Lancet Infectious Diseases study shows that dogs meet WHO threshold for distinguishing infected from non-infected based on smell.
For many of us, dogs provide love and companionship.
For others, they offer that and then some, serving as eyes for those with vision impairment or as a form of therapy for those with PTSD. Still more canines play important roles in law enforcement.
Now, though, our 4-legged friends may have a future in the detection of vector-borne infectious diseases—at least if the findings of an analysis published in The Lancet Infectious Diseases are any indication. For the project, a team of investigators from Britain and The Gambia worked with 2 dogs specifically trained to recognize the body odor produced by malaria parasites in people infected by the mosquito-borne disease.
What they discovered is that these dogs were certainly capable of learning new tricks.
“Canines are a technology,” said T. Craig Angle, PhD, MEd, co-director, Canine Performance Sciences, School of Veterinary Medicine, Auburn University. “They are a rapidly programmable, real-time [tool], capable of detecting odors at the part-per-trillion level.”
Of course, not every Fido or Rover is up to the task, notes Angle, whose team was not involved in the project in The Gambia. Indeed, the researchers there worked with dogs trained by a Britain-based organization called, fittingly enough, Medical Detection Dogs, which touts “dogs saving lives” for people with cancer and certain neurological disorders. According to Angle, training for dogs used in medical detection, and specifically infectious disease identification, depends on a number of factors, including the pathogen and host characteristics. Angle’s team is made up of researchers with expertise in veterinary medicine, infectious disease, pathobiology, microbiology, chemistry, and animal behavior “to figure out how to create a reliable, safe, successful, and sustainable canine bio detection program.”
To assess the potential role of these canines in the detection of malaria, the authors of The Lancet Infectious Disease report collected foot odors from 175 Gambian school children between 5 and 13 years of age by fitting them with specially designed socks, which they wore overnight. In all, 30 of the children had asymptomatic falciparum malaria as confirmed via microscopy. The researchers then assessed diagnostic accuracy of the 2 trained dogs in a blinded study performed roughly 18 months after the socks were collected from the children.
The dogs were able to distinguish differences in sample sock odors between malaria-infected and uninfected children with sensitivities of 73.3% and 70.0% and specificities of 91.0% (85·2—95·1) and 90.3%, respectively. The investigators also noted that both dogs “responded in the same way” to 80.6% of the samples and 6 uninfected samples were indicated as positive by both dogs, possibly due to uninfected children sharing beds (a common practice among Gambian children) with infected siblings or relations. In all, agreement among the 145 uninfected samples was 89.7%, the team reported.
“These results are broadly in line with [the World Health Organization’s] criteria for the procurement of rapid diagnostic tests, in which the test should be able to detect at least 75% of Plasmodium falciparum samples at densities of 200 parasites per μL or greater, with specificity of no less than 90%,” investigators concluded. “In our study, 60% of the samples were below this parasite density. Including only samples with 200 parasites per μL or greater increased the mean sensitivity of both dogs to 81.8%, above the threshold required for…malaria diagnostics.”
None of which is a surprise to investigators such as Angle and his colleagues at Auburn. “The K9 partners we utilize in our research are…purpose-bred elite canines that have extraordinary capabilities above and beyond the general [dog] population,” he told Contagion®. “They come from our breeding program, which has made genetic advancements in detection dogs for about 20 years now. As we push the boundaries of operational performance and olfaction, we find what it takes to be true detectors and discriminators of disease. We were the first to show that canines can not only detect a particular virus but they can actually discriminate [between] different viruses, [and] we have since shown that they can do this with bacteria.”
Although much more work needs to be done before canine detection can be implemented in infectious diseases—including larger studies demonstrating diagnostic accuracy and the development of enhanced training methods—the findings to date show that dogs may be ready to join the fight. And there’s even video evidence on YouTube.