Gene Expression Classifier Boosts Bronchoscopy Lung Cancer Detection

Combining bronchoscopy and gene expression analysis could improve the detection of lung cancers

medwireNews: Applying a bronchial airway gene expression classifier to epithelial cells collected from the mainstem bronchus during bronchoscopy may aid the diagnosis of patients with suspected lung cancer, US researchers suggest.

“We found that the gene-expression classifier had high sensitivity across different lesion sizes, locations, stages, and cell types of lung cancer”, say Avrum Spira, from the Boston University Medical Center in Massachusetts, and co-authors in The New England Journal of Medicine

When applied to current or former smokers with suspected lung cancer participating in the Airway Epithelial Gene Expression in the Diagnosis of Lung Cancer (AEGIS) trials, the combined technique was 96% sensitive for lung cancer in the 298 AEGIS-1 patients and 98% sensitive in the 341 AEGIS-2 patients. By comparison, bronchoscopy alone was 74% and 76% sensitive for lung cancer, respectively.

The classifier alone was 86% sensitive for cancer in 57 AEGIS-1 patients with a non-diagnostic bronchoscopy result and 92% sensitive in 63 AEGIS-2 patients with non-diagnostic bronchoscopy.

When combining patients with and without a diagnostic bronchoscopy test, the classifier plus bronchoscopy was more sensitive than bronchoscopy alone for detection of lesions smaller than 3 cm in diameter, peripheral tumours and cancer in patients without hilar or mediastinal adenopathy. 

And the gene expression classifier alone or in combination with bronchoscopy showed “consistently high” sensitivity for cancer detection regardless of tumour size or location, histology and adenopathy, varying between 94% and 100%. 

For 101 patients with a pre-test physician-assessed intermediate probability of lung cancer, bronchoscopy was nondiagnostic for 83% of the group despite a 41% prevalence of cancer, the researchers observe.

But for these patients, the gene expression classifier had a negative and positive predictive value of 91% and 40%, respectively. And for patients with indeterminate nodes and a low or intermediate physician-assessed likelihood of cancer, the classifier was 88% sensitive for malignancy with a negative predictive value of 94%.

“These findings suggest that this classifier has the potential to assist in clinical decision making for patients with an intermediate probability of cancer, in whom the prevalence of lung cancer is 41% but the sensitivity of bronchoscopy is only 41%”, Avrum Spira et al write.

A negative classifier score may help avoid unnecessary invasive testing in patients with intermediate probability of cancer but could lead to delayed diagnosis in the small number of patients with lung cancer, the researchers admit.

“However, patients with a nondiagnostic bronchoscopic examination and a negative classifier score would probably undergo active surveillance with the use of imaging, which is the standard practice when an immediate invasive strategy is not used”, they explain. “This would identify lesion growth and trigger additional invasive testing to establish a definitive diagnosis.”


Silvestri GA, Vachani A, Whitney D,et al. A Bronchial genomic classifier for the diagnostic evaluation of lung cancer. N Engl J Med 2015; Advance online publication 17 May. DOI: 10.1056/NEJMoa1504601

medwireNews ( ) is an independent medical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2015