#19,022
Based on limited surveillance and reporting systems (see NAS : Diagnostic Tools, Gaps, and Collaborative Pathways in Human H5N1 Detection (Rapid Expert Consultation), the available evidence suggests that HPAI H5N1 viruses are not currently transmitting efficiently (or often) among humans.
While somewhat reassuring, we've also seen evidence to suggest that mild, or asymptomatic cases are likely being missed (see MMWR: Serologic Evidence of Recent Infection with HPAI A(H5) Virus Among Dairy Workers and JAMA Open: Asymptomatic Human Infections With Avian Influenza A(H5N1) Virus Confirmed by Molecular and Serologic Testing).
While the WHO, PAHO, and the ECDC (see ECDC Pre-pandemic Guidance: Strategies to Fight Avian and Swine flu in Humans) have all called for increased vigilance, there appears to be some reluctance among many member countries - and their agricultural interests - to comply.
Meanwhile, we continue to see studies and preprints that suggest that HPAI H5Nx viruses are continuing to accrue mammalian adaptations (see here, here, here, here, and here).
To this rapidly expanding list we can add a new preprint from researchers in Switzerland who conducted experiments to test the replication of the bovine (B3.13) H5N1 virus in human nasal epithelial cells - and found they replicated remarkably well - even at low temperatures.
That said, after 24-hours, the human innate immune system was able to substantially reduce replication via IFN-λ (`lambda interferons'), which may explain its relatively mild presentation.
While lengthy, and somewhat technical, this is a fascinating report and many will want to read it in its entirety. I've reproduced the abstract and some highlights below.
I'll return with a bit more after the break.
Bovine-derived influenza A virus (H5N1) shows efficient replication in well-differentiated human nasal epithelial cells without requiring genetic adaptation
Etori A. Moreira,Samuel Constant, Charlene Constant, Lisa Butticaz, Michele Wyler, Teodora David, Peter M. Grin, Charaf Benarafa, Volker Thiel,Marco P Alves, Gert Zimmer
doi: https://doi.org/10.64898/2026.01.16.699876
This article is a preprint and has not been certified by peer review
Abstract
Highly pathogenic avian influenza H5N1 viruses of clade 2.3.4.4b have caused widespread avian mortality and sporadic mammalian infections, raising concerns about their potential for efficient replication in the human population. Efficient replication in the human upper respiratory tract is considered a key barrier to transmission.
Here, we demonstrate that an H5N1 virus isolated from bovine milk in Texas in 2024 (H5N1Tex/24) replicates as efficiently as the 2009 pandemic H1N1 virus (H1N1HH4/09) in well-differentiated human nasal epithelial cells.
These cells express both avian- and human-type influenza receptors, indicating receptor adaptation is unnecessary for entry. H5N1Tex/24 replicates effectively at 33 degrees Celsius, reflecting nasal cavity temperature, whereas earlier avian H5N1 strains require 37 degrees Celsius, suggesting that H5N1Tex/24 has acquired another key adaptive feature to the human upper respiratory tract.
H5N1Tex/24 remains sensitive to interferon-λ (IFN-λ) despite inducing low cytokine levels. Notably, no known mammalian-adaptive mutations such as PB2-E627K were detected. These findings suggest that H5N1Tex/24 possesses intrinsic traits enabling efficient replication in the human upper airways, a critical step toward potential airborne transmission, underscoring the need for vigilant surveillance.
(SNIP)
Overall, our results show that bovine-derived H5N1Tex/24 replicates at high titers in primary human nasal epithelial cells even though it lacks many canonical markers of adaptation to mammals. This observation contrasts with the currently low number of confirmed human infections. There are several possible explanations for this discrepancy.
- First, the sensitivity of H5N1Tex/24 to the antiviral effects of MxA may limit virus dissemination to the lower respiratory tract and thus disease severity.
- Second, human infections may go undetected because they are completely asymptomatic or because they are associated with only mild symptoms.
- Third, efficient airborne transmission of H5N1 requires that HA induces membrane fusion at pH values significantly lower than pH 6.076,77. However, the HA protein of bovine H5N1 retains typical avian characteristics, with fusion triggered at approximately pH 6.046,78.
- Finally, pre-existing immunity to influenza viruses may provide some partial protection against bovine H5N1. Cross-reactive antibodies against the NA protein of human H1N1 viruses can particularly inhibit avian N1 sialidase activity, thereby potentially limiting replication of clade-2.3.4.4b viruses.
The recent fatal human case caused by a H5N5 HPAI virus in the United States is alarming, as there is probably no pre-existing immunity to the N5 antigen in the human population.
Overall, our results show that H5N1Tex/24 has a remarkable ability to replicate in primary human nasal epithelial cells. Since well-differentiated nasal epithelial cells represent a relevant model for the human upper respiratory tract, it is not unlikely that the virus can also replicate efficiently in vivo in humans
This study confines itself primarily to the milder B3.13 `bovine' H5N1 virus (with a brief mention of HPAI H5N5), but there are also concerns over other genotypes (and subclades) including the D1.1 genotype, which burst onto the scene in the fall of 2024, and has caused more severe (and 2 fatal) infections in humans.
The exact number of human infections with the D1.1 genotype is unknown, since not all of the (now, roughly 6 dozen) North American human cases have been fully characterized.
A study, published 2 months ago (see J.I.D.: Avian influenza virus A(H5N1) genotype D1.1 is better adapted to human nasal and airway organoids than genotype B3.13) described this ambiguity:
A total of 53 strains were identified, of which 6 strains had 2 sequences deposited. These 53 strains were collected from patients between March 28, 2024 and February 12, 2025 (Supplementary Table S3). Of these 53 strains, 22 (41.5%) were assigned to genotype B3.13, 8 (15.1%) were assigned to D1.1, 1 (1.9%) was assigned to D1.3, and 22 (41.5%) could not be assigned to any genotypes according to GenoFLU version 1.06 (https://github.com/USDA-VS/GenoFLU).
That study attributed D1.1's increased severity to their findings that the D1.1 genotype replicates better in lab-grown nasal and lung tissues than the bovine B3.13 strain, and it binds more tightly to human‑type (α2,6-linked SA) receptors.
While direct comparisons between these studies are difficult due to differences in methods and materials, both show that - compared to older H5Nx strains - both the B3.13 and D1.1 genotypes have become much better adapted to the human nasal passage.
And that's a trend we really shouldn't sneeze at.
