杏吧原创

IT鈥橲 an ordinary day at the Banfield pet hospital in Carmel, Indiana. Gracie has a bad case of fleas, Punky has an ear infection, Ruby patiently waits to be spayed and Sheila has a swollen tongue after eating a bee. As veterinarian Kelly Rickabaugh and her staff move from dog to dog, dispensing medicine, cuddles and the occasional treat, preventing an epidemic of human disease isn鈥檛 high on their list of priorities.

But as they log every last detail of the pampered pets鈥� care onto the computer, they are helping to do exactly that. Banfield, a chain of US pet hospitals, has one of the biggest animal health databases in the world and, thanks to a new pet surveillance programme developed at Purdue University in West Lafayette, Indiana, the information is being put to work for humans too. Even the most innocuous feline fevers and canine complaints could serve as warnings of an impending outbreak of a serious human disease.

About two-thirds of human diseases 鈥� including emerging diseases such as West Nile virus (WNV), SARS, and Ebola 鈥� are zoonotic, meaning they can jump between animals and people. So a good way to keep track of what diseases are out there, and could be about to infect us, is to keep track of what鈥檚 jumping into animals.

Animals have long provided us with an early warning of disease, whether it鈥檚 rats dying of bubonic plague or dairy cattle with tuberculosis. But while most countries routinely check farm animals and make efforts to keep tabs on wildlife, until recently no one had come up with the seemingly obvious idea of keeping an eye on pets and zoo animals which, after all, are already under close, loving and often daily scrutiny.

鈥淭here is an army of astute diagnosticians out there that hasn鈥檛 been brought into the system,鈥� says veterinary pathologist Tracey McNamara, whose observation of an unusual crow die-off when she worked at Bronx Zoo in New York city contributed to the discovery of a WNV outbreak in 1999. 鈥淭hese are people who day in, day out look for zoonotic diseases because of their concern for the animals.鈥�

In 2004 Larry Glickman, a veterinary epidemiologist at Purdue who had long seen the potential value of the Banfield database, received a $1.2 million grant from the US Centers for Disease Control to start mining it for useful information. The result is the National Companion Animal Surveillance Program.

Banfield started 50 years ago as a single practice in Portland, Oregon, and began to expand dramatically during the 1990s when it started opening in-store hospitals at PetSmart, the pet-care superstore chain. Today, Banfield has more than 500 hospitals with 2.5 million dogs, 800,000 cats and thousands of birds, ferrets, rats, mice, rabbits, snakes and lizards on its books. Few, if any, medical databases can match its depth and breadth, with detailed electronic records on an estimated 2 per cent of the US pet population.

鈥淭he beauty of the system is you don鈥檛 have to chase animals,鈥� says Glickman. 鈥淚 don鈥檛 have to do anything to get this data. They come to you and tell you what their problem is.鈥�

Pets deliver useful information even when they are healthy. Roughly 1.5 million dogs and 200,000 cats get regular check-ups through Banfield鈥檚 鈥渨ellness鈥� programme. That alone generates at least a million blood samples annually, which can be used for all kinds of studies, including tests for zoonotic pathogens that do not necessarily trigger symptoms in animals. Dogs, for example, can become infected with WNV and develop antibodies, but don鈥檛 usually get sick. So analysing blood samples, especially from dogs too young to have had previous exposure to the virus, could provide an easy way to assess its prevalence in an area and give an early warning of what鈥檚 out there. It鈥檚 easier to sample living dogs than hunt for the standard sign of a West Nile outbreak: dead birds.

鈥淎 good way to keep track of what鈥檚 out there is to keep track of what鈥檚 jumping into animals鈥�

Glickman鈥檚 programme has already produced useful results. Last year the team started studying leptospirosis, a bacterial disease that infects dogs and humans as well as rats, cattle, horses and many other species, causing symptoms ranging from fever and nausea to kidney damage and liver failure. Severe cases in humans are known as Weil鈥檚 disease and have a fatality rate of up to 40 per cent. It is arguably the world鈥檚 number one zoonotic disease, based on geographical spread and incidence, says Glickman.

Growing epidemic

For the past 30 years there has been a veterinary vaccine which kept dogs healthy while also creating a disease barrier to protect people. But in the mid-1990s, new and virulent subgroups not covered by the vaccine appeared and started to spread across the US. By plotting lab test results collected between 2002 and 2004 on a map, Glickman has been able keep track of the spread of the new strains. 鈥淭here is this huge, growing epidemic and now for the first time we have a way of monitoring it,鈥� says Glickman. It turns out that the riskiest places are new housing developments built on old farms and wooded areas.

The database could also be used to keep track of outbreaks of flea and tick-borne diseases, such as Lyme disease and Rocky Mountain spotted fever. Dogs and cats are notorious flea and tick collectors who work cheaper than field biologists and deliver the goods directly to the vet. In 2004, Banfield clinics diagnosed over 60,000 pets with fleas or ticks. Glickman wants to see the parasites and their microbial contents sent to labs for testing, to track any diseases they carry.

Since the pets鈥� home addresses are in the database, it would be pretty straightforward to map the positive results and identify local hotspots of tick-borne disease. And if you know where the danger areas are, you can start working on prevention with bug sprays and warning owners to check their pets more closely for parasites.

