© 2005 American Thoracic Society
Can the Electronic Nose Really Sniff out Lung Cancer?To the Editor:"Extraordinary claims require extraordinary proof" said Carl Sagan, so the claim by Machado and colleagues that a breath test with an electronic nose detected lung cancer (1) merits close scrutiny. What biomarkers of lung cancer did the electronic nose detect in breath? Table 4 from Machado and coworkers' article, a list of 11 volatile organic compounds (VOCs) in the breath of some patients with lung cancer, does not answer this question. First, the VOCs in Table 4 were not identified with the electronic nose but with a different assay employing gas chromatography and mass spectroscopy. The authors present no evidence that the electronic nose detected these VOCs in breath. Second, there is no evidence that the VOCs in Table 4 were biomarkers of lung cancer, because they were reported only in a subset of eight patients, without statistical analysis of differences between patients with cancer and control subjects. Third, the authors state: "Of VOCs present, many have been previously reported in different concentrations in patients with lung cancer compared with those without lung cancer." In fact, these "many" numbered only three: acetone, pentane, and benzene, previously identified by our group (2, 3) and by Gordon and coworkers (4) as candidate breath biomarkers of lung cancer. Fourth, the authors present no evidence that the electronic nose can detect VOCs in breath. More than 3,000 different VOCs have been observed in breath, and most of them are present in picomolar concentrations (5), which translates approximately into parts per trillion. It is unlikely that the electronic nose employed by Machado and coworkers, or indeed any electronic nose in current use, possesses sufficient sensitivity to detect these VOCs. Fifth, tobacco smoke contains most of the VOCs listed in Table 4—for example, benzene and toluene (6)—and their concentrations in smokers' breath could feasibly fall within the detection limits of the electronic nose. The experimental design did not control for the potential confounding effects of age and tobacco smoking. In the training set, patients with lung cancer were significantly older than controls and had smoked longer (p < 0.001). In the validation set, the mean age of patients with lung cancer was 61.1 yr (SD = 13.1) (my calculation, using data in Table 2). Consequently, the patients with cancer were, on average, 24 yr older and had smoked more than the largest set of control subjects, the nonsmoking healthy volunteers. The authors have not proven that the electronic nose predicted lung cancer because their findings were apparently skewed by age and tobacco smoking.
Menssana Research, Inc., Fort Lee, New Jersey FOOTNOTES Conflict of Interest Statement: M.P. is President and CEO of Menssana Research, Inc., a company that develops breath tests for diseases including lung cancer. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
