In early tests, the approach produced fewer false positives than would be expected with the commonly used PSA test, French researchers report.
The concept isn’t new. Other researchers have reported varying degrees of success using dogs to detect cancers of the skin, lung, and bladder, says researcher Pierre Bigot, MD, of Tenon Hospital in Paris.
The theory is that many tumors release chemicals with distinct odors that can be picked up by dogs, whose sense of smell is much more sensitive than that of humans, he tells WebMD.
More accurate prostate cancer tests are sorely needed, says Anthony Y. Smith, MD, chief of urology at the University of New Mexico in Albuquerque.
While the widely used PSA test picks up a lot of cancers, it also has a high false-positive rate, he tells WebMD.
“If all the men with high PSA scores go on to have biopsies, fewer than one-third will actually have cancer,” Smith says.
Plus, many men with early prostate cancer are unnecessarily treated because existing tests can’t distinguish between life-threatening and slow-growing tumors, he says.
In the United States, one man in six will receive a diagnosis of prostate cancer during his lifetime, but a much smaller proportion — one in 35 — will die from the disease, according to the American Cancer Society.
Smith moderated a news briefing on the findings at the American Urological Association annual meeting.
For the new study, researchers led by Jean-Nicolas Cornu, MD, also of Tenon Hospital, trained a Belgian Malinois — a shepherd breed used for detecting bombs and drugs — to identify urine from patients with confirmed prostate cancer and then to discriminate those samples from urine from healthy men. After about a year of training, the dog was put to the test. During 11 runs, the dog faced six urine samples, only one of which came from a man with prostate cancer. Its mission: To sit in front of the urine it considers cancer.
In 66 tests, the dog was correct 63 times. There were three false positives, in which the dog mistakenly identified samples from healthy men as being cancerous. And there were no false negatives.
And one of the three false positives might not have been that false; when the man who provided the urine sample had another biopsy, he turned out to have prostate cancer, Bigot says.
Other dogs are now being trained, he says.
The low false-positive rate “is pretty spectacular,” Smith says.
“But this is a very small study,” and it remains to be seen if the findings will hold up in other studies, he says.
Skeptics are concerned that the animals may be picking up on subconscious signals from researchers, among other things, Smith says.
The next step is to figure out what chemicals or combination of chemicals the dog is sensing, he says.
If the approach does pan out, don’t look for dogs running around hospitals, sniffing urine samples. That would be impractical and prohibitively expensive, Bigot says.
But if researchers can identify which chemical the dog is reacting to, they hope to develop an “electronic nose” for more accurate prostate cancer detection, he says.
This study was presented at a medical conference. The findings should be considered preliminary as they have not yet undergone the “peer review” process, in which outside experts scrutinize the data prior to publication in a medical journal.