Short Report: Pre-Existing Cross-Reactive Antibodies to Avian Influenza H5N1 and 2009 Pandemic H1N1 in US Military Personnel

We studied cross-reactive antibodies against avian influenza H5N1 and 2009 pandemic (p) H1N1 in 200 serum samples from US military personnel collected before the H1N1 pandemic. Assays used to measure antibodies against viral proteins involved in protection included a hemagglutination inhibition (HI) assay and a neuraminidase inhibition (NI) assay. Viral neutralization by antibodies against avian influenza H5N1 and 2009 pH1N1 was assessed by influenza (H5) pseudo- typed lentiviral particle-based and H1N1 microneutralization assays. Some US military personnel had cross-neutralizing antibodies against H5N1 (14%) and 2009 pH1N1 (16.5%). The odds of having cross-neutralizing antibodies against 2009 pH1N1 were 4.4 times higher in subjects receiving more than five inactivated whole influenza virus vaccinations than those subjects with no record of vaccination. Although unclear if the result of prior vaccination or disease exposure, these pre- existing antibodies may prevent or reduce disease severity. Outbreaks of 1997 avian influenza H5N1 and 2009 pandemic (p) H1N1 in humans have provided an opportunity to gain insight into cross-reactive immunity. The US military periodi- cally collects and stores serum samples from service members linked to medical records. 1 We measured cross-reactive anti- bodies in stored serum to avian influenza H5N1 and 2009 pH1N1 from US military personnel and identified factors associated with presence of neutralizing antibodies. Two hundred archived serum samples were obtained from the US Department of Defense Serum Repository. They were representative of a wide cross-section of active military person- nel at the times of collection, whereas specific geographic infor- mation was not available on the individual selected; the cohort represents the general US military population, which is deployed throughout the United States and globally. Fifty sam- ples each were selected from four birth cohorts: (1) < 1949, (2) 1960-1965, (3) 1966-1971, and (4) 1972-1977. Within each cohort, 25 samples were collected in the year 2000 (before the introduction of intranasal live attenuated influenza vaccine (LAIV)), and 25 samples were collected in 2008 (where 51% ofdonorshadreceivedLAIV).IthasbeensuggestedthatLAIV elicitscross-reactiveimmunity. 2,3 Thesampleswereallcollected before the outbreak of 2009 pH1N1, and there have not been any reported outbreaks of H5N1 in US military personnel. Assays used to measure antibodies included a hemagglutina- tion inhibition (HI) assay and a neuraminidase inhibition (NI) assay. 4 Viral neutralization by antibodies against H5N1 and 2009 pH1N1 was assessed by influenza (H5) pseudotyped lentiviralparticle-based(H5pp) 5 andmicroneutralizationassays, respectively. Electronic medical and vaccination records from the Defense Medical Surveillance System (DMSS), which cap- tured records before the serum sample date, were linked to samples andcomparedwiththeinvitroresults. 1 The odds ratios (ORs) and 95% confidence intervals (95% CIs) of univariate and multivariate binary logistic regression analyses were used to determine the association between donor characteristics and positive antibody responses. A multiple logistic regression model was constructed, and it included inde- pendent variables with a P value of < 0.05 in univariate logistic regression. A P value of < 0.05 was considered to indicate statistical significance. SPSS 12.0 for Windows (SPSS Inc., Chicago, IL) was used to perform all statistical analysis. Cross-reactivity is summarized in Table 1. Although HI assay titers to H5N1 were uniformly low (0.5%), neutralizing antibodies were considerably higher: 14% for the more sensi- tive H5pp assay 5 and 22.5% for the NI assay. H5pp and NI antibody titers to H5N1 were evenly distributed among birth cohorts and did not differ substantially based on history of vaccination or prior respiratory infections. Of those individ- uals with neutralizing antibodies to H5N1 (N = 28), 32.1% also had neutralizing antibodies to pH1N1, whereas 19.3% of those individuals with any H5N1-specific antibody response also had neutralizing antibodies to pH1N1 (Table 1). As with H5N1, samples with positive HI titers were low for 2009 pH1N1 at 5.5%, whereas neutralizing antibody titers were higher, with 16.5% positive in the microneutralization assay but only 9% positive in the NI assay. Positive neutrali- zation titers were less evenly distributed among birth cohorts, with only 4% positive in the 1972-1977 birth cohort, whereas 30% were positive in the 1960-1965 cohort. Like H5N1, pos- itive antibody titers to 2009 pH1N1 did not differ substantially based on history of vaccination or prior respiratory infections. Of those individuals with neutralizing antibodies to pH1N1 (N = 33), 27.3% also had neutralizing antibodies to H5N1, whereas 28.9% of those individuals with any pH1N1-specific antibody response also had neutralizing antibodies to H5N1. Univariate associations between the prevalence of cross- reactive antibodies to H5N1 and 2009 pH1N1 and indepen- dent variables, including year of birth, serum collection year, sex, and seasonal influenza vaccination history, are shown in Table 2. The odds of having cross-neutralizing antibodies against 2009 pH1N1 were threefold higher in those donors who received inactivated whole influenza virus vaccine more than five times than those donors with no record of vaccina- tion (95% CI = 1.1-8.9). In relation to subjects born between