PHILADELPHIA — McBaine, a bouncy black and white springer spaniel, perks up and begins his hunt at the Penn Vet Working Dog Center. His nose skims 12 tiny arms that protrude from the edges of a table-size wheel, each holding samples of blood plasma, only one of which is spiked with a drop of cancerous tissue.
The dog makes one focused revolution around the wheel before halting, steely-eyed and confident, in front of sample No. 11. A trainer tosses him his reward, a tennis ball, which he giddily chases around the room, sliding across the floor and bumping into walls like a clumsy puppy.
McBaine is one of four highly trained cancer detection dogs at the center, which trains purebreds to put their superior sense of smell to work in search of the early signs of ovarian cancer. Now, Penn Vet, part of the University of Pennsylvania’s School of Veterinary Medicine, is teaming with chemists and physicists to isolate cancer chemicals that only dogs can smell. They hope this will lead to the manufacture of nanotechnology sensors that are capable of detecting bits of cancerous tissue 1/100,000th the thickness of a sheet of paper.
“We don’t ever anticipate our dogs walking through a clinic,” said the veterinarian Dr. Cindy Otto, the founder and executive director of the Working Dog Center. “But we do hope that they will help refine chemical and nanosensing techniques for cancer detection.”
Since 2004, research has begun to accumulate suggesting that dogs may be able to smell the subtle chemical differences between healthy and cancerous tissue, including bladder cancer, melanomaand cancers of the lung, breast and prostate. But scientists debate whether the research will result in useful medical applications.
Dogs have already been trained to respond to diabetic emergencies, or alert passers-by if an owner is about to have a seizure. And on the cancer front, nonprofit organizations like the In Situ Foundation, based in California, and the Medical Detection Dogs charity in Britain are among a growing number of independent groups sponsoring research into the area.
A study presented at the American Urological Association’s annual meeting in May reported that two German shepherds trained at the Italian Ministry of Defense’s Military Veterinary Center in Grosseto were able to detect prostate cancer in urine with about 98 percent accuracy, far better than the prostate-specific antigen (PSA) test. But in another recent study of prostate-cancer-sniffing dogs, British researchers reported that promising initial results did not hold up in rigorous double-blind follow-up trials.
Dr. Otto first conceived of a center to train and study working dogs when, as a member of the Federal Emergency Management Agency’s Urban Search and Rescue Team, she was deployed to ground zero in the hours after the Sept. 11 attacks.
“I remember walking past three firemen sitting on an I-beam, stone-faced, dejected,” she says. “But when a handler walked by with one of the rescue dogs, they lit up. There was hope.”
Today, the Working Dog Center trains dogs for police work, search and rescue and bomb detection. Their newest canine curriculum, started last summer after the center received a grant from the Kaleidoscope of Hope Foundation, focuses on sniffing out a different kind of threat: ovarian cancer.
“Ovarian cancer is a silent killer,” Dr. Otto said. “But if we can help detect it early, that would save lives like nothing else.”
Dr. Otto’s dogs are descended from illustrious lines of hunting hounds and police dogs, with noses and instincts that have been refined by generations of selective breeding. Labradors and German shepherds dominate the center, but the occasional golden retriever or springer spaniel — like McBaine — manages to make the cut.
The dogs, raised in the homes of volunteer foster families, start with basic obedience classes when they are eight weeks old. They then begin their training in earnest, with the goal of teaching them that sniffing everything — from ticking bombs to malignant tumors — is rewarding.
“Everything we do is about positive reinforcement,” Dr. Otto said. “Sniff the right odor, earn a toy or treat. It’s all one big game.”
Trainers from the center typically notice early on that certain dogs have natural talents that make them better suited for specific kinds of work. Search and rescue dogs must be tireless hunters, unperturbed by distracting environments and unwilling to give up on a scent – the equivalent of high-energy athletes. The best cancer-detection dogs, on the other hand, tend to be precise, methodical, quiet and even a bit aloof — more the introverted scientists.
“Some dogs declare early, but our late bloomers frequently switch majors,” Dr. Otto said.
Handlers begin training dogs selected for cancer detection by holding two vials of fluid in front of each dog, one cancerous and one benign. The dogs initially sniff both but are rewarded only when they sniff the one containing cancer tissue. In time, the dogs learn to recognize a unique “cancer smell” before moving on to more complex tests.
What exactly are the dogs sensing? George Preti, a chemist at the Monell Chemical Senses Center in Philadelphia, has spent much of his career trying to isolate the volatile chemicals behind cancer’s unique odor. “We have known for a long time that dogs are very sensitive detectors,” Dr. Preti says. “When the opportunity arose to collaborate with Dr. Otto at the Working Dog Center, I jumped on it.”
Dr. Preti is working to isolate unique chemical biomarkers responsible for ovarian cancer’s subtle smell using high-tech spectrometers and chromatographs. Once he identifies a promising compound, he tests whether the dogs respond to that chemical in the same way that they respond to actual ovarian cancer tissue.
“I’m not embarrassed to say that a dog is better than my instruments,” Dr. Preti says.
The next step will be to build a mechanical, hand-held sensor that can detect that cancer chemical in the clinic. That’s where Charlie Johnson a professor at Penn who specializes in experimental nanophysics, the study of molecular interactions between microscopic materials, comes in.
He is developing what he calls Cyborg sensors, which include biological and mechanical components – a combination of carbon nanotubes and single-stranded DNA that preferentially bond with one specific chemical compound. These precise sensors, in theory, could be programmed to bind to, and detect, the isolated compounds that Dr. Otto’s dogs are singling out.
“We are effectively building an electronic nose,” said Dr. Johnson, who added that a prototype for his ovarian cancer sensor will probably be ready in the next five years.
Some experts remain skeptical.
“While I applaud any effort to detect ovarian cancer, I’m uncertain that this research will have any value,” said Dr. David Fishman, a gynecologic oncologist at Mount Sinai Hospital in New York City. One challenge, he notes, is that any cancer sensor would need to be able to detect volatile chemicals that are specific to one particular type of cancer.
“Nonspecificity is where a lot of these sort of tests fail,” Dr. Fishman said. “If there is an overlap in volatile chemicals — between colon, breast, pancreatic, ovarian cancer — we’ll have to ask, ‘What does this mean?’ ”
And even if sensors could be developed that detect ovarian cancer in the clinic, Dr. Fishman says, he doubts that they would be able to catch ovarian cancer in its earliest, potentially more treatable, stages.
“The lesions that we are discussing are only millimeters in size, and almost imperceptible on imaging studies,” Dr. Fishman says. “I don’t believe that the resolution of the canine ability will translate into value for these lesions.”
McBaine remains unaware of the debate. After correctly identifying yet another cancerous plasma sample, he pranced around the Working Dog Center with regal flair, showing off his tennis ball to anyone who would pay attention. In an industry saturated with hundreds of corporations and thousands of scientists all hunting for the earliest clues to cancer, working dogs are just another set of (slightly furrier) researchers.