Highly pathogenic avian influenza (HPAI) A(H5N1) emerged in 1997.1 Since then, it has spread globally by migratory birds, resulting in infections in animals on every continent. HPAI A(H5N1) clade 2.3.4.4b emerged in 2021 and resulted in fatal infections in poultry as well as terrestrial and marine mammals.1 In early 2024, influenza A infection was first recognized in dairy cows with mastitis in Texas. Infection in dairy cows is now widespread in the United States, affecting more than 875 herds in 16 states. Most cow infections are genotype B3.13, whereas most outbreaks in wild birds and poultry are genotype D1.1.

Against this background, more humans have come into contact with HPAI A(H5N1). Investigators now report in the Journal a series of human cases from the United States and Canada.2,3 The former series involves 46 case patients with generally mild, self-limited infection with A(H5N1): 20 with exposure to poultry, 25 with exposure to dairy cows, and 1 with undefined exposure.2 Among case patients with occupational exposure, the use of personal protective equipment (PPE) was not universal. 

Most case patients presented with conjunctivitis, almost half with fever, and a minority with mild respiratory symptoms, and all recovered. The only hospitalization occurred in the case patient with undefined exposure, although hospitalization was not for respiratory illness. Of cases with sequenceable virus, most were B3.13; four cases in patients with poultry exposure were D1.1.

In Canada, a 13-year-old girl with mild asthma and obesity presented with conjunctivitis and fever and had progression to respiratory failure.3 She received intubation and venovenous extracorporeal membrane oxygenation. After treatment that included oseltamivir, amantadine, and baloxavir, she recovered. Notably, genotype D1.1 was detected; sequencing of one isolate from the lower airways that was collected 8 days after the onset of symptoms showed three mutations potentially associated with enhanced virulence and human adaptation: E627K in the polymerase basic 2 gene and E186D and Q222H in the H5 hemagglutinin gene. It is unclear whether these mutations were present in the infecting virus or emerged during the course of the patient’s illness.

These reports show several critical features of the threat of HPAI to human health, and how we might respond. First, collaboration among investigators in human and veterinary medicine, public health leadership, health care providers, and occupational authorities (especially agricultural), exemplified by the case series in United States, is paramount. Cases of H5N1 respiratory illness have been detected because of a standard surveillance approach aimed at detecting novel (nonseasonal) influenza. This approach involves cultivating trust not only among numerous entities but with people seeking care for symptoms of concern, including conjunctivitis.

Second, the mutations evident in the Canadian case highlight the urgent need for vigilant surveillance of emerging mutations and assessment of the threat of human-to-human transmission. The One Health paradigm is foundational to this outbreak, yet to date, genomic sequencing data that have been collected from animals frequently lack critical metadata. Without information pertaining to where and when isolates were collected, the data cannot be linked phylogenetically to other reported sequences, which limits insight into how the virus is spreading. Such data would also provide opportunity for early detection of mutations that might portend avidity for human respiratory epithelium, which may require as little as one mutation.4

Third, we must continue to pursue development and testing of medical countermeasures. Fortunately, current vaccine candidates neutralize the circulating strains in vitro, and these strains so far are susceptible to antiviral agents. Studies have shown the safety and immunogenicity of A(H5N1) vaccines and the need for a two-dose prime–boost approach and use with adjuvants.5 Work is ongoing to complete candidate vaccines for clinical use, if needed. Furthermore, studies are ongoing to develop messenger RNA–based A(H5N1) vaccines and other novel vaccines that can provide protection against a broad range of influenza viruses, including A(H5N1). 

Circulating isolates are susceptible to all approved neuraminidase inhibitors, adamantanes, and baloxavir marboxil.6,7 The Canadian case showed higher viral loads in the lower airway and very prolonged shedding, despite therapy, which highlights the potential need for longer therapy. Recently, the Centers for Disease Control and Prevention (CDC) issued emergency-use instructions for oseltamivir that recommend longer durations of oseltamivir therapy for persons hospitalized with novel influenza viruses and twice-daily dose administration for prophylaxis. Baloxavir marboxil is not recommended for monotherapy in hospitalized patients because of concern for resistance emergence.8 Resistance to oseltamivir has occurred frequently among patients hospitalized for A(H5N1) infection.9 Combination of two agents may reduce this risk.10

Fourth, precautions to prevent infection are critical, including the use and optimization of PPE in occupational settings and education about the risk of contact with sick birds and animals. PPE use can be challenging in settings where dust, milk, and feathers may easily contaminate surfaces. Many environments are not air-conditioned and during the summer may be exceptionally warm, and the use of large fans may facilitate airborne particles.

Although these reports help to define some aspects of the clinical course of infection in the current H5N1 outbreak, many questions remain. The type of clinical presentation and severity may depend on the host, the route and inoculum of exposure, and the virus clade. For example, severe cases to date have been associated only with the wild-bird clade, D1.1. Historically, H5N1 infection in Southeast Asia and Egypt has been associated with a case fatality rate of approximately 50%. Whether host immunity (e.g., changes in population-level immunity to the neuraminidase component [N1] of seasonal influenza virus), route of exposure, or other changes in the virus itself might be partly responsible for the lower case fatality rate in North America is unknown.

The CDC still designates the risk of HPAI A(H5N1) to most Americans as low. We do have candidate vaccines and antivirals available to try to mitigate severe influenza in the case of wider spread. That said, a balance between enhanced vigilance and “business as usual” is needed. The past weeks have seen more cases detected in more states as well as more persons with respiratory illness acquired through exposure to poultry or from an unknown source. Without a clearer understanding of the extent of exposure, infection, viral evolution, and transmission, we will be unable to properly protect our communities from a pathogen that has proven to be a formidable challenge to human and animal health.