Using breath samples from 40 healthy volunteers and 56 lung-cancer patients, the group used the sensors to identify which biomarkers would collectively act as an accurate sign of lung-cancer signature. After training the sensors to identify the signature and testing it again, Haick and his colleagues found that their device could reliably differentiate between cancerous and healthy breath. They're now testing the device on a larger group of people in various stages of the disease and believe they'll be ready to start clinical trials within two or three years.

"Any advance in the area of developing sensors for breath research is exciting to me, and I think this is certainly an advance," says Peter Mazzone, a lung cancer specialist and breath-analysis researcher at the Cleveland Clinic in Ohio. "This was a very well-done, very promising study. I don't know if it is accurate enough to use in clinical practice, yet it's very exciting to see another promising sensor system."

Preliminary tests indicate that the gold-nanoparticle sensors can not only differentiate among stages of lung cancer, they can detect distinct signatures for other ailments, such as liver failure. Haick's group has even tested the electronic nose above colonies of cells grown in culture. This study found that while the sensor was able to sniff out compounds already known to be in breath, other lung-cancer-associated VOCs weren't detected. "Obviously, something is going on in the body to metabolize and create additional VOCs," Haick says. He's now working to figure out precisely what that is, in the hopes that it could provide new insight into lung cancer and how to treat it.