An experimental Alzheimer's drug slowed progression of the disease in a small study, researchers reported Friday, offering a glimmer of hope after a string of failures.

In an early-stage trial of 166 patients, the drug reversed buildup of beta amyloid in the brain and reduced cognitive decline, with higher doses and longer treatment resulting in more improvements, according to results presented at a scientific meeting in Nice, France.

Beta amyloid, a protein fragment that creates plaque in the brain, was observed with PET scans. Cognition was measured with tests used by neurologists.

"Either result alone would have been exciting," said Alfred Sandrock, chief medical officer for Biogen Idec Inc., the Swiss company that is developing the drug.

The results for aducanumab, previously known as BIIB037, are positive enough that Biogen is skipping midstage testing and moving directly into final-stage trials that will begin this year, Sandrock said. The company's shares rose to record highs on Friday's news.

Biogen is pushing rapidly ahead into an area where other drugmakers have tried and fallen short in final-stage trials.

Alzheimer's is the most common form of dementia, affecting more than five million Americans, according to the Alzheimer's Association. The disease is progressive, causing symptoms that include memory loss, confusion, and behavior changes.

With no drug to cure Alzheimer's or lessen its advance available for sale, a working treatment offers a huge opportunity for drugmakers.

Biogen's trial may have avoided pitfalls that tripped up other drugmakers because it focused on patients in the early stage of Alzheimer's disease, and made sure they had beta amyloid in the brain to verify they had Alzheimer's, Sandrock said.

After a year at the highest dose of treatment, PET scans showed that plaque was reduced to a level that was "very close to the upper limit of normal," he said. At the lowest dose, the reduction wasn't statistically significant.

On cognition, based on what's called the Clinical Dementia Rating, patients in the placebo group worsened by an average 2.04 points after one year, compared with 1.7 points among patients who took the lowest dose and 0.59 points for the highest dose.

"We are excited about it," said Keith Fargo, director of scientific programs and outreach at the Alzheimer's Association.

But the study was small. "We have to do the big phase 3 study to see if it's actually going to work. There have been a lot of promising phase 1 and phase 2 trials that didn't work out," Fargo said.

The drug had side effects, notably brain swelling known as amyloid-related imaging abnormalities, or ARIA, that can cause headaches. Among patients carrying the ApoE4 gene, which is strongly linked with Alzheimer's, 55 percent of patients taking the trial's highest dose reported ARIA, compared with 5 percent among those in the lowest-dose group. That resulted in 35 percent of the ApoE4 carriers on the highest dose discontinuing treatment.

Colin Masters, a laureate professor at the University of Melbourne who has studied Alzheimer's disease for more than 30 years, attended the conference in France, where the results were presented. "It's a major advance in confirming amyloid beta is the right target," he said by phone.