequences were used to construct phylogenetictrees. Meanwhile, we tried to isolate the viruses with MDCK and chick embryo.Results:(1) A total of305samples were collected and28.5%(87/305) were influenzaA positive. The positive rate of influenza A in samples from LPMs was higher thanthat of confined poultry feeding operations (31.9%vs.10.4%，χ2=9.16，P=0.002).（2）Six samples can be subtyped: three H9N2, one H5N8, one H11N2and one H5N1.Phylogenetic analyses demonstrated that the H5N8and H11N2were reassortant AIVs.（3）Three isolates were obtained: A/Chicken/1/Wuxi/2013(H9N2),A/Chicken/2/Wuxi/2013(H9N2) and A/Chicken/1/Wuxi/2013(H5N1).Poultry can be infected with various AIVs and occupational populations exposedto poultry are at high risk of AIVs infection. Human infection with AIVs can besymptomatic or even fetal, however, asymptomatic infection with subtypes of H5, H7,H9and H11can also be found in occupational population. The special susceptibilityof pigs to infection with swine, avian and human influenza viruses makes them veryimportant as intermediate hosts in the transmission of influenza viruses. Researchershave demonstrated that pigs can be infected with HPAI H5N1and LPAI H9N2.Approximately1%of swine-exposed occupational population in Guangdong, Fujianand Zhejiang were infected with H9N2. To explore the status of influenza A infectionamong poultry and swine occupational population, to explore the relationship betweeninfluenza A infection and their demographic and epidemiological characteristics, andto identify the risk factors of AIVs infection and to provide valuable basis for theprevention and control of AIVs, we carried out a cross-sectional serological survey ofinfluenza A viruses among occupational populations in cofined poultry (swine)feeding operations, slauterhouses, markets and backyard raising farms in Wuxi City. Atotal of1271subjects including498poultry-exposed population,608swine-exposedpopulation and165general controls were recruited. Venous blood was collected andserum was abtained for testing of antibodies against seasonal influenza A/H1N1(sH1N1), seasonal influenza A/H3N2(sH3N2), HPAI H5N1, LPAI H9N2and H7N1 by hemagglutination inhibition (HI) assay and microneutralization (MN) assay.Results:(1) Seropositive rate of sH1N1of the poultry-exposed population,swine-exposed population and general controls were17.7%(88/498),10.9%(66/608)and27.3%(45/165), and the difference was statistically significant (χ2=29.00,P=0.000). Seropositive rate of sH3N2of the three populations were17.3%(86/498),6.4%(39/608) and8.5%(4/165), and the difference was statistically significant (χ2=34.29，P=0.000).(2) The ratio of MN titer≥1:80and≥1:160against H5N1amongpoultry-exposed population was14.6%and8.4%, respectively, and the ratio amongswine-exposed population was0.2%and0.2%, respectively. In the controls, only1subject (0.6%) had an MN titer of1:80against H5N1and no higher titer was detected.Multiple logistic regression showed that poultry-exposed population working inmarkets (OR=57.3,95%CI7.0-466.5), chronic respiratory diseases (OR=12.8,95%CI1.6-102.5) and an age of40-60(OR=3.1,95%CI1.5-6.6) were risk factors for anMN titer≥1:80against H5N1.(3) We further tested the76sera with MN titiertitieagainst H5N1and found that3subjects fulfilled the WHO case definition ofH5N1(MN titer≥1:80and HI titer≥1:160against H5N1). They were2poultrysellers in the market and1worker from cofined poultry feeding operation, consisting0.6%(3/498) of the poultry-exposed population.(4) There were5subjects who hadMN titer≥1:10against H9N2, two of them were poultry sellers in the market, twowere pork sellers in the market and1was slaughterhouse worker. The seropositiverate against H9N2in poultry-exposed population, swine-exposed population andcontrols was0.4%(2/498),0.5%(3/608) and0, respectively.(5) No subject with HItiter≥1:10against H7N1was found in this study.Avian influenza A (H7N9) virus has emerged in China in March2013and spreadrapidly since then. Exposure to poultry or poultry related environment such as livepoultry markets have been recognized as the primary risk factor of H7N9infection.During the H7N9epidemic, is there subclinical infection of H7N9amongpoultry-exposed population? Are poultry workers and non-poultry workers in livepoultry markets and poultry workers outside live poultry markets at equal risk ofgetting infected? What is the relationship between H7N9infection and demographiccharacteristics, chronic conditions and occupational behaviors? Based on thesescientific issues, we conducted a cross-sectional serological survey of influenza A(H7N9) viruse among poultry-exposed population in Wuxi City. A total of1588subjects including300poultry workers and878non-poultry workers,245poultry-exposed workers beyond live poultry markets, as well as165controls wererecruited. Sera were primarily screened for specific antibody against the H7N9virusby HI assay using turkey red blood cells. Sera with HI titers≥1:20were tested byMN assay for confirmation. An MN titer≥1:20was defined as positive. To explore possible confound of cross-reactions from previous infection with human seasonalinfluenza viruses and other avian influenza virus of H7subtype, we also testedantibodies against sH1N1, sH3N2and LPAI H7N1and added the results into thebivariate analyses. Results:(1) For the preliminary antibody screening using HI assay,twenty-seven (1.7%) sera had HI titers≥1:20. Confirmed by MN assay,8subjectsfrom8different live poultry markets were seropositive. Seven were live poultrysellers and one was a bean curd seller. Anti-H7N9seropositive rate among poultryworkers and non-poultry workers was2.3%(7/300) and0.1%(1/878), respectively.None of them had any manifestations indicating ILI in the past12months.(2) Nosubject with HI titer≥1:10against H7N1was found.(2) Bivariate logistic regressionshowed that the risk of having elevated antibodies against H7N9of poultry workers inlive poultry markets was higher than that of non-poultry workers in live poultrymarkets (OR=20.95，95%CI2.57-171.01). No significant differences were foundamong different genders, age groups, education background and financial status.Neither sH1N1nor sH3N2showed statistical relationship with H7N9infection.Conclusion:In examining poultry specimens, the Life Technologies VetMAXTM-Gold SIVDetection Kit and the Quidel Molecular Influenza A+B assays have comparabledetection parameters with the WHO H7N9qRT-PCR assay. Infections of AIVs inpoultry in Wuxi were diverse and there were different recombinants. Infections ofAIVs in occupational populations are common. There is relatively high level of H5N1infection rate and low level of H7N9and H9N2infection rate in poultry-exposedpopulation. There also is low level of H5N1and H9N2infection in swine-exposedpopulation. These infections are possibly asymptomatic or mild. Poultry-exposedpopulation, especially those in live poultry markets and those with chronic respiratorydiseases are the key population of AIVs prevention. It is essential to improvemanagement and disinfection of live poultry markets and promote the use of personalprotective equipment (PPE). Systematic and continuous surveillance of AIVsinfection in poultry and in occupational populations are of great importance for publichealth.Innovations of the research:(1) Using a combination of HI assay and MN assay,we proved that H7N9subclinical infection did exist in poultry-exposed humanpopulation in LPMs and that those outside the LPMs were not infected;(2) We foundsimultaneous infection with different AIVs including H5N1, H7N9and H9N2in thepoultry-exposed human population;(3) The infection rate of H5N1in poultry exposedpopulation was high.(4) There was low level of H9N2infection in swine exposedpopulation;(5) Infections of AIVs in poultry were diverse and there were differentrecombinants. The main limit of the research is that time of AIVs infection can not beconfirmed in a cross-sectional seroepidemiological study, thus only possible previousinfection can be proved.