Exhaled breath analysis in suspected cases of non-small-cell lung cancer: a cross-sectional study

Abstract

Lung cancer is one of the most frequently diagnosed cancers worldwide and is still the leading cause of cancer-related deaths. There is a considerable interest in finding diagnostic methods in the disease’s earliest stages. A complementary approach to imaging techniques could be provided by exhaled breath gas phase and exhaled breath condensate (EBC) analysis. The aim of this study was to quantify various biomarkers in the exhaled breath gas phase and EBC in suspected cases of non-small-cell lung cancer (NSCLC). The study involved 138 subjects with suspected lung cancer, 71 of whom had a subsequent diagnosis of NSCLC. The diagnostic power of a combination of hydrogen peroxide (H2O2)-EBC, and exhaled pentane, 2-methyl pentane, hexane, ethyl benzene, heptanal, trans-2-nonenal in distinguishing NSCLC and non-NSCLC subjects was poor-to-fair (area under the curve (AUC) = 0.68), similar to that of smoking history alone (expressed as pack-years, AUC = 0.70); a further improvement was observed when smoking history was combined with exhaled compounds (AUC = 0.80). The diagnostic power was increased in those patients with little or no past smoke exposure (AUC = 0.92) or where past smoke exposure was up to 30 pack-years (AUC = 0.85). Exhaled substances had a good accuracy in discriminating suspected cancerous cases only in those subjects with a modest smoking history (≤ 30 pack-years), but the inclusion of other exhaled biomarkers may increase the overall accuracy, regardless of tobacco smoke.