Canine Scent Detection Published Research

Canine Scent Detection: Breast and Lung Cancer

In an international collaboration, the Pine Street Foundation conceived of and obtained grant support for a rigorously designed diagnostic study aimed at accurate early detection of lung and breast cancers.

Diagnostic Accuracy of Canine Scent Detection of Lung and Breast Cancers in Exhaled Breath

The following research was published in the March 2006 issue of Integrative Cancer Therapies, a peer-reviewed journal.

Download a full, free copy of this study by clicking here: PDF [104kb]

Precisión del diagnostico de Detección Canina a través del olfato en las fases tempranas y terminales del cáncer de mama y de pulmón: PDF [258kb]

Lung cancer, when symptoms arise, is usually diagnosed at Stage III or IV, when the prognosis is rarely good. Treatment is significantly more effective at early stages (I or II) when the tumor is smaller and has not yet spread. Detecting lung cancer in its early stages is difficult, so therefore developing a feasible and effective early detection method is the subject of a considerable amount of research around the world. In terms of breast cancer, while detection at early stages is comparatively easier, finding methods to detect it even earlier are worthwhile.

In an international collaboration, the Pine Street Foundation conceived of and obtained grant support for a rigorously designed diagnostic study aimed at accurate early detection of lung and breast cancers. We invited scientist Professor Tadeusz Jezierski, ScD, of the Polish Academy of Sciences, to serve as principal investigator for this study.

A dog’s nose, considered by both dog trainers and chemists alike to be one of the world’s most powerful olfactory sensor, was the “medical device” used in this research. In a study of 86 people (55 with lung cancer and 31 with breast cancer), five professionally trained scent dogs accurately distinguished between breath samples from diseased patients and those from 83 healthy controls. The dogs’ ability to correctly identify or rule-out lung and breast cancer, at both early and late stages, was around 90%. A more exact and specific discussion of the sensitivity and specificity of the dogs’ abilities is available in the full study.

This work is based on the hypothesis that cancer cells emit different metabolic waste products than normal cells. The differences between these metabolic products are apparently so great that they can be detected by a dog’s keen sense of smell, even in the early stages of disease.

In carefully controlled conditions, the dogs were presented with breath samples of both cancer patients and healthy controls. The dogs were trained by professional dog behaviorist Kirk Turner and the methods used in the field research were developed by Professor Jezierski. Not only did the dogs perform exceptionally well, they did so consistently over a lengthy four month investigation of 12,295 separate scent trials – each one documented on videotape.

What is important about this study is that (1) ordinary dogs, with no prior scent discrimination training, could be rapidly trained to identify lung and breast cancer patients by smelling samples of their breath, when compared to blank unused sample tubes; (2) dogs could accurately and reliably distinguish breath samples of lung and breast cancer patients from those of healthy controls; and (3) the dog’s diagnostic performance was not affected by disease stage of cancer patients, age, smoking, or most recently eaten meal among either cancer patients or controls.

We are actively seeking funding to launch new studies that will build upon this research. This pilot work using canine scent detection demonstrates the validity of using a biological system to examine exhaled breath in the diagnostic identification of lung and breast cancers. Future work will closely examine the chemistry of exhaled breath to identify which chemical compounds can most accurately identify the presence of cancer. A prospective cohort study, comparing both biological (canine scent detection) and chemical methods to conventional diagnosis within a given set of individuals, should help answer the question of whether these approaches can lead to changes in how cancer in diagnosed in clinical practice.

Another goal is to study canine scent detection of ovarian cancer, a disease that is currently very difficult to detect at its earliest stages.

All subjects provided written informed consent and our protocol and patient recruitment materials were approved by an independent Institutional Review Board (Independent Review Consulting, Corte Madera, Calif.). Additionally, our animal training and handling methods were designed in consultation with two veterinarians, an independent dog trainer, and the dog owners themselves, all of whom approved our methods. We used food reward-based positive reinforcement training, also known as clicker training; no compulsion training methods were ever used. None of the animals were ever harmed, either physically or psychologically, in the course of this research; the health, safety, and happiness of all the animals involved was always our first priority. We did not provide any compensation to subjects for providing breath samples or to dog owners for volunteering use of their dogs in the study.

This research was made possible through the generous support of the MACH Foundation (Fairfax, Calif.) and Frank and Carol Rosenmayr (Kentfield, Calif.). Their support allowed us to invite Dr. Jezierski to work with us for four months and to provide our team of trainers and volunteers with methodical guidance and close supervision. Study sponsors had no role in study design, data collection, data analysis, data interpretation, manuscript preparation, or the decision to publish. We would also like to acknowledge Guide Dogs for the Blind (San Rafael, Calif.) for their help in making dogs available for this study.

There are many ways you can help us to further research in this emerging field:
» Funding. Support from individuals is incredibly important. If you are interested in continuing this research, please make a donation. 92% of our funding comes from individuals like you. Additionally, should you know of any foundations or individuals who might also be interested in supporting us, please pass the word along.
» Replicate Our Study. For researchers, we actively encourage you to learn more about our current study and to build and improve upon it through new research.
» Get the Word Out. The more people who know about this work, the better. Please share us with your friends and colleagues.

Was this research published in a peer-reviewed journal?
Yes. Our research was rigorously peer-reviewed prior to publication in Integrative Cancer Therapies, which is a peer-reviewed journal indexed in MEDLINE.

