Avian Influenza Outbreak

Less than a week ago, I interrupted my blog series on a comprehensive update on COVID-19 to bring you breaking news about an outbreak of highly pathogenic avian influenza (HPAI) that was occurring in cattle on dairy farms in the U.S., including Idaho. This was especially of note and concern because:

1. While avian influenza viruses can incidentally infect mammals, we have typically believed these infections to be sporadic and involving relatively few mammalian species. Over the past two years, the extent of infection and breadth of species of animals infected (cats, dogs, foxes, tigers, leopards, coyotes, bears, seals, dolphins) has been quite alarming relative to potential ecological impacts and threats to food supply (readers likely have witnessed first-hand the rise in egg prices resulting from the culling of large numbers of chickens when outbreaks have occurred on poultry farms);

Breaking News As I am writing this blog piece, I received an alert that 532 dead Adelie penguins were found last month and “thousands” more are feared dead on Heroina Island in Antartica. Initial tests for H5N1 infection were inconclusive, so samples have been sent to specialized laboratories for further testing. Source: Reuters’

2. Cows have never before (to our knowledge – other than in controlled animal studies) been infected with this virus (influenza A(H5N1) – see my prior blog post for an explanation of what the virus designation represents);

3. We have previously believed that in cases of animals being infected, that each case was the result of contact with a sick or dead bird or contamination of the animal’s water or food source by an infected bird’s excrement containing the virus, as opposed to forward transmission between and among animals. The circumstances of the recent outbreaks on dairy farms have heightened suspicion of potential transmission between cattle, but this, as well as many other important questions remain unanswered.

Breaking News Again, even as I am writing this, I just received another alert that the USDA (Dept. of Agriculture) representatives were reported to have shared at an international meeting that they believe that the virus is not spreading from cow-to-cow directly, because the virus is concentrated in the cows’ udders. Instead, they believe that the virus is spreading indirectly from cow-to-cow by the suction cups on the milking machines, although they have not ruled out transmission from milk droplets on clothes or gloves or workers. Source: Science. Note: I have not seen any reports of the virus being detected or recovered from the suction cups, but I am sure that the animal health investigators have sent samples off for testing.

Before we begin to get into the details, here is the bottom line. I agree with the CDC’s current assessment that based upon the facts as we know them today, the public health risk appears to be low. However, three important caveats:

1. “Low” risk does not mean “no” risk.

2. Something has clearly changed, and the evidence of these changed circumstances is unfolding very quickly, so this is not a stable situation, and as a result, our risk assessment could change very suddenly. (As you can see from above, things are changing or happening even while I am in the process of writing this update.) Long time readers of my blog know that how a virus manifests itself in a population is a result of a number of interacting factors – biological changes in the virus (e.g., new mutations to the virus that may impact its transmissibility, virulence, tropism (the specific receptor(s) the virus uses to get into the cells of the host and as a consequence, the type of tissues and organs the virus affects, or immune escape), human behavior changes (for example, increase in large scale travel activities such as spring break or holidays, school beginning or ending, etc.), environmental changes (especially for seasonal viruses, but also changes in our interactions with animals (such as animal fairs, markets, or deforestation) or vectors (such as mosquitoes, ticks, fleas), and the background population immunity (e.g., whether there is some prior exposure and durable immunity or whether there may be cross-reactive protection from a similar virus or prior strain of that same virus).

3. We need to approach this with facts and scientific evidence. I have read and heard many reactions to these developments at both ends of the extremes – there is no reason for any concern (something like, “there is one person affected who has pink eye.”) to the other extreme that is doom and gloom and a certain apocalyptic outcome with at least half of the population dying off. Neither assessment is realistic. It is important that we look to the facts, put this in perspective, and then identify the questions that need to be answered to allow us to better assess the risk, be transparent, clear and frequent in health agency communications, and reexamine our plans. It was precisely for developments such as this that Dr. Epperly and I wrote our book that was released almost exactly one year ago on lessons learned from the COVID-19 pandemic and how we should plan for future pandemics (whether they materialize or not). https://www.press.jhu.edu/books/title/12896/preparing-next-global-outbreak

What new or additional information do we have?

