In this study, the longitudinal nature of the JH LIISA 75 + study cohort with participants enrolled for multiple consecutive influenza seasons has enabled us for the first time to directly assess interseason antibody decline and explore factors that contribute to pre-existing humoral immunity in community-dwelling older adults over 75 years of age. This was made possible by the availability of data on both current season pre-vaccination and prior season post-vaccination HAI titers measured using the prior season vaccine strain antigens when the corresponding vaccine strains in the vaccine formula changed between the two consecutive study seasons. Our results demonstrate that while interseason antibody decline is evident, the rates of such declines are highly variable depending on vaccine strains and influenza seasons. Moreover, we have identified both interseason antibody decline and prior season post-vaccination HAI antibody titers as major factors that impact pre-existing humoral immunity in this highly vaccinated older adult population.
Intra-(within)season decline of strain-specific HAI antibody titers has been recognized since the first successful trial of inactivated influenza vaccine in 1943 in which Francis and colleagues reported that vaccine-induced HAI antibody titers declined by about one-third 4–5 months after vaccination [21]. However, studies have also showed that vaccine-induced HAI antibody titers could be maintained at high levels in persons ≥ 60 years of age for at least 4 months (summarized in [14]). While interseason antibody decline is likely, its accurate assessment has been challenging. This is because the conventional approach is to employ current season vaccine strain antigens to measure current season pre-vaccination HAI antibody titers, instead of the prior season vaccines with which individuals were immunized. Such conventionally measured pre-vaccination HAI antibody titers do not accurately represent “residual” antibody titers from prior season vaccination, particularly when virus strains in the vaccine formula change. Whether this would lead to overestimating interseason antibody decline had not been determined. Here we assessed current season pre-vaccination HAI titers using the conventional approach for the evaluation of interseason antibody decline, compared with the specific strain approach outlined above. As shown in Supplement Table 1, except for IBV between 2015–2016 (0.67, p = 0.12), ratios of current season pre-vaccination HAI antibody titer over prior season post-vaccination HAI antibody titer were lower than those in Table 2, ranging from 0.59 to 0.17. This translates to statistically significant interseason antibody declines from 41–83% (p < 0.01 for all), after adjusting for age at prior study season, sex, race, and education. In addition, we evaluated and compared conventionally measured current season pre-vaccination HAI antibody titers over those measured using prior season vaccine strain antigens. As shown in Supplement Table 2, conventionally measured pre-vaccination HAI antibody titers were lower than those measured using prior season vaccine strain antigens. Specifically, ratios of conventionally measured pre-vaccination HAI antibody titers over those measured using prior season vaccine strain antigens for IAV-H3N2 in 2015 and 2016, IBV in 2015, and IAV-H1N1 in 2017 were 0.36, 0.56, 0.41, and 0.86, respectively, all statistically significant (p < 0.01, 0.03, 0.01 and 0.04, respectively), after adjusting for age, sex, race, and education. These results were confirmed when we conducted the same analysis in a larger sample consisting of all participants with both sets of pre-vaccination HAI antibody titer data available in all individual study seasons (Supplement Table 3). Taken together, results from these additional analyses support the notion that the conventional approach, i.e., measuring pre-vaccination HAI antibody titer using current season vaccine strain antigens, overestimates the rate of interseason antibody decline; the explanation for such overestimation is its inaccurate assessment of residual HAI antibody titers from prior season vaccination. An exception is the IBV strain of Victoria lineage in the 2016 season as conventionally measured pre-vaccination HAI antibody titers against this vaccine strain were almost two-fold higher than the residual antibody titers measured using prior season IBV strain antigens, as shown in Supplement Table 2 (unadjusted) and Supplement Table 3. The reason for this exception is not entirely clear, but may be related to the fact that IBV strains are highly conserved and the repeated exposure to similar strains over the years has instilled robust humoral immunity in older adults [22, 23]. It may also be related to an unusually high cross reactivity of pre-vaccination HAI antibody against the 2016 vaccine IBV strain. As such, the interseason decline of HAI titers against IBV between 2015–2016 was not statistically significant, particularly when current season pre-vaccination HAI antibody titers were measured using the conventional approach (Supplement Table 1).