Perhaps the most useful and dramatic contribution the surveillance programme could make right now would be to provide an early warning of an outbreak of bird flu. Last year, veterinary pathologist Thijs Kuiken of the Erasmus Medical Centre in Rotterdam, the Netherlands, discovered that domestic cats can catch the H5N1 virus from infected food and can then pass it from cat to cat, possibly through grooming or shared bowls (Science, vol 306, p 241).

So far there is no evidence of cat-to-human transmission, but a cat experimentally infected with a milder strain of human flu developed antibodies and shed the virus for days. 鈥淭his indicates that human-to-cat transmission of influenza virus is possible,鈥� says Kuiken. There is no reason to think the reverse route wouldn鈥檛 be just as viable, in which case it鈥檚 possible that a human outbreak would be heralded by a cluster of sneezing cats. And in some respects that鈥檚 good news. Sick people often delay seeing a doctor, while ailing pets are whisked right off to the vet鈥檚. In terms of surveillance, pets can often be the better bet.

But pets are not the only creatures being enlisted as sentinels. Over at Cornell University in Ithaca, New York, researchers are doing the same with zoo animals.

The Zoo Network was set up in 2001 in response to the appearance two years earlier of a new and frightening killer. In the summer of 1999, McNamara, then at the Bronx Zoo in New York city, noticed that crows on the zoo grounds were dying of an unusual encephalitis, or inflammation of the brain. Around the same time, some of New York鈥檚 human residents also started dying of an unusual encephalitis. They were initially diagnosed with a viral disease called St Louis encephalitis. But McNamara suspected that the crow and human outbreaks were linked: though birds can become infected with St Louis encephalitis, they don鈥檛 die of it. Investigators from the Centers for Disease Control and Prevention eventually confirmed that the human and bird outbreaks had a common cause 鈥� a related virus called West Nile.

Catching bird flu

Surveillance of zoo animals played a major role in the West Nile story. By comparing fresh blood samples from animals in the New York area with stored samples from the same animals 鈥� zoo blood samples are banked for years 鈥� researchers were able to confirm that the virus hadn鈥檛 been seen in that area before. Zoo vets were also the first to publish reports of neurological damage in infected animals, beating their counterparts on the medical side by more than a year.

The WNV outbreak showed how useful zoo animal surveillance could be, since not only are their blood samples stored, but the animals are almost always autopsied at death, providing a wealth of baseline data. So instead of waiting for the next animal to drop dead, why not actively use zoo animals as an early warning system for disease? Most zoos are located in urban areas so the animals will potentially be exposed to many of the same pathogens as people. And zookeepers continually check their charges for signs of illness. In fact, the first indication that felines could catch bird flu came from sick tigers in a Thai zoo last year. The Zoo Network鈥檚 lab at Cornell University is now screening for a variety of pathogens, including bird flu.

So far, about 170 US zoos have sent samples to the network, and not only from zoo residents. 鈥淲e also do passive surveillance, which is whatever runs in the door,鈥� says Dominic Travis, a veterinary epidemiologist at Chicago鈥檚 Lincoln Park Zoo who, along with McNamara and virologist Amy Glaser at Cornell, helped start up the network. 鈥淲e sample them all: birds and squirrels that die on zoo grounds, wildlife brought into the zoo hospital, animals confiscated by the US Fish and Wildlife Service.鈥� Eventually, the hope is to expand the network to include organisations dealing with sick and injured wildlife as well.

鈥淚t鈥檚 possible that a human outbreak of influenza would be heralded by a cluster of sneezing cats鈥�

Ultimately, though, this approach will only succeed if doctors and vets start working together. The Canary Database, launched in August this year by Yale University School of Medicine, is an attempt to make that happen. In general, medical journals focus on human cases, while animal studies are published in veterinary and wildlife journals, and there is little crossover, or cross-readership. So the project鈥檚 director, Peter Rabinowitz, and his team combed through thousands of papers on the occurrence of infectious disease in animals, analysing each for its relevance to human health. The database now contains more than 1000 papers linking animal and human health. As well as monitoring disease outbreaks, Rabinowitz also sees the value of animals for monitoring bioterror events. 鈥淎n agent could get into the animal population and spread,鈥� he says. If the worst should happen, our modern-day miners鈥� canaries might give us a warning.

Meanwhile, Mother Nature is busy cooking up new threats of her own. The Zoo Network recently isolated several strains of a new reovirus from dead crows. Some reoviruses can infect mammals, including humans, causing respiratory diseases and gastroenteritis: many people infected with the SARS virus during the Chinese outbreak in April 2004 also had a reovirus infection. So far, there is no specific threat to us, but just knowing what is out there is likely to pay dividends. 鈥淲hen [the coronavirus causing] SARS emerged, there weren鈥檛 a lot of coronavirus sequences from wildlife to help us figure out where the virus came from,鈥� notes Glaser. 鈥淭hat鈥檚 the power of surveillance 鈥� to see what鈥檚 out there before it bites us.鈥�

The only sure thing is that something will bite us: there will be a next time. But maybe, with a little help from zoo animals, pets and their owners, we will at least get a lead on it.