Vorapaxar has been dogged by bleeding concerns since January 2011, when a safety committee overseeing the 26,449-patient study said the new type of anti-platelet drug was not appropriate for patients who had suffered a stroke, due to bleeding risk.

But Merck was allowed to press ahead with the trial, called TRA-2P, after discontinuing treatment for patients with a history of stroke. It continued to test patients who were in stable condition after having heart attacks or being diagnosed with clogged leg arteries, to see if vorapaxar could prevent future heart attacks and stroke.

The drugmaker last month provided limited data from the study, saying patients were benefiting from vorapaxar. But it noted they were experiencing excessive bleeding, as was seen in a separate earlier trial called TRACER that involved almost 13,000 patients who derived no protective benefit from vorapaxar.

Merck officials, in an interview, declined to comment on whether the company will seek approval of vorapaxar, given the mixed results of the latest trial and failure of the earlier TRACER study. They also would not say whether Merck might conduct a costly new study assessing only patients with heart-attack histories - the group that was best protected by the drug in the most recent study.

"We are extremely excited by the results," said Francis Plat, Merck's head of cardiovascular clinical research. "This is a positive trial."

But ISI Group analyst Mark Schoenebaum said he continued to have dim hopes for the medicine despite its effectiveness in the study.

"We have zero in our model for it," he said, referring to his sales expectations for vorapaxar. He said the drug, even if approved, would only be used by a "minimal" number of patients because of bleeding risks.

HEART ATTACK PATIENTS BENEFIT MOST

Researchers on Saturday said vorapaxar, when used for an average of 30 months with standard anti-platelet medicines like aspirin and Bristol-Myers Squibb Co's Plavix, met its main trial goal. It did so, they said, by causing a 13 percent reduction in the combined risk of cardiovascular death, heart attack and stroke compared to those receiving standard treatment and a placebo.

"This is the first time we've seen the addition of any new type of anti-platelet drug provide additional long-term protection among stable patients with a history of heart attack," said Dr. David Morrow, a cardiologist from Brigham & Women's Hospital in Boston who led the trial.

He said vorapaxar was not deemed effective in protecting patients that had previous strokes or clogged leg arteries. But he said it was responsible for an impressive 20 percent reduction in combined cardiovascular death, heart attack and stroke among the almost 18,000 patients in the trial who were enrolled after having heart attacks.

"This is the most important finding in the trial - that there is this clinically compelling benefit in this group with previous heart attacks," Morrow said in an interview.

Morrow added that more than 8 million Americans have a history of heart attacks and another 1 million suffer heart attacks each year. "So, it's a broad group of patients."

But he said 2.4 percent of patients in the latest trial with prior strokes experienced brain bleeding, compared with 0.9 percent of those receiving placebos.

"There's clearly a tradeoff here between the protective benefit and the risk of bleeding," Morrow said. Should the drug be approved, he said the challenge for doctors would be to withhold it from patients at likely higher bleeding risk - including those with a history of stroke and those over 75 years of age.

Morrow was to present his full findings Saturday at the annual scientific meeting of the American College of Cardiology in Chicago.

Some 4.2 percent of all patients enrolled in the study who took vorapaxar had moderate or severe bleeding, significantly higher than the 2.5 percent in the placebo group. Roughly 1 percent taking the Merck drug had bleeding in the brain, twice the percentage of those receiving only standard treatment.

Risk of brain bleeding was far less pronounced among patients who had no history of stroke - with an incidence of 0.6 percent for those on vorapaxar, compared with 0.4 percent in the placebo group.

Although the incidence of overall severe bleeding among patients with no stroke history was statistically similar in the vorapaxar and placebo groups, moderate bleeding was twice as frequent in those taking vorapaxar - at 2.7 percent.

Merck's Plat said Merck will continue discussions with the researchers for both vorapaxar studies, as well as outside experts, to assess the drug's potential, and in which populations. But he declined to comment when asked when Merck might signal next steps, if any, for the drug.

ISI Group's Schoenebaum predicted Merck will decide within the next several months whether to continue development.

Merck obtained vorapaxar through its $41 billion acquisition in late 2009 of rival Schering Plough. At the time it was considered the crown jewel of Schering Plough's drug pipeline.

(Editing by Paul Simao)

• A Permanent Economic Emergency