Your breath can tell a lot more about you than whether you brushed your teeth this morning or have been drinking alcohol. Researchers at the Cleveland Clinic have developed a sensor that can identify telltale chemicals in the breath of people with lung cancer. Lung cancer is difficult to detect in its early stages and difficult to treat when diagnosed in its late stages. Researchers hope that with further development, the breath sensor, an array of chemically sensitive dye spots, will help catch the disease earlier.
Lung cancer causes 160,000 deaths a year in the United States–more than any other cancer. Current diagnostic techniques including CT scans and needle biopsies are invasive and expensive and have a high risk of complications, says Peter Mazzone, a pulmonologist at the Cleveland Clinic. “We need an easy-to-use test for finding early-stage lung cancer,” he says.
Some of the products of metabolism, called volatile organic compounds, are carried in the breath and can serve as biomarkers. Cancer cells make different groups of these volatile compounds than normal cells do. Researchers have known since the mid-eighties that these differences can be detected on lung-cancer patients’ breath using a combination of gas chromoAtography and mass spectrometry. But Mazzone says applying these sophisticated analytical techniques to cancer diagnosis is expensive–it requires trained technicians and large machines–and they have not proved accurate enough for clinical use.
The Cleveland lung-cancer sensor is a disposable piece of paper called a colorimetric array. The paper has 36 chemically sensitive dye spots that change color when they interact with compounds in the breath. Changes in color are read by a flatbed scanner, which sends the images to a computer for analysis.
In a proof-of-principle study, Cleveland Clinic researchers analyzed the breath of 143 patients, some healthy, some known to have lung cancer, and some with other lung diseases. Patients breathed into a machine that kept their exhalations at body temperature and circulated them over the sensing array. The doctors modeled the patterns of color changes in the sensor array characteristic of lung cancer, then used the model to try to detect lung cancer in the remaining patients.
The Cleveland device is nonspecific: the researchers did not look for particular compounds in the breath but for patterns of color changes. The best sensors yet for detecting cancer on the breath are also nonspecific: dogs’ noses. A study published last year showed that trained dogs can distinguish between the breath of healthy people and those with lung cancer with 99 percent accuracy. The Cleveland device can detect lung cancer in about three out of four breath samples. “We hope our sensing technology can get close to the nose and can get as accurate results as the dogs do,” says Mazzone.
To improve the accuracy of breath-sensing devices, researchers need to know more about what they’re looking for. The metabolic reactions that make the characteristic compounds found in the breath are complex, says Nicholas Broffman, executive director of the Pine Street Foundation, the cancer-research center in San Anselmo, CA, whose scientists demonstrated dogs’ cancer-sensing ability. “We need to see which compounds are unique to cancer and which are not,” he says, in addition to finding out which compounds are unique to which cancers. Lung cancer isn’t the only cancer that causes detectable changes in the breath. “You don’t want to biopsy everywhere,” says Broffman.
Mazzone expects that future advances in mass spectrometry will make it possible to identify which compounds are characteristic of cancer patients, and of patients with different kinds of cancer. “When we know truly what the chemicals are, then we can go back to easier-to-use sensors [like the calorimetric array] and fine-tune them,” he says, in the hopes of increasing their accuracy. For example, the researchers might find that a group of alkanes occurs only in the breath of ovarian-cancer patients and tailor the dye spots on a colorimetric sensor to these compounds.
The Cleveland Clinic study “validates the idea that cancer could have a smell,” says Broffman. Noting that doctors historically sniffed patients’ breath for yeasty or sweet smells to diagnose tuberculosis and diabetes, Broffman says that “medicine is coming full circle.”
DeepMind’s cofounder: Generative AI is just a phase. What’s next is interactive AI.
“This is a profound moment in the history of technology,” says Mustafa Suleyman.
What to know about this autumn’s covid vaccines
New variants will pose a challenge, but early signs suggest the shots will still boost antibody responses.
Human-plus-AI solutions mitigate security threats
With the right human oversight, emerging technologies like artificial intelligence can help keep business and customer data secure
Next slide, please: A brief history of the corporate presentation
From million-dollar slide shows to Steve Jobs’s introduction of the iPhone, a bit of show business never hurt plain old business.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.