An experimental drug from Bristol-Myers Squibb and Pfizer reduces the risk of stroke by more than half compared with aspirin, with no significant rise in major bleeding, researchers said on Tuesday.

The results boost prospects for apixaban, one of several new anticoagulant pills aiming to displace the old, problematic heart drug warfarin as a stroke preventer for patients with a common heart arrhythmia called atrial fibrillation (AF).

Industry analysts believe the market for products to replace warfarin, which was originally developed as rat poison, could eventually be worth more than $10 billion annually and possibly as much as $20 billion.

German company Boehringer Ingelheim is currently leading the race with Pradaxa, which will be reviewed by U.S. regulatory advisers on September 20.

But heart specialists meeting in Stockholm said a clear winner had yet to emerge. Stuart Connolly of McMaster University in Canada, who presented the apixaban findings to the European Society of Cardiology congress, said the results with apixaban were truly impressive.

Medical experts had been expecting a good result from the 5,600-patient AVERROES study, after the drug's makers halted the Phase III trial in June due to clear evidence that it prevented stroke. The exact results, however, had been keenly awaited.

Connolly said apixaban reduced the rate of strokes by 52 percent compared with aspirin in the patients studied, who were selected because they were unsuited to warfarin treatment. The annual rate of strokes in patients on apixaban was 1.5 percent compared to 3.3 percent for those taking aspirin.

Patients with AF are at risk of blood pooling in an upper chamber of the heart, which can cause clots and stroke.

An estimated 40 to 50 percent of AF patients are either intolerant to warfarin or refuse to take the medicine, leaving them dependent on less-effective aspirin.

The annual rate of major bleeding was not significantly different between patients on apixaban and aspirin, at 1.4 and 1.2 percent respectively, and there was also no significant difference in clinically relevant non-major bleeds.

We didn't expect it to look as clean on bleeding, Connolly said. For these patients, apixaban appears to have a very favorable risk-benefit profile.

The results from the AVERROES trial are only part of the apixaban story, however, since Bristol and Pfizer are also conducting a second trial, called ARISTOTLE, comparing the drug head-to-head with warfarin, for which results are due next year.

Connolly said it now looked highly likely that the ARISTOTLE trial would also be positive.