Pandemic Flu Risk Raised by Lax Hog-Farm Surveillance
The great lesson of the 2009 influenza pandemic was that new, deadly flu strains wouldn’t necessarily emerge from the pathogenic hotbox of an Asian animal market. They could start in the western world’s own backyard, percolating from the incubators of modern farms.
Yet despite the fact that pig farms hosted, and arguably fueled, the evolution of the last pandemic strain, pigs are largely overlooked in influenza surveillance. Researchers who track flu’s evolution — in people, poultry, and even camels and whales — in order to understand its behavior lack reliable data on a vast, teeming viral reservoir. It’s a hog-sized blind spot.
“If we want to know what’s coming in the flu world, it’s important to know what’s out there,” said Stephen Morse, a Columbia University epidemiologist and co-director of USAID’s Emerging Pandemic Threats program. “By the time you see human cases, it’s really too late.”
A striking example of swine surveillance shortcomings comes from the most recent pandemic, which emerged in April 2009 and had global infection rates of about 15 percent before vaccines were available.
Though 2009 H1N1, the strain that became globally known as swine flu, may not have jumped directly from pigs to humans — it could have spent some time in another species, such as chickens — the likelihood is high, and there’s no question its ancestors spent years on North American pig farms.
‘We don’t really know what’s out there and where it is.’
– Peter Daszak
If virologists had samples of that pandemic flu’s most recent ancestors, collected from pigs near the time it started infecting humans, they could compare it to gene sequences from the human strain. From that comparison they might learn which mutations made it so contagious, thus helping predict when other, not-yet-pandemic strains could break out.
That can’t be learned simply by looking at 2009 H1N1 alone: Flu mutates rapidly, and the number of changes its genes accumulate in just a few months makes it nearly impossible to say which mutations were key. The only other way to gain comparable insight is for researchers to genetically engineer more-pathogenic flu strains, a highly controversial approach that might not produce realistic mutants.
“It was the first time in history that we could get the ancestor of a pandemic and see what animal changes were needed for it to become a human pandemic,” said Raul Rabadan, a Columbia University biomedical informaticist who helped identify the pandemic strain’s origins. “Getting data from the experiments that nature does for us is essential.”
But when Rabadan’s team looked for swine flu samples gathered from Mexican pigs in 2009, just before or right after the human pandemic, at a time when near-relatives would be expected to circulate on local farms, they found just a single sequence from one pig.
It was a strangely low number. Though it’s theoretically possible that H1N1 was nowhere to be found in Mexico’s swine herds, it’s much more plausible that almost nobody was looking, or that they didn’t report what they found.
Smithfield Farms, the world’s largest pork producer and the owner of a factory farm near the flu outbreak’s epicenter, claimed to have privately tested its herd and found no flu, but the testing wasn’t independently confirmed and conflicts of interest were blatant.
The Pan American Health Organzation, the international health agency responsible for sampling swine after the outbreak, doesn’t appear to have done so. Sheer post-outbreak confusion and personnel shortages are a possible reason, but there was also political pressure to dissociate the flu from swine farming. (Neither Smithfield nor PAHO replied to repeated requests for comment.)
“All we have is that one isolate from the beginning of the pandemic. It seems to give more information about the pandemic’s origin, but it’s only one isolate. It’s only one datapoint,” said Rabadan. With that sequence his team puzzled out a pandemic family tree that dated a last swine ancestor of the human pandemic strain to October 2008, but it was a low-resolution finding. More specific insights into the pandemic flu’s crucial genetic jump were lost.
“In this day and age, with such incredible advances in sequencing and surveillance technology, you’d think that a question like this, we’d be able to answer it quickly,” said National Institutes of Health epidemiologist Martha Nelson. “The fact that it remains outstanding is tremendously frustrating.”
Yet if PAHO and Smithfield, and arguably the Mexican and U.S. governments, failed in sampling swine during the outbreak, their neglect was hardly atypical. Despite widespread recognition that industrial hog farms, in which animals are raised in crowded, overmedicated conditions, are evolution-accelerating incubators of new diseases that are shipped transcontinentally and overseas, swine flu surveillance has traditionally lagged behind human and bird surveillance.
Mutant Bird Flu Study Fuels Fears of Airborne H5N1
Controversy Continues After Engineered-Bird-Flu Study Published
Lab Security Standards Raise Chances of Mutant-Bird-Flu Catastrophe
Controversial Mutant Flu Research Paused
“It’s not like there was some cover-up,” said Morse. “There was very little sampling going on in pigs in Mexico, and in Latin America as a whole.” In the United States, said Morse, there had been a “patchwork of surveillance.”
Since 2009, swine flu surveillance in the United States and some other countries has improved, at least somewhat. “There are some companies that have been incredibly active,” said Nelson. “It’s increased a lot since the pandemic. It’s increased in the South American countries.” The federal Centers for Disease Control and Prevention, along with the Department of Agriculture, have also pushed for more testing on farms.
Despite these efforts, the patchwork remains. In the first three quarters of 2011 — the most recent data available — the National Animal Health Laboratory Network, a collaboration of federal and state laboratories that offer animal disease testing services to farmers, tested 1,426 U.S. swine herds. The total number of herds is close to 70,000.
Part of the reason so few herds are tested is the nature of swine flu itself. Though pigs are susceptible to flu infections, the symptoms are typically less severe than in humans. “If the animals aren’t even particularly sick, you may not even know it’s going on,” said Gregory Gray, head of the University of Florida’s Global Pathogens Laboratory.
If infections are noticed, there’s still few incentives for farmers. Testing is entirely voluntary, and while infected animals may not seriously threaten a farm, news that flu’s been found can ruin business. Even if people can’t get flu from eating bacon, they’ll still stop buying it. Testing’s benefits for pork producers are minimal.
“People get leery when they see that their livelihoods could become affected,” said Lisa Becton, director of Swine Health Information and Research at the National Pork Board.
For now, the compromise solution is to keep testing data anonymized so that flu strains can’t be traced anywhere more specific than to a state. That, however, limits the sequences’ scientific value. This dilemma remains unsolved, though it’s less pressing than the need for U.S. swine surveillance to be routine and part of a larger, global surveillance network. In many other countries, swine surveillance is even patchier.
“From the pathogen’s point of view, it’s circulating in the whole U.S. population, moving across borders easily through trade and migrating birds. It moves globally every year,” said disease ecologist Peter Daszak, president of the EcoHealth Alliance. “We don’t have a good picture of the whole story. That’s a huge problem. It’s like feeling around for the elephant in a dark room. We don’t really know what’s out there and where it is.”
Daszak’s group is part of the OneHealth Initiative, a strategy for coordinating disease surveillance efforts around the world. That level of of testing and information-sharing is necessary to develop the ability to predict which viruses will become dangerous, he said.
For now that’s impossible, but “in ten years, we should be able, when we got to China or Malaysia and find a virus in a bat or a tree shrew, to get a sequence and know whether it would be pathogenic if it infected people,” Daszak said. “This should be a race-to-the-moon type project.”
Image: Rennett Stowe/Flickr