It was known that man’s best friend had a highly developed sense of smell, which is used to detect explosives, drugs and fraudulently imported food in luggage. However, a few years ago, it was discovered that dogs could even detect different human diseases, including cancer, COVID-19, the warning signs of an epileptic seizure, even abnormal blood sugar levels in people. diabetics. Researchers from Quebec and France are currently carrying out studies aimed at scientifically confirming the ability of dogs to identify people with breast cancer or infected with COVID-19 and to determine what would be the best protocol to adopt so that these detections are as reliable as possible.
Since 2016, a team of researchers from the Institut Curie, in France, led by Isabelle Fromentin, has been training dogs to screen for breast cancer from sweat samples as part of a research program called KDOG which is funded by the Royal Canin Foundation – from the French dog and cat nutrition company – and the dog handler company Seris. During a first proof of concept, the dogs passed this screening with a sensitivity (ability to detect positive samples) of 93% and specificity (ability not to mark a negative sample as positive) of 97%. Building on this success, the team of Mme Fromentin has launched a clinical study involving 441 women selected by seven partner hospitals, some of whom have breast cancer and others are healthy.
Study participants should put a compress on their chest for a minimum of six hours, overnight. They must follow a very specific protocol which consists of washing and wiping their hands and breast with neutral soap and sterile wipes. They are also asked to avoid perfumes, which could contaminate the compress, not to smoke and not to eat too spicy, “to stay away from Mr. or Mrs., even from their cat or their dog during the night,” and to sleep if possible in clean sheets ”, specifies Philippe Mont, technical coordinator of the canine section of the KDOG program.
The compresses thus impregnated with women’s sweat are then presented to the previously trained dog. Each sample is placed behind a cone that the dog comes to smell. During a screening session, the dog faces four cones which he must sniff and mark the sample as positive, by sitting in front of the cone, or negative, by going directly to sit on it. a board at the end of the line of cones. “The screening of clinical samples takes place double-blind, that is to say that no one knows in which cone the positive sample is located, except the statisticians of the Institut Curie. Ultimately, the dog handler rewards the dog, regardless of the answer he gives. It is during training that the reward is only granted if the dog has succeeded in detecting the positive sample, because during the training sessions it is the person who deposited the positive sample behind a of the four cones which warns the master whether he should reward the dog or not ”, explains Mr. Mont.
In Quebec, the team of Professor Éric Troncy, from the Faculty of Veterinary Medicine of the University of Montreal, which participates in KDOG, exposes the dogs to about fifteen samples of sweat from women at a time. “It looks a lot more like the reality on the ground that the dogs who have been trained to screen for [infections au] SARS-CoV-2, ”notes Mr. Troncy.
To detect people with COVID-19, Mr. Troncy’s team presents the dog with a swab soaked in sweat collected under the armpits, on the face or around the neck. “The exhaled air would also be interesting as a sample for COVID-19 – and lung cancer – but the risk that we would contaminate ourselves was too great”, specifies Mr. Troncy before pointing out that at the airport of In Helsinki, travelers would take a sample of their sweat and give it to the lab technicians, who then submitted it to the dogs on the other side of a door.
In all the studies that have been published on COVID-19, dogs were able to detect the infection with a sensitivity ranging from 92% to 95% and a specificity close to 100%. “In comparison, the PCR test loses sensitivity depending on when the person was exposed to the virus. All researchers have witnessed dogs that have [déclaré positives] of infected people whose PCR test was negative but turned positive three days later. PCR tests also present a high risk of false positives [dus à une] contamination or poor performance of the test. And as for rapid antigen detection tests, their sensitivity is around 60% and their specificity around 80%. So the dog is much better! Raises Mr. Troncy.
Specific volatile organic compounds
But what exactly does the dog detect with his nose? “Whether in cancer or COVID-19, cells are attacked, and it is suspected that the fruit of this cellular activity leads to the production of volatile organic compounds. The body then gives off these specific odors (especially through sweat) that dogs can detect, ”says Mr. Troncy.
Knowing that each disease results in the release of specific volatile organic compounds, the researchers used chromatographs to find the olfactory signature of cancer present in their samples. They thus observed that an amalgam of different chemical molecules contributed to this specific odor. “We do not yet know exactly which molecules the dog detects to signify that a sample is positive,” admits the veterinarian.
In what context would the scout dogs be used after the clinical studies are completed? Dogs are a detection tool “at the same time fast, sensitive, specific and flexible, because a dog can be re-trained to detect a new scent.” If next year a new variant of SARS-CoV-2 appears that is not detected by the PCR test, we will just have to expose the dog to the odor generated by this new variant, and he will then be able to. detect, ”says Éric Troncy.
In the case of breast cancer, “the dog is not going to replace the mammogram or even the biopsy, but it could be used upstream to [faire un dépistage chez les] people who are reluctant to come to the hospital for a mammogram or have a sample taken. Even though we have a national breast cancer screening program, only 50% of Canadian women take it. This is totally insufficient, ”points out Mr. Troncy.
“The idea is also not to compete with mammography, which is extremely reliable. On the other hand, we hope in the long term that the use of dogs will be less expensive than mammography, but above all that it will be easier to access. This is a less restrictive method that is completely painless, safe and easy to do, so it could allow us to [tester les] women who live in villages far from major centers. Detection by dogs could also be useful in developing countries that cannot afford to have as many mammograms as we do here, ”explains Mont. “On the other hand, all these countries have dogs that dog handlers could train for medical detection,” adds Mr. Troncy.
“When a dog tells us that a sample is positive, we won’t go and tell the lady: you have cancer. We will tell her instead that she has a high risk of having one and we will invite her to go for a mammogram, ”warns Mr. Mont.