Study: Poultry most likely to bring H5N1 to Americas

Dec 5, 2006 (CIDRAP News) – Poultry infected with H5N1 avian influenza pose the greatest risk of bringing the disease to the Americas, according to a new study by British and US researchers that challenges US efforts to detect flu in migratory birds.

Once on this continent, avian flu is likely to spread to migratory birds that will cross US borders—but the greatest risk will be birds from Central and South America that are not sampled in current wild-bird testing, the researchers said.

The study, to be published in Proceedings of the National Academy of Sciences, employs a complex analytical method that compares the migratory routes of wild bird species thought to be the main reservoirs of avian flu with data on legal trade in poultry and wild birds and avian-flu gene sequences deposited in the public database GenBank.

Plotting those pieces of data against each other allowed the researchers to hypothesize whether migratory birds, wild bird trade, or poultry were responsible for H5N1 influenza's past spread across the globe, as well as to model its possible future paths.

Heading their conclusions: The combination of poultry trade and bird migrations allowed the virus to spread much farther than either would have allowed on its own.

Heading their predictions: The greatest threat to the continental United States will be the arrival of avian flu in Central and South America—where poultry trade is less restricted than in North America—via live poultry imports from countries where avian flu has affected either domesticated or wild birds. Strict regulation of poultry trade across US borders will not be adequate protection, they concluded.

"The question is not just who you trade with, but who your neighbors trade with," A. Marm Kilpatrick, PhD, a senior research scientist with the Consortium for Conservation Medicine and the lead author of the study, told CIDRAP News.

The Consortium is a New York-based non-profit supported by six institutions: the Cummings School of Veterinary Medicine at Tufts University, the Johns Hopkins University School of Public Health, the U.S. Geological Survey's National Wildlife Health Center, the University of Pittsburgh School of Public Health, the University of Wisconsin-Madison Nelson Institute for Environmental Studies, and the Wildlife Trust. Other authors came from the Royal Society for Protection of Birds and the Smithsonian Institution.

Kilpatrick said the researchers' analytical method allowed them to theorize about which population—poultry, migratory birds such as ducks and swans, or traded wild birds such as parrots and birds of prey—was responsible for the spread of H5N1 influenza across Asia and into Europe.

Poultry played a greater role than wild birds in distributing H5N1 through Asia, they found, but migratory birds that picked up the virus from poultry carried it westward, introducing it to 20 of the 23 European countries where it has been found. In Africa, they suggested, both poultry and wild birds played a role, along with poultry products such as chicken droppings bought for fertilizer and fish feed.

The findings challenge previous conclusions on the routes by which some countries were infected. For instance, plotting genetic sequences from H5N1 isolates against migratory routes revealed that bird flu arrived in Turkey, the first European-region country to be affected, not through previously blamed poultry imports from Thailand but via migratory birds winging from Russia.

The researchers' method—which combined estimates of "infectious bird days" (the product of the number of birds entering a country, the prevalence of infection in those populations, and the number of days birds are likely to shed virus) with data on trade and migration from U.S. and international agencies—does not consider the possible influence of the illegal trade in poultry and wild birds, an omission that Kilpatrick acknowledged is a weakness.

But the analysis points so strongly to the influence of legal trade in spreading the pathogen that it argues for implementing trade controls, he said.

"Although the risk of H5N1 introduction into the mainland United States by any single pathway is relatively low, the risk of introduction by poultry to other countries in the Americas, particularly Canada, Mexico and Brazil, is substantial unless all imported poultry are tested for H5N1 or trade restrictions on imports from the old world are imposed," the report says.

The argument over the relative roles played by poultry and migratory birds in spreading H5N1 has been bitter, with agricultural interests defending poultry and conservation groups contending that wild birds are victims rather than disease vectors. The researchers' conclusions are likely to find favor with conservation groups, and appear to accord with past observations by avian virologists that migratory-bird importation to the United States is unlikely because flyways and feeding grounds allow relatively little overlap for viral exchange.

But the research implicitly challenges the focus of the $29 million migratory-bird testing effort being conducted in the United States by the departments of Interior and Agriculture. Since April that effort has tested more than 21,000 samples from wild birds in the United States, primarily in Alaska, without finding any high-pathogenic avian flu.

Because the wild birds sampled to date have shown such low prevalence of all avian flu strains—2.6% among Alaskan isolates, according to the National Wildlife Health Center in Madison, Wis.—surveillance should refocus on dead birds, the researchers said.

But scientists at the National Wildlife Health Center—which leads the US sampling effort but is also a coalition partner of the Consortium for Conservation Medicine—said Monday's study lacks enough data to persuade them to shift their efforts. In particular, they said a decision by the authors to exclude shorebirds from their analysis leaves out important information, because shorebirds congregate in large groups that facilitate viral exchange more than individual encounters do.

"A model is only as good as the assumptions you make and the data you put into it," said Leslie Dierauf, VMD, the center's director. "There may be better data we can obtain on trade in domestic fowl. There is certainly in my mind at this point not good enough data for migratory birds."

Nevertheless, Dierauf, who reviewed the paper a year ago when it was in draft form, said the analysis raises questions that are vital for successful avian flu prevention and control.

"I am not certain [the paper] makes a significant advance in knowledge, but I do know it sets a number of scientific matters on the table that we all need to look at, no matter whether we are looking from the wild-bird perspective or the poultry perspective or the trade perspective," she said. "That is very good."

Kilpatrick AM, Chmura AA, Gibbons DW, et al. Predicting the global spread of H5N1 avian influenza. Proc Nat Acad Sci 2006 (published online Dec 7) [Abstract]

See also:

Olsen B, Munster VJ, Wallensten A, et al. Global patterns of influenza A virus in wild birds. Science 2006 Apr 21;313(5772):384-8 [Full text]

This week's top reads