This article is more than 5 years old. Information may no longer be current.
Gene-expression classifier improves detection of lung cancer with bronchoscopy
Click Here to Manage Email Alerts
DENVER — The use of a gene-expression classifier improved the diagnostic efficacy of bronchoscopy for the detection of lung cancer in patients with suspicious pulmonary lesions, according to the results of two prospective studies.
Bronchoscopy of pulmonary lesions is frequently nondiagnostic, according to study background. Although many lesions are benign, patients often undergo further invasive testing for diagnostic purposes.
The Percepta Bronchial Genomic Classifier (Veracyte) measures epithelial cells collected from mainstem bronchus to determine the probability of lung cancer. The classifier is based on the expression of 23 genes and patient age.
“Pulmonary physicians are seeing an increasing number of smokers referred to our practices because they have a suspicious lesion on their CT scans,” Avrum Spira, MD, MSc, professor of medicine at Boston University School of Medicine, told HemOnc Today. “Presently, it is very difficult for us to make the right decision as to which of those lesions should undergo an immediate surgical or needle biopsy and which can be followed noninvasively, through surveillance.”
Spira and colleagues evaluated data from current or former smokers undergoing bronchoscopy for suspected lung cancer at 28 centers in the U.S., Canada and Ireland for the Airway Epithelial Gene Expression in the Diagnosis of Lung Cancer (AEGIS) trials. The AEGIS-1 trial included 298 patients and the AEGIS-2 trial included 341 patients (total, n = 639).
Spira and colleagues presented results from these trials — which also were published in The New England Journal of Medicine — at the American Thoracic Society International Conference.
Lung cancer diagnoses occurred in 74% in the AEGIS-1 trial population and 78% in the AEGIS-2 trial population.
Bronchoscopy had a diagnostic sensitivity for the detection of lung cancer of 74% (95% CI, 68-79) in the AEGIS-1 trial and 76% (95% CI, 71-81) in the AEGIS-2 trial. Forty-three percent of bronchoscopic procedures were nondiagnostic — which included 120 patients who were ultimately diagnosed with lung cancer — and 35% of patients who underwent further invasive procedures had benign lesions.
In the AEGIS-1 trial, the gene-expression classifier demonstrated an area under the receiver-operating-characteristic curve (AUC) of 0.78 (95% CI, 0.73-0.83), a sensitivity of 88% (95% CI, 83-92) and a specificity of 47% (95% CI, 37-58). In the AEGIS-2 trial, the gene-expression classifier had an AUC of 0.74 (95% CI, 0.68-0.8), a sensitivity of 89% (95% CI, 84-92) and a specificity of 47% (95% CI, 36-59).
The combination of the gene-expression classifier with bronchoscopy resulted in a sensitivity of 96% (95% CI, 93-98) in the AEGIS-1 trial and 98% (95% CI, 96-99) in the AEGIS-2 trial regardless of lesion size and location.
Among 101 patients with an intermediate pre-test risk of cancer, the negative predictive value of the classifier was 91% (95% CI, 75-98) for those who had inconclusive bronchoscopy results.
The researchers acknowledged limitations of their study, including the inability to measure 155 patients for the gene-expression classifier due to insufficient RNA. Further, researchers lost 9% of patients to follow-up and 5% to a lack of definitive diagnosis after 12 months.
“CT scanning is a very good tool, but it picks up a lot of abnormalities and lesions, many of which are not cancer,” Spira said. “Still, it raises a lot of anxiety in both patients and physicians. This test could immediately provide physicians with information to better decide which patients should go immediately to get a biopsy and those who can be comfortably monitored. In particular, negative test results would give the physician a 90% confidence that what they are seeing on the CT scan is not a lung cancer.” – by Cameron Kelsall
For more information:
Avrum Spira, MD, MSc, can be reached at the Boston University School of Medicine, 72 E. Concord St., Boston, MA 02118; e-mail: aspira@bu.edu.
Disclosure: Spira reports receiving personal fees from Allegro Diagnostics, Inc. and Veracyte, Inc. He also reports several patents pending related to the content of the study. Please see the full study for a list of all other researchers’ relevant financial disclosures.
Click Here to Manage Email Alerts