Can Bees Be Trained to Sniff Out Cancer?

Some insects, such as bees, have a sense of smell so acutely sensitive that they can locate the faintest of odors in a room, even if it consists of only a few molecules. But scientists are particularly intrigued by the fact that these bugs can even be taught to detect various chemicals, from methamphetamines to ingredients in explosives. They've even been shown to effectively diagnose diseases like tuberculosis and diabetes.

U.K.-based product designer Susana Soares has created a simple, elegant way of harnessing bees to screen for a number of diseases, including cancers, like tumors of the lung and ovaries. Her glass apparatus, called "Bee's," features a large chamber and a smaller connected chamber housed within it. After training the bees to associate a specific chemical odor with a food reward, such as sugar, the insects are released into the diagnostic device through an opening. Patients would simply blow into the smaller compartment and wait to see if a swarm gathers toward something alarming in the person's breath.

The project, part of her master's thesis at London's Royal College of Art, began in 2007 when Soares came across research on bees and their phenomenal olfactory abilities. After talking to researchers in the field, she learned that certain diseases, such as lung cancer, noticeably alter the composition of bodily fluids, producing odorous compounds that show up in urine and sometimes blood. Some investigators have even been experimenting with various sensory methods to home in on these "biomarkers." In Philadelphia, for instance, scientists have trained mice to identify the scent of lung cancer. Trained dogs have also been used to sniff out ovarian cancer. Others have focused on replicating these animal abilities in electronic nose devices that are calibrated to pick up these biomarkers undetectable to human noses.

Insects offer key advantages over mammals and electronics, however, because of their antennae. For example, electronic nose devices have trouble detecting an odor amid more complicated conditions, like when there's a greater mixture of gases, as is found in human breath. And studies have revealed that sniffer dogs identify odors correctly only about 71 percent of the time, while also requiring at least three months' training. Bees, in contrast, have achieved an accuracy rate of 98 percent and can be trained in about 10 minutes.

In developing "Bee's," the Portuguese native needed something that enabled the user to easily transport bees into the instrument and safely suck them back out using a vacuum. The source material also had to be malleable enough to shape into a system with well-defined pathways that don't impede their movement. She eventually settled on glass as the material because of its flexibility and transparency. “To know the results of a breath test, you'd have to see the behavior of the insects," she says. "Everything is about their behavior."

Prototypes have undergone field testing, and although it didn't find any instances of cancer, it did turn up a case of diabetes that was later confirmed. It’s unlikely, though, that the concept will amount to anything beyond being an exhibition curiosity. While there was a brief period in which she felt ambitious enough to reach out to potential collaborators, the process proved so time consuming and unfruitful that she ultimately gave up. The only organizations that seemed even remotely interested in her idea were a handful of charities. So for now, "Bee's" exists as one of those purely academic exercises to show, as she puts it, the "symbiotic relationship" humans have with nature and how "technology and science can better foster these relationships."

"I think there's only four labs in the world doing research into insects for disease screening, which shows you that this approach doesn't go over well in the western world," says Soares. "Medical and health technologies are a big business, and the bottom line is they just don't see how something like this can be profitable."

Glen C. Rains, an agricultural professor at the University of Georgia, largely concurs, though he adds that there are more complex issues besides economics. The entomologist, as well as licensed beekeeper, has dealt with numerous challenges while developing a similar device called the Wasp Hound, which uses a batch of five wasps to detect the presence of bedbugs. Rains' system is a bit more elaborate in that it uses a camera to record the wasps' behavior. The data is then fed into software that analyzes these movements to determine if the bugs actually did indeed detect these unwanted guests. After over a decade of development, Rains has forged a partnership with Bennett Aerospace, an engineering firm, to refine the technology for large-scale real applications.

"The whole notion is definitely something people find fascinating," he says. "But once you get into how it would work or how they make money, there's no model for how it would be done."

(Credit: Susana Soares)

By Tuan C. Nguyen

SMITHSONIAN.COM

DECEMBER 13, 2013