Coalition for Advanced Cancer Treatment and Prevention


Search the site

From CAT scan to Dog scan:

Are Lassie, Rin Tin Tin, and the Electronic Nose

the Future of Cancer Detection?

P. Anthony Chapdelaine, Jr., MD, MSPH, Exec. Dir./Sec.*

The aging of the world’s population, especially in the United States and Europe, has contributed to the increased prevalence and incidence of cancer, resulting in a rapid increase in the cost of health care treatment, medications, and insurance. Cancer prevention is still the medical community’s long-term goal, but in the United States achieving this goal will require a retooling of the “health care” industry away from its lucrative “sick care” business model (treatment of disease, illness and disability) to a model that emphasizes population and clinical education and guidance, social modeling and marketing, and prevention research into both genetic and environmental interactions leading to disease.

Such a shift from treatment to prevention is unlikely to occur in the foreseeable future. In the interim we need more accurate and less costly cancer screening methodologies (in terms of predictive value, risk-benefit, and cost-per-detected-cancer). The earlier we can detect cancer the easier and less risky our treatments, the more effective our premature death prevention, and the lower our overall “health care” costs. To do this, however, requires that we detect these various cancers using better (and preferably non-invasive) screening and diagnostic methods than we presently use.

Currently, our most widely-used population-based screening methods – mammography, colonoscopy and fecal occult blood tests, Pap smears, and PSA tests – are inadequate, being surrounded by controversy regarding their accuracy, reliability or practical clinical use, as well as uncertainty regarding the risk-benefit recommendations for their use.1, 2 These problems have been well-described by the United States Preventive Services Task Force (USPSTF) (which reports its findings annually to Congress and is an independent, volunteer group of experts in prevention, primary care, and evidence-based medicine formed in 1984 and now supported by the United States governmental Agency for Healthcare Research and Quality).1, 2 For instance, the PSA blood test used for prostate cancer screening has an overall “false positive” rate of over 75% (meaning the test is wrong three-fourths of the time when it says a man has prostate cancer). Similarly, a woman who has an annual mammogram for ten years in-a-row will be told almost half the time (false positive rate of 46%) that she has breast cancer when she really does not. This problem is typical in cancer screening and places doctors in the challenging position of trying to decide whether to recommend that their patients undergo further, usually invasive and potentially harmful, diagnostic testing – which for some patients can lead to life-long physical and emotional difficulties.

One underused, infrequently studied, and potentially superior method for early detection of diseases (of which most doctors and hospitals apparently are unaware) is the use of canine scent detection.3-9 Occasionally, other species have been studied for their ability to detect disease through scent, such as rats or bees to detect tuberculosis-causing bacteria.10-12 A recent analysis of the number of mammalian and amphibian genes coding for olfactory receptors showed that elephants ranked highest with nearly two-and-a-half times more genes than dogs (which ranked ninth, behind rats, cows, Chinese softshell turtles, horses and others).13 While still undergoing research, it is thought that the higher the number of olfactory receptor genes the better the ability to detect a greater number of odors. On the other hand, although dogs may not detect as many odors as some on the list, dogs often show a greater sensitivity (are better at detecting much smaller quantities) than those ranked higher with regard to number of genes coding for olfactory receptors.13

Compared to the 5 to 6 million scent (olfactory) receptors of humans, dogs have around 150 to 300 million receptors and are up to fifty times more sensitive to scent; thus, humans show the ability to detect a few parts per billion, but dogs show the ability to detect as few as one part per trillion.6 Thus, there are many reasons, both theoretical and practical, why canines remain the best species for studying early cancer detection through olfactory signals.

