The primary endpoint for the IgA analysis was the ratio of influenza-specific IgA against A/H1N1, A/H3N2, or B strains in the vaccine to total IgA antibody. Geometric mean titers (GMTs) of absolute strain-specific IgA and total IgA were also evaluated at all time points. For strain-specific and total IgA, values for samples with no IgA were selleck chemicals imputed as 50% of the minimum detectable value. Detailed methodologies and specific reagents used for this analysis are available in Supplementary Text 1. Serum antibody titers were evaluated by HAI assay using standard methods, as previously described [14] and [20]. Seronegative subjects were

defined as those with a prevaccination HAI antibody titer of 4 or less; seropositive subjects were those with a titer greater than 4. An HAI response was defined as a 4-fold increase from prevaccination to postvaccination. For descriptive purposes, the IgA response was categorized using 3 measurements: the percentages of subjects with ≥2-fold and ≥4-fold increases in the ratio of strain-specific to total IgA from baseline and the geometric mean fold rise (GMFR) in the ratio of strain-specific to total IgA from baseline. Results were evaluated separately for each study. The correlation between nasal IgA and serum HAI antibody

responses was Nutlin-3a chemical structure evaluated across studies for each influenza type/subtype. To examine the relationship between IgA and the incidence of influenza illness, geometric mean postvaccination IgA ratios were compared between subjects with culture-confirmed influenza illness and those without evidence of culture-confirmed influenza illness. Influenza illness was evaluated for any influenza strain regardless of antigenic match to the vaccine as well as due to vaccine-matched strains. LAIV and placebo recipients were evaluated separately for each study. Additionally, given

the small size of the immunogenicity DNA ligase cohorts in each study and the similarities in the design of the studies, a pooled analysis of all 3 studies was conducted to increase the statistical power to detect an effect. Only studies with at least 1 case of influenza illness were pooled. Statistical comparison tests were conducted at the significance level of 0.05 using Fisher’s exact test for the proportion of subjects with a ≥2-fold increase in titers and using the two-sample t-test for GMFRs and geometric means. In year 1, there were 183 (107 LAIV, 76 placebo), 101 (64 LAIV, 37 placebo), and 333 (226 LAIV, 107 placebo) subjects in studies 1, 2, and 3, respectively, with IgA data available for analysis. In year 2, there were 175 (94 LAIV, 81 placebo), 41 (24 LAIV, 17 placebo), and 791 (528 LAIV, 263 placebo) subjects in studies 1, 2, and 3, respectively. In each study, LAIV and placebo recipients were well-matched in regards to age and sex.