Together we will beat cancer

A dog's nose

Canine cancer detection isn't practical

Man’s best friend was in the news yesterday, as scientists in Japan reported that an 8-year-old Labrador had been successfully trained to distinguish breath and stool samples from healthy people from those with bowel cancer.

We’ve written at length before about previous studies investigating whether dogs can detect various types of cancer. Overall, only a handful of dogs have been successfully trained, with varying degrees of success.

For obvious reasons, it’s highly unlikely that canine cancer detectors will ever be a practical solution for the NHS. As the scientists responsible for this new research point out:

“It may be difficult to introduce canine scent judgement into clinical practice owing to the expense and time required for the dog trainer and for dog education. Scent ability and concentration vary between different dogs and also within the same dog on different days.”

As a solution, the scientists suggest that research should be directed towards identifying the smelly molecules produced by tumours, and developing “electronic noses”. Such devices – which are currently being worked on by scientists around the world – would be able to detect cancer in a highly accurate and reproducible way without the need for feeding and walkies, although they may not be quite as cute.


Sonoda, H., et al (2011). Colorectal cancer screening with odour material by canine scent detection Gut DOI: 10.1136/gut.2010.218305


Robert February 29, 2012

I am a canine trainer and I can say for many years I have heard how the military ( huge budgets) and domestic law enforcement ( of which I have been a part of) has been working on , developing, and dreaming of quality replacements to canines in the areas of scent detection for many purposes. Dispute the millions spent and the 2nd rate results, the “dog” continues to be the most cost effective and most proficient asset in every area. Don’t “poo poo” the dogs abilities and results. Yes they work cheap and thus there will be money to be
Made by competent trainers and not huge research firms, but the results will be cheap for the consumer and available in the “now”.

Amy February 19, 2012

It’s disappointing to read: “it’s highly unlikely that canine cancer detectors will ever be a practical solution” when dog have been used around the world to effectively identify explosives, drugs, human remains, blood and even bed bugs and mold. Reminds me of all the naysayers who thought man would never fly or even better when the US over engineered a pen for space when Russia used a pencil. When will we learn that not everything needs to be engineered?

Jared April 26, 2011

If one person could benefit from an early dog-detection, then I say its worth training 1,000,000,000 dogs. It looks like this was written by someone who works in cancer research. I believe the only reason we don’t have cancer screening dogs at every clinic now, is because people see their research dollars being diverted to training dogs. Since everyone in the medical industry is in it to make money, there is no way they will accept dog screening.

Allan Parsons March 21, 2011

As a sufferer from terminal Colorectal Cancer (CRC) myself, I obviously have a strong interest in the subject. Further, as previously an Analytical Chemist by profession, I was intrigued to see that it appears likely that there are Volatile Organic Compounds (VOC’s) that are indicative of Colorectal Cancer and to see that dogs can apparently be trained to detect these.

Whilst like other comments I have seen, I think that, in the longer term, the use of dogs for this purpose, is not altogether a practical proposition. However, the very fact that they can do it with accuracies, apparently even better than colonoscopy, indicates that there are VOC’s present that could be detected using Analytical Chemistry techniques. For, such techniques can these days, be even more sensitive than our canine friends and are certainly much more specific (i.e. they could be set up to look for single specific compounds or combinations of two or more compounds in amongst a myriad of others).

It thus seems to me that it would not be an insurmountable problem, using positive and negative samples from CRC patients and Controls, to fairly soon, identify these VOC,s and, from this, to set up a relatively easy screening test for CRC.

The analytical technique that I feel, as an Analytical Chemist with more than 40 years experience, would almost certainly be best suited for this and one that is readily available in many laboratories, would be High Resolution, Gas Chromatography/Mass Spectrometry (HiRes-GC/MS), possibly in combination with a variety of other analytical techniques. This has the ability to detect any volatile compounds at extremely low levels and, once a target molecule has been identified, with extreme specificity. I would have thought that using this together with modern pattern recognition techniques, it should be possible, within a fairly short time span, to produce a screening test that would be quite cheap, very easy to carry out using relatively cheap equipment and which would be very specific.

I realise that it is quite possible that such work is already in progress, but in my fairly short research into this, I have not yet found any mention of it, so thought it might be an idea to raise it. I have seen that researchers are looking into identifying various proteins etc, that might be linked to CRC and other cancers, but of course, these are much larger molecules and would not be volatile and so are very unlikely to be the compounds that dogs are detecting.

So, is anyone out there looking at this angle ? I, as an analytical chemist who used similar techniques to help identify unknown compounds such as metabolites (i.e. breakdown products) of various candidate pharmaceuticals, pesticides etc., feel confident that similar techniques could be used here. If this possibility is NOT currently being researched, the question must be, why not !

george woolley February 1, 2011

I have had dogs for many years and all I have to say is my dogs for the main have been Parson Jack Russells`and for the most part they have a fantastic sence of smell .My Dog now will follow a smell for miles over open land .And can you tell me of any human who can smell urin and tell if the dog that weed is in season /male of femail.Dont knock the dog they leave us standing .