Are the dogs smelling cancer or just people?
Dogs and patients did not meet directly. We gathered breath samples in specially designed tubes and then, at a separate location, presented those tubes to the dogs for training and testing. Additionally, during the entire double-blinded testing phase, upon which our data is based, all breath samples sniffed by dogs, for both cases and controls, were from completely different subjects not previously encountered by the dogs during training or single-blinded testing.

What about the smells of smoking?
Smokers were present in both lung cancer patients and the control groups. The strong results we found remained even after accounting for confounding by smoking in our analysis.

What about the effect of chemotherapy on body odor?
Breath samples were taken before starting conventional treatment.

Weren’t the cases and controls very different from each other?
Yes. This is correct and a valid criticism of our study. In the long course of developing a new diagnostic method, the first step is to see if the method can distinguish known cases from controls. This preliminary step is then followed by testing in which the new method is compared head-to-head with existing diagnostic methods and then examining people whose true health status is not known at the time of the study.

Do you really expect dogs to be used in hospitals to detect cancer?
Our study provides compelling evidence that cancers hidden deep within the body can be detected simply by examining the odors of a person’s breath. The fact that it was dogs who did this does not detract from the novelty of our findings. The dog’s brain and nose is currently one of the most sophisticated odor detection devices on the planet…technology now has to rise to meet that challenge and it remains to be seen whether chemical analysis can meet the level of the dogs.

What are the next steps?
The direction our research is taking, and should take, is toward the development of an “electronic nose”. This sort of device would be much more likely to be incorporated into clinical practice. In the end, diagnostic breath analysis deserves further rigorous study, both through additional work with dogs and through chemical analysis towards the goal of understanding precisely what compounds it is that the dogs are detecting by scent. These are the questions our next generation trial will address, and for which we are currently seeking funding.

What sorts of dogs did you use in this study? What were your selection criteria?
Five dogs, ages 7-18 months, were chosen out of a total of 13: three Labrador retrievers (two males and one female) and two Portuguese water dogs (one male and one female). Dogs were provided by local dog owners and by Guide Dogs for the Blind in San Rafael, Calif. Our selection criteria called for dogs over six months old with basic obedience training typically given to household pets, as defined by the American Kennel Club, who were judged by the experimenters to be eager to sniff objects and respond to commands.

What were the names of the dogs?
Kobi: Labrador retriever (yellow)
Isabelle: Portuguese water dog
Court: Labrador retriever (black)
Estelle: Labrador retriever (yellow)
Django: Portuguese water dog

How can I get my dog involved in your research?
Should we receive funding to continue research in this field, we will likely be recruiting additional dogs from the San Francisco Bay Area. An announcement will be made through our newsletter, so be sure to sign up to receive email updates. Additionally, the Pine Street Foundation is available to discuss this research with individuals or groups who may be interested in pursuing a collaboration leading toward grant-funded clinical research.

Can I have your dogs screen me to see if I have cancer?
Unfortunately, it is not currently possible to be “screened” by the dogs involved in our research. At this time, it is also not recommended that dogs be used as a primary screening method for cancer. It is our hope, however, that the concepts explored in our research may help develop more accurate cancer screening methods in the future. Nevertheless, numerous anecdotal reports have been published and televised documenting individual cases in which dogs began to display persistent and animated behavior around specific body locations on their owners. These behaviors, on subsequent medical evaluation, proved to be accurate, and in some cases life-saving, early warning signs of cancers such as those of the breast and skin (melanoma). Therefore, should your dog display such behavior, we do recommend medical follow-up.

How can I train my dog to detect cancer?
The methods used in our study are similar to how you might train your dog to sit or roll over. We specifically used “clicker training”, a reward-based method of teaching in which the dog is given food treats for offering the desired behavior. What is more difficult is obtaining breath samples from cancer patients; there are various state and federal requirements regarding medical research that must be followed. There are also important ethical and scientific issues that must be addressed before engaging in this type of training. For example, what would a person do with the dog after training is complete? Would the dog be used to screen people for cancer? How would the dogs’ accuracy and consistency be determined? What would a person who is screened then do with the information gleaned from the dogs? Could or should the results be trusted? Worse than a false positive (where a dog indicates there is cancer present when no disease actually exists) would be a false negative (where the dog indicates no cancer is present when in fact there is disease). Because of the various logistical and ethical considerations associated with this type of research, it is recommended that individuals engaging in this type of training follow appropriate scientific and ethical guidelines such as those documented in our study. Additionally, the Pine Street Foundation is available to discuss this research with individuals or groups who may be interested in pursuing a collaboration leading toward grant-funded clinical research.

Have comments or additional questions? Please contact us.

One reply on “Canine Scent Detection: Breast and Lung Cancer”

Outstanding work! I am not surprised that man’s best friend could come to his rescue yet again.

I am interested in receiving your email updates regarding this information and future studies. Please put me on your list.

I am also interested in promoting your research to save human lives while at the same time helping to save dog lives. I see it as possible to do both at the same time.

Please call me to discuss this matter.

Brian Bouchard

PS: I know you have stated that you do not publish email addresses. I wanted to be certain that you knew we wish NOT to have any solicitations to this email address by anyone. This address is for private correspondence only. Thank you.

Leave a Reply

Your email address will not be published. Required fields are marked *