A Bronchial Genomic Classifier for the Diagnostic Evaluation of Lung Cancer

In order to improve the diagnostic accuracy of bronchoscopy for lung cancer, epithelial cells were collected from normal appearing mucosa in the mainstem bronchus of patients suspected of having lung cancer during diagnostic bronchoscopy.  The samples were evaluated with a gene expression classifier.  43% of the standard examinations were nondiagnostic.  The gene expression classifier had an accuracy of 74% to 78%.  The combination of gene classifier and standard bronchoscopy had a sensitivity of 96% to 98%.  The gene classifier improved the diagnostic accuracy of bronchoscopy in intermediate-risk patients.