(Adds details from Biogen call, analyst comment)
By Julie Steenhuysen
CHICAGO, July 25 (Reuters) - Biogen Inc and partner Eisai Co Ltd said patients with early-stage Alzheimer's disease treated with their experimental drug BAN2401 experienced 30 percent less cognitive decline than those who got a placebo in a highly anticipated midstage trial.
The second-highest dose of the drug also showed some benefit, but did not meet statistical significance, according to data presented on Wednesday at the Alzheimer's Association International Conference in Chicago.
Investors, who had bid up Biogen shares earlier this month after the companies said the drug was effective at the highest dose, were selling off shares after presentation of the details. Biogen shares fell more than 11 percent in extended trading to $339.
Alzheimer's experts, however, were more positive.
"Overall, it's a shot in the arm for the field," said Dr. Ronald Petersen, from the Mayo Clinic in Rochester, Minnesota.
"It rejuvenates some of the enthusiasm for attacking amyloid, that it is possible and may be successful," he said of the theory that removing toxic deposits of the protein beta amyloid from the brain will disrupt Alzheimer's progression.
BAN2401 demonstrated a dose-dependent reduction in amyloid plaques that was statistically significant at all doses tested, researchers said.
There is a desperate need for an Alzheimer's treatment that works after dozens of failures of experimental drugs. The most common form of dementia affects nearly 50 million people worldwide and is expected to rise to more than 131 million by 2050, according to Alzheimer’s Disease International.
The BAN2401 trial involved 856 patients with early Alzheimer's disease who had beta amyloid in their brains as confirmed by brain scans. The highest does of 10 mg/kg bi-weekly - one of five doses tested - was given to 161 patients.
One key concern was that for safety reasons a regulator outside the United States stipulated that carriers of a gene mutation called APOE4 associated with earlier onset of Alzheimer's not receive the highest dose of the drug. These people had been more prone to a brain swelling side effect in earlier trials known as amyloid-related imaging abnormalities-edema, or ARIA-E.
That resulted in 26 people being moved out of the highest dose cohort. Analysts on a Biogen conference call asked if that imbalance might confound the results. Dr. Alfred Sandrock, Biogen's chief medical officer, said a subgroup analysis was underway to look at the issue.
Baird analyst Brian Skorney said the difference resulted in a "huge imbalance."
"We expect a sell-off tomorrow as the irrational exuberance burns off, but would be looking for an opportunity to be buyers if it goes too far," he said of Biogen stock.
The companies have said they are planning large late-stage trials in which they must replicate positive results.
The results, based on traditional statistical methods after 18 months of treatment, followed 12-month results that failed to meet the study's main cognitive decline goal based on a complex predictive Bayesian statistical model.
Dr. Julie Schneider, an Alzheimer's expert with Rush University Medical Center, said more work is needed to understand how much of a clinical effect the drug has.
"We don't know how many went on to get Alzheimer's dementia. I think there's still a lot of questions," she said.
Alzheimer's experts at the briefing for journalists expressed concern about the measurement devised by Eisai - a compilation of conventional cognitive assessment tools called ADCOMS. The Japanese drugmaker felt a new measurement approach was needed to capture early cognitive changes in patients at the beginning stages of Alzheimer's decline.
They also raised concerns that while the drug slowed cognitive decline, such as memory loss, the disease was not arrested and patients continued to experience decline over the 18-month treatment period.
Nevertheless, using a more conventional measure of cognition known as ADAS-cog, the highest dose of the treatment led to a statistically significant 47 percent reduction in cognitive decline at 18 months compared with patients taking a placebo.
"I'm impressed," Dr. Howard Fillit, chief science officer of the Alzheimer's Drug Discovery Foundation, said of the results.
Fillit said a 30 percent reduction using the more traditional ADAS-cog scale, which is a hard target to budge, is considered clinically meaningful.
"This result might be clinically meaningful if it's reproduced," he said. "I'm a skeptic, but I think there's something here." (Reporting by Julie Steenhuysen Editing by Bill Berkrot)