According to research conducted over the last decade dogs can be trained to detect chemicals produced by cancer cells and can determine whether medical samples of urine, breath, stool and tissue are cancerous or non-cancerous with great accuracy.3, 6, 14 The reason for this lies in the chemicals produced by cancer cells, known as Volatile Organic Compounds (VOCs), which appear to play a significant part in differentiating (separating) normal from cancerous cells.6, 7, 15 When a mutation in a cell causes a disruption of cellular genetic or protein replication and function, the outer “fatty” lipid membrane of the cell undergoes what is called “peroxidation” (like turning rancid) which then releases organic chemicals into the air (the VOCs evaporate and leave a detectable scent). These VOCs are then detected by the extremely sensitive olfactory (scent) detection organs of canines. The encouraging results from studies on canine olfactory detection of disease (cancer, infections, metabolic and endocrine disorders) over the last decade spurred the concurrent search to develop an electronic sensing device (commonly called an “electronic nose”) which is described in more detail below.3-7, 16, 17

One of the earliest reviews of research evidence supporting the use of canines in cancer detection was that of Moser and McCulloch in 2010.3 In their analysis of five studies on canine olfactory cancer detection that had passed reasonable scientific scrutiny they concluded that using dogs for cancer screening held promise, especially when canines were exposed to breath samples which seemed to give better results than urine samples.3

The latest and better-conducted research on VOCs emitted by samples of blood, urine, stool and breath has resulted in the expansion of the list of cancers (prostate, lung, ovarian, breast, colorectal, and melanoma) amenable to screening by canine olfactory detection.7-9, 17-21

One problem hampering widespread clinical use of canines for olfactory cancer detection has been the widely-variable study results, an ongoing challenge originally observed by Moser and McCulloch who largely attributed the inconsistent results shown by different studies to the lack of standardization in researchers’ training and testing methodologies.3, 6 The problems encountered in standardizing the canines’ training approaches includes: the blinding of the dogs and trainers to prevent unconscious cues from the trainers; the small number of dogs used in the studies (usually 1 to 5) which limits the reliability and reproducibility of training methods; the accuracy and practicality (reliability, reproducibility, sensitivity and specificity, and the positive and negative predictive values) of the tests themselves; the costs, practicality and acceptance of canines in clinical or laboratory settings; and the time required for initial training (which can range from months to two years), the necessity for periodic retraining (which varies with researcher from weekly to every two months), the temperament and motivation of different canines; and the variability of results found across studies and for individual canines. 3-7, 16, 19, 22-26

Most researchers now agree that the criteria for selecting the best dogs for this work includes: the dog should be young or easy to train; friendly and well-behaved (calm or good temperament); and a good work-ethic (dependable and enjoy herding or hunting).27 This leads to using certain breeds (German Shepherds, Collies, Labrador and Golden Retrievers, Poodles, Portuguese Water Dogs, and mixes of these) although using any breed that meets the above criteria could be used.27

Despite these objections, various solutions for the problems encountered in training and testing canines have been proposed: genetic testing to select canines having a particular chromosomal pattern that maximizes olfactory “binding” of molecules; improved sample collection and storage (for example, collecting breath samples for both control and cancer patients at the same site); better canine training; simultaneous (bimodal) studies using both “electronic sensors” and canine olfactory detection; and standardization of testing, analysis and reporting.6, 24, 25

Another solution for these problems involves developing a reliable, chemical-based electronic sensor device (so-called “electronic nose”) which can match the current and undeniable superiority of canine cancer-detection using scent; however, the difficulty in generalizing the results of electronic nose studies, the scarcity of studies using these devices, and the inferior detection thresholds of the current electronic noses compared to those of canines, make electronic noses a potentially useful but still long-term goal.4, 6, 16, 19, 25, 28 Nonetheless, the practicality of a mass-produced, accurate, reliable, low-cost mechanical medical device that can be used world-wide in all clinical settings, including third-world countries, is considered an appropriate, achievable, and ultimate goal by most scientists in this field.4, 6, 16, 24, 27

Currently, the European Union is spending a lot of money developing electronic noses, such as the NaNose™ breathalyzer technology developed by Professor Hossam Haick of the Technion, which in early research gives a ninety percent accuracy rate in identifying the VOCs from breath samples that distinguish between malignant and cancerous lung lesions, results which begin to approach those of canines’ ability to detect lung cancer in breath samples.19, 21, 22, 24, 25 Until fully tested and implemented the clinical use of electronic noses is years away, since canines – with their superior ability to correctly detect disease (predictive value) – still outperform the best electronic noses and likely will do so for many years.3, 4, 6, 7, 15, 16, 19, 20, 22

