Now, we’ve arrived at a point in time where the published results from a number of these articles are being questioned, and the broad consensus is that the raw data from these articles needs to be reanalyzed by independent investigators. It’s not just an academic interest that drives this call for reanalysis. It’s because these drugs remain in wide use based on the efficacy and safety results published in those articles – often certified by the FDA in the drug approval process.
Pharmalot: WSJBy Ed SilvermanJan 29, 2015
With considerable fanfare, the Institute of Medicine released a long-awaited report last month praising the virtues of sharing clinical trial data. This is an important, but also contentious issue because without access to such information it can be virtually impossible for independent researchers to verify results that can lead to improved treatments, better health care and lower costs. Although the report is nothing more than a set of recommendations, the IOM effort is, nonetheless, seen as a needed step toward prodding industry and academia to release detailed data. Many drug makers, in particular, have been reluctant to grant much, if any, access over concerns about relinquishing trade secrets and compromising patient privacy, among other things. But some have taken steps to do so.
There is, however, one area where some experts say the report falls short and underscores ongoing difficulties in sharing trial information. While the IOM offers specific suggestions for releasing data from future trials, the institute did not provide a framework for obtaining retrospective data. This remains an unresolved issue for the pharmaceutical industry, especially in light of various safety scandals and ensuing litigation that revealed data for some products were never fully published or disclosed. “Clinical trials underpin the approval of medicines that are in use today, so if we don’t have access to that older data, how can we verify accuracy and conduct independent analyses?” says Peter Doshi, an assistant professor of pharmaceutical health services at the University of Maryland. “It’s a basic issue surrounding data transparency.”One IOM committee member, Ida Sim, a professor at the University of California, San Francisco, says the report does suggest sharing retrospective data, but on a case-by-case basis and for studies that may influence clinical care. Why? Informed consent must be obtained from people who participated years earlier. Often, researchers have scattered. “We have to be pragmatic and realistic.”
Some say convenience and liability are also factors. “Obviously, it’s less burdensome for companies to only have to do so going forward, especially if there’s a product on the market that’s raising concerns,” says Diana Zuckerman, who heads the National Center for Health Research, a non-profit think tank. “I see it as a way to protect companies from disclosing embarrassing and expensive information.”
Murray Stewart, however, argues progress is being made. As chief medical officer at GlaxoSmithKline, he oversees an effort to disclose data going back 15 years and has convinced nine other drug makers to similarly provide access through a jointly run website. But most will not provide retrospective data. “We’ve shown it can be done,” he says. “But it’s a long journey” getting others on board. The inconsistent approaches underscore a key issue – finding a universal gatekeeper to vet researcher requests and corporate concerns. Harlan Krumholz, a Yale University cardiologist, is trying to offer a template with the Yale Open Data Access project, which last month signed an agreement with Johnson & Johnson to provide data for diagnostic and device products, but only as of 2014. A deal that J&J signed with YODA last year covers drugs and does not have a cut-off date for data requests. So far, J&J has not refused any request, but Krumholz says any J&J refusal would be disclosed on the site. “We are open science advocates,” he says, “and are not interested in giving anyone a pass.”
Meanwhile, some say the FDA has been a hindrance. While European regulators have a policy that allows researchers to make requests, the FDA does not. An FDA spokeswoman says the agency published a notice seeking comment about ways to make trial data available. But that was nearly two years ago. She says reactions are being evaluated and next steps are being considered. “I think FDA should make as much information available as possible, as soon as possible, particularly for approved drugs, including any safety information that might not be in the primary trials,” says Steve Goodman, a Stanford University School of Medicine professor. “What we need to avoid is the situation where the FDA is in possession of information that could materially impact or alter the assessment of the relative benefits and harms of a drug, and that is not easily available to other scientists or the public.” And “information they have on ‘failed’ drugs can often have an impact on marketed drugs or on other new drugs in the same class, so such information can be quite important as well, even though currently little or no information about such drugs can be divulged.”