A team of researchers from Duke University Medical Center and the Duke University Institute for Genome Sciences and Policy have developed a gene expression-based prognostic profile to help clinicians decide which early stage non-small cell lung cancer patients would benefit from chemotherapy, but would otherwise be advised against it.
The test’s intended use is the inverse of Genomic Health’s Oncotype Dx, which helps clinicians decide which patients can safely avoid painful chemotherapy. “The idea of the lung Metagene test is to identify high-risk groups in stage 1 non-small cell lung carcinoma—those patients who would currently not receive chemotherapy,” Anil Potti, lead author of the study and an assistant professor of medicine at the medical center, told Pharmacogenomics Reporter in an interview last week.
After a prospective clinical trial validation, the Duke group is “absolutely” planning to seek FDA in vitro device clearance for the test, said Potti. The group will begin offering the test from Duke in January as a part of its clinical trial, but it has no definite plans to commercialize the test with any diagnostics company.
About “50 percent of [early stage non-small cell lung cancer] patients will have a recurrence within five years,” Potti said. Instead of improving patients’ quality of life, which is what Oncotype does, the Lung Metagene Predictor test is aimed at improving survival by finding patients who need chemo, he added. “Recurrence in non-small cell lung cancer equates to death.”
But like Oncotype, the Metagene test predicts recurrence of cancer after surgery based on the gene expression signature of a tissue sample. Metagene, which runs on an Affymetrix array, examines the expression of 133 genes to predict recurrence within two years versus which patients will live beyond five years, said Potti. In retrospective analyses, the test has displayed an accuracy of about 90 percent.
Instead of looking at individual expression levels, the test examines the expression levels of groups of genes with similar expression levels, an approach that adds statistical power, Potti said. The group’s research appears in the Aug. 10 issue of the New England Journal of Medicine.
The Metagene test will be validated in an approximately 1,200-patient prospective clinical trial with the Cancer and Leukemia Group B national clinical trials cooperative group. It will be the first to select treatment options for lung cancer patients. “We’re in the process of negotiating with the FDA to get an IDE” for the test in anticipation of the multi-site trial, Potti said. The group is hoping to get its exemption before the end of the multi-site clinical trial, he said.
The test should also be useful beyond stage 1 patients—it also applies to stage 2 and stage 3 patients, said Potti. Down the road, the Duke researchers may conduct a trial to determine if the test can identify stage 2 patients — who are typically given chemotherapy but may not need it because they are at very low risk of recurrence, he said.
The Metagene test’s expression profile was developed through retrospective studies of three datasets, an 84-sample set from the CLGB, a 44-sample set from the American College of Surgeons, and a 20-sample set from the Mayo Clinic.
The Duke research group also has a paper under review at Nature Medicine that describes gene expression-based profiles that predict response to a panel of chemotherapy agents, said Potti. “We just need to get [the current clinical trial] ironed out so that we have the process in place,” he said. “And then we can run multiple trials using the same kind of platform example, where we can just validate the use of markers that predict chemotherapy response.”
A trial validating a drug-response panel might begin within about a year, Potti estimated.