July 30 (Bloomberg) -- An experimental Alzheimer's drug made by Wyeth and Elan Corp. was linked to a brain-swelling side effect in a study that showed no benefit for the majority of patients. Elan fell 37 percent and Wyeth dropped 12 percent in after-market trading yesterday.

The treatment, called bapineuzumab, slowed memory loss better than existing treatments for some patients screened with a genetic test. It didn't help patients who carried a gene known as ApoE4, a group that makes up about half of Alzheimer's patients. The research was reported yesterday at the International Conference on Alzheimer's Disease in Chicago.

Bapineuzumab is the first treatment shown to help patients by removing spaghetti-like clumps of protein that accumulate in their brains. To get that result, Elan and Wyeth had to exclude 65 percent of the patients in the study, those who had the gene. Of the remaining patients, 47 patients received the drug and 32 were given a placebo.

``Whenever you go back and look retrospectively, you find things that could be a mirage,'' said Michael Krensavage, founder of Krensavage Asset Management in New York. He said the number of patients in the final analysis was too small to make a clear judgment on bapineuzumab's effectiveness. ``The drug very well may work, but there isn't the data to prove that.''

Both groups tested with the drug reported serious side effects, among them a brain-swelling condition called vasogenic edema. Patients receiving higher doses also didn't record stronger benefits, researchers said at the meeting. Usually a drug that works will have stronger effects in higher doses, Krensavage said.

Shares Tumble

Wyeth, based in Madison, New Jersey, dropped $5.61 to $39.50 at 7:49 p.m. in extended trading on the New York Stock Exchange. Elan's American depositary receipts tumbled 37 percent to $21.35.

Within its limitations, the trial offers hope to the world's 18 million Alzheimer's sufferers, whose only available medicines temporarily treat the symptoms of the disease and not the underlying condition.

Patients without the gene improved 5 points on a 70-point test widely used in Alzheimer's drug trials. The test measures memory, language use, and attention. The leading Alzheimer's drug, Pfizer Inc.'s Aricept, improved performance an average 2.5 points in previous tests, said Ronald Black, assistant vice president of neuroscience at Wyeth.

Sales Estimates

If approved, bapineuzumab may reach yearly sales of as much as $8 billion by 2015, Cowen & Co. analyst Steven R. Scala in Boston said in a note to clients before the announcement. Wyeth and Elan, of Dublin, equally share the profits and expenses of the drug, he said.

Twelve cases of vasogenic edema were seen in the trial of 234 patients, 10 of which occurred in those carrying the gene. Both cases in the non-carriers were in the highest-dose group, as were eight of those with the genetic variant. Six of the patients went back on the treatment after the edema disappeared.

``Most of the edemas were detected on our routine MRIs, and didn't have symptoms,'' said Gordon Francis, the senior vice president of global clinical development for Elan. There were also small bleeds in the brains of ``three or four patients, with no overt hemorrhage,'' he said.

About 12 percent of patients had back pain, 11 percent had anxiety and 10 percent had vomiting, the study found. Three deaths among patients taking the drug were not considered related to the treatment, the researchers said.

Treatment Procedure

The drug is given as an intravenous infusion over the course of an hour, about four times a year, Black said. Six doses were given in the study. About 95 percent of the patients in the study took another drug as well, such as Pfizer's Aricept or Forest Laboratories Inc.'s Namenda, Black said.

Bapineuzumab didn't show a statistically valid benefit for patients under the original design of the study. The company separated the patients with the genetic marker and changed the study to find patients who responded, Black said.

The number of people with Alzheimer's will almost double to 34 million by 2025, according to the World Health Organization. In the U.S., about one in eight baby boomers, the generation born from 1946 to 1964, will get dementia from the disorder, according to the Alzheimer's Association in Chicago.

Elan and Wyeth are competing with Eli Lilly & Co. to develop the first drugs that reduce the buildup of protein strands, known as beta amyloid, found in the brains of Alzheimer's patients.

Lilly Results

Indianapolis-based Lilly also will present results from a similar test in patients today at the same meeting in Chicago. Both drugs are in final tests before regulatory approval.

Tests have yet to show whether preventing or removing amyloid plaque will slow Alzheimer's. The strategy failed in another study presented yesterday. Patients taking Flurizan, made by Myriad Genetics Inc. and H. Lundbeck A/S, had a similar decline in mental and physical function at the end of the 18- month clinical trial as those receiving a fake drug. The drug triggered more anemia, infections and ulcers.

The companies halted development of Flurizan when the initial results were released in June. The study, the largest comparing an active control to a placebo for Alzheimer's, was in the final stage of testing needed for regulatory approval.

The drug, similar to the one being developed by Lilly, takes a different approach than bapineuzumab. It prevents the production of toxic forms of amyloid by blocking a critical enzyme, called gamma secretase.

Other Therapies

``Just because the Flurizan Phase III clinical trial failed doesn't mean that other amyloid-targeted therapies in the clinical trial pipeline aren't valid,'' said Samuel Gandy, chair of the Alzheimer's Association's medical and scientific advisory council and associate director of the Mount Sinai Alzheimer's Disease Research Center in New York.

Wyeth needs the drug to replace earnings from its heartburn pill Protonix and antidepressant Effexor, which generate a quarter of revenue and face generic competition in two years. Wyeth says it has more than 350 scientists working on Alzheimer's and has spent more than $500 million on it since 2001.

Elan has cut costs to boost investment in Alzheimer's, which consumes half the company's $335 million annual research budget.

To contact the reporters on this story: Tom Randall in New York at trandall6@bloomberg.net; Elizabeth Lopatto in Chicago at elopatto@bloomberg.net.

Last Updated: July 30, 2008 00:01 EDT