In my post last night on avian influenza, I mentioned a recently published article (yesterday) that I thought should give us pause about many of the assumptions being made about the current H5N1 outbreak in dairy cattle, and that should cause us to increase and improve our response to containing this outbreak. Today, we will review that study in greater detail.
That article – “A human isolate of bovine H5N1 is transmissible and lethal in animal models” https://www.nature.com/articles/s41586-024-08254-7 – was published in Nature on October 28, 2024.
In this study, the investigators examined the H5N1 virus circulating in dairy cattle that had infected a farm worker. This is very important because we need to know whether this was just an isolated occurrence that happened by chance related to the closeness of the farm worker and length of time working with these infected animals that represents a very low risk to the general population that is not at such occupational risk, or alternatively, whether the virus is evolving in such a way as to be able to more efficiently spread to humans, and even more concerning would be if the virus is developing more effective transmission from infected humans to their human close contacts. Thus far, we have seen no evidence of the latter, which would be the most concerning as to risk to the general population and as to pandemic potential.
Somewhat surprising (at least to me) was that the virus could grow well on human lung cells (alveolar epithelial cells), but it did not grow well in eye cells (corneal epithelial cells) – both cells have the receptor type used by avian influenza viruses [my surprise being in part because the most common presentation of infected farm workers has been conjunctivitis, an infection of the lining over the eye].
We don’t test these viruses on human subjects for ethical reasons, so animal models are used, including mice. However, ferrets are one of the best animal models because they most closely mirror influenza infections in humans, so they were used in this study, as well.
When directly infected with the virus recovered from the farm worker’s eye, the infection caused rapid and deadly infections in the mice and ferrets, in fact, all of the ferrets died. This tells me that we may very well be lulled into thinking that the current H5N1 only causes mild illness in humans since that is what we have observed in the 36 humans that are known to have been infected thus far. But we need to keep in mind that is likely only because it infected their eyes and not their lungs. Our upper airways have different receptors that are not amenable to H5N1 binding, so the virus is not finding a way to get to our lungs just yet. However, to me, this means we must contain the spread of this infection in cattle and other animals so that the virus does not continue to evolve in such a way as to be able to infect our upper airways and eventually our lungs.
Another very concerning finding was that the recovered H5N1 virus was able to spread (though not terribly efficiently) through the air 17 – 33% of the time from directly infected ferrets to healthy ferrets in a separate, but nearby cage in the same room. Five of the six infected ferrets who became infected from these respiratory droplets died. This is concerning because the USDA has been telling us that cow-to-cow and cattle-to-poultry transmission is likely occurring predominantly through fomites (virus on inanimate objects such as workers clothing or the equipment used to milk the cows) and not through respiratory means, and it is further concerning, because even though not highly transmissible through respiratory droplets, this represents an increase in transmissibility compared to prior strains of the bovine H5N1 virus.
We know from our prior studies of avian influenza viruses that while avian influenza viruses generally only cause sporadic infections in mammals, including humans, that a mutation named PB2-E627K (the PB2 is a designation for one of the H5N1 proteins, the E and K specify the amino acid substitution (lysine) that is present in the protein, and the 627 specifies the exact location in the amino acid sequence of the protein where the mutation occurs. This mutation has been known to increase transmissibility of avian influenza viruses in mammals, increase the polymerase (the enzyme necessary to replicate (make more of) the virus’ proteins) activity, and to increase virulence. Thus, there is concern that the H5N1 virus circulating cattle might be evolving to better infect mammals, including humans.
Another mutation was also observed – PB2-M631L. This mutation increased polymerase activity in human cells, and this is another concerning indicator of potential evolution of the virus to adapt to enhanced mammalian transmission.
Fortunately, the mutations have not resulted in resistance to all of our available antiviral medications for influenza A, although the virus has become less sensitive to the antiviral medication that has been our standard treatment for these patients.
My take-a-way from this research is that we are likely underestimating the risks of the current outbreak of H5N1 among dairy cattle in the U.S. both in terms of the potential for spread to humans and the risk for severe disease. This does not mean that either of these are imminent, but there is a progression of warning signs that I look for with any outbreak of a novel virus.
