Cancer-sniffing dogs: How canine scent detection could transform human medicine

September 16, 2019
Ed Kane, PhD

Ed Kane, PhD, is a researcher and consultant in animal nutrition. He is an author and editor on nutrition, physiology and veterinary medicine with a background in horses, pets and livestock. Kane is based in Seattle.

dvm360, dvm360 December 2019, Volume 50, Issue 12

With dogs amazing olfactory capabilities, researchers are optimistic that they will one day play a regular in early disease detection.

Special characteristics give dogs an extraordinary level of sensitivity to detect even minute amounts of a particular odorant, an ability estimated to be one million times more efficient than in humans. Human cancers are devastating for many families throughout the world. According to the American Cancer Society, the most common cancers diagnosed in men are prostate, lung and colorectal cancers. Women experience breast, lung and colorectal cancer more than any other type, with breast cancer accounting for 30% of new cases. A total of 1.8 million new cancer cases of all types and nearly 607,000 deaths from cancer are expected to occur in the U.S. in 2019.

As these statistics reveal, there is a compelling need to utilize early screening strategies and novel approaches for early detection. And there is increasing evidence that dogs may have a significant role to play in the early diagnosis of human cancer.

The canine nose is uniquely suited to detecting scents far beyond human capability. Dogs' ability to detect odorant molecules is supported by several anatomic factors, such as the extraordinary dimension of their olfactory epithelium (up to 170 cm2 vs. 10 cm2 in humans), the huge number of olfactory receptors they possess (over 200 million vs. 5 million in humans), the dense innervation of their olfactory mucosa and their ability to “sort” meaningful incoming odors from those that are unwanted or unnecessary.1 These characteristics give dogs an extraordinary level of sensitivity to detect even minute amounts of a particular odorant, an ability estimated to be one million times more efficient than in humans.1

Dogs detecting human cancer

The following is a timeline of studies and case reports from the medical literature in which dogs have detected human cancers:

1989: A 44-year-old woman's border collie-Doberman cross continuously sniffed at her left thigh, which was later biopsied and a malignant melanoma diagnosed.2

2001: A 66-year-old man's Labrador retriever repeatedly sniffed at his leg through his trousers; he was found to have a basal cell carcinoma.3

2004: Two dogs, a 4-year-old standard schnauzer and a 6-year-old golden retriever, were trained to identify melanoma tissue samples hidden on the skin of healthy volunteers. One of the schnauzers positively identified samples at first tested negative, but further histologic examination revealed a small number of cancerous cells.4

2004: Six dogs of varying breeds and ages were trained to detect the urine of patients with bladder cancer. The dogs correctly determined bladder cancer urine in 41% of the cases.5

2006: Researchers used a food-reward system to train five household dogs to identify exhaled breath samples of lung and breast cancer patients, distinguishing them from healthy controls. The sensitivity and specificity were 99% for lung cancer patients and 88-98% for breast cancer patients; results were remarkably similar across all four stages of disease.6

2008: A dog was taught to identify ovarian carcinoma samples consisting of 31 different histopathological types of various grades and stages. In double-blind tests, the dog was capable of correctly identifying all cancer samples with 100% sensitivity and 97.5% specificity, as well as discriminating ovarian carcinomas from other gynecological carcinomas with 100% sensitivity, 91% specificity.7

2010: Two dogs were trained to detect ovarian cancer from normal ovarian tissue and distinguish blood plasma of patients with ovarian carcinomas. Tissue test sensitivity was 100% and specificity 95%; blood plasma sensitivity was 100% and specificity 98%.8

2011: A Belgian Malinois was clicker-trained to scent and identify prostate cancer patients from their urine. The dog correctly identified cancer in 31 of 33 patients, with 91% sensitivity and specificity.9

2011: A Labrador retriever was trained to scent-detect colorectal cancer from breath and watery stool samples. Compared with colonoscopy, dogs showed their ability to detect cancer from breath samples at 91% sensitivity and 99% specificity. With stool samples, sensitivity was 97% and specificity 99%. Accuracy was high even for early cancer.10

2012: Trained dogs successfully detected lung cancer from human breath with sensitivity of 90% and specificity of 72%.11

2015: Two 3-year-old explosion-detection German shepherds were trained to identify human prostate cancer from specific volatile compounds in urine samples from 362 patients with prostate cancer. For the first dog, sensitivity was 100% and specificity 98%; for the second dog, sensitivity was 99% and specificity 98%.12

2017: Researchers investigated the feasibility of whether dogs could use olfactory cues to discriminate urine samples from dogs with diagnosed urinary tract transitional cell carcinoma versus control dogs.13

2017: Researchers investigated the detection of hepatocellular carcinoma from human breath using canine olfaction. Results showed an accuracy rate of 78%.14

Applying dogs' innate ability

“Several things are happening now regarding the research into canine scent detection of human cancer,” says Michael McCulloch, LAc, MPH, PhD, a researcher with the Pine Street Foundation in San Anselmo, California.

