In December 2024, the prevalence of seroprotection was low, especially for A/H1N1, even among healthcare workers who received the vaccine. Nevertheless, there was a dose-response relationship between post-vaccination HI titers against A/H1N1 and influenza protection during the high circulation period of A/H1N1, showing significant protection even at HI titers of 10 and 20.

Our finding of 12.7% of seroprotection against A/H1N1 among vaccinated individuals in the 2024/2025 season was substantially lower than the results of the 2009–2011 clinical trials on the immunogenicity of QIIV (80–100%) [3], while it is compatible with a low prevalence among vaccinated Tokyo residents in the 2023/2024 season (4.0%) [2]. These results suggest that the immunogenicity of the influenza vaccine has decreased during the long-term COVID-19 pandemic. This may be due, at least in part, to a decline in influenza-specific immune memory over the extended period without a major influenza epidemic. Current inactivated influenza vaccines without adjuvants cannot induce robust T cell-dependent memory B cells (MBCs) [10], which play a crucial role in antibody production [11]. Thus, the antibody response to these vaccines primarily depends on pre-existing MBCs [12, 13]. Natural influenza infection can strongly boost MBCs, although this memory fades with time [14]. Most of the present study participants (92.3%) were not diagnosed with influenza during the last four seasons (2020/2021 to 2023/2024). In addition, 6% of the vaccinated participants had contracted influenza A within the last two seasons, and they had significantly higher HI titers against A/H3N2 than those without a recent infection history but only marginally higher against A/H1N1, indicating that most previous influenza A cases were likely due to A/H3N2. These findings suggest that nearly all participants may not have had the opportunity to enhance their influenza A/H1N1-specific immune memory for several seasons.

Another possible reason for the low immunogenicity is repeated annual vaccination. In this study, approximately 90% of participants received influenza vaccine in both the current and previous seasons, and they had 11-30% lower HI titers than those vaccinated only in the current season. This finding is consistent with previous immunologic studies [15, 16] and is compatible with vaccine effectiveness (VE) studies showing a 7-18% lower VE against influenza infection in individuals vaccinated in both seasons compared to those vaccinated only in the current season [17]. This phenomenon could be explained by the antigenic distance hypothesis [18]; pronounced negative interference occurs when previous and current vaccine strains are antigenically similar. The 2023/2024 and 2024/2025 QIIV vaccines used for the NCGM in-house vaccination program contained the same antigens except for A/H3N2 (Supplementary Table 1). Notably, A/H3N2—the only strain that changed from the previous season—elicited the highest antibody titer after vaccination compared to the other three antigens (GMTs: 42.4 vs. 10.8-23.8). However, these data should not be interpreted as evidence against vaccination can be skipped. According to VE studies [19-21], vaccination in both the current and previous seasons showed higher effectiveness against influenza-related hospitalizations than vaccination in either season alone, supporting current recommendations for annual influenza vaccination.

Among vaccinated participants, we found that a higher HI titer against A/H1N1 was associated with greater protection against infection during the A/H1N1 outbreak in a dose-response manner. This finding is reasonable because the A/H1N1 antigen of the 2024/2025 vaccine was antigenically well-matched to the current circulating virus in Japan (clade: 6B.1A.5a.2a) [22]. An HI titer of 40 has been accepted as a threshold corresponding to a 50% reduction in the risk of contracting influenza [9]. However, we found that an HI titer of 10–20 against A/H1N1 conferred 47.3%–57.9% protection against infection, which is consistent with a meta-analysis showing that the HI titer corresponding to 50% protection is as low as 17 (95% CI: 9–28) [23]. In the present study, 61.3% of vaccinated individuals had HI titers ≥10 (Supplemental Figure 4); thus, more than half of vaccine recipients may have achieved moderate protection.

This study boasts several key strengths, including a relatively large cohort, a high participation rate of 84%, and the availability of HI antibody titers measured around 4 weeks after the vaccination and immediately prior to the influenza outbreak. We should also acknowledge limitations. First, we did not measure pre- and post-vaccination HI titers from the same participants, preventing within-subject comparison and limiting the accurate immune response to vaccination assessment. Second, we did not collect a history of influenza diagnosis before the COVID-19 pandemic. Although pre-COVID-19 studies suggested that natural infection-induced antibody immunity to HA is long-lived (over 50 years) [24], our data did not clarify whether this remains the case following a prolonged actual absence of influenza exposure, including asymptomatic infections. Third, while information on the type of influenza (A or B) was available in both the history and follow-up surveys, data on specific subtypes, such as A/H1N1 and A/H3N2, were not collected in this study. Regarding history, since both A/H1N1 and A/H3N2 predominated in the past two seasons (Supplemental Figure 2), it is not possible to estimate which subtype virus represents the history of influenza A. For the follow-up survey, the cases were most likely due to A/H1N1, which accounted for 94.5% of sequenced influenza samples in Japan during the follow-up (Supplemental Figure 2). Fourth, we did not assess other humoral or cell-mediated immune indicators. For example, neutralizing antibodies directly reflect the functional ability to block viral infection [24], and cell-mediated immune responses are crucial for long-term and cross-protective immunity [11]. Fifth, we did not conduct active surveillance of influenza infection during the follow-up, possibly underestimating the total cases. Finally, this study included only healthcare workers from a single medical center in Tokyo, most of whom received annual influenza vaccinations and had experienced fewer infections over at least four seasons. In addition, no severe cases were observed. Caution should be exercised when generalizing our findings to different populations, particularly those at high risk of severe influenza, such as children, older adults, and patients with immunosuppressive conditions.

In conclusion, HI titers against A/H1N1 were extremely low even in vaccinated healthcare workers after a prolonged period without a major influenza epidemic. Nevertheless, higher post-vaccination HI titers, even at modest levels below 40, were associated with greater protection against influenza infection, supporting the benefit of influenza vaccination in the post-COVID-19 era. Future studies should explore the mechanisms behind the low immunogenicity after vaccination and address improving the immunological response and effectiveness of the influenza vaccine.