(Corrects by moving and reworking material previously in sixth paragraph to third paragraph, showing 19 percent improvement related to effectiveness goal, not bleeding events, and also met primary goal) (Adds further doctor commentary in paragraphs 10-11)

By Kim Dixon

ORLANDO, Fla., Nov 4 (Reuters) - Eli Lilly and Co's <LLY.N> investigational anti-clotting drug was better able to prevent heart attacks than the standard treatment Plavix, but the drug led to excess bleeding, a study released on Sunday found, a shortcoming that could curb its usefulness.

The widely anticipated findings of Lilly and Daiichi Sankyo Co's <4568.T> bid to take on Bristol-Myers Squibb's <BMY.N> Plavix also found the drug, called prasugrel, was significantly more effective in limiting death from cardiovascular events, nonfatal heart attack and strokes.

That 19 percent improvement met the trial's primary goal of improved effectiveness.

But the 32 percent increase in major hemorrhage seen with the drug, tested in heart patients with symptoms from severe chest pain to heart attacks, could hobble its chances for regulatory approval and hoped-for blockbuster sales, according to analysts.

"This is going to be a very troubling finding with the FDA (U.S. Food and Drug Administration)," said Gordon Tomaselli, a professor of medicine at Johns Hopkins Medicine. "I suspect they'll need more trials to address this."

The 13,600-patient trial found 2.4 percent of patients on prasugrel suffered at least one major bleeding event, compared with 1.8 percent of patients on Plavix, which Bristol markets with Sanofi-Aventis <SASY.PA>.

Results were published in a special issue of the New England Journal of Medicine (NEJM), timed for release at the American Heart Association's annual meeting in Orlando, Florida.

There was no statistical difference in overall mortality seen with the drug.

Analysts had been fretting that prasugrel would cause more bleeding in the trial, called Triton, especially after Lilly in recent weeks stopped giving the drug to patients in two small trials without fully explaining its reasons.

Noted cardiologist Eric Topol, former cardiology chief at the Cleveland Clinic, was charged with discussing the study following the AHA presentation, noted some shortcomings.

For example, the patients had their anatomy screened prior to entry. In addition, the study's definition of bleeding doesn't incorporate all bleeding events.

Patients in the trial were "artificial in terms of practice," and "very different from the real world," Topol said, saying the drug proved "incremental efficacy."

SEEKING TO CHALLENGE PLAVIX

Lilly and Daiichi have been developing prasugrel in hopes it will become an alternative to Plavix, a drug with a combined $6.5 billion in annual sales for Bristol and Sanofi that makes it one of the world's top-selling medicines.

Both Plavix and prasugrel prevent clots by keeping blood cells called platelets from sticking together.

Lilly said they are still on track to file an application to the FDA by the end of the year.

One of the trial's investigators downplayed the bleeding risk, by pointing to an analysis of subgroups of patients who may be harmed from the drug, including those who have had a previous stroke and the elderly.

"I believe we've isolated where that (risk) is coming from," said Elliot Antman, professor of medicine at Harvard University said.

These groups, some of which may do better on a lower dose, made up about 20 percent of the study's patient population.

However, this type of post-study analysis usually carries much less weight with regulators than statistical trial results based on previously specified study goals.

Deepak Bhatt, who wrote an editorial accompanying the study in the NEJM, said the study is important because it validates the "more is better" approach to preventing blood clots in patients with heart disease.

"This is the first validation that more powerful oral anti-platelet drugs further reduce coronary complications," Bhatt said.

Prasugrel proved significantly better at limiting death from cardiovascular causes, nonfatal heart attacks or strokes. About 9.9 percent of prasugrel patients suffered one of those events, compared with 12.1 percent among the Plavix group.

Patients on prasugrel also had significant reductions in heart attack, urgent surgeries and blood clots within stents, tiny mesh tubes used to prop open arteries.

(Additional reporting by Ransdell Pierson in New York)

AlertNet news is provided by