H5N1 bird flu virus particles found in pasteurized milk but FDA says commercial milk supply appears safe

WASHINGTON — Testing conducted by the Food and Drug Administration on pasteurized commercially purchased milk has found genetic evidence of the H5N1 bird flu virus, the agency confirmed Tuesday. But the testing, done by polymerase chain reaction, or PCR, cannot distinguish between live virus or fragments of viruses that could have been killed by the pasteurization process.

The agency said it has been trying to see if it could grow virus from milk found to contain evidence of H5N1, which is the gold standard test to see if there is viable virus in a product. The lengthy statement the agency released does not explicitly say FDA laboratories were unable to find live virus in the milk samples, but it does state that its belief that commercial, pasteurized milk is safe to consume has not been altered by these findings.

“To date, we have seen nothing that would change our assessment that the commercial milk supply is safe,” the statement said.

The document was long on assurances but short on details of what has been undertaken or found. It does not specify how many commercial samples were taken or in how many markets, nor does it indicate what percentage of the samples were PCR-positive for H5N1. The statement did not indicate if the testing suggested the amounts of viral genetic material in the milk were low or high.

Furthermore, the statement did not reveal if the milk products were purchased in parts of the country where outbreaks have occurred, or in areas where cows haven’t been seen to have been infected.

The FDA did say it is testing milk from infected animals, in the processing system, and from store shelves with the goal of producing a “large representative national sample” of the milk supply. Beyond assessing the safety of the supply, such a sample could enable the agency to construct a more robust picture of H5N1’s spread among dairy cows nationally.

The agency emphasized that testing of commercially available milk is ongoing, and this includes efforts to discern any potential differences between different dairy products, such as cream and whole milk.

Michael Osterholm, director of the University of Minnesota’s Center for Infectious Diseases Research and Policy, called the findings “unsurprising,” saying that PCR testing of milk could turn up genetic evidence of an array of other pathogens that would have been killed by pasteurization, such as listeria or salmonella.

“I still believe that the risk is very low for any live virus activity in milk,” Osterholm said, though he noted the FDA’s statement included too few details to show that. “It’s kind of an obtuse document.”

The FDA, which has been dodging questions for some time about the work it is doing to assess the safety of the milk supply, said it did the commercial survey as part of its efforts to assess the federal-state milk safety system in the context of the current outbreak of H5N1 bird flu in dairy cows in at least eight states across the country. As of Monday, 33 outbreaks in herds have been confirmed.

“Some of the samples collected have indicated the presence of [H5N1] using quantitative polymerase chain reaction (qPCR) testing,” the FDA wrote in the statement.

It reiterated that it believes the pasteurization process is “very likely” to inactivate H5N1. “Data from previous studies that serve as the underpinnings of the FDA’s current milk supply safety assessment show that pasteurization is very likely to effectively inactivate heat-sensitive viruses, like H5N1, in milk from cows and other species,” the FDA wrote, though the agency acknowledged that no studies have been published on the impact of pasteurization on H5N1 viruses in milk.

Some academic centers have begun that work, however. John Lucey, director of the Center for Dairy Research at the University of Wisconsin-Madison, told STAT in an email that preliminary testing of milk conducted at the university has indicated that pasteurization is effective in killing the virus. But he declined to share further details. The U.S. Department of Agriculture “is currently doing their own trials and I expect to hear from them shortly on their results on pasteurization which should be viewed as the official testing,” he wrote.

The virus, which has been causing outbreaks around the world for more than a quarter century, had previously not been seen to infect cows.

Pasteurization is a process of heating raw milk to kill problematic microbes. There are several ways it can be done; some involve pushing milk rapidly through a hot pressurized tube, while others heat milk over longer time periods in large vats.

The extent of microorganism destruction depends on the combination of temperature and holding times in the pasteurizer as well as the specific properties of the pathogen and how many of them are present in the milk. Higher levels of bacteria or viruses will take more time to kill. Minimum temperature and time requirements that dairy processing plants use in the U.S. are based on studies showing what it takes to kill the most heat-resistant pathogen found in milk: Coxiella burnetii, the bacteria that causes Q fever.

“It’s a very tough microbe,” said Terry Lehenbauer, a bovine disease epidemiologist and director of the Veterinary Medicine Teaching and Research Center at UC Davis. “As a result, if we know that pasteurization can effectively kill that indicator organism, we know that all the other pathogens that are less durable will also be eliminated.”

The majority of milk in the U.S. is treated with a high-temperature, low-time technique, where the milk is heated to 162 degrees Fahrenheit for 16 seconds. Because it is a thermal process and doesn’t involve the physical removal of any bacteria or viruses, dead genetic material left behind could be detected by a method that picks up small fragments of DNA or RNA, such as qPCR.

“It really doesn’t tell you anything about aliveness or the capacity of the virus to infect,” said Jennifer Cruickshank, a dairy extension specialist at Oregon State University who specializes in genomics. “It really only tells you that in the case of influenza virus, this particular fragment of RNA was present.”

To date, there have not been any published studies about the effectiveness of pasteurization on the H5N1 virus in milk. But researchers who spoke to STAT said that inactivation of this virus is expected based on prior work that has shown it is relatively sensitive to heat. A review published in 2022 by researchers in Germany, found that at a temperature of 140 degrees Fahrenheit, influenza viruses, including H5N1 strains found in chicken meat, feces, and eggs could be reduced by four orders of magnitude within 30 minutes.

But milk is a very different substance, a suspension of water, milk fats, and proteins, whose complex kinetics can sometimes complicate the pasteurization process. Studies conducted by the USDA in the ’70s and ’80s found that those milk fats could act like protective bubble wrap around the virus that causes foot and mouth disease — making it more difficult for the heat from pasteurization to penetrate and inactivate the virus. There’s no indication that is true with H5N1 viruses. And Lehenbauer pointed out that avian influenza is in a class of viruses encased in an envelope, which, counterintuitively, makes them much less hardy compared to non-enveloped viruses, like the one that causes foot and mouth disease. But as with any emerging pathogen, it’s worth knowing for sure.

“Pasteurization kills much sturdier viruses than influenza, so we expect it to work,” said Andrew Pekosz, a molecular microbiologist who studies respiratory viruses at Johns Hopkins Bloomberg School of Public Health. “But it’d be great to have the data.”

This story has been updated.