For more than two decades, researchers have been unable to settle an important question: Can anti-depressant medications stimulate the growth of breast and ovarian cancers?
Some studies pointed to a link, but others did not.
Now a re-examination of the available evidence has cast a new - and disturbing - light on the previous research. Many studies that seemed to absolve the drugs of blame were carried out by researchers with close ties to the pharmaceutical industry, according to a report published this week in the online journal PLoS (Public Library of Science) One.
"I think that's an important piece of information," said the lead author of the paper, Lisa Cosgrove of the Edmond J. Safra Center for Ethics at Harvard University in Boston.
For the review, Dr. Cosgrove and colleagues amassed a total of 61 studies, which included both laboratory and epidemiological research.
They then conducted separate searches on the principal investigators for each of the studies, looking for drug-company connections. "It was a lot of legwork," said Dr. Cosgrove, noting that, in the past, such ties were not always publicly reported.
A clear pattern emerged. "None of the researchers with industry affiliation reported a positive association" between antidepressants and the risk of cancer, she said.
There were more mixed findings among researchers free of corporate ties. "Approximately 43 per cent of researchers without industry affiliation reported a positive association."
A closer examination of these studies - using meta-analysis, which pools the data - suggests the risk is real, but not very large. Women with a history of antidepressant use faced an 11-per-cent increased chance of developing breast and ovarian cancer, compared with those who had not taken these medications. In the case of breast cancer, that means taking antidepressants would raise an average woman's lifetime risk to 13.8 per cent from 12.5 per cent, Dr. Cosgrove said.
The risk is primarily associated with just one class of antidepressants known as selective serotonin reuptake inhibitors, or SSRIs, which includes Prozac (generic name, fluoxetine), Paxil (paroxetine) and Zoloft (sertraline). "All studies but one in the SSRI analysis found a positive association," she said.
Tricyclics, an older class of drugs, don't appear to be as problematic. "There was a slight increase in risk but it was not statistically significant," she said.
"I wouldn't want any woman who is taking an antidepressant to abruptly stop," cautioned Dr. Cosgrove. "She should talk over that decision with her doctor."
Although the drugs appear to boost the odds of getting cancer by only a small amount, she said women should at least be informed of the potential risk. And that risk doesn't come to light in industry-connected studies.
"We are at a point where public trust in biomedicine is rapidly decreasing as a result of conflicts of interest."
However, Dr. Cosgrove doesn't believe there is some covert "Machiavellian" plot by industry to cover up all unfavourable study results. She noted that the vast majority of researchers who receive industry funding consider themselves to be unbiased.
Over time, though, close industry ties can have a "pernicious" effect on the attitudes and actions of doctors and scientists. "Researchers might not be aware of the subtle ways they can be influenced, " she said.