For this reason researchers like Dr. Guest, co-founder of the United Kingdom charity Medical Detection Dogs, and Dr. Michael McCulloch, research director of the California-based Pine Street Foundation, believe that canines have proven themselves better than current medical screening for many cancers, and that the funding to train large numbers of dogs for this purpose should be undertaken immediately, since it may be another decade or two before a reliable and accurate electronic nose will be available for widespread use.15, 19, 27

We can conclude from all this that for some cancers dogs can be used right now for the initial cancer screening, while for other cancers dogs can be used for detecting relapse (by detecting early-on a cancer that returns after treatment). Also, when using dogs along with conventional non-invasive cancer screening, the overall accuracy and reliability of the testing increases beyond each method alone, which results in giving doctors more confidence when advising patients to either undergo more invasive testing or to simply “wait and watch” and do non-invasive retesting later.

As an example of using dogs in a clinical setting consider that Medical Detection Dogs in Great Britain has trained canines to reliably screen prostate cancer, resulting in an accuracy rate above 90% and a specificity of 95% (false positives of less than 5%) – giving vastly better results than achievable when using standard PSA with digital rectal exam screening.2, 27 Medical Detection Dogs has demonstrated the clinical practicality of canine cancer screening by using its trained dogs to screen for prostate cancer from among several hundred urine samples taken from symptomatic patients and which had been sent to the charity from nearby Milton Keynes Hospital, giving the hospital doctors additional information (beyond symptoms, PSA and ultrasound) to use in determining which patients to send for biopsy.27

Dog scans have already joined CAT scans in a handful of medical clinics around the world and promise to improve early cancer detection and treatment outcomes.

* The Coalition for Advanced Cancer Treatment and Prevention, a Project of The National Fund for Alternative Medicine


