While Sanofi-Aventis says patients on antidepressants should not take diet-drug rimonabant (Acomplia / Zimulti), the FDA says data it obtained for 2004-2006 showed that 30 percent of the 580,000 patients prescribed diet pills each year received a concurrent prescription for an antidepressant drug.

The fact that many people who purchase prescription diet drugs appear to also be candidates for antidepressants is a key problem the FDA is grappling with as it decides whether to approve rimonabant, which it has concluded is effective as a weight loss drug but with worrisome psychiatric side-effects.

Sanofi-Aventis, in a briefing paper prepared for the June 13th meeting of an FDA advisory panel that will air these issues, concedes that in its clinical trials "depressive events" and suicidal thoughts were more frequent for the patients on rimonabant.

Sanofi also says it excluded from its Phase III clinical trials patients with a history of severe depression or with severe psychiatric illness, and said antidepressant treatment was not permitted.

But that may only add to concerns, since it would appear to indicate the higher incidence of psychiatric disorders seen in patients taking Acomplia in the clinical trials was occurring among patients without a history of depression. So what happens when patients on antidepressants start taking rimonabant?

Sanofi's answer to that question is "there is limited data in patients taking antidepressant medication in combination with rimonabant; therefore, the use of rimonabant is not recommended in these patients."

That's all well and good, but what is the likelihood that depressed obese patients will not succeed in obtaining rimonabant -- which would be known as Zimulti in the United States -- if it is approved for sale?

Sanofi has an answer for that, too, proposing a Risk Minimization Action
Plan (Risk MAP) to "reduce the possibility of rimonabant use inconsistent with the labeling in patients with diseases, conditions, or concomitant therapy that raise identified or potential safety concerns."

The French pharmaceutical company claims that in Europe, where Acomplia has been on sale for almost a year, fewer than 5 percent of the more than 100,000 patients who have been prescribed rimonabant have been given it in a manner not " consistent with product labeling."

Even if the experts on the FDA advisory panel believe the 5 percent figure, they will need to decide as part of their "risk-benefit" analysis whether finding out what happens when more than 25,000 U.S. patients take rimonabant together with antidepressants is a risk the United States wants to take.