Immunogenicity and Cross-Reactivity of 2009–2010 Inactivated Seasonal Influenza Vaccine in US Adults and Elderly

The campaign of 2009–2010 Northern Hemisphere seasonal vaccination was concurrent with the 2009 H1N1 pandemic. Using a hemagglutination inhibition (HAI) assay, we evaluated the immunogenicity and cross-reactivity of 2009–2010 inactivated trivalent influenza vaccine (TIV) in US adult and elderly populations. Vaccination of TIV resulted in a robust boost on the antibody response of all subjects to seasonal A/Brisbane/59/2007 (H1N1) and A/Uruguay/716/2007 (H3N2) with over 70% of recipients reaching a seroprotective titer of 40. B/Brisbane/60/2008 was the least immunogenic among the three seasonal vaccine strains with <30% of TIV recipients reaching a seroprotective titer of 40. TIV vaccination also induced a moderate boost on the pandemic specific antibody responses. Twenty-four percent of adults and 36% of elderly reached a seroprotective HAI titer of 40 or more against pandemic A/South Carolina/18/2009 (H1N1) after receiving TIV compared to 4% and 7% at the beginning of vaccination, respectively. In addition, 22% of adults and 34% of elderly showed an increase of 4-fold or more in A/South Carolina/18/2009 specific HAI titers after TIV vaccination. The pandemic specific cross-reactive antibodies strongly correlated with the post-vaccination HAI titers against the seasonal H3N2 vaccine strain in all subjects.


Introduction
The newly emerged 2009 pandemic H1N1 viruses have the hemagglutinin (HA) gene derived from the classical swine lineage, and are genetically and antigenically distinguished from recently circulating seasonal H1N1 influenza viruses [1]. The majority of the population had no prior exposure to the 2009 pandemic H1N1 viruses and thus had little pre-existing immunity against these viruses except those over the age of 60 years [2]. Despite the dominance of 2009 pandemic H1N1 viruses in the Northern Hemisphere, sporadic infections caused by seasonal influenza viruses were also reported. Hence, WHO and CDC recommended that the public seek both seasonal and pandemic vaccines for the 2009-2010 flu season. Thanks to extensive media coverage of the 2009 pandemic and heightened public awareness of the potential risk of influenza, a quite percentage of people have followed the recommendation and taken the 2009-2010 seasonal vaccine before the pandemic monovalent vaccine became widely available. However, the effectiveness of this vaccination strategy was unclear, especially with regard to potential impact on prevention of 2009 H1N1 pandemic. In this study, we evaluated the immunogenicity of 2009-2010 Northern Hemisphere inactivated trivalent influenza vaccine (TIV) and its effects on the development of cross-reactive antibody response to the current pandemic influenza as measured by hemagglutination inhibition (HAI) assay.

Immunogenicity of 2009-2010 seasonal TIV
There were no significant differences in the baseline geometric mean of HAI titers (GMTs #20) among all the subage groups against each of the three current seasonal vaccine strains  Figure 1A and 1B). Overall, .70% of all subjects had an HAI titer of 40 against both H1 and H3 seasonal strains after receiving TIV and $65% of them showed a 4-fold or more increase in the post-vaccination titers ( Figure 1D and 1E). An HAI titer of 40 has been suggested as a seroprotective measure associated with at least 50% reduction in the risk of influenza infection or diseases in humans [3][4][5]. On the contrary, both adult and elderly groups responded significantly less robustly to B/Brisbane/60/2008 than to the other two type A vaccine strains after TIV vaccination ( Figure 1C and 1F). Fewer than 35% of the TIV recipients reached a seroprotective titer of 40 or more toward B/Brisbane/ 60/2008. Only 38% of adults and 19% of elderly showed a $4fold rise in B/Brisbane/60/2008 specific HAI titers after TIV vaccination ( Figure 1F).  (

