The strains of highly pathogenic avian influenza H5N1 currently circulating in the United States are significantly better adapted to replicate in human nasal tissue than older versions of the virus — a finding that carries important implications for pandemic risk assessment. At the same time, one of the dominant genotypes appears to trigger a notably muted immune response despite replicating vigorously, raising new questions about how the virus interacts with human biology.
Those are the central findings of a new study published in the May 2026 issue of Emerging Infectious Diseases by researchers at the National Institute of Allergy and Infectious Diseases (NIAID), part of the National Institutes of Health. The study compared how multiple H5N1 strains — including historical isolates and contemporary genotypes linked to U.S. human and animal cases — replicate and provoke immune responses in a laboratory model of the human nasal epithelium.
Why the Nasal Epithelium Matters
The distinction between upper and lower respiratory tract infection is central to the question of pandemic potential. Seasonal influenza viruses replicate efficiently in the upper respiratory tract, which is why they spread so readily between people. H5N1, historically, has preferentially replicated deep in the lungs, which partly explains why it causes severe disease but does not transmit efficiently between humans.
If contemporary H5N1 strains are gaining the ability to replicate in the upper airway, that shift could represent a step toward greater transmissibility.
What the Study Found
The NIAID team used a commercially available tissue model derived from primary human nasal epithelial cells to test eight influenza A virus isolates, including two seasonal strains and six H5N1 isolates spanning three genotypes: B3.13, B3.6, and D1.1. All are variants of clade 2.3.4.4b, the lineage responsible for the ongoing outbreak in U.S. poultry and dairy cattle.
The results showed that contemporary H5N1 isolates with known mammalian adaptations — particularly mutations in the polymerase basic 2 (PB2) protein, including the well-characterized E627K substitution — replicated more efficiently in nasal tissue than a historical H5N1 strain isolated in Vietnam in 2004. One contemporary isolate, A/Texas/37/2024, a B3.13 genotype virus from a human case, replicated more efficiently in nasal tissue than even the seasonal influenza strains tested.
Notably, this pattern reversed when the same research group previously tested the viruses in a lower respiratory tract model: in that setting, the historical H5N1 strain outperformed contemporary isolates. Taken together, the findings suggest that current H5N1 viruses have adapted toward the upper airway in ways their predecessors had not.
A Blunted Immune Warning Signal
The immune response findings added another layer of concern. B3.13 genotype viruses — responsible for many of the early U.S. human cases — induced significantly lower interferon-stimulated gene responses than D1.1 genotype viruses, historical H5N1, and even seasonal flu strains. This dampened antiviral response occurred despite high levels of viral replication, a combination the researchers flag as potentially significant.
In contrast, the D1.1 genotype, which has been associated with more severe human illness including a hospitalized case in Wyoming, triggered stronger and earlier interferon responses. The isolate from that severe case also induced higher immune gene expression than one from a milder D1.1 case in Nevada, suggesting the immune response profile may correlate with clinical severity.
A Possible Protective Factor
Despite the concerning replication data, the researchers point to a potentially important brake on transmission: existing immunity from seasonal influenza. Prior exposure to H1N1 influenza has been shown in ferret models to substantially reduce nasal shedding of contemporary H5N1 virus and to prevent infection in contact transmission settings. Given that H1N1-like viruses continue to circulate seasonally, and that serosurveillance of exposed dairy farm workers found high prevalence of neutralizing H1N1 antibodies, the authors suggest population-level immunity may be offering a degree of protection that explains why efficient nasal replication has not yet resulted in human-to-human spread.
Implications for Preparedness
Since March 2024, 70 human cases of H5N1 have been reported in the United States, the result of sporadic spillover from infected poultry and dairy cattle. The emergence of new genotypes — particularly D1.1, first detected in dairy cattle in January 2025 and subsequently found in humans — has intensified scrutiny of how these viruses are evolving in mammalian hosts.
The NIAID findings underscore that characterizing H5N1 behavior in relevant human tissue models is essential to staying ahead of potential pandemic risk. The study’s authors call for additional research to clarify how replication efficiency and immune evasion interact to affect transmission, and note that the role of pre-existing immunity warrants continued surveillance as the virus continues to evolve.
Sources and further reading:
Flagg M, Winski CJ, Brackney BG, et al. Replication Efficiency of Contemporary Highly Pathogenic Avian Influenza A(H5N1) Virus Isolates in Human Nasal Epithelium Model. Emerging Infectious Diseases, 1 May 2026.