Dogs Sniff Out Bladder Cancer

They detect telltale chemicals in urine samples, researchers contend

THURSDAY, Sept. 23, 2004 (HealthDayNews) -- British researchers say they've taught dogs to sniff out the scent of bladder cancer in urine, a feat that might someday lead to better diagnostic technology.

The work was inspired by several reports that have appeared in medical journals and the popular press about dogs picking up distinctive smells associated with cancer, said Caroline M. Willis, director of research at the Amersham Hospital department of dermatology. One item that especially aroused interest was a report of a woman who sought medical help after her dog became unusually interested in a skin lesion that proved to be malignant melanoma.

The Amersham study began in a somewhat unscientific way, enlisting six dogs that were pets of hospital staffers. The dogs included three spaniels, who turned out to be the best sniffers; they are widely used to detect illicit drugs, Willis said.

"Our original intention was to train the dogs to detect skin cancer, but that was quite difficult to do," she said. "So we decided to test for the abnormal chemicals released by bladder cancer into urine, which is easy to collect.

A report on the research appears in the Sept. 25 issue of the British Medical Journal.

What developed into a seven-month training program began with the dogs sniffing urine samples from bladder cancer patients, Willis said. "Once they got the idea, they were given simple discrimination tasks, detecting urine from bladder cancer patients as compared to normal urine," she said.

The dogs were taught to lie down next to a urine sample identified as being from a cancer patient. Each time a dog made a correct selection, it got a bit of food as a reward.

"Then we went much further, training the dogs to recognize the specific smell of cancer, not just one associated with secondary effects such as inflammation," Willis said. "Then we ran this series of tests designed to rigorously prove that they were detecting a specific kind of cancer."

The result was an overall success rate of 41 percent, with 22 successes in 54 tries. It is "a first, very tentative step into the whole area" of cancer detection by smell, Willis said.

The research effort now has turned to skin cancer, Willis said. "But we haven't quite finalized how we would do that. We can't present the dogs with fresh tissue, so we have to find ways to capture the smell," she said.

Willis does not foresee a role for dogs in cancer diagnosis. "Where dogs can be helpful is to help identify specific chemicals that are characteristic of bladder cancer," she said. "We may be able to use that information to develop instrumentation to identify cancers."

Tim J. Cole, a professor of medical statistics at the Institute of Child Health in London, wrote an accompanying editorial in the journal that was cautiously praiseful. The study, Cole wrote, "gives the lie to Dorothy Parker's epigram," which said, "You can't teach an old dogma new tricks."

The most intriguing finding of the study, Cole said, was that of a supposedly cancer-free patient whose urine sample was consistently identified as cancerous by the dogs. Even though other tests were negative, "the consultant was sufficiently impressed by the dogs' performance to test the patient again and found a kidney carcinoma," he wrote.

As for the potential usefulness of dogs in cancer diagnosis, Cole said, "I deliberately didn't say anything about that in my commentary." There is a chance the work could lead to new diagnostic tools, he said.

"It is a fascinating speculation," he added. "But I wouldn't want to be dogmatic."

More information

The conventional approach to bladder cancer diagnosis and treatment is described by the National Cancer Institute.

Related Stories

No stories found.
logo
www.healthday.com