  1. United States Preventive Services Task Force (USPSTF) Accessed November 14, 2016.
  2. Hoffman R, “Screening for Prostate Cancer,” 2016. Available at: November 14, 2016.
  3. Moser E, McCulloch M, “Canine Scent Detection of Human Cancers: A Review of Methods and Accuracy,” J Vet Beh, 2010, 5(3), Pgs. 145-152.
  4. Bijland LR, et al, “Smelling the Diagnosis: A Review on the Use of Scent in Diagnosing Disease,” Neth J Med, 2013, 71(6), Pgs. 300-307.
  5. Godfrey A, “Canine Scent Detection of Human Cancers: Is This a Viable Technique for Detection?” Vet Nurs J, 2014, 29, Pgs. 392–394: doi:10.1111/vnj.12200.
  6. Gould O, Smart A, et al, “Canine Olfactory Differentiation of Cancer – A Review of the Literature,” 2015 Oct, Cornell University Library,
  7. Taverna G, Tidu L, Grizzi F, “Sniffing Out Prostate Cancer: A New Clinical Opportunity,” Cent European J Urol, 2015 Oct, 68(3), Pgs. 308–310: doi:10.5173/ceju.2015.593.
  8. Taverna G, Tidu L, Torri V, et al, “Olfactory System of Highly Trained Dogs Detects Prostate Cancer in Urine Samples,” J Urol, 2015, 193(4), Pgs. 1382-1387.
  9. Taverna G, Tidu L, Grizzi F, Stork B, et al, “Highly-trained Dogs’ Olfactory System for Detecting Biochemical Recurrence Following Radical Prostatectomy,” Clin Chem Lab Med (CCLM), 2016 Mar, 54(3), e67-70: doi:10.1515/cclm-2015-0717.
  10. Suckling DM, Sagar RL, “Honeybees Apis mellifera Can Detect the Scent of Mycobacterium tuberculosis,” Tuberculosis (Edinb), 2011 Jul, 91(4), Pgs 327-328: doi:10.1016/
  11. Poling A, Mahoney A, et al, “Using Giant African Pouched Rats to Detect Human Tuberculosis: A Review,” The Pan Afr Med J, 2015, 21, article # 333: doi:10.11604/pamj.2015.21.333.2977
  12. Tiantian G, et al, “Multi-odor Discrimination by a Novel Bio-hybrid Sensing Preserving Rat’s Intact Smell Perception in Vivo,” Sensors and Acutators B, 2016 Mar, 225, Pgs. 34-41.
  13. Niimura Y, et al, “Extreme Expansion of the Olfactory Receptor Gene Repertoire in African Elephants and Evolutionary Dynamics of Orthologous Gene Groups in 13 Placental Mammals,” Genome Res, 2014 Sept, 24(9), Pgs. 485-496: doi:10.1101/gr.169532.113.
  14. Brooks S, Moore D, et al, “Canine Olfaction and Electronic Nose Detection of Volatile Organic Compounds in the Detection of Cancer: A Review,” Cancer Investigation, 2015 Oct, 33(9), Pgs. 411-419.
  15. Mazzatenta A, et al, “Pathologies Currently Identified by Exhaled Biomarkers,” Respir Physiol Neurobiol, 2013, 187(1), Pgs. 128-134.
  17. Capelli L, Taverna G, et al, “Application and Uses of Electronic Noses for Clinical Diagnosis on Urine Samples: A Review,” Sensors (Basel), 2016, Oct, 16(10), Pg.1708: doi:10.3390/s16101708.
  18. Ehmann R, Boedeker E, et al, “Canine Scent Detection in the Diagnosis of Lung Cancer: Revisiting a Puzzling Phenomenon,” Eur Respir J, 2012, 39, Pgs. 669-676.
  19. Kwak J, Gallagher M, Ozdener, et al, “Volatile Biomarkers from Human Melanoma Cells, J Chromatography B, 2013, 931, Pgs. 90-96.
  20. McCulloch M, Turner K, et al, “Lung Cancer Detection by Canine Scent: Will There Be a Lab in the Lab?” Eur Resp J, 2012, 39, Pgs. 511-512: doi:10.1183/09031936.00215511.
  21. Hovrath G, et al, “Cancer Odor in the Blood of Ovarian Cancer Patients: A Retrospective Study of Detection by Dogs During Treatment, 3 and 6 Months Afterward,” BMC Cancer, 2013 Aug, 13:396: doi:10.1186/1471-2407-13-396.
  22. Shamah D, “Mobile SniffPhone Will Detect Cancer On a User’s Breath,” 2015 Feb. Available at: November 14, 2016.
  23. Amundsen T, Sundstrom S, et al, “Can Dogs Smell Lung Cancer? First Study Using Exhaled Breath and Urine Screening in Unselected Patients with Suspected Lung Cancer,” Acta Oncol, 2014 Mar, 53(3), Pgs. 307-315: doi:10.3109/0284186X.2013.819996.
  24. Elliker KR, Sommerville BA, et al, “Key Considerations for the Experimental Training and Evaluation of Cancer Odour Detection Dogs: Lessons Learnt from a Double-blind, Controlled Trial of Prostate Cancer Detection,” BMC Urology, 2014, 14(1):22: doi:10.1186/1471-2490-14-22.
  25. Hackner K, Errhalt P, et al, “Canine Scent Detection for the Diagnosis of Lung Cancer in a Screening-like Situation,” J Breath Res, 2016 Dec, 10(4), Pgs. 409-461: doi:10.1088/1752-7155/10/4/046003.
  26. Boedeker E, et al, “Sniffer Dogs as Part of a Bimodal Bionic Research Approach to Develop a Lung Cancer Screening,” Interact Cardiovasc Thorac Surg, 2012 May, 14(5), Pgs. 511–515: doi:10.1093/icvts/ivr070.
  27. Jezierski T, Walczak M, et al, “Study of the Art: Olfaction Used for Cancer Detection on the Basis of Breath Odour. Perspectives and Limitations,” J Breath Res, 2015 Mar, 9(2):027001. doi:10.1088/1752-7155/9/2/027001
  28. Schubert M, “Animal Instincts,” 2016. Available at: Accessed November 14, 2016.
  29. Gordon RT, Schatz CB, et al, “The Use of Canines in the Detection of Human Cancers,” J Alt Comp Med, 2008 Jan, 14(1), Pgs. 61-67: doi:10.1089/acm.2006.6408.