While details about the ongoing H5N1 influenza outbreak in dairy cattle in the US have been really sparse, a new pre-print about the early H5N1 cases in Texas dairy herds posted on bioRxiv earlier this week (Hu et al. 2024) has provided some important information.

  • A pre-print is a non-peer reviewed version of a paper that’s typically posted online on an open access server. This situation highlights the value of being able to release pre-prints: it allows good labs to get good data out faster to let people know, and to get quick feedback before (or while) going through the process to have the study peer reviewed and published in a properly refereed journal. Unfortunately, there’s also a lot of complete crap on pre-print websites, posted by dodgy groups to get publicity or advance their specific cause, when there’s no hope of their material ever being published in a reputable journal.

We have to be careful about relying on information from pre-prints, but with this research group (from Iowa State University) and the quality of the data in this report, I’m not worried about this one. Here are some highlights:

H5N1 disease in dairy cattle

The report describes what has largely been reported to date for clinical signs in cattle: Infected animals had decreased appetite, decreased milk production and abnormal-looking (thickened) milk. Clinical signs peaked about 4-6 days after they started, and cows got better on their own within 10-14 days. That’s pretty “flu-like” when you consider how other strains of flu present in other species (including humans). Affected cows were most often older cows in the middle of their lactation periods. The problems seem to have started in late February 2024.

H5N1 detections in other species on farm

Dead birds and cats were reported on and close to affected farms. Not long after the virus was detected in the cows, it was also isolated from a local skunk, and then from a dairy worker with conjunctivitis. The pre-print mentions virus from “humans in Texas during March 2024”, but to date only one human infection associated with the cattle outbreak has been reported. It’s not clear whether there are more unreported cases, or whether “humanS” was a typo. (Let’s hope it’s a typo and there was just one human infection.)

Identification and characterization of H5N1 flu virus from dairy farms

The virus was identified in milk samples from affected cattle and in lung and brain samples from dead farm cats. The virus was classified as clade, which is the main clade that’s been circulating in wild birds in the area (and across North America). The genomes of the viruses from the cattle, cats, local wild birds and the person with conjunctivitis were nearly 100% identical and shared a common ancestor, confirming they are linked (see figure below and at the bottom of the post).

More detailed study of the virus genome indicated that it was a reassortment of the H5N1 genotype B3.7 and a low pathogenicity avian influenza (LPAI) virus. The B3.7 ancestor emerged in 2023 from an H5N1 virus and a different LPAI strain. This highlights the tendency of influenza viruses to evolve and mix with other flu viruses (reassort), and is a major reason we do surveillance testing of flu viruses. This current strain is a bit of a nothingburger for humans, but we’re worried about what it could become if it keeps spreading, spilling over to other species and reassorting.

Surveillance for important genetic mutations in H5N1 flu from cattle

This is still an avian influenza virus. It can spillover into mammals but isn’t (yet) well adapted for sustained transmission between mammals. There are some genetic markers in flu viruses that indicate a greater ability to infect mammalian cells and therefore spread between mammals, including people. The H5N1 isolates from the cattle and cats on these farms had some changes that might increase their ability to infect people, but none contained the mix of other mutations that are also important for this. So this strain, while potentially able to spread cow-to-cow via milk, doesn’t have the range of genetic signatures (yet) that would suggest it’s going to be a problem for humans.

It’s good to see some more details about what’s been going on, and it’s good to see that the virus hasn’t yet evolved to something that would be expected to cause widespread problems in mammals. The lack of epidemiological data and information about how this virus has spread between dairy farms is still a big issue. It’s been a textbook example of how NOT to do outbreak communications, which is incredibly frustrating, and has hampered risk assessment and contingency planning. We’ll hopefully get more relevant information in the near future, as some scientific reports like this one start to come out.