Can Dogs Detect Cancer? - We Save Dogs Life

Can Dogs Detect Cancer?


THERE HAVE BEEN many anecdotal reports about dogs detecting cancer, such as the case of Arlene Goldberg of Chicago. She owes her life directly to her Cocker Spaniel, Duffy, who developed an annoying habit of jumping up on her and sniffing and nipping at a mole on the back of her shoulder near the base of her neck. Duffy was so persistent that Arlene mentioned it to her doctor, who removed the mole and had a biopsy done. A few days later Arlene was back in the hospital to have the surrounding skin removed because the mole turned out to be a virulent form of melanoma that could have killed her if it had metastasized. The original set of medical reports analyzing case studies like Arlene’s first appeared in the respected medical journal the Lancet in 1989. Since then, a number of controlled experimental studies have confirmed that dogs can detect cancer as well as, or even better than, traditional medical screening ­techniques.
One of the first studies was conducted by the Florida dermatologist Armand Cognetta, who decided to investigate possible medical uses of dogs in cancer detection. In 1996, Cognetta borrowed a seven-year-old Schnauzer named George (a recently retired bomb-sniffing canine), with the goal of determining whether the dog could consistently sniff out melanoma, in both tissue samples and people. Normally, a handheld microscope is used to diagnose potential skin cancer, and a biopsy usually follows. Because observation with a microscope is only about 80 percent effective in early diagnosis (meaning that one in five diagnoses will be wrong), further tests are generally conducted to confirm the cancer. Using tissue samples of melanomas from research institutes and a hospital, Cognetta trained George to find a tube containing a melanoma sample. George eventually became quite proficient, so that when the melanoma sample was placed in one of ten holes in a large rectangular box, with the other holes containing normal tissues, the dog could find the melanoma 99 percent of the time. Cognetta then allowed George to “examine” actual patients, which is considerably more difficult. Eventually, George discovered melanoma in four (possibly five, depending on how you look at the results) of seven patients. These results were interesting but far from conclusive, so a set of larger and more controlled studies seemed to be called for.
Carolyn Willis of Amersham Hospital in England, along with her associates, conducted a meticulously controlled, double-blind study showing that dogs can be trained to recognize and indicate bladder cancer. They used sets of six urine samples belonging to patients who were either healthy or suffered from another disease, plus a sample from patients with bladder cancer. Neither the researchers doing the testing nor the dogs being tested had any way of knowing in advance which sample was cancerous until after the dogs made their choices.
In one instance the dogs kept identifying a sample that medical staff had asserted was cancer-free. Willis said, “The trainers just couldn’t train the dogs past this particular sample at all. They were really getting quite desperate that this wasn’t going to work. Because the dogs consistently went for that one sample, we went back and conferred with a specialist.”
Andy Cook, one of the trainers, describes what happened next. “The hospital had seen our dogs’ work and had got confidence in our dogs, sent it off for further tests. They were completely blown away when it came back that this patient not only had cancer on his kidney but there was also bladder cancer.”
Some dogs seem to spontaneously recognize cancer, but recent work shows how quickly and effectively dogs can be trained to find this malignancy. Michael McCulloch of the Pine Street Foundation in San Anselmo, California, together with his colleagues, needed only three weeks of training to teach five pet dogs to detect lung or breast cancer by sniffing the breath of participants. The trial itself involved eighty-six cancer patients (fifty-five with lung cancer, and thirty-one with breast cancer) and a group of eighty-three healthy patients. In the study, the dogs sniffed breath samples captured in special tubes. Dogs were trained to sit or lie down directly in front of a test station with the cancerous sample. The results were spectacular, showing that dogs can detect breast and lung cancers with an average of better than 90 percent accuracy.
In a new study, Hideto Sonodo of Kyushu University in Japan collected breath and stool samples from patients with colorectal cancer, also known as bowel cancer. The researchers placed one cancerous sample and four noncancerous samples in storage containers, and in a series of “sniff” tests they commanded a Labrador Retriever to search for the container that held the cancerous sample. The dog sniffed out the cancerous sample in thirty-three out of thirty-six breath tests, and in thirty-seven out of thirty-eight stool tests. That’s almost as accurate as a colonoscopy test for colorectal cancer.

 Furthermore, in some cases the samples had come from patients with early stages of bowel cancer, which are notoriously difficult to detect.
A trained dog has the potential of screening more than twelve thousand people for cancer in its lifetime, making the investment of time to produce cancer-detecting dogs good economics, as well as good medicine. Perhaps sometime in the future, that “lab test” you get for possible cancer may well come in the form of some educated sniffing by a Labrador Retriever.

Aucun commentaire