We compared humoral neutralization of 2 H5N1 viruses, A/dairy_cattle/Texas/24-008749-002/2024 (2.3.4.4b) and A/Cambodia/NPH230776/2023 (2.3.2.1c), in serum samples alongside a seasonal influenza A(H1N1) virus isolate, A/Wisconsin/67/2022, before and after QIV in participants from the University College London Hospitals–Francis Crick Institute Legacy study cohort (https://clinicaltrials.gov/study/NCT04750356) (Appendix Figure). The study included 61 adults (median age 49 [range 38–58] years); 44 (72%) were women, and 17 (28%) were men. Thirty (49%) adults were vaccinated in only the 2021–22 season, but all 61 were vaccinated in >1 study season (2021–22, 2022–23, and 2023–24). Median sampling duration after vaccine dose for 2021–22 was 81 (interquartile range [IQR] 61–81) days, for 2022–23 was 67 (IQR 38–68) days, and for 2023–24 was 77 (IQR 44–77) days. Twenty-seven (44%) participants reported a single underlying condition (Table).

Figure

Figure. Live-virus neutralization of preseasonal and postseasonal quadrivalent influenza vaccination in a study of influenza A(H5N1) virus neutralization by quadrivalent seasonal vaccines, United Kingdom, 2021–2024. Violin plots show paired serum samples...

In line with their effectiveness against seasonal influenza, each QIV generated a statistically significant boost in serum neutralization of A/Wisconsin/67/2022 in each season tested (p = 0.003–0.007 by χ² test) (Figure). In contrast, HPAI H5 virus neutralization in our cohort of healthy adults was blunted or absent. In prevaccine serum samples, we detected neutralization above background against A/Cambodia/NPH230776/2023 in a few samples but did not detect any neutralization against A/dairy_cattle/Texas/24-008749-002/2024 isolate (Figure). No seasonal QIV resulted in a cross-neutralization boost against either HPAI H5 virus (Figure).

Ongoing adaptation of HPAI H5N1 clade 2.3.4.4b virus in cows and other mammal hosts found on dairy farms, including rodents and cats, substantially increases the risk for a major HPAI H5N1 virus epidemic or pandemic in humans (2,4). The paucity of human serologic memory against either H5N1 virus strain raises the potential for widespread vulnerability to infection within the adult population. We observed a predictable boost to neutralizing titers against the contemporary seasonal influenza A(H1N1) virus (A/Wisconsin/67/2022) that was absent for the 2 clinically relevant H5N1 viruses tested in our high-throughput neutralization assay (Appendix). Neutralizing antibody titers have long been used as a correlate of protection against seasonal influenza (12); thus, our observations suggest seasonal QIVs are unlikely to offer adequate serologic protection against H5N1 virus.

Immunity against influenza evolves throughout the lifespan, and early infection exposures influence subsequent antibody responses after infection and vaccination (13). Few participants in our study had detectable neutralizing titers above background to the 2.3.2.1c A/Cambodia/NPH230776/2023 virus and none to the 2.3.4.4b A/dairy_cattle/Texas/24-008749-002/2024 virus. Together with our observed lack of QIV boosting, our results suggest that strategies reliant on existing population-level or QIV-based immunity against H5N1 virus infection must be approached with caution.

One limitation of this study is the lack of in vivo challenge to test for cross-protection. Some studies have reported transient protection against H5N1 challenge after transferring QIV-vaccinated human serum to mice, which was not accurately predicted by in vitro assays, including virus neutralization assays (14,15). Cross-neutralization might also occur in the absence of nAbs, but without in vivo testing, we cannot conclusively determine the extent to which the QIV might provide protection against H5N1 virus. A second limitation is that our cohort, predominately working age, healthy adults receiving occupational QIV, do not represent a population at high risk for severe influenza disease and death. However, they represent an immunocompetent population and would be expected to have the most robust detectable immunity. Third, although we tested 2 H5N1 viruses associated with recent human disease, despite extensive efforts, we could not eliminate background signal in our assay. Thus, we were unable to fully quantify neutralization at lower titers and opted to describe the range within which we detected background signal. As of 2025, no neutralization titers were available from postinfection serum samples in dairy farm workers to further refine that cutoff (6). Further investigation is required to address the issue of background signal. However, other studies suggest nonspecific inhibition by human serum as a possible explanation for low-level readouts for protection (15). Finally, our use of whole virus to assess nAb titers did not allow determination of the extent to which hemagglutinin- or neuraminidase-specific antibodies might have contributed to overall neutralization. However, the high-throughput live-virus neutralization we describe (Appendix) is a highly valuable tool for pandemic preparedness, offering a method for near real-time analysis of serum-based immunity to emerging viruses in large cohorts.

The effectiveness of QIV against influenza A(H5N1) virus remains uncertain, and clarification on the extent of cross-protection in humans is urgently needed. Considering that uncertainty, timely and effective deployment of targeted vaccines would be crucial during widespread influenza A(H5N1) outbreaks. To reduce risks for severe illness and death requires enhanced measures that mitigate the spread of HPAI H5N1 viruses to humans, accelerated pipelines for H5-directed influenza vaccines, and systems that rapidly and equitably reach clinically vulnerable persons worldwide (2).