Hear those cries for publication of all clinical trial data? They're getting louder, thanks to new research in the British Medical Journal. According to the study from the German Institute for Quality and Efficiency in Health Care, Pfizer's antidepressant reboxetine (Edronax) is "ineffective and potentially harmful''--but no one knew it until now because nearly three-quarters of trial data was never before published.
Edronax, which has been on the European market for a decade, is no better at treating depression than placebo, the researchers found after analyzing the published and unpublished data. The studies that did see the light of day overestimated Edronax's efficacy by 115 percent compared with placebo and 23 percent compared with other antidepressants, the BMJ article states.
Pfizer spokesperson Chris Loder tells Bloomberg that the company is reviewing the BMJ study, but that reboxetine is "an effective treatment option" for patients with depression. FDA never approved the drug, and now European regulators will now need to reconsider their assessment, the Guardian reports.
So what about unpublished data? In 2007, the U.S. Congress decreed that all clinical trials must be registered. Europe lacks similar rules, the new study's authors point out. But even if it had those rules, is simple registration enough to make sure all data is aired? Even in this study, when researchers asked Pfizer for data that hadn't been released, the company at first only handed over "data that from our point of view are suited for a benefit assessment."
Pfizer later turned over most of the rest, the Guardian reports. But as a BMJ editorial points out, getting drugmakers to cough up unpublished data has required "dogged, even heroic, persistence"--even, as in this case--when researchers already know that the data exists. The BMJ is so hyped up about the problem that it's devoting a special issue to the subject next year.