The threshold question is whether the virus is capable of infecting humans (the answer for H5N1 is yes) and is there already existing immunity in the population (likely no, although there is one study showing that people infected with the 2009 pandemic influenza A virus might have some degree of cross-reactive antibodies, but keep in mind that that particular pandemic did not cause a large proportion of the U.S. population to become infected, so this may not be a significant mitigating factor.)
The progression of risk factors in my mind then goes as follows:
Is there a large and sustained outbreak of the virus in its natural host (in the case of H5N1 that would be wild aquatic birds and ultimately domesticated birds and poultry)? Yes
Next level of concern: Do humans have significant interactions with those natural hosts? Yes, with respect to poultry farm workers and those who cull flocks in which infection has occurred.
Next level of concern: Have the infections extended to other animal species, especially mammals? Yes.
Next level of concern: Do humans have significant interactions with any of those species? Yes.
Next level of concern: Have there been any documented spillover (zoonotic) infections from infected animals to humans? Yes.
Next level of concern: Is the virus showing genetic evidence of evolution to more efficient transmission to humans? Yes.
Next level of concern: Is there evidence that the virus can be transmitted by respiratory droplets and/or aerosols? Possibly.
Next level of concern: Have any of the humans infected by animals infected other humans (in other words, is there evidence of onward human-to-human transmission)? Thus far, no.
So, I cannot predict whether this virus will cause an epidemic or pandemic in humans, but certainly, to me, the warning signs and risk factors have steadily been accumulating. This is the time to contain the spread of the infection, and frankly, we are failing to do so. The number of cattle herds infected continues to grow, the number of poultry flocks infected continues to grow, and we are detecting more human infections in the past few weeks than we did for the entire year up to then. We are failing to contain this epizootic (epidemic in animals) and if we don’t get serious about containing this outbreak, we do so at the potential peril to the entire world.
To elaborate on a point that I have made previously, as well as in yesterday’s blog post, this is also an increasingly dangerous time for us not to intervene aggressively. The reason is that our human seasonal influenza season is upon us. Soon we can expect to see many humans, and presumably those who work on farms with cattle, poultry and swine getting infected with our seasonal H1N1 and H3N2 viruses. With large numbers of cattle infected with H5N1, we now have the potential to infect those cattle with the human H1 and H3 viruses. We now know that infected cattle have very high levels of the H5 virus in their utters and that those utters have the cell receptor types needed for avian influenza viruses, as well as those needed for human influenza viruses. That means that cattle could be infected with both. That then increases the risk for a “reassortment” event.
Influenza viruses have eight segments of genetic material. When a host is infected with two different influenza viruses, the viruses can swap one or more of those eight segments. This results in a reassortment – an influenza virus with some genetic material from the avian virus and some genetic material from the human virus. This can result in changes to the intrinsic properties of the virus and could result in increase transmissibility, infectiousness, virulence and/or an enhanced ability for the virus to transmit from humans to other humans. This has been at play in at least three prior influenza pandemics. It is especially of concern because, as we have seen in trying to identify cases of avian influenza infections in farm workers, many workers do not seek medical attention when ill because they may not have health benefits, but even more often, they are concerned that missing work will result in loss of pay or even loss of their jobs. Conversely, as we get to this flu season and need to reduce the risk that infected farm workers with seasonal influenza will transmit those viruses to animals, we will be undermined by the same challenges that will result in workers working even while sick. To address this would require symptomatic screening of workers, rapid testing, and isolation of those who have influenza, but this would be a huge challenge without cooperation from farmers, and likely significant incentives from the government.
Thus, it is imperative that we increase testing to identify infected animals and humans, increase the genetic surveillance of the virus to detect these genetic changes that may increase these risks, increase testing of close contacts to ensure that we are not seeing human-to-human transmission, increase containment efforts of infections within and between herds, increase the availability of testing to health care providers because it may be difficult to distinguish an avian influenza infection from the human seasonal influenza that we see at this time of the year, increase the cooperation of farmers, and increase our research into effective vaccines and antivirals.
Dr. Pate's blog 