Skip to main content


Return to Transcripts main page


Cancer-Detecting Dogs

Aired February 7, 2006 - 18:00:00   ET


JONATHAN MANN, CNN HOST (voice-over): Old dogs, amazing new trick. Researchers in California say dogs may be able to smell cancer before humans can detect it.

MICHAEL MCCULLOCH, PINE STREET FOUNDATION: The results were so high, we were just astounded.


MANN: Hello and welcome.

This is something you'll probably find hard to believe. We're wondering about it ourselves. But let's start with something that doctors have known for a long time, people can smell sick. Diabetes can give you an odor like nail polish remover. People with liver disease will give off a whiff of ammonia. And now a team of researchers says that dogs, with their much more powerful sense of smell, can actually detect cancer.

We'll meet the man who led it study in a moment.

On our program today, going to the dogs.

Elizabeth Cohen has this look.


ELIZABETH COHEN CNN CORRESPONDENT (voice-over): We live with them, play with them and rely on them. But a new study is making extraordinary claims that just may change the way you thing of your four-legged friend and that curious wet nose.

That's because dogs could be the newest weapon in the war against cancer. Researchers in California say they trained five dogs to smell the disease on a person's breath with an amazing degree of accuracy. 99 percent of the time with lung cancer. 88 percent of the time with breast cancer. Results that are raising hopes, creating international headlines and making stars out of the dogs involved in the study.

Michael McCulloch was lead researcher.

(on camera): Were you surprised by how accurate the dogs were?

MCCULLOCH: We were very surprised by how accurate they were. The dogs were spot on. They were identifying who had cancer and they were also saying who didn't.

COHEN (voice-over): Dogs diagnosing cancer? Sure, it sounds crazy, but is it that far-fetched? Dogs' sense of smell is legendary, so strong, so reliable that we count on it to sniff out bombs, detect drugs and find the missing and deceased when no human can. So researchers like McCulloch say it's entirely possible that sometimes dogs know our bodies better than we do.

MCCULLOCH: Because the dog may be telling the person something about them that they don't know yet.

COHEN: This is Koby, a walking, wagging tumor detective, and one of the five dogs McCulloch and his team trained to sniff out cancer.

How did they do it? We asked McCulloch and his team to stage a sample test so we could see for ourselves. It starts with five people, four healthy and one with cancer, exhaling into plastic tubes like these.

Inside the tubes, fibers capture microscopic particles from their breath. The tubes are then placed in bowls one yard apart from each other while dog and handler wait outside. The rest is up to Koby.

Time after time after time. Six times out of six attempts, Koby gets it right. Sitting at the cancer sample to mark his discovery.

(on camera): These rates are actually higher than mammograms, higher than Pap smears.

MCCULLOCH: Well, the results were so high, we were just astounded.

COHEN (voice-over): Koby's owner, Maria, was equally impressed.

(on camera): Does it give you a new appreciation for a dog's powers?

MARIA FRIANEZA, TEST DOG'S OWNER: Oh, definitely. Definitely. You always hear that dogs have this amazing sense of smell, but you just never realize how amazing it is until tests like these are done.

COHEN (voice-over): And with 20 to 40 times as many smell receptors in their noses as we have, some researchers believe that a dog's sense of smell may be 10,000 to 100,000 times stronger than ours.

(on camera): Could any household dog be trained to do this?

MCCULLOCH: I believe almost any dog has the hardware, the nose and the brain, to be able to smell things accurately. What really makes the difference is the willingness of the dog to learn and the work together with people.

COHEN (voice-over): No one knows exactly what the dogs are smelling when they stop at a cancer sample, but experts say it probably has something to do with tiny biochemical markers emitted by cancer cells. But some wonder, 88 percent accuracy? 99 percent accuracy? Those numbers are almost unheard of in medicine.

One renown dog trainer said he seriously questions the findings.

LAWRENCE MYERS, DVM, PH.D., AUBURN UNIV.: I'm excited about the findings, but cautiously optimistic at best. A little skeptical at this point.

COHEN: Larry Myers, a veterinarian and professor at Auburn University, has been training detection dogs for 25 years. He says it takes 13 weeks to train dogs to sniff out bombs and doubts any dog could be trained to detect cancer in just three weeks, a claim the study makes.

MYERS: Three weeks is awfully fast. It's not like pushing a button and seeing that it works or doesn't.

Dogs require training. Dogs require maintenance. They're not the panacea. They're just one part of the tool kit in trying to find things.

