(CNN) — Smell is a primary human sense, key to our survival.
Like a super-sensitive human nose, an experimental technology can “smell” and identify the chemical composition of a person’s breath and then diagnose up to 17 potential diseases, according to the scientists who developed it.
These researchers, led by Hossam Haick of the Technion-Israel Institute of Technology, say their Na-Nose, which uses nanorays to analyze breath, can identify Parkinson’s disease, various cancers, kidney failure, multiple sclerosis and Crohn’s disease with 86% accuracy.
“I would say our technology in many cases (is) equivalent to the accuracy of the currently available invasive technology,” Haick said, adding that for some diseases, including gastric cancer, Na-Nose has a “much higher” accuracy rate than currently available technologies. And, unlike most screenings, including standard blood tests, breath analysis technology is noninvasive — a benefit most patients would appreciate.
The theory behind the technology is that each of us has a unique chemical “fingerprint.” Each disease also has a particular chemical signature, which can be detected on our breath. The Na-Nose technology, which consists of a sensor chamber with a breathing tube and software, is able to detect this precise chemistry of disease by interpreting the impact on our usual chemical fingerprint.
Seven companies have licensed the underlying research for the technology from Technion in hopes of creating a commercial product, Haick said. He hopes that the companies, each specializing in a different application, will translate the science and technology from the lab to mass production.
One application, for example, would turn smartphones into “sniffphones” that would monitor our health routinely.
But with further testing and regulations to meet, neither the Na-nose device nor any variations will be available on the market — or in our doctor’s offices — for a number of years, said Haick.
Dogs, flies and rats
Though Na-Nose may seem revolutionary, smell was recognized as a potential diagnostic tool in antiquity.
“The ancient Greeks used breath and urine scent to diagnose disease,” said Dr. Mangilal Agarwal, director of the Integrated Nanosystems Development Institute and an associate professor at Richard L. Roudebush VA Medical Center in Indianapolis. “Thucydides said there was a specific scent to plague victims in Athens, and Hippocrates cataloged a specific disease because it caused bad breath and bad-smelling sweat.”
Agarwal, who is not involved with the Na-Nose technology, said he is working on a number of projects that analyze scents to diagnose diseases, including hypoglycemia (low blood sugar), prostate cancer and breast cancer.
“Breath has the scents or volatile biomarkers necessary to identify many diseases,” he said. “We know this from canines who can detect hypoglycemia and epileptic seizures, fruit flies (and canines) that can detect cancer, and from giant rats that detect tuberculosis in Africa.”
“Dr. Haick’s group is certainly ahead of our group in terms of getting close to doctor’s visit tests,” Agarwal said, adding that an important aspect to breath analysis is that it “excels at capturing changes in human health in a noninvasive manner.”
“Quick diagnosis can help in identifying the most appropriate treatment response,” he said. He added that prostate cancer grows on a longer time-scale, but the prostate biopsy is such “a sufficiently unpleasant experience” that a noninvasive test would be beneficial and lower health-care costs.
The high accuracy claims of Haick’s research group is “very reasonable, if the signal is not masked by environmental fluctuations in some manner,” Agarwal said, though he cautions that some “breath-based tests have had difficulty duplicating results in different regions, likely because the sensor has difficulty adjusting to different background air signals.”
Other scientists raise additional concerns.
Not ready for prime time?
Dr. George Preti, a faculty researcher at Monell Chemical Senses Center, a nonprofit scientific institute in Philadelphia, said it’s hard to distinguish body chemicals from environmental chemicals in breath samples because “most of the compounds detected in breath are also detected in room air and their levels are similar to each other.”
Until scientists “understand the origin and biochemical pathways leading to disease-related” markers in human breath, reliable results from a diagnostic breath test will be difficult to achieve, he stated in a recent review of studies.
In fact, there are more than a few issues that must be addressed before effective technologies will be produced, according to Dr. Lisa Spacek, an adjunct assistant professor at Johns Hopkins School of Medicine, and Terence Risby, professor emeritus at the Johns Hopkins University Bloomberg School of Public Health.
Using breath to diagnose disease first requires a profile of breath molecules for normal health to be established, Spacek and Risby say in a recently published paper. These must take into account variables such as age, gender, ethnicity and body mass index.
Researchers also need to investigate the factors that might contaminate breath results, such as what someone ate within eight hours of breath collection or whether they used a mouth rinse, say Spacek and Risby. Another issue: How do you store breath that is not immediately analyzed?
Advances in instrumentation, particularly portable monitors, is one factor inspiring and enabling the new research into breath analysis.
Though the field is growing and results are promising, translation of the work into meaningful tests is another matter: “I take every claim by manufacturers … with a grain of salt,” Risby wrote in an email.
Today’s widespread interest in breath analysis stems from the relatively recent discovery — within the past 20 years or so — that nitric oxide, a common pollutant, works as a signaling molecule in the cardiovascular system, Risby observes. The three scientists who made the discovery won a Nobel Prize for their efforts in 1998.
So despite ancient roots, Risby says, “clinical breath analysis remains in its infancy.”