The epidemiological investigation is unfolding at a rapid rate, and the following are the updates that have been made public.

1. The USDA has stated that roughly 10% of the cattle on the affected dairy farms appear to be infected (not clear whether this was found to be the case at each dairy farm or whether there were wide ranges of numbers of infected animals and this is an average). While HPAI infections in domestic birds result in very high rates of mortality, thus far, I have not seen any reports of deaths among infected cows. Cows do get sick, as manifested by weight loss, decreased milk production and a change in color and consistency of their milk to more resemble colostrum, but there are already reports of cows recovering from the illness. Antibody testing is being done to determine just how much of herds have been infected by the virus.

2. The USDA has also indicated that there are no concerns for the general public because:

• Sick cows are being isolated from the remainder of the herd;

• Their milk is being discarded and not entering the commercial market;

• Even if some milk gets through to the commercial markets, the pasteurization process reliability kills influenza viruses.

• As of April 3, 2024 at 12 p.m. (ET), there are outbreaks at 15 dairy farms involving the following states (reference 2):
  • Idaho
  • Kansas
  • Michigan
  • New Mexico
  • Ohio
  • Texas

• Confirmation of A(H5N1) infection in cattle is not made until the sample is confirmed at the National Veterinary Services Laboratories (NVSL). The dates of confirmation are as follows:
  • Texas (1^st dairy farm) 3/25/24
  • Texas (2^nd dairy farm) 3/26/24
  • Kansas (1^st dairy farm) 3/26/24
  • Kansas (2^nd dairy farm) 3/26/24
  • Texas (3^rd dairy farm) 3/26/24
  • Texas (4^th dairy farm) 3/27/24
  • Texas (5^th dairy farm) 3/27/24
  • Texas (6^th dairy farm) 3/27/24
  • Michigan (1^st dairy farm) 3/29/24
  • Texas (7^th dairy farm) 3/30/24
  • Texas (8^th dairy farm) 3/30/24
  • New Mexico (1^st dairy farm) 4/1/24
  • Idaho (1^st dairy farm) 4/1/24
  • Kansas (3^rd dairy farm) 4/1/24
  • New Mexico (2nd dairy farm) 4/1/24
  • Ohio (1^st dairy farm) 4/2/24

Breaking News While writing this update, I just received an alert that A(H5N1) has just been detected at another dairy farm in Texas bringing the number of dairy farms in the US with outbreaks to 16 and making this the 9^th dairy farm in Texas. Source: BNO News

• There appear to be very high levels of virus in the cow’s milk. This is a bit surprising in that HPAI typically causes either respiratory or gastrointestinal (or both) infections in animals, which would not provide an obvious explanation as to how the virus would end up in the milk.

The CDC (Reference 4) issued an update regarding the human infection that occurred in a farm worker on one of the Texas dairy farms. Here is what we know from that:

1. The worker presented with signs of conjunctivitis – irritation or infection of the inner lining of the eyelids – the only sign reported by this patient being redness of the eye.

2. The patient is or was being treated with a commonly used influenza A antiviral medication.

3. The CDC reports that human infections with avian influenza A viruses are uncommon, but do occur sporadically across the world. It further reports that the CDC has been monitoring for illness among people exposed to H5 virus-infected birds since outbreaks were first detected in U.S. wild birds and poultry in late 2021. 

4. This is the second known human case of infection from this virus (clade 2.3.4.4b) in the U.S., with the first being in 2022 in a Colorado in a worker involved in culling poultry at a site of an outbreak. That patient’s only presenting symptom was fatigue and he has subsequently recovered from the illness.

5. The CDC indicated that the low levels of virus obtained from the eye specimen and the negative throat swab results indicate that it is very unlikely that the person’s respiratory tract was infected.

What have we learned from the genetic sequencing from the affected cows and the infected human?

1. This virus in an RNA virus, so if you have been following my blog, you will not be surprised that it mutates frequently.

2. Many mutations are of no significance, some are detrimental to the virus and those viruses will generally lose in competition to more fit forms of the virus, so they disappear over time, and then some may be advantageous to the virus (increase viral fitness through increasing transmissibility, receptor binding or evading immune defenses).

