http://healthland.time.com/2011/08/18/a-new-way-to-detect-lung-cancer-dogs-can-sniff-it-out/ They're man's best friend, but dogs, it turns out, may also be a doctor's newest secret weapon for detecting cancer. German researchers report in the European Respiratory Journal that dogs can be trained to detect lung cancer by sniffing human breath. The scientists worked with an admittedly small number of canines — just four, including two German shepherds, a Lab and an Australian shepherd — but the dogs had good success. They were able to suss out cancer in 71 out of 100 breath samples from lung cancer patients, and were able to correctly identify 93% of cancer-free samples, giving them an impressively low rate of false positives. That's better than the imaging tests that most physicians currently use to detect lung cancer. WebMD reports that in a recent study, longtime smokers who went in for annual CT scans of their lungs cut their risk of dying from lung cancer by only 20%. How does Fido do it? It's no secret that dogs have an acute and sophisticated sense of smell, and the scientists believe the canines are picking up on very subtle changes in certain volatile organic compounds in the breath, which may change when cancer is present. (In the study, patients exhaled into glass tubes filled with odor-capturing fleece, and the dogs were given these to sniff.) The dogs were also able to distinguish between patients with lung cancer and those with chronic obstructive pulmonary disease, as well as between smokers and nonsmokers. “Our results confirm the presence of a stable marker for lung cancer,” wrote lead author Thorsten Walles of Schillerhoehe Hospital, in a statement. “This is a big step forward in the diagnosis of lung cancer, but we still need to precisely identify the compounds observed in the exhaled breath of patients.” While there has been some precedent for disease-sniffing dogs — in other studies, they have had anywhere from 40% to 90% success in accurately identifying cancers including bladder and colon cancers, not to mention their ability to sniff out low blood sugar levels among diabetics — more work needs to be done before dogs become part of a clinical workup. For one thing, there is the question of how the dogs are trained and for how long. In the German study, Walles worked with the dogs for nine months to recognize the scent of cancer. It's also not clear exactly what odor the dogs are detecting. As Walles noted, “It is unfortunate that dogs cannot communicate the biochemistry of the scent of cancer!” If they could, we might have one more way to pick up the first signs of the disease and help patients to receive treatment earlier, when it may do the most good. Read more: http://healthland.time.com/2011/08/...g-cancer-dogs-can-sniff-it-out/#ixzz1VQNhpHGP
Ok, but what is the percentage rate of correct diagnosis from annual CT scans or other current methods of detection? Sounds like they're comparing a 'staged' or 'experimental' detection/success rate, of dogs on fixed subject types, with the rate at which detection through CT scans results in lowered risk of dying in real world scenarios. I'm not even sure what they're getting at. Maybe it's me. Do the dogs detect better than the docs, or not? Strange wording/analysis/comparison.
