LURKING BENEATH some of the biggest prescription drug scandals of our time is the specter of unflattering data that was known to pharmaceutical companies and could have revealed problems sooner had it been made public. Now, an international federation of drug manufacturers, whose members include the industry’s main players, has pledged to submit most clinical trial results for publication in peer-reviewed journals, whether the outcome is positive or negative.
This step should inject transparency into the now-murky process of data disclosure. But it will work only if there is a concerted effort by companies, journals, and regulators to ensure that all trials are published, and in a timely manner. Disclosure of unflattering data doesn’t just provide a fuller picture of how useful and safe a drug might be in treating a given medical condition; it also provides some guidance to researchers studying similar drugs and similar illnesses.
Until now, the industry’s track record has left much to be desired. Currently, companies seeking approval for new drugs must supply basic information about all their completed studies with ClinicalTrials.gov, a government website. But fewer than half of the studies registered there are published in journals. One-quarter to one-half of clinical trials for drugs that gained Food and Drug Administration approval remain unpublished several years later. A much higher proportion of the unpublished studies turn out to have unfavorable results for the marketed treatment.
In line with the new policy, adopted in June by the trade associations for drug manufacturers in Europe, Japan, and the United States, companies have agreed to submit results from “phase III’’ clinical trials (the most expensive kind, in which a drug is tested in large groups of people for safety and efficacy), as well as other smaller but medically relevant trials, to peer-reviewed journals. The commitment is to submit studies within 12 months, and no later than 18 months, of a drug either being approved or discontinued from development.
The success of this plan depends on the willingness of peer-reviewed journals to seek and accept more submissions of the not-so-sexy type — studies showing no difference between treatments. It isn’t unheard of: according to The New England Journal of Medicine, about 30 percent of the studies it published in 2008 and 2009 were negative. If established journals are unable to publish all this new data, it’s possible that new peer-reviewed forums will emerge, either in print or on the web.
Also, the FDA should take a stronger stance against companies that avoid publication of unflattering results. It should be more vigilant, too: More than one study has found a surprising number of clinical trials whose results, as registered with ClinicalTrials.gov or as sent to the FDA, did not fully match what was later published in a journal; the statistics looked better, or certain failures were omitted, in the published version. The FDA should do a better job at policing companies’ publication records, especially now that the industry has made a formal pledge to publish.