The Science of Olfactory Diagnosis

The smell of rotten apples on a patient’s breath doesn’t positively indicate diabetes, and fishy breath doesn’t positively indicate liver disease–any more than the smell of Smirnoff positively indicates alcoholism–but it can give a physician a pretty good starting point for further testing. “Olfactory diagnosis” has always been a valuable tool for physicians but, of course, the limited abilities of even the most talented human noses have severely limited the range of usefulness for the practice. What if a doctor was able to identify and track changes in the more than 3,000 chemicals which are exhaled or transpired or otherwise exuded from the human body?

As with most tasks requiring a keen sniffer, this job was first farmed out to man’s best friend, and dogs have proven effective in detecting certain forms of cancer. But it’s an impractical solution, at best. Qualifying and training sniffer-dogs is an expensive and lengthy proposition, and dogs, like all living things, are susceptible to error. Moreover, as with drug or cadaver sniffing dogs, the cancer-sniffing dogs can only offer a positive or negative, without degree or nuance. Dr. Boguslaw Buszewski of Nicolaus Copernicus University in Torun, Poland, likens the use of diagnostic dogs to checking for fever by touching a patient’s forehead. Excessive warmth will indicate illness, but only by measuring temperature with a thermometer can the severity of the condition be discovered. So, is there a breath-analysis equivalent of the thermometer?

Yes. In the 1970s, a Nobel-prize-winning chemist named Linus Pauling conducted the first scientific analysis of human breath. Using a technique called gas chromatography, he was able to separate complex mixtures and isolate some 250 organic compounds regularly exhaled from the lungs. A second procedure, mass spectrometry, then allowed him to identify the isolated components by molecule weight. Just this month, Michelle Gallagher, of the Monell Chemical Senses Center, concluded a study confirming the effectiveness of these combined techniques in diagnosing disease. Working with samples of air taken from just above the regions of known tumors and comparing it with air taken from the same regions in healthy volunteers, she discovered that both groups contained the same chemicals but in differing quantities. From this data, she created a “biomarker profile” which should allow the studied diseases to be reliably diagnosed in their early stages.

Though effective, Dr. Gallagher’s process can take up to two days to carry out, an inconvenience which Dr. Buszewski hopes to eliminate. He is currently developing a device comprised of ultra-thin silica columns coated with special polymers which bind to particular compounds. The idea is that this device will be passed through a waft of human breath and the molecules of interest will be immediately captured and analyzed. The patent is pending and some fine-tuning is in the works to widen the range of biomarker profiles recognized by the device, but, once perfected, it should shorten the testing time from two days to less than one hour, opening the door for complex “olfactory diagnosis” to become standard practice.