Diet drug taranabant, which performed well in early stage clinical trials and appeared on track to be submitted for FDA approval later this year, did not live up to expectations in its Phase III trial and had worse side effects than anticipated, according to an analyst for investment bank Leerink Swann.

Analyst Aileen Salares based her research report on an abstract of data from the two-year Phase III trial that has been submitted in advance of the annual scientific meeting of the American College of Cardiology scheduled for later this month in Chicago.

Taranabant's developer, Merck, said it could not comment on the study until data from the trial is formally presented at the meeting.

But Salares said the 2 milligram dose of the taranabant -- the dose many observers expected to be submitted to the FDA -- produced only a 4 percent weight loss, failing to reach FDA's 5 percent weight-loss threshold for approval of a new diet drug.

She added that the safety profile of the 2 milligram dose of the drug -- which like Sanofi's rimonabant (Acomplia / Zimulti), acts on the CB-1 receptors -- was also "not nearly as clean as expected" with twice as many participants dropping out of the trial because of psychiatric side effects as patients taking a placebo.

The side effects included suicidal thoughts reminiscent of the psychiatric side-effects which led to the decision by an FDA advisory panel last summer to unanimously recommend against approval of rimonabant.

While a higher 4 milligram dose of taranabant also tested in the Phase III trial achieved the desired 5 percent weight loss, Salares said it was associated with a considerably higher incidence of psychiatric side effects than the 2 milligram dose.