Novel Classifier Test Helps Reduce Invasive Procedures in Lung Cancer Diagnosis

In a large prospective multicenter set of two clinical validation studies, a new bronchial genomic test was found effective in helping to prevent unnecessary invasive, often risky and expensive procedures during diagnosis of suspected lung cancer cases.

The two studies, the AEGIS I and AEGIS II trials, assessed clinical performance of the Percepta Bronchial Genomic Classifier (BGC) test from Veracyte, Inc. (South San Francisco, CA, USA), a pioneer of molecular cytology in diagnostics. Findings demonstrated the ability of the genomic test to help patients more safely avoid unnecessary procedures to examine suspicious lung nodules or lesions initially found on computed tomography (CT) scans.

The trials involved 639 patients (298 in AEGIS I, 341 in AEGIS II), at 28 centers in the US, Canada, and Ireland, who were undergoing bronchoscopy to non-surgically assess lung nodules. Among patients initially assessed as having “intermediate” risk of malignancy who then had a bronchoscopy that was non-diagnostic, Percepta BGC had a negative predictive value (NPV) of 91%, demonstrating the test’s ability to reclassify these patients as “low risk” with high accuracy. Percepta BGC and bronchoscopy had a combined sensitivity of 97%, compared to 75% for bronchoscopy alone. Percepta BGC’s specificity was 47% in both studies, and its sensitivity, alone and in combination with bronchoscopy, was consistently high, regardless of nodule size and location, and cancer stage and type.

“These data suggest that the BGC enables physicians to confidently identify patients who are at low probability for having lung cancer following an indeterminate bronchoscopy result,” said Avrum Spira, MD, MSc, professor of medicine at Boston University School of Medicine and co-inventor of the genomic test, “Many of these patients can then be followed with CT scans, rather than undergoing harmful, invasive—and potentially unnecessary—follow-up procedures. The need for improved lung cancer diagnosis becomes increasingly acute given new screening programs that will expand the number of nodules found.” In 2015 over 8 million Americans at higher risk for lung cancer became eligible for annual screening with low-dose CT (LD-CT) through new private-insurer and Medicare coverage requirements.

Lung nodules and lesions found on CT scans are often small and located at peripheral sites in the lung, making them hard to diagnose without invasive biopsies. Bronchoscopy is currently the safer, least-invasive diagnostic option, but often produces inconclusive results. In the AEGIS trials, 43% of bronchoscopies were non-diagnostic for lung cancer, with 64% of patients undergoing invasive follow-up procedures, among which 35% of patients had benign nodules or lesions.

“These rigorous studies represent the largest evaluation to date of patients undergoing bronchoscopy, and establish the performance and utility of our Percepta Bronchial Genomic Classifier in lung cancer diagnosis,” said Bonnie H. Anderson, president and CEO, Veracyte, “By helping to prevent unnecessary invasive diagnostic procedures, the Percepta test should provide very tangible benefits for patients and physicians, while removing costs from the healthcare system.”

The test is a 23-gene molecular classifier using proprietary genomic technology to detect gene expression changes that can be detected in cells obtained from brushings taken during bronchoscopy from the mainstem bronchus. It is designed to determine likelihood of cancer without the need to sample the nodule or lesion directly.

The report describing the AEGIS I and II studies, by Silvestri GA, and was published online May 17, 2015, in the New England Journal of Medicine (NEJM).

Related Links:
Veracyte