3. The viruses that infected the cows and the human are the same version of virus (we refer to these as clades, which you can think about as being different variants if we were talking about SARS-CoV-2 viruses). The virus has mutated over the years, so that is why this current clade designation reflects a number of previous clades.

4. We can look at the phylogenetic trees for this virus (you probably have seen these, but not known what they are called or what they mean, if you have followed the developments with SARS-CoV-2 over the past four years. These are diagrams that plot out the various versions of the virus starting (usually at the left side of the diagram) with the original form (wild-type) or at least the first discovered form of the virus and then those new versions of the virus with the fewest mutations will be closer and those with the most mutations will be further away from the original form of the virus. When the collection of mutations has been found to be significant, then the SARS-CoV-2 virus was assigned a new variant name or in the case of this A(H5N1) virus, it is assigned to a new clade).

5. In this case, we can look at a phylogenic tree for the changes in each of the viruses’ major proteins that we are interested in. We previously discussed the HA (hemagluttinin) protein and its role (especially in virus attachment to the host cell and fusion with the host cell’s membrane in order to allow the virus to enter the cell). This virus has the H5 subtype protein, and the phylogenic tree shows that there have been relatively minor mutations to this protein (that is good, because this protein is a vaccine target because neutralizing antibodies are made to this protein as a result of infection). We can look at the tree for the NA (neuraminidase) protein (important in facilitating the release of newly formed viral progeny from the cell), in this case, the N1 protein, and see that it too, has relatively minor mutations, though certainly more than have occurred within the H protein (again, good news, because the H1 subtype is also a target for vaccines, but also of some of the antivirals we use).

However, there is a mutation required to the PB2 protein (not a vaccine target) that is necessary for transmission to mammals, though not sufficient in and of itself to allow for transmission in humans, and we see that has occurred, but only in the sample from the human, suggesting that this was an “in-host” mutation (occurred in the human after infection during translation and replication within human cells rather than in any of the cattle or prior bird samples). However, there are many more mutations to this protein and far more “divergence” (distance away from the prior forms – due to the large number of mutations), even from versions of the virus detected in January of this year in birds from other U.S. states (CA, CO, MN, WI, ME). Perhaps one of these explains the increased transmission in mammals.

6. The CDC issued a “Technical Update: Summary of Genetic Sequences of Highly Pathogenic Avian Influenza A(H5N1) Viruses in Texas” (Reference 5 on April 2, 2024). CDC confirmed that sequencing identified that the virus infecting the farm worker was from clade 2.3.4.4b and is closely related to viruses detected from the dairy cattle in Texas. However, the CDC did note that the human isolate did have the mutation mentioned above in #5 referred to as PB2 E627K that is associated with viral adaptation for transmission to mammalian hosts. Reassuringly, the CDC did not see markers that would suggest that the virus has become resistant to the oral antiviral agents we have available to treat this viral infection. More good news was the fact that the sequence suggested that the candidate vaccine viruses that have already been provided to vaccine manufacturers are a good match.

Important Article

A very timely article appeared in the April 2024 issue of Emerging Infectious Diseases (Reference 7) entitled: “Divergent Pathogenesis and Transmission of Highly Pathogenic Avian Influenza A(H5N1) in Swine. The investigators assessed the susceptibility of swine to avian and mammalian HPAI H5N1 clade 2.3.4.4b strains (the strains identified above on the dairy farms in cows and in the human). They demonstrated that all sources of the virus replicated (produced new viruses indicating infection) in the lungs of pigs and caused lesions (areas of infection that can be seen on imaging or under a microscope) consistent with influenza A infection. Interestingly, and concerningly, viral replication in the nasal cavity of the pigs was only observed when the source was mammalian isolates of virus.

What are the important questions to ask and why?

1. Seroprevalence testing results in dairy herds. This is the antibody testing we are waiting on that I mentioned above. When domestic birds get A(H5N1), its pretty hard to miss because most, if not all of the birds die and we can easily spot the carcasses. However, it doesn’t appear yet that cows get particularly and obviously ill in all cases, so the antibody testing can give us information as to how wide-spread this viral infection is in cattle. We also don’t know whether asymptomatic infections in cattle occur and if so, whether they, too, have high levels of virus in their milk.