COHEN: A respected cancer researcher says he too is skeptical. Donald Berry, the head of biostatistics at MD Anderson Cancer Center in Houston has authored more than 200 articles on cancer. He also reviewed McCulloch's study.

DONALD BERRY, MD, MD ANDERSON: It may be true. I would be astounded if it were true. It's not impossible. It's just quite unlikely.

COHEN: But wait, is this just Western establishment medicine looking down their noses at a study done by a small alternative medicine clinic? That's what Nancy Best thinks.

NANCY BEST, CANCER SURVIVOR: I'm sitting here alive today to tell you that if it weren't for Mia, I'd be gone.

COHEN: Mia is Nancy's dog, an untrained yellow lab who she says sniffed and sniffed at Nancy's right breast until she finally paid attention.

BEST: Mia came running in and jumped up on my lap and dove with her nose into my chest, and that's when I found the lump, because it hurt when she pressed her nose there.

COHEN: Sure enough, a lab confirmed Nancy had cancer, stage two carcinoma in the exact spot where Mia had sniffed. Nancy needed surgery and chemotherapy.

(on camera): That must have blown your mind.

BEST: That blew my mind away. When the diagnosis came back positive. And then it really hit me, that this is what she'd been trying to tell me all along, was that I had cancer and I just wasn't listening.

COHEN (voice-over): That was six years ago. Today Nancy is cancer free, she says because of the early detection.

(on camera): Did Mia save your life?

BEST: Yes, she did. I know she did.

COHEN (voice-over): Researchers admit there is a lot more work to be done, but if dogs can actually sniff out cancer before it spreads, it would certainly give new meaning to the term man's best friend.

Elizabeth Cohen, CNN, San Enselmo (ph), California.


MANN: Quite a story. We take a break. When we come back, following your schnauzer's schanz. You'll be amazed where it might lead.

Stay with us.


MANN: A dog's nose doesn't just dominate its face. It dominates its brain too. They have smaller brains than people do, but the portion of it involved with the sense of smell is four-times bigger than ours. Dogs can smell things we can't, they can distinguish between smells better than we can. And they can follow smells to their source.

Welcome back.

Scientists have been trying to find ways to understand and exploit dogs' noses for years. British researchers use them to try to detect bladder cancer in patient's urine. The dogs didn't do all that well, but their success rate was dramatically better than just random chance would indicate.

One of the many remarkable things about the California research is the success rate, in fact higher, as we've heard, than even some trained doctors and manage with proven traditional tests.

Joining us now to talk about all of this is Michael McCulloch. You may recognize him as the lead researcher featured in Elizabeth Cohen's report this evening.

Thanks so much for being with us. Congratulations to you and the dogs. Let me just ask you first of all, whose idea was this? How did you even think to try this?

MCCULLOCH: Well, the first published report of a dog detecting cancer in a human was a 1989 case report in the British medical journal called "Lancet."

We then looked into the matter a little bit further and began a collaboration with animal trainers. We were interested, really, in training the dogs to detect cancer that ordinarily are not very easy to find, lung cancer really being one important example.

MANN: Now, anyone who is watching this is probably going to be stunned, and then they're probably going to be a bit skeptical. So forgive me for asking one or two skeptical questions. Was there anything different about either the people or the samples, the people who really had cancer, the people who didn't have cancer, or the samples that they gave you or how you treated the samples once you got them?

MCCULLOCH: To answer the first point, that in terms of the differences between the cases or cancer subjects and the healthy controls, we set up in our analysis to have them be as close as possible in background factors like whether they smoked or not, their most recently eaten meal, whether they had any other conditions that could cause changes in the odor of the breath. For example, dental infection, sinus infection. So we tried to keep those as balanced as possible.

In the cases where there was a difference, for example in lung cancer, we had slightly more smokers in the cases than in the controls, then you basically include that as a factor in the statistical analysis, to balance out for that.

And even with including the factor of smoking, the results still held out as strong as they did, and that's why we were so astounded.

MANN: Okay, one more skeptical question, and I hope you'll forgive, but with the recent memory, the recent example of stem cell research in South Korea fresh in everyone's mind, is there any possibility that there is fraud involved? That someone is fooling you? You're fooling us? That someone is going to get rich because of this?

MCCULLOCH: Well, since we were able to control all aspects of the study, we don't think so.

Additionally, our vision for the future of this work is not to have a dog in every hospital but rather what the dogs have done in this study is they've really opened to door that says looking at the exhaled breath as a source of diagnostic information has value.

