With an eye to advancing ‘personalized medicine’, clinicians have tried to predict who will respond to certain therapies using biomarkers gleaned from tests that probe genomics, proteomics and other branches of biomedicine. But according to a report from the US Institute of Medicine (IOM), such ‘omics-based’ tests require greater regulatory oversight and more transparent data-sharing before they should be allowed to move from the lab to the clinic.
“Nothing short of patient safety and public trust are at stake,” IOM committee chairman Gilbert Omenn, a computational biologist at the University of Michigan Medical School in Ann Arbor, said in a press conference this afternoon.
The report was commissioned by the US National Cancer Institute and the US Food and Drug Administration (FDA) after a genomic test developed by Anil Potti and his colleagues at Duke University in Durham, North Carolina proved to be worthless. The test, reported in and then retracted from Nature Medicine, supposedly assessed the molecular traits of a malignant tumor to determine which chemotherapy would be most effective; it also triggered a storm of lawsuits and resignations. The IOM was tasked with determining how such a faulty test ever got through the regulatory and institutional review board systems.
The centerpiece of today’s nearly 300-page report, called Evolution of Translational Omics, is the finding that institutional review boards are not adequately assessing clinical protocols for diagnostic tests and the FDA is not demanding sufficient evidence of safety and efficacy before allowing the tests to hit the market. The report cites six case-studies of commercially-available genomic and proteomic tests, including two faulty tests for ovarian cancer, Lab-Corp’s OvaSure and Correlogic’s OvaCheck, which were licensed and, in the case of OvaSure, even sold to patients before they were shown to be ineffective. “This is a wake-up call for people doing translational research,” Omenn said. “There’s a big incentive to get tests to patients, but it’s got to be done right.”
The report also calls for public disclosure of all data and algorithms from studies related to genomic tests being developed for clinical settings. “It’s very important to release data and computer code to address issues of reproducibility,” committee member Nathan Price, a biostatician at the Institute for Systems Biology in Seattle, told reporters at the press briefing. “When you start to compare across many different studies, your idea of whether the test is effective gets better.”
Photo courtesy of the Institute of Medicine