First, a multinational consortium that includes researchers in Chile, France and the United Kingdom set out to understand the threshold of canine detection of odorant molecules. To study this, researchers wafted amyl acetate into a training space at precisely measured concentrations. From this process investigators have concluded that dogs' olfactory detection threshold ranges between 40 parts per billion and 1.5 parts per trillion.15

“It has been determined that the scientific possibility of dogs detecting odorant target molecules at very low levels is very strong,” states McCulloch. “[This provides] very strong evidence for the detection capability of the dog's nose with the brain hard-wired into it.”

The first published account of canine cancer detection was that of a woman in the U.K. whose dog spontaneously indicated cancer in the owner.2 “This paper was the first to give us the idea that if a dog can spontaneously offer the behavior, then we can train it to reproduce that behavior,” McCulloch says.

Dogs have successfully identified odorant-potential molecules in exhaled breath, transdermal excretions, fecal samples, urine, blood components and pathology samples.

“In our 2006 study, we were able to demonstrate strikingly high sensitivity and specificity in both lung and breast cancers,” McCulloch says.6 “The difference in sensitivity for lung vs. breast cancer may be due to the breast tissue being deeper within the body, whereas a lung tumor is essentially a surface tumor. In lung cancer patients, the tumor cells are coming into contact with the air a person breathes.”

McCulloch says the next step is for someone to turn scientific results into a commercial service. “The existing sensitivity and specificity data suggest that it's scientifically feasible, and what's remaining now is for entrepreneurs collaborating with scientists to demonstrate the commercial feasibility of the project,” he notes.

McCulloch and his team are about to publish the results of a double-blinded trial of ovarian cancer detection. While he can't reveal the actual results while publication is pending, he does say the outcomes are impressive. He is also undertaking a meta-analysis to determine whether it's possible to identify the factors influencing success one way or the other in published research to date.

He adds that formally trained veterinarians are involved in training dogs to detect cancer by scent across the globe-Costa Rica, Canada, Slovenia, Finland, Norway, the U.K., France, Germany, the Netherlands and Poland. “Though already much has been accomplished, a lot still needs to be done, and there is a lot of opportunity that remains,” McCulloch says.

Additional possibilities

While cancer detection is leading the way, exceptional canine olfaction could also potentially be used to detect other pathologies that have characteristic odor signatures, researchers say. These include urinary tract infections, gastrointestinal disorders such as gastritis due to Heliobacter pylori, diabetes, psychological disorders, endocrine diseases such as Cushing's, thyroid disorders, cirrhosis, Candida esophagitis, sinusitis and alcohol abuse.1

It's possible that one day dogs will be used to assist physicians in early detection screening of multiple diseases and especially advancing cancer research. In this way, they continue to be our protectors.


1. Lippi G, Cervellin G. Canine olfactory detection of cancer versus laboratory testing: myth or opportunity? Clin Chem Lab Med 2012;50(3):435-439.

2. Williams H, Pembroke A. Sniffer dogs in the melanoma clinic? Lancet 1989;1:734.

3. Church J, Williams H. Another sniffer dog for the clinic? Lancet 2001;358:930.

4. Pickel DP, Manucy GP, Walker DB, et al. Evidence for cancer olfactory detection of melanoma. Appl Anim Behav Sci 2004;89:107-116.

5. Willis CM, Church SM, Guest CM, et al. Olfactory detection of human bladder cancer by dogs: Proof of principle study. BMJ 2004;329:712.

6. McCulloch M, Jezierski T, Broffman M, et al. Diagnostic accuracy of canine scent detection in early- and late-stage lung and breast cancer. Int Cancer Ther 2006;5(1):1-10.

7. Horvath G, Jarverud S, Horvath I. Human ovarian carcinomas detected by specific odor. Integr Cancer Ther 7(2):76-80.

8. Horvath G, Anderson H, Paulsson G. Characteristic odor in the blood reveals ovarian carcinoma. BMC Cancer 2010;10:643.

9. Cornu JN, Cancel-Tassin G, Ondet V, et al. Olfactory detection of prostatic cancer by dogs sniffing urine: a step forward in early diagnosis. Eur Urol 2011;59(2):197-201.

10. Sonada H, Kohnoe S, Yamazato T, et al. Colorectal cancer screening with odour material by canine scent detection. Gut 2011;60(6):814-819.

11. McCulloch M, Turner K, Broffman M. Lung cancer detection by canine scent: will there be a lab in the lab. European Resp J 2012;39(3):511-512.

12. Taverna G, Tidu L, Grizzi F, et al. Olfactory system of highly trained dogs detects prostate cancer in urine samples. J Urol 2015;193(4):1382-1387.

13. Dorman DC, Foster M, Fernhoff KE, et al. Canine scent detection of canine cancer: a feasibility study. Vet Med Res Rep 2017;8:69-76.

14. Kitiyakara T, Redmond S, Unwanatham N, et al. The detection of hepatocellular (HCC) from patient's breath using canine scent detection: a proof-of-concept study. J Breath Res 2017;11(4):046002.

15. Concha AR, Guest CM, Harris R, et al. Canine olfactory thresholds to amyl acetate in a biomedical detection scenario. Front Vet Sci 2018;5:345.

Ed Kane, PhD, is a researcher and consultant in animal nutrition. He is an author and editor on nutrition, physiology and veterinary medicine with a background in horses, pets and livestock. Kane is based in Seattle.

download issueDownload Issue : dvm360 December 2019