Several years ago scientists at Elan, an Ireland-based drug company, and at U.S.-based Wyeth Pharmaceuticals developed a vaccine that showed promise in slowing the advance of Alzheimer's disease. The approach was to expose patients to a tiny amount of beta amyloid—the rogue protein thought to trigger the sticky plaques that accumulate in the brain. The exposure would prompt the body's immune system to raise its own disease-fighting antibodies to destroy the protein. But in January 2002, months into the clinical trials, it became apparent that serious brain swelling had developed in about 6 percent of patients. The trial was halted, hopes were dashed, and the researchers went back to square one.

Now investigators are recruiting people for a new trial. This time they will deliver the antibody itself to patients who have mild to moderate stages of the disease. Giving the antibody directly should not activate an immune response, says Dale Schenk, Elan's chief scientific officer. A faulty immune response is what causes dangerous swelling. And “it doesn't take much antibody” to see if the new approach is working, Schenk says.

Other scientists have confirmed that the antibodies protect against plaque buildup. Schenk says subjects in the earlier trial who showed an elevated antibody response performed significantly better on memory tests. Evidence from autopsied brains of some of those who died also indicated reduced plaque formation.

The new study will test different doses for safety. And scientists should be able to tell if the antibodies alleviate some of Alzheimer's devastating mindrobbing symptoms. Eli Lilly and others are also working on antibody treatments, but none have reached patients yet. Of the few medicines federally approved to treat Alzheimer's, most improve symptoms temporarily by boosting a brain chemical that is key to memory and learning.