Combining exhaled-breath analysis data with clinical parameters to improve the diagnosis of lung cancer

Introduction: Lung cancer remains a leading cause of cancer mortality. Exhaled-breath analysis of volatile organic compounds (VOC’s), reflecting pathological processes, might detect lung cancer at an early stage, possibly leading to improved outcomes. Combining breath patterns with clinical parameters may improve the accuracy to diagnose lung cancer.

Methods: In a multi-center study 144 subjects diagnosed with non-small cell lung cancer (NSCLC) and 146 healthy subjects breathed into the Aeonose™ (The eNose Company, Zutphen, Netherlands). The diagnostic accuracy, presented as Area under the Curve (AUC) of the Aeonose™ sec was compared with the diagnostic accuracy when combined with clinical parameters in a multivariate logistic regression analysis.

Results: Confirmed NSCLC patients (67.1 (9.0) years; 57.6% male) were compared with controls without NSCLC (62.1 (7.1) years; 40.4% male). The AUC of the absolute Aeonose™ value obtained by a trained neural network was 0.76 (95% CI: 0.71-0.82). Adding age, number of pack years, and presence of COPD to this absolute value of the Aeonose™ from the neural network resulted in an improved performance with an AUC of 0.86 (95% CI: 0.81-0.90). By choosing an appropriate threshold value in the ROC-diagram of the multivariate model, we observed a sensitivity of 95.7%, a specificity of 59.7%, and a positive and negative predictive value of 69.5% and 92.5%, respectively.

Conclusion: Adding readily available clinical information to the absolute obtained value of exhaled-breath analysis with the Aeonose™ improves the diagnostic accuracy to detect the presence or absence of lung cancer.