Can Dogs Sniff Out Cancer?

There's no question dogs are an integral part of the fabric of America. There are more than 65 million dogs – and two in every five American households have dogs as pets. They and their owners pride themselves on their special gifts: brilliant, goofy, gravity-defying dogs.

As different as they are, they share one common trait. They know their owners. In fact, chances are, if you own a dog, your pet is paying closer attention to you, your state of mind, and even your health than you pay to yourself.

"When I come in, in the evening, my dog always comes over, and just gives a sniff. Why is she doing that? To make sure it's me? Or to find out where I've been," Correspondent Morley Safer asked Dr. Donald Broom, a researcher at Cambridge University Veterinary School who specializes in those noses.

"It would be of interest to see what you've been doing. There would be a lot of information on what you've been doing. Your dog knows a lot about you," says Broom, laughing.

"Their world is a very complex, olfactory world. They are surrounded by a vast array of different things that we are really hardly aware of. It's full of colors. It's full of brightnesses, which are smells."

Broom and fellow researcher Barbara Somerville have set out to prove that the value of that sense of smell may be far greater than anyone realizes. They believe that dogs, with their extraordinary noses, can smell cancer. And they're not alone.

In September, a leading medical journal in Britain, the BMI, gave its blessing. The journal published the results of the first ever meticulously controlled, double blind, peer-reviewed study on the subject, stating, "The results are unambiguous. Dogs can be trained to recognize and flag bladder cancer."

Bee is one of the dogs used in the study. She's a working cocker spaniel trained to smell the odor of the chemical that's in cancer, in this case, bladder cancer. But can she really?

60 Minutes asked dog trainer Andy Cook, who assisted in the BMI study, to run a test for us. There were six urine samples belonging to patients who are either healthy or suffer from some other disease, and one sample belonging to a patient who actually has bladder cancer. The test was conducted at a hearing-aid dog center near Amersham, England, where the original study took place.

Bee's challenge was to find the cancerous sample.

"It's very much free expression. We let her do what she wants. We don't push her into an indication," says Cook. "We don't try and direct her. We don't try to push her into a decision. She's thinking about them."

Bee's been trained to lie down next to the cancerous sample. And she found it. We asked her to do it again, just to make sure. And once again, she nailed it.

In the actual scientific study, six dogs, including Bee, had to distinguish the cancerous sample from the six non-cancerous samples. Dr. Carolyn Willis, a research dermatologist who assisted in the study, says that neither the researchers nor the dogs had any way of knowing in advance which sample was cancerous.

"All the way along, it was blinded so that I would code the samples. And then they would be taken to a completely different building, and those coded samples would be put in a certain position along the line-up," says Willis. "Nobody at any one time knew which was the bladder cancer sample."

Not until after the dogs made their choices. One dog failed completely, but two picked out the cancerous sample 60 percent of the time. The overall average was 41 percent success. That percentage may seem small, but Willis says it amounts to a major success for the dogs.

"The 41 percent, as far as I'm concerned, was a remarkable result," says Willis. "And it was highly statistically significant."

It was significant because it meant that the dogs actually smelled the cancer, and were not merely guessing. And there was an even more startling success story, when one of the non-cancerous control samples caught the interest of the dogs.

As demonstrated in 60 Minutes' recreation of the study, the dogs kept identifying a sample that medical staff had assured the trainers was cancer free. The trainers were dismayed by the dogs' performance and thought the test a total failure.

"The trainers just couldn't train the dogs past this particular sample at all. And they were really getting quite desperate that, in fact, they wouldn't, that this wasn't going to work," says Willis. "Because they consistently went for this sample, we went back and conferred with a specialist."

"The hospital had seen our dogs' work and had got confidence in our dogs, sent it off for further tests," says Cook. "And they were completely blown away when it came back that this patient not only had cancer on his kidney but it was bladder cancer."

Those results impressed the British medical community, and made headlines in England and the United States. But they came as no surprise to Somerville and Broom, who are working now on their own study.

"We've got 16 cases of cancer picked up by a pet dog. And in every case, the dog has shown signs of being anxious and upset," says Somerville. "Now what's going on in the dog's mind, I don't know. But there is some change, which it clearly thinks is threatening its owner."

Somerville says that, in at least one case, a dog detected cancer that had been missed by a doctor. "One of the three breast cancers, which we've had picked up by dogs, turned out to be a very, very small focus of malignancy, undetectable unless screened. And this was removed, and the dog immediately lost interest," says Somerville. "But three months later, it began sniffing, snuffling and becoming agitated again when sitting on her lap. So, she shot back to the hospital, and lo and behold, they had missed a tiny bit of cancer."

Just because your dog is sniffing you, it doesn't necessarily mean she's detecting cancer. Safer's elderly dog, Dora, is constantly checking out him, as well as perfect strangers. It's probably just old-fashioned curiosity, Safer says, but for a handful of pet owners in this country and England, there is no question that their animals have a special gift.

Gill Lacey is one of them. Trudi, his Dalmatian, smelled trouble 25 years ago.

"One particular day, I noticed when she walked past me, she came towards me, sniffing at my leg. And I thought I'd just spilled something," recalls Lacey. "But when I looked, she was sniffing at a tiny mole on my leg."

That mole turned out to be a malignant melanoma, a deadly form of cancer, if not discovered early. Doctors removed the mole and a mass of tissue around it, and when Lacey left the hospital a month later, Trudi confirmed that the cancer was gone.

"Although they'd already said to me that it was clear, I felt reassured that it really was," says Lacey, who believes Trudi saved his life. "I'm convinced of that."

There are an increasing number of similar stories. Time and again, dog owners in England and the United States report much the same behavior. But until very recently, the medical establishments in both countries mostly ignored or dismissed such anecdotes.

Before her involvement in the study, Willis, a research dermatologist, was one of the skeptics. "Anecdotal reports on their own really don't prove very much at all," she says. "They are quite useful indicators of perhaps something to look at. But on their own, they don't provide any sort of proof of any particular phenomenon."

But she became convinced, in large part, because of a simple medical fact -- diseases do give off odors, and dogs, at least theoretically, can smell them.

"Back in the sixth century, I think Hippocrates was describing fruity smells associated with people with diabetes. And musty smells associated with liver disease," says Willis.

If dogs can recognize such odors, the implications for medicine could be enormous. Those noses might provide early detection that science cannot yet achieve. For a disease like prostate cancer, for instance, current detection through blood tests can be notoriously inaccurate.

No one has been more obsessed with the possibilities than Dr. John Church, the driving force behind the British medical journal study. A retired orthopedic surgeon, he believed for a decade that dogs could detect cancer.

He says that there was an element of skepticism, but the patients who were recruited were all "tickled to death because this was something brand new."

But now, in the wake of his study, he feels vindicated. "This is a first step in the right direction. I would say it is a great breakthrough in the sense that this is the first such presentation of a rigorous study of this type," says Church. "We regard this as a great breakthrough."

And there are more studies on the way. In California, there's a test of dogs' ability to detect lung cancer, and back in England, with the help of a cancer researcher at Cambridge, Broom and Somerville are finishing up their own study on prostate cancer.

Final results are expected this year, but early tests show very high success rates. Meanwhile, the Amersham team is planning to move ahead on further research -- with their handpicked roster of specialists, of course. The team includes Biddie and Tangle, Oak and Dill, Bee, and a couple of pre-med rookies, Briar and Daisy.

"I personally see a day when you could use dogs to detect disease," says Church. "You've got a marvelous asset. You've got a wonderful tool."