2. It will be important to see if virus will be recovered from the milking machine to determine its possible role in transmitting the virus.

3. Given the extent of the genetic mutations to the PB-2 protein, we need to focus efforts on studying these mutations to learn which ones are meaningful and if one or more of these mutations accounts for the increased transmission to the broad range of mammals we are seeing infected.

4. It would be very useful for researchers to take milk from infected cows, pasteurize it, and then check the pasteurized milk for any viable virus to confirm USDA’s assertion that pasteurization completely inactivates the virus. Similarly, it would be good to test unpasteurized cheeses for the presence of the virus or to see if the virus in infected cows’ milk can remain viable and infectious after going through the cheese-making process.

5. An important question to ponder is whether we are doing enough surveillance for avian influenza on farms and in other animals, for example beef cattle (I am not very knowledgeable about animal medicine, but I don’t know why beef cattle would be less susceptible to infection than dairy cattle, infections in beef cattle may be more likely to go undetected because we don’t have milk production changes to alert us to illness, and further, it is my understanding that beef cattle spend more time outside than do dairy cattle) and particularly, whether we need to increase our monitoring in pigs given the outbreaks on dairy farms. As mentioned in my previous post, although there is some question as to whether pigs really are able to transmit influenza strains to humans more easily than birds are, it has long been hypothesized that they do, and there certainly are some reasons to believe that is the case. If the avian influenza can directly infect pigs (see “Important Article” above and Reference 7), or given we now know that avian influenza can directly infect cows, and if some of these farms also raise pigs, it is reasonable to ask whether pigs may be getting infected directly from cows, or more likely from shared food or water supplies. Given the high levels of virus in cow’s milk, there is also a reasonable concern as to whether the virus is getting aerosolized during the milking process (usually twice a day as I understand it). I also have read that some farms use pressure washing to clean floors, and this is a process very likely to aerosolize viruses, in fact, this was thought to likely contribute to spread of SARS viruses in the animal markets in China. Perhaps consideration should be given to keeping cows and pigs separated and not allowing them to share the same water or food sources.

General recommendations:

1. The CDC should update its Influenza Pandemic Plan based on learnings from the past four years.
2. HHS should be honest with hospitals and health care providers as to whether the Strategic National Stockpiles have been replenished.
3. The FDA should do an assessment as to the risk of drug shortages for treatments that would be necessary were A(H5N1) to develop pandemic potential, and an assessment of how much and how fast the production of influenza antivirals could be ramped up in the event of human outbreaks or a pandemic.
4. Hospitals should update their own pandemic plans in light of learnings from the COVID-19 pandemic.

Even though at this time, a pandemic threat from this virus seems low, that could change. Even if that threat remains low, it is a great opportunity to think through pandemic planning scenarios and game theory in light of the potential for a different influenza virus to cause a pandemic and with the experiences and increased knowledge we have now having gone through a very serious and prolonged pandemic.

References:

1. U.S. dairy farm worker infected as bird flu spreads to cows in five states | Science | AAAS
2. APHIS (Animal and Plant Health Inspection Service), USDA (U.S. Department of Agriculture), EMRS (Emergency Management Response System). Highly Pathogenic Avian Influenza (HPAI) Detections in Livestock | Animal and Plant Health Inspection Service (usda.gov).
3. Highly Pathogenic Avian Influenza A (H5N1) Virus Infection Reported in a Person in the U.S. | CDC Online Newsroom | CDC
4. Highly Pathogenic Avian Influenza A (H5N1) Virus Infection Reported in a Person in the U.S. | CDC Online Newsroom | CDC.
5. https://www.cdc.gov/flu/avianflu/spotlights/2023-2024/h5n1-analysis-texas.htm.
6. https://www.statnews.com/2024/04/03/h5n1-bird-flu-in-cows-risk-to-humans/.
7. https://wwwnc.cdc.gov/eid/article/30/4/pdfs/23-1141.pdf.