Putting Molecular Profiling to the Test

Although some early studies of the use of molecular profiling to guide cancer treatment have been promising, few randomized trials have assessed the benefits of the approach in specific kinds of tumors or across cancer types. As a result, insurance companies often won't pay for it, says Richard Schilsky, MD, chief medical officer of the American Society of Clinical Oncology in Alexandria, VA.

“There are people who rightly say, ‘Show us the evidence before we adopt this in clinical practice,’” says Schilsky.

Such evidence may soon be forthcoming, thanks to a randomized trial that will determine whether the approach actually extends survival in patients with cancer. Dubbed IMPACT2, the trial is led by Apostolia Tsimberidou, MD, PhD, of The University of Texas MD Anderson Cancer Center in Houston, and builds on its predecessor, IMPACT1. In that study of 1,144 people with advanced or metastatic cancer that advanced after standard therapy, Tsimberidou and colleagues found that tailoring patients' treatment to abnormalities in a handful of their tumors' genes prolonged survival. For example, among individuals who carried one genetic abnormality, 27% of patients showed a response to tailored treatment, whereas only 5% of patients responded to unmatched therapy. However, the study was not randomized.

For IMPACT2, Tsimberidou and colleagues plan to recruit about 1,400 patients whose tumors have metastasized. They will profile the patients' cancers using an assay from Cambridge, MA–based Foundation Medicine that can detect changes in more than 300 genes.

The researchers will then divide up the patients based on the availability of matched therapies: Two groups of patients will play no further part in the trial—either because FDA-approved treatments already exist for their specific genetic alterations, or because no potential therapies for those aberrations are being tested. A third group will include patients whose tumors harbor alterations that may be targeted by an investigational therapy.

The researchers will then subdivide the third group. Patients will be randomly assigned to receive either the therapy matched to their tumors' genetic abnormalities or a standard treatment chosen by their doctor without considering genetic information. Patients will receive the matched therapy in two ways: Either they will enter clinical trials evaluating new treatments targeted to that genetic change, or they will be given drugs that have been approved to target the same alteration in another disease.

“We hope to demonstrate that patients treated with a targeted therapy selected on the basis of a genetic alteration in the tumor have longer progression-free survival than those whose treatment is not selected on the basis of genomic analysis,” says Tsimberidou.

Other randomized trials of molecular profiling, including studies organized by the NCI and the Institut Curie in France, have already begun or are in the works, which means researchers may soon have better evidence about its effectiveness in guiding therapy. “We should be encouraged that these studies are beginning,” says Schilsky. “They need to be done.”