Discussion
The effectiveness of seasonal vaccines largely relies on how well selected vaccine strains represent circulating influenza viruses in the environment. Antigenically poorly matched vaccine strains could lead to significantly reduced protection in vaccinated population as shown in the 2007-2008 flu season [6]. Hence, it is necessary to continue annual evaluation of influenza vaccine strains as currently led by WHO through its collaborative centers over the world. As one of the WHO collaborative centers, each year our laboratory will prescreen human vaccinated serum samples collected from US seasonal clinical trials and select those having post-vaccination titers $40 against all three vaccine strains for distribution to other collaborative centers for further evaluation. In January of 2010, 24 pairs (before and after seasonal vaccination) each of selected human clinical samples representing adult and elderly populations in US were distributed for annual vaccine strain evaluation. Because of the predominance of 2009 pandemic H1N1 viruses in circulations, however, the antibody responses to the seasonal H1 vaccine strain were not considered [7]. By the same token, the cross-reactivity of the 2009-2010 seasonal TIV toward the pandemic H1N1 viruses was not evaluated [7].
But it does not necessarily imply that the 2009-2010 seasonal TIV played no role in protecting against 2009 H1N1 pandemic, since a suboptimal vaccine could still provide some protection against influenza-related hospitalization, especially in persons with high risk medical conditions [8,9]. In addition, a good percentage of US population has received 2009-2010 seasonal TIV before the pandemic monovalent vaccine became widely available. Thus, it was of value to evaluate the impact of seasonal vaccination on the development of cross-reactive antibodies against pandemic viruses. With the expanded sample sizes including all the subjects enrolled in US seasonal trials, we showed in the present study that both adult and elderly groups had increased antibody responses to pandemic vaccine strain A/California/07/2009 (H1N1) and a newly isolated variant A/South Carolina/18/2009 (H1N1) after receipt of 2009-2010 TIV. This is unlikely due to prior infection of   [12]. Using a highly sensitive pseudovirus based influenza HA neutralization assay, Labrosse et al. (2010) found that the subjects with 2008-2009 seasonal vaccination showed significantly higher neutralizing titers against pandemic A/ California/04/2009 (H1N1) than the subjects without the same seasonal vaccination [13]. They also found a strong correlation between the neutralizing titers against A/Brisbane/59/2009 (H1N1) and A/California/04/2009 (H1N1) [13]. Interestingly, we found in the present study that the post-vaccination HAI titers against A/California/07/2009 (H1N1) and A/South Carolina/ 18/2009 (H1N1) strongly correlated with their seasonal H3 specific antibody response instead seasonal H1 specific antibodies following 2009-2010 TIV administration. Additionally, it was out of many experts' initial anticipation that a single 15 mg HA dose of pandemic monovalent vaccine was found to be sufficient enough to induce a robust antibody response and a priming-boosting regimen was unnecessary in most of vaccinees, despite these swineorigin H1N1 viruses were new to the public [12,14,15]. This may be partially attributed to the priming effect of the current and prior seasonal vaccinations, since a substantial portion of subjects enrolled in the 2009 pandemic monovalent clinical trials had already received the 2009-2010 seasonal vaccine [14].
A moderate boost on cross-reactive response to a novel influenza virus by seasonal vaccination may not lead to a complete protection against a pandemic, but might reduce the burden of infections substantially in affected subjects, which is often seen in young children receiving partial vaccination [16][17][18]. A surveillance study on military service members stationed in the US soil has found that prior seasonal vaccinations between 2004-2009 were positively associated with protection against clinically apparent and laboratory-confirmed 2009 pandemic H1N1 illness [19]. This study along with another study led by L. Garcia-Garcia in Mexico City also found that the cross-protective effect of seasonal vaccination was more obvious in subjects with severe symptoms or hospitalization than those with mild outcome [20]. Controversies also exist. The Canadian public petitioner D.M. Skowronski and her team have claimed that an increased risk of 2009 pandemic H1N1 illness was present in local communities after receipt of seasonal vaccines [21,22]. However, the bias or confounding factors could not be ruled out as these always being the major concern for observational studies. Interestingly, another cohort study on hospital nurses from Ontario, Canada has reported a possible positive effect of the 2008-2009 TIV on reducing risk of 2009 pandemic infections though the sample size was small [23]. A more recent study by Cowling et al. (2010) has found no direct link between the 2008-2009 Southern Hemisphere seasonal TIV and increased risk of 2009 H1N1 pandemic infection in children at ages of 6-15 years from Hong Kong [24].
In addition to the pandemic specific cross-reactive antibody responses elicited by 2009-2010 TIV, our data also indicated that B/Brisbane/60/2008 was significantly less immunogenic than the other two type A vaccine components that, only 33% of adult and elderly subjects reached a seroprotective titer of 40 or more against B/Brisbane/60/2008 after TIV vaccination. The low immunogenicity of B/Brisbane/60/2008 was likely due to that it was a new strain unlike Nevertheless, ours and others' studies suggest that annual seasonal vaccination play an important role in protecting the public not only against seasonal flu but also a pandemic. In the United States, the seasonal influenza vaccination coverage jumped significantly during the 2009-2010 flu season, especially in children and adults aged 18-49 years without high risk medical condition [25], which should be greatly attributed to the joint campaign by governments and media. Continuous seasonal vaccination on individuals and expanding vaccination coverage in the community could potentially reduce disease burden from routinely circulating or newly emerging influenza viruses in future.

Ethics statement
All the human pre-and post-vaccinated serum samples involved in the present study were analyzed anonymously at CBER/FDA.   The immunogenicity of 2009-2010 seasonal TIV was evaluated by HAI assay using 4 HA units of viruses and 0.5% turkey erythrocytes according to the procedures described previously [26,27]. Sera were pre-treated with receptor-destroying enzyme (RDE, Denka-Seiken, Tokyo, Japan) at 37uC for 18 h followed by heat-inactivation at 56uC for another 30 min. RDE-treated sera were then serially two-fold diluted starting from 1:10 dilution prior to HAI assay. A titer of 5 was assigned if negative HAI reactions occurred at $1:10.

Data analysis and statistics
All subjects including adults and elderly were mixed together and grouped according to the ages with 10-year-intervals with demographic characteristics shown in Table 1. Three immunogenicity end points were chosen for the final data analysis: the geometric mean of HAI titers (GMTs); the seroprotection rate (the proportion of subjects with HAI titers $40); and the seroconversion rate (the proportion of subjects with a $4-fold rise in HAI titers). A correlation analysis between seasonal strain specific HAI titers and pandemic specific HAI titers after 2009-2010 TIV administration was performed using nonparametric Spearman's r test by JMP (Version 7). A p value of ,0.05 was considered statistically significant.