To that end, then, we have begun a collaboration with chemists at the University of Maine, Dr. Taraz Saluki (ph), and animal research specialists at Florida State University, Dr. James Walker, where we're going to work to replicate these findings, because that's really then the next step in the scientific process, is if you can replicate a discovery. And then we're going to look at the chemistry of the breath to see what is it exactly that the dogs are cuing on. That would be an important discovery.

MANN: Do you have any hunch about that? Because that's an obvious question. What is it they're smelling?

MCCULLOCH: Well, we have several hunches. Number one, it may be that, because cancer sells behave differently than normal cells, they may excrete different kinds of biomarker compounds that then make their way into the breath. That's one possibility.

Another possibility is that because cancer can be a destructive process, then as cells are destroyed and that material gets into the blood stream, it also can get into the breath. We really don't know the answer to that. We just have a hunch. And we're hoping that in our upcoming collaboration with Dr. Saluki (ph), for which we've already submitted funding proposals, that we'll be able to discover the answer to that important question.

MANN: One of the talented people I work with who is on the other side of the camera just said before you came on the air, "My dog," referring to her dog, "My dog is too dumb to do this."

Were all those dogs particularly smart? Were they unusual dogs? Or were they just household pets that people are used to hitting with newspapers every once in a while?

MCCULLOCH: Exactly. They were -- some people have said that it was a very short training time, and we think that's true. Several reasons may be, one, we began with just ordinary household puppies, and a puppy, really, in a household environment grows up in a happy environment and they're very eager to interact with humans and to learn from humans. So that's one possibility as to why they were so good. But, yeah, basically we just had ordinary household puppies with only basic puppy training and no prior professionally run detection training.

MANN: Nothing special about the dogs. Let me ask you, is there anything special, do you think, about cancer? Or could dogs smell other diseases?

MCCULLOCH: Well, I think that dogs can smell other diseases. I think that's going to be an important question to pursue. I think other diseases, certainly, like diabetes is a well-known example, other diseases do have evidence of their presence in the breath.

And so again, really, we're looking forward. What the dogs have done is they've opened the door to a whole new way of looking at diagnoses, and that is can you analyze the breath using either chemical means, like a chemical analysis, or biological means, like the dog, to detect disease.

And we don't know yet which one is going to be better.

MANN: I don't mean to be flip, but do you foresee a future where there are going to be dogs in every doctors office, in every hospital?

MCCULLOCH: Well, one of the questions that we're anticipating is in our collaboration with the chemists and the animal specialists, we're going to be -- in our prior study, the dogs set the bar at the 99 percent level. And now as we collaborate with the chemists, we're basically looking to see can technology reach that bar? Can technology reach that challenge that the dogs have set for us?

And so to answer your question there, we're really looking ahead to see in the next phase of our research, where we compare chemistry analysis of the breath and the dogs, what are we going to do with the answer if the dogs were still better. It will basically say that technology still has more work to do.

MANN: One last question for you. What happened to the dogs? Are they back fetching people's slippers?

MCCULLOCH: Yes, they're back at home. They're very happy. And, you know, one astounding study that I can tell you about Koby, whom you met earlier, and that is that Koby, now two-and-a-half years after completing the field work for this project, he still retained the learning, and that's an astounding observation that animal scientists should take note of.

MANN: An amazing piece of research. Michael McCulloch, good luck with your work. And thanks so much for being with us.

MCCULLOCH: Thanks so much for speaking with us, Jon.

MANN: Coming up on INSIGHT, when it comes to smell, it's a dog's world. Some lessons the pups can teach the people when we come back.


MANN: Wasps also have an extraordinary sense of smell and they can even be trained, in a way, more quickly than dogs. Researchers at the University of Georgia have found they can be taught to associate a particular smell with food. Then when they find the target smell again, they behave as if it's mealtime.

Welcome back.

We're all affected by smells and we, people, are clearly missing most of the best ones. Animals are much more sensitive to odors than we are, but researchers are finding out about where our senses of smell can lead us.

Joining us now to talk about some of the research and results is Pam Dalton, an experimental psychologist at Monell Chemical Senses Center in Philadelphia.

Thanks so much for being with us.

Someone watching all of this and finding out that maybe dogs might be able to diagnosis cancer is wondering why doctors can't use their noses better. Could people be doing more of the kind of work that we've seen in wasps and dogs?

PAM DALTON, MONELL CHEMICAL SENSES CTR.: Well, I think it's important to realize that historically there have been physicians who have used their sense of smell to diagnosis patients' maladies prior to the development of many of our standard laboratory tests.

