#18,119
While the respiratory route is the most common way that influenza is transmitted, we've seen plenty of evidence over the years that influenza A viruses - and particularly avian influenza viruses - can transmit via the ocular or even the gastrointestinal route.
Many of the H5 wild mammal deaths in the wild have been carnivores, that were likely exposed after eating infected birds or mammals. California Condors, and other scavengers, have been similarly affected, and there are a number of anecdotal reports of humans becoming infected through consuming under cooked poultry in Asia.
Over the years we've also seen reports of conjunctivitis, particularly (but not exclusively) linked to H7 avian flu infection (see MMWR: Mild H7N3 Infections In Two Poultry Workers - Jalisco, Mexico).
More recently, all 3 of the cattle-linked H5N1 infections in dairy workers have presented with ocular symptoms.
To better understand this route of infection, and the potential for it to be transmitted onward, CDC researchers have conducted some preliminary experiments using ferrets and the same H5N1 virus recovered from a human case last year in Chile (A/Chile/25945/2023).
Last year the CDC published their Genomic Analysis of H5 Virus In Human in Chile, citing two noteworthy `mammalian adaptations' found in its PB2 gene.
- The PB2-D701N mutation, which has been linked to increased severity of infection (see here and here), and has been previously seen in both human H5N1 and H5N6 infections.
- The PB2-Q591K mutation, which enhances polymerase activity and virus replication of avian viruses in mammals (see PB2-Q591K Mutation Determines the Pathogenicity of Avian H9N2 Influenza Viruses for Mammalian Species).
This tag-team of D701N and Q591K have been previously seen in zoonotic H7N9 viruses (see Genomic Signatures for Avian H7N9 Viruses Adapting to Humans), and together may compensate for the lack of an E627K mutation which increases viral replication at lower temperatures found in mammals.
Earlier this year, in Emerg. Microbes & Inf: HPAI H5N1 Virus Isolated from Human in Chile Causes Fatal Disease & Transmits Between Co-housed Ferrets, we looked at a study that found that this virus transmitted readily between co-housed ferrets.
Today's findings are based on an initial ferret infection via the ocular route, and are published today in an early release in the EID Journal. While you'll want to read the report in its entirety, briefly they found.
- Ocular inoculation of a clade 2.3.4.4b highly pathogenic avian influenza A(H5N1) virus caused severe and fatal infection in ferrets
- At either a high (106 PFU) or low (103 PFU) challenge dose, all ferrets inoculated by the ocular route became productively infected
- Infectious virus was detected in either nasal or rectal wash samples in all co-housed contact animals on at least 1 day
Our finding that a clade 2.3.4.4b H5N1 virus isolated from a human can exhibit a virulent and transmissible phenotype after nontraditional inoculation, even with a low dose of inoculum, underscores the public health threat posed by those IAVs.
Excerpts from the report follow. I'll have a brief postscript after the break.
Fatal Infection in Ferrets after Ocular Inoculation with Highly Pathogenic Avian Influenza A(H5N1)
Abstract
Ocular inoculation of a clade 2.3.4.4b highly pathogenic avian influenza A(H5N1) virus caused severe and fatal infection in ferrets. Virus was transmitted to ferrets in direct contact. The results highlight the potential capacity of these viruses to cause human disease after either respiratory or ocular exposure.
In recent years, clade 2.3.4.4b highly pathogenic avian influenza A(H5N1) viruses have exhibited substantial host expansion, geographic spread, and reassortment with other circulating influenza A viruses (IAVs) in birds, resulting in epornitics on all continents and virus detection in an expanding group of mammals (1). Human cases of H5N1 clade 2.3.4.4b virus infection have been reported, typically following direct exposure to infected animals, contaminated environments, or both (2).
A 2.3.4.4b highly pathogenic avian influenza A(H5N1) virus was isolated from a human patient in Chile during 2023 (A/Chile/25945/2023 [Chile/25945]) (3) and caused severe and fatal disease in ferrets intranasally inoculated with 106 PFU of virus. Transmission of virus to animals housed in close contact was also reported (3), highlighting the pandemic potential of clade 2.3.4.4b viruses.
