#18,741
Even though we are now more than 14 months, and over 1,070 infected dairy herds, since the discovery of HPAI H5N1 in dairy cows there is still much we don't know about how the virus is spreading. Testing of cattle has been limited, is generally voluntary, and is focused almost exclusively on lactating dairy cows.
A recent study suggests the virus is far more widespread in livestock than has been reported (see Nature: A Mathematical Model of H5N1 Influenza Transmission in US Dairy Cattle) and serological data suggest some human infections are likely going unreported (see MMWR: Serologic Evidence of Recent Infection with HPAI A(H5) Virus Among Dairy Workers).
Many dairy farmers - fearing loss of income or the stigma of infection - simply prefer a `Don't test, don't tell' strategy. We've also seen farm workers reluctant to report illnesses, or be tested for the virus, over fears of losing their jobs (see EID Journal: Avian Influenza A(H5N1) Virus among Dairy Cattle, Texas, USA).
Over the past year we've also seen spillovers of H5N1 into goats, alpacas, pigs, and most recently a sheep in the UK. Despite very limited surveillance, we've also seen hundreds of peridomestic animals (cats, foxes, mice, etc.) infected in and around dairy farms and in the wild.The USDA's Dairy Herd Status Program website hasn't been updated in 5 weeks (Apr 25th), and continues to show just 100 herds (out of an est. 36,000) from 18 states enrolled in the voluntary herd monitoring program.
It seems likely we are only seeing the tip of the H5N1 iceberg.
While the prevailing theory is that HPAI mainly affects lactating dairy cows - and is due to the virus's affinity to bovine mammary cells - very little testing of beef cattle has been undertaken.
At least one study (see Virology: Detection of Antibodies Against Influenza A Viruses in Cattle) has recently reported that bulls and steers were just as likely to carry antibodies to (non-HPAI H5) IAV as cows and heifers.
Until now, evidence of IAV viremia in cows has been scant. Admittedly, most studies have focused on viral detection in milk, mammary tissues, and less often; respiratory tissues.Today we have a dispatch, published on Friday in the CDC's EID journal, which raises new questions on how HPAI might be spreading among cattle, as it finds evidence of viremia in a number of cows they tested.
If today's findings prove to be commonplace, it raises new questions as to how the virus may be spreading, and even potential risks to the food supply
Volume 31, Number 7—July 2025
Dispatch
Evidence of Viremia in Dairy Cows Naturally Infected with Influenza A Virus, California, USA
Jason Lombard , Chloe Stenkamp-Strahm, Brian McCluskey1, and Blaine Melody
Abstract
We confirmed influenza A virus (IAV) by PCR in serum from 18 cows on 3 affected dairy farms in California, USA. Our findings indicate the presence of viremia and might help explain IAV transmission dynamics and shedding patterns in cows. An understanding of those dynamics could enable development of IAV mitigation strategies.
In March 2024, the United States Department of Agriculture’s National Veterinary Services Laboratories (NVSL) confirmed avian influenza virus (IAV) A(H5N1) clade 2.3.4.4b in dairy cows in Texas, USA (1,2). That subtype was further characterized as genotype B3.13. Since that detection, >1,070 herds in 17 states have been affected; most of those herds are in California (3). Clinical signs observed have been variable, but fever, nasal discharge, loss of milk production, and mastitis are common (4).
Experimental and field investigations into the transmission dynamics, pathogenesis, and epidemiology of H5N1 virus in cows are ongoing. Researchers have inoculated cows via the intramammary route in 3 studies, and resulting clinical signs were similar to those from field reports of affected cows, including severe disease requiring euthanasia. Viral RNA was found in milk samples in all studies (5–7) but in blood samples in only 1 study (7).
Early in the ongoing outbreak, nasal swab, whole blood, serum, and milk samples were collected from affected dairies in Texas, New Mexico, Kansas, and Ohio. Viral RNA was detected in nasal swab (10/47 cows), whole blood (3/25 cows), serum (1/15 cows), and milk (129/192 cows) samples (4).
We sampled cows using a serial sampling design early in the outbreak on affected dairy farms in California. We report detection of IAV RNA in serum samples from lactating dairy cows.
(SNIP)
Conclusions
Our results suggest that a percentage of lactating cows on dairies affected by H5N1 virus experience viremia before or during the peak of clinical cases in the herd. We detected viral RNA in serum of each PCR-positive cow at a single sample date. Viremia therefore appears to be transient, but the duration is unknown because cows were not sampled daily.
Although the finding of viremia does not specify how IAV made it to the bloodstream, virus present in circulation suggests that multiple exposure pathways might be possible, including oral and respiratory routes. Intramammary inoculation studies have shown viral RNA to be in multiple tissues at necropsy (6,7), although viremia had not been consistently detected.
Viremia enables virus to reach many tissues in the body, including the kidneys, which is evident in this study given detectable RNA in urine samples. That process raises concern for food safety and whether viremia could lead to the presence of H5N1 virus in meat from culled dairy cows.A study that evaluated condemned carcasses found viral RNA in 1 of 109 total samples (8). Further, an aborted fetus from farm B, not from a known viremic cow, was positive for H5N1 virus in lung and brain tissue by PCR and immunohistochemical staining. H5N1 virus can move into the reproductive tract and is associated with abortion, which also has implications for the use of fetal serum products.
All cows in this study had IAV detection in serum and milk, so it is unclear whether intramammary infection led to viremia or viremia led to intramammary infection. Three cows classified as healthy had viral RNA detected in serum, and 1 of the 3 had viral RNA detected in serum and urine. The relationship between viremia and clinical signs is therefore unclear, although we might have sampled those cows before the onset of clinical disease.
Determining whether the viremia we detected is a rare event is crucial. Viremia has only been detected in 1 previous H5N1 intramammary experimental infection study and 1 field study. To clarify the frequency of viremia, more studies that evaluate IAV RNA in the serum of cattle should be completed.
The prevalence of viremia detected in the Jersey breed herd compared with the 2 Holstein breed herds suggests breed differences in susceptibility to viremia from IAV infection might be involved, but differences in cow selection by farm might have affected prevalence. We also recommend a genetic comparison of viral strains collected between studies and between states.
Viremia in California dairy cows could be the result of viral evolution, because viremia was not well documented previously, and experimental studies used a strain of H5N1 virus from the early stages of this outbreak. Further in-depth studies that include viral sequencing are necessary to strengthen the evidence supporting our conclusion.
In summary, findings of IAV in serum of cows on farms in California indicates the presence of viremia and could help explain viral transmission dynamics and shedding patterns in cows. Understanding such dynamics could help in development of mitigation strategies to prevent transmission and spread of IAVs, including H5N1 virus.
Dr. Lombard is a veterinary epidemiologist at Colorado State University, Fort Collins, CO, USA. His research interests are infectious diseases of cattle.
This study hints at something we've discussed often. The unseen impacts of long chains of infection over time; in cattle, other livestock, and in peridomestic animals (mice, cats, birds, etc.) that may be exposed to infected cattle.
Like a classic serial passage experiment (see above), the virus that emerges at the end of a long chain of infections may differ greatly from the wild-type virus at the start. Most will be evolutionary duds, but occasional a more biologically `fit' virus will emerge.
Sadly, our current passive and selective (usually voluntary) testing is unlikely to show us what changes may be occurring in real time.
Which places us at very real risk of being caught short when something more virulent, or more transmissible, finally does emerge.
