Genetic analysis reveals H5N1 flu virus outbreak in cows likely started earlier than thought

The H5N1 bird flu outbreak in dairy cows in the United States has likely been going on for months longer than was previously realized, and has probably spread more widely across the country than the confirmed outbreaks would imply, according to an analysis of genetic sequences that were released Sunday by the U.S. Department of Agriculture.

The genetic data point to a single spillover event that probably occurred in late 2023, Michael Worobey, an evolutionary virologist at the University of Arizona, told STAT on Tuesday.

“The bad news is it looks like this is well entrenched and has been in cattle for a long time and … probably very, very, very widespread,” said Worobey, who worked on the analysis with a number of scientists in the U.S. and Europe.

He suggested the outbreak needs to be taken more seriously than it has been until now, especially given the amount of exposure humans have with cattle. “We need to just study the hell out of it for starters … and see if we can close the gap on what is happening and what we know.”

Though there were reports of a mysterious illness affecting dairy cows in Texas as early as February, the USDA first confirmed an outbreak of H5N1 in dairy cows in that state in late March. Since then, the number of confirmed outbreaks has climbed to 33 herds in eight states. One person — a farm worker who developed conjunctivitis — has been confirmed to have been infected.

The USDA has been unclear about whether these outbreaks are all linked — the result of the movement of cattle, farm equipment, or workers — or whether there have been multiple spillover events where infected wild birds have transmitted the virus to cows. Last week, it told STAT it could find links between the infected herds in Texas, Kansas, and New Mexico; it is also known that cows in the first infected Michigan herd detected had been brought in from Texas. But USDA hasn’t been able to draw a line between those outbreaks and others in Idaho, Ohio, North Carolina, and South Dakota.

The genetic sequences cluster too closely for this to be anything other than a single spillover, Worobey said. “If it were jumping from bird to cattle over and over again … I wouldn’t think you would just get this same very reduced genetic diversity where each of the internal segments are showing the same pattern.”

The prospect of an outbreak that has been underway for months longer than had been previously known is not reassuring to those who follow the dangerous H5N1 virus.

“If that’s true, it’s been flying under the radar for a really embarrassingly, frustratingly long time,” Worobey said. “And we have no idea how much it’s spreading asymptomatically and how widespread it is. And we’re trying to deal with something long after the horse has bolted.”

Tom Peacock, an influenza virologist at the Pirbright Institute, a British organization that focuses on controlling viral illnesses in animals, concurred with Worobey’s read of the data.

The USDA has previously reported that it believed that virus from an infected cattle herd had found its way into a nearby chicken operation in Michigan. The genetic sequences for cows, poultry, and other infected species that were released by the USDA suggest that hypothesis is correct, Peacock said. “If you look at all the cattle sequences together, they all cluster, as do the cats and the chickens and the grackles and stuff.”

“The thing that doesn’t fit that picture is the human case,” he said.

The genetic sequence from the human case, which occurred on an unidentified farm in Texas, is sufficiently different from the cattle sequences that it can’t be easily linked to them, he said. The differences suggest that the individual was either infected in a separate event — maybe not via a cow, but through contact with infected wild birds — or that there might have been another line of viruses in cattle early on and it has since died out.

“It’s basically too distant a cousin to be connected directly to this outbreak, which either means it’s a second spillover or there was an early bifurcation of the cattle sequences,” Peacock said.

The 239 genetic sequences the USDA shared on Sunday did not include what’s known as metadata — information on where the sample that generated the sequence was taken, what part of the body of the infected animal it was taken from, or when precisely the collection occurred. They simply state “USA” and “2024,” which limits how well outside scientists can interpret what they are seeing.

Peacock said it would have been helpful to know whether any of the cattle sequences had been generated from samples taken on the farm where the infected worker was thought to have been exposed to the virus. But that information is not available.

Asked whether the analysis of the genetic data increases his sense of the risk H5N1 poses to humans, Worobey suggested he was uncomfortable with the knowledge that H5N1 seems to be spreading in mammals, calling that unprecedented.

Having the virus in a mammalian species with which people have frequent contact gives H5N1 more opportunities to acquire the mutations needed to be able to evolve to be able to infect people, or “more shots on goal,” Worobey said. “That’s bad.”