Late-breaking abstract: New algorithm of lung cancer diagnosis by analysis of exhaled breath with electronic nose and multifactorial logistic regression method

Abstract

Background Exhaled breath of lung cancer patients contains unique pattern of volatile organic compounds (VOCs) which can be distinguished by analysis with electronic nose. Objective The aim of our study was to develop optimal diagnostic algorithm by multifactorial logistic regression (MLRA) analysis and test its diagnostic potential in patients with lung cancer. Methods Exhaled breath of lung cancer patients (cancer group) and mixed group of patients (COPD, asthma, pneumonia) and healthy volunteers (no cancer group) was examined. Exhaled air was collected using standardized method and sampled by electronic nose (Cyranose 320). Optimal detector parameter combination and mathematical model for discrimination of lung cancer was computed by MLRA backward step-wise method in smokers, ex-smokers and nonsmokers. Sensitivity, specificity, positive (PPV) and negative predictive value (NPV) of the algorithms in the training sample of each group was calculated. Results Total 474 patients, out of them 282 lung cancer patients and 192 patients with different lung diseases and healthy volunteers were recruited in the study. 129 were nonsmokers, 135 ex-smokers and 210 smokers.