You are here

Reanalysis of 1990s Trial Data Shows Antidepressant Risk Outweighs Benefit

RIAT initiative seeks to resurface unpublished or misreported trial data

A re-review of a trial from the 1990s studying the efficacy of paroxetine and imipramine (Study 329, funded by GlaxoSmithKline [GSK]) concluded that neither paroxetine nor high-dose imipramine showed efficacy for major depression in adolescents and that there was an increase in harm with both drugs.

This re-review is part of an initiative called “restoring invisible and abandoned trials” (RIAT) in which an international group of researchers calls on investigators of abandoned (unpublished) or misreported trials to publish undisclosed outcomes or correct misleading publications. If funders and investigators fail to undertake to correct a trial that has been identified as unpublished or misreported, independent groups were encouraged to publish an accurate representation of the clinical trial based on the relevant regulatory information.

RIAT suggests that access to primary data from trials has important implications for both clinical practice and research, including that published conclusions about efficacy and safety should not be read as authoritative. The reanalysis of Study 329 illustrates the necessity of making primary trial data and protocols available to increase the rigor of the evidence base.

According to the Boston Globe, the new analysis, published in BMJ, confirms the link to suicidality and indicates that the drug wasn’t helpful in the first place. It caps more than a decade of recriminations, lawsuits, and government probes that cast a harsh spotlight on GSK.

The study will increase pressure on the pharmaceutical industry to open its books. “This analysis has provided a window into a fundamental problem with the way in which we evaluate medicines,” said Peter Doshi, an assistant professor of pharmaceutical health services at the University of Maryland and an associate editor at BMJ.

Researchers must be able to replicate results and analyses of clinical trials; if they can’t conduct these kinds of exercises, outstanding questions about the safety or effectiveness of a drug may linger.

“This is not really about any one drug or any one trial or any one condition,” said Dr. David Healy, one of the researchers who analyzed the GSK trial data and a professor of psychiatry at Bangor University in Wales. “It’s really about access to data. Because if you don’t have access, you can’t see how a company can manipulate data and, therefore, miss the harms” a drug may cause.

GSK did not dispute the new analysis. Instead, the company maintains that the findings “appear to be in line with the long-standing view” that paroxetine  and similar antidepressants cause an increased risk in suicidal behavior in youngsters, according to a company spokeswoman. “This is widely known” and has appeared on product labeling for years, she wrote in an email.

To its credit, GSK provided the researchers with some 77,000 pages of detailed patient data that was contained in case report forms. In response to a settlement with the U.S. government, the drug maker established a mechanism for researchers to seek access to its trial data and convinced several other companies to join its effort. But while some other drug makers have created access schemes, they’ve done so in varying degrees because these are voluntary efforts. Not all companies, for instance, are willing to make older data available, given the logistics and expense involved in digging up files. And in general, the pharmaceutical industry still balks at releasing data that it says would compromise trade secrets.

A key question is the extent to which dissecting the Paxil trial data will spark similar efforts by other researchers. Although there is growing interest in re-examining scientific research (especially as medical journals make more retractions),  the Paxil study required nearly two years of effort on the part of a large team of unpaid researchers.

Even if a new look at old data simply confirms original results, doctors and scientists must be able to show that the data can stand up to scrutiny.

“I don’t think it’s as grim as it seems, but it’s no small task” to attempt such an undertaking, said Dr. John Nardo, a retired psychiatrist in Jasper, Georgia, who was one of the authors of the BMJ paper. “I guess every little bit counts, and I hope this is enough impetus to get a few more little bits out there.”

Sources: BMJ; September 16, 2015; Boston Globe; September 20, 2015.

Recent Headlines

Despite older, sicker patients, mortality rate fell by a third in 10 years
Study finds fewer than half of trials followed the law
WHO to meet tomorrow to decide on international public heath emergency declaration
Study of posted prices finds wild variations and missing data
Potential contamination could lead to supply chain disruptions
Declining lung cancer mortality helped fuel the progress
Kinase inhibitor targets tumors with a PDGFRA exon 18 mutation
Delayed surgery reduces benefits; premature surgery raises risks
Mortality nearly doubled when patients stopped using their drugs