1206 Application of metabolomic science in the detection of lung cancer: a case-control study. (24th April 2018)

• Record Type: Journal Article
• Authors: Huang, Chi-Hsiang; Zeng, Chian; Chang, Che-Jui; Peng, Hsin-Yi; Wang, Yi-Chia; Yang, Hsiao-Yu
• Is Part Of: Occupational and environmental medicine. Volume 75(2018)Supplement 2
• Page Start: A122
• Page End: A122
• Publication Date: 2018-04-24
• Subjects: metabolomics -- lung cancer -- electronic nose; Medicine, Industrial -- Periodicals; Environmental health -- Periodicals; 616.980305
• DOI: 10.1136/oemed-2018-ICOHabstracts.346
• Languages: English
• ISSNs: 1351-0711

Abstract

Abstract : Introduction: Lung cancer is the leading cause of cancer death in the world. The challenge of screening for early stage lung cancer is still unresolved. The exploration of metabolites in breathe using sensor array technique may become a powerful screening tool to solve the problem. Methods: We conducted a prospective study to enrol cases of lung cancer and controls who received surgery for gall bladder stone, hernia, hemorrhoid resection, and thoracoscopic surgery in the same hospital between July 2016 and June 2017. The alveolar air of subjects were collected under the guidance of mainstream carbon dioxide analyzer. An electronic nose composed of 32 carbon nanotubes sensors was used to measure the VOCs of the alveolar air. The diagnostic accuracy was analysed by linear discriminant analysis (LDA) using the pathological reports as the reference standard. Results: After excluding 2 subjects with technical problems in sampling, 12 subjects with cancers in other sites, benign lung tumour, or metastatic lung cancer, 5 subjects received chemotherapy, 5 subjects with diabetes, 2 subjects with asthma, and 2 subjects with chronic obstructive pulmonary disease, a total of 17 cases and 105 controls were used in the final analysis. We randomly split the data into 80% for model building (training set) and 20% for validation (test set). By LDA, the accuracy, sensitivity, specificity, false positive rate, false negative rate, and ROC-AUC were 96.9%, 75.0%, 100.0%, 0%, 25%, and 0.98 (95% CI: 0.96 to 1.00) in the training set, and 84.0%, 80.0%, 85.0%, 15.0%, 20.0% and 0.84 (95% CI: 0.62 to 1.00) in the test set. Conclusion: The use of sensor array technique to explore the metabolites in breathe may become a powerful tool in the screening for lung cancer. Standardised procedures to eliminate confounding factors are warranted before clinical application.