Because of the clinical and immunological importance of pre-existing humoral immunity characterized by pre-vaccination HAI antibody titers, we sought to identify factors contributing to antibody status. One accepted contributing factor is interseason antibody decline, but our results document that less interseason antibody titer decline is associated with higher current season pre-vaccination HAI antibody titers. Nonetheless, such an association was not consistent across all vaccines strains or study seasons evaluated. Our study participants are drawn from a highly vaccinated older adult population with repeated annual vaccinations throughout many prior influenza seasons. In this context, it is likely that contribution of interseason antibody decline from any particular season vaccination, e.g., the immediate prior season in this case, to the pre-vaccination HAI titers, may represent a portion of the potentially accumulative effect and can be impacted by other covariates and, therefore, may vary. On the other hand, the prior season post-vaccination HAI antibody titer appears to be a strong and consistent contributing factor across all vaccine strains and seasons studied. Theoretically, a high post-vaccination antibody titer in the prior season represents a high starting point for interseason antibody decline, leading to a high residual antibody titer for the next influenza season. Results from further analyses indicate independent effects of both interseason antibody decline and prior season post-vaccination HAI antibody titers, highlighting the importance of both. Nevertheless, further longitudinal studies are warranted to determine whether a low rate of interseason antibody decline and/or a high prior season post-vaccination antibody titer serve as a predictor for a robust pre-existing humoral immunity and whether this protects better against influenza as hypothesized. Additionally, it may be that these parameters could be used as an indicator of certain unmeasured intrinsic immune system characteristics, which might include the maintenance of memory B cells and plasma cells for low-level antibody production to counteract antibody waning and provide for further understanding of influenza immunity to identify vulnerable populations who are unable to mount robust humoral immune responses to vaccination.
The strengths of this study include: 1) a well characterized population of community-dwelling older adults over 75 years of age with participants enrolled for multiple consecutive influenza seasons; 2) a unique opportunity of available pre-vaccination HAI antibody titers measured using prior season vaccine strain antigens for in-depth characterization of interseason antibody decline and contributing factors to pre-existing humoral immunity; and 3) ability to exclude breakthrough influenza cases identified through vigorous post-vaccination influenza surveillance to minimize their impact on pre-existing humoral immunity in the following season and interseason antibody decline.
The study also has limitations. For example, our data do not address vaccine effectiveness, a direct and important clinical outcome measure that would require a large sample size. However, strain-specific HAI antibodies are considered to be the major immune mechanism mediating vaccine-induced protection against influenza infection [24] and serve as the basis of age-specific immunogenicity criteria employed by the regulatory committees of the European Medicines Agency (EMA) and the US Federal Drug Administration (FDA) for the approval of new influenza vaccines [25]. Another limitation is the relatively small study sample size. While JH LIISA 75 + cohort enrolled significant number of participants in each study season, the sample size for this study is limited by the requirement for participation in two consecutive study seasons. To partially address this limitation, the analysis of comparisons of two sets of pre-vaccination HAI antibody titers presented in Supplement Table 2 was repeated and validated among all participants who completed individual study seasons with a larger sample size (Supplement Table 3). Finally, given the nature of the hemagglutination inhibition assay, while pre-vaccination HAI antibody titers measured using prior season vaccine strain antigens are more likely representative of residual antibodies from prior season vaccination, “cross-reactivity” cannot be completely excluded, i.e., such measurements may also indicate HAI antibodies cross-reactive to viral strain antigens from prior influenza vaccinations and/or breakthrough infections accumulating over many previous influenza seasons. Despite these limitations, findings from this study provide a framework for considering a more accurate assessment of residual antibodies from prior season vaccination, interseason antibody decline, and factors contributing to pre-existing humoral immunity in this highly vaccinated older adult population. They also provide initial evidence suggesting that the extent of interseason HAI antibody decline is not as pronounced as conventionally believed to be in this subset of older adults. To further address intra- and interseason waning of vaccine-induced humoral immunity as well as pre-existing humoral immunity in older adults, more in-depth investigations including detailed intra-seasonal HAI antibody analyses with frequent sampling after vaccine administration until the beginning of the next influenza season in longitudinal studies as well as vaccine clinical effectiveness are indicated.