Now, obviously, this isn't something that everyone has done or can do, but with training and certain chemical markers, like that which is present in diabetes or lung cancer or certain infections, interested individuals probably can train themselves to pick up on some of these smells. It is the case, however, that dogs, having more receptors for smell than people do, are probably able to do it at much lower concentrations, and so they might be able to diagnose a disease at an earlier stage.

MANN: Well, let me ask you about people then. How much more information is out there that is available to us if we were just a little bit more alert to it? How much information are we in fact getting that we don't know we're getting because of our sense of smell operating at a level we're not conscious of?

DALTON: Well, every time we're breathing, we're bringing in information about smell, and it's really interesting. A lot of it goes past us without our actually consciously registering it until and unless we're surprised by it or we smell something that we don't expect in a certain area.

And that's when all of the sudden our senses become alert, aroused, and we orient ourselves and say what is that.

So I think there is a great potential for people to use their sense of smell in far more ways than they presently do, and all it requires is the willingness to do it and a little bit of training.

MANN: Any examples come to mind? Any research into any possibilities that comes to mind?

DALTON: Well, for example, we recently published some work that suggests that women of reproductive age are able to actually develop heightened sensitivity to odors that they are repeatedly exposed to, that their sensitivity increased in some cases 1,000 fold.

Now, this might have obvious evolutionary advantages if you think about needing to find food or avoid danger during this period when women are caring for their offspring, or any female mammal, for example.

But it just shows that with the right equipment and the right set of circumstances in the environment, people can in fact increase their sensitivity to odors that they're paying attention to.

MANN: Well, on a related subject and in a serious vein, one reads every once in a while about the interaction between men and women being colored by smell. Is there anything to that?

DALTON: Well, I think there is still a great deal about the biochemical and odor signals that people might be giving off that some people may simply be more attuned to.

I mean, studies that look at these kinds of chemical communication between people often find that some women or even some men are better able to read emotional states that are given off by certain body odors and that sort of thing. So I think the ability to do this might be present in everyone, but it simply requires both a willingness to actually pay attention to these odors and learn about them.

MANN: How about people's emotional states? Everyone has had the familiar sensation of coming home after a long absence or going to a childhood home and smelling something again and having a rush of really indescribable emotion tied to that smell. How well is that process understood? Or is that just simply memory operating at a very basic level?

DALTON: Well, it's thought that olfaction, or sense of smell, may have some privilege communication with our emotional system. Maybe because of proximity in the brain, that the part of the brain that processes odor information actually is routed through the limbic system, where emotions and emotional memories are processed. So it may simply be something that has evolved over time based on this physical structure of the brain.

But it is common. Everyone has had such an experience, whether it's the remembrance of a childhood classroom or a former boyfriend or, you know, your aunt's living room, for example. And the emotion that is generated and the feelings that come with it are often profound and palpable. And people feel as though they've been transported back to that situation again. And it's something that everyone may not have multiple examples, but they certainly have had that experience at least once. Very powerful.

MANN: Let me ask you just one last question to bring us full circle. Obviously this isn't your primary field of study, but do you think it's possible dogs really can detect cancer?

DALTON: Oh, I think the surprising finding about this is really their accuracy rate. I don't doubt for a minute that there are biochemical markers that produce odor signals that could be picked up by dogs, maybe even by people, and hopefully some day by technology such as electronic noses, which are instruments that are based on the structure of the olfactory system but might be more amenable to use in a hospital or a clinic setting. And in fact research is going on at the present time in developing electronic nose technology to diagnosis disease based on exhale breath or even skin samples.

MANN: And you wouldn't have to take it for a walk. Pam Dalton of the Monell Chemical Senses Center in Philadelphia, thanks so much for talking with us.

DALTON: Thank you.

MANN: Just a final word before we go, we've been telling you about a dog's impressive sense of smell, but consider the rat. The rat can smell in stereo. A study published in the "Journal of Science" said that rat's two nostrils send independent and slightly different signals to the rodent's brain, so the rat can immediately identify the direction, say, of food it may want to find or the predator it would like to flee, a judgment that can be made in as little as 1/20 of a second. And, yes, rats are being tested for the ability to smell disease. One species is already being used in Africa to detect tuberculosis, an amazing story as well for another day.

That's INSIGHT for today. I'm Jonathan Mann. More news after this.



© 2007 Cable News Network.
A Time Warner Company. All Rights Reserved.
Terms under which this service is provided to you.
Read our privacy guidelines. Contact us. Site Map.
Offsite Icon External sites open in new window; not endorsed by
Pipeline Icon Pay service with live and archived video. Learn more
Radio News Icon Download audio news  |  RSS Feed Add RSS headlines