Although the eyes represent a secondary mucosal surface that is susceptible to respiratory virus exposure (4), as evidenced by recent reports of conjunctivitis in 2 dairy workers exposed to clade 2.3.4.4b H5N1 virus (5), risk assessment approaches for clade 2.3.4.4b H5N1 viruses to date have been limited to standard intranasal inoculation (3,6) and have not evaluated the capacity of those viruses to cause disease after alternative portals of entry. To investigate relative similarities between ocular and respiratory exposure to H5N1 virus, we assessed the severity and kinetics of disease after ocular exposure of ferrets to Chile/25945 virus and compared our findings with a previously published assessment of animals intranasally inoculated with this virus at the Centers for Disease Control and Prevention in 2023 (3).
To assess disease severity and transmissibility under different exposure concentrations, we inoculated ferrets by the ocular route with either a high (106 PFU) or low (103 PFU) dose of Chile/25945 virus (7).
At either challenge dose, all ferrets inoculated by the ocular route became productively infected, reaching mean maximum weight loss of 12.7% (high dose) and 13.2% (low dose) and mean maximum rises in temperature of 2.4°C (high dose) and 2.0°C (low dose). Humane endpoints were reached on postinoculation days 5–7 in 3/3 (high dose) and 2/3 (low dose) animals (Figure 1, panel A). During necropsy, we detected infectious virus throughout the respiratory tract and in several extrapulmonary tissues (including from the gastrointestinal tract, central nervous system, and ocular system) (Figure 1, panel C), consistent with the highly virulent phenotype observed after high-dose intranasal inoculation of ferrets (3). One ferret survived the challenge with a serologic titer of 160.
To assess if ferrets inoculated by the ocular route were as capable as intranasally inoculated ferrets to transmit Chile/25945 virus in a direct contact setting (3), we cohoused a serologically naive ferret with each inoculated ferret 24 hours after inoculation. To assess virus replication within and beyond the respiratory tract, we collected nasal wash, conjunctival wash/swab, and rectal swab samples from inoculated and contact animals.
All ferrets inoculated by the ocular route with a high dose of virus had detectable infectious virus in nasal wash (mean peak titer 5.3 ± 0.2 log10 PFU/mL), conjunctival wash/swab (4.4 ± 0.9 log10 PFU/mL), and rectal swab (3.6 ± 0.3 log10 PFU/mL) samples (Figure 2, panel A).
The magnitude and frequency of viral titers in these specimens was reduced, but still present, in animals inoculated with a low dose of virus (Figure 2, panel B). Infectious virus was detected in either nasal or rectal wash samples in all contact animals on at least 1 day; infectious virus was not detected in conjunctival wash samples from any contact animal, possibly resulting from less overall virus shedding than in inoculated animals. Humane euthanasia because of severe disease was warranted for 4/6 contact animals (Figure 1, panel B); the other 2 survived, 1 of which exhibited a low level of seroconversion (hemagglutination inhibition titer 20).
Our finding that a clade 2.3.4.4b H5N1 virus isolated from a human can exhibit a virulent and transmissible phenotype after nontraditional inoculation, even with a low dose of inoculum, underscores the public health threat posed by those IAVs.
The ocular surfaces may be exposed to infectious virus from the environment by several means (e.g., airborne particles, physical transfer from contact with fomites, and splashing liquids). Furthermore, circulation of tear fluid between ocular and nasopharyngeal tissues via the lacrimal duct offers an opportunity for infectious virus to spread from the respiratory tract to the ocular system (8), in agreement with successful H5N1 viral isolation from both conjunctival and nasopharyngeal swab specimens from a human with conjunctivitis (5).
Conjunctivae may be exposed to virus by direct contact (e.g., virus-contaminated hands), indirect contact (e.g., virus-contaminated fomites), or after deposition of virus-laden droplets or aerosols onto the ocular surface (4), permitting opportunities for H5N1 virus to establish a productive infection in humans even in the absence of an ocular tropism.
Considering the myriad ways humans may be exposed to IAVs, our study supports the need to consider nontraditional inoculation modalities in risk assessment activities and supports consideration of using eye protection in potentially contaminated environments (9).
Dr. Belser is a microbiologist in the Influenza Division, National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention, Atlanta, GA. Her research interests include the pathogenicity, transmissibility, and tropism of influenza viruses, with an emphasis on ocular tropism and ocular exposure.
Presumably similar experiments will be conducted on the A/Texas/37/2024 virus in the weeks ahead, as the CDC puts together a risk assessment. Given what we've seen, I've added some basic eye protection to my PPEs (inexpensive non-medical grade lab-goggles).
