Abstract
Computer tomography (CT), a gold standard screening tool for lung cancer in high risk symptomless subjects, has relatively high associated costs and risks with low incidence of cancer detection. There is a need for a tool to predict subjects with a higher probability of early stage lung cancer (ESLC - Staged I or II by 7th ed TNM system) prior to CT screening. We previously demonstrated that exhaled breath volatile organic compound (eVOC) profile differs between lung cancer and healthy smokers. Aim: To determine if eVOC profile of subjects diagnosed with ESLC, without symptoms (ESLC-S) differs from that of those with symptoms (ESLC+S) and that of healthy smokers (HS). Method: ESLC+S (n= 53), ESLC-S (n=29) and HS (n=188) subjects provided a breath sample after breathing filtered air for 5 mins for analysis on a 32 sensor Cyranose 320 (Sensigent). Sensor data were log transformed and reduced to principal components for canonical discriminant analysis to determine differences between groups. Accuracy (CVV) was calculated using SPSS V17.0 with leave one out cross validation. Sensitivity, specificity, PPV and NPV were calculated. Results: The eVOC profile distinguished between ESLC-S and HS (p<0.0001), CVV 91%, sensitivity 38%, specificity 100%, PPV 100% and NPV 91 %. The difference in eVOC profile between ESLC+S and ESLC-S was not significant (P=0.08). Conclusion: These data indicate the ability of exhaled breath VOC profiles to distinguish people with asymptomatic early stage lung cancer from healthy smokers, and therefore support its potential as a non-invasive screening tool for early detection of lung cancer. Supported by The Prince Charles Hospital Foundation and Private Practice Trust Fund.
- © 2014 ERS