From SCIENTIFIC AMERICAN
BY DIANE MARTINDALE
Last fall a clinical trial got under way in the U.S. and Europe to test a hugely touted vaccine designed to reverse the course of Alzheimer's disease. In February that landmark trial came to an abrupt end after 15 patients fell ill with brain inflammation. The vaccine's maker, the Irish drug company Elan, has stopped giving the shots to its 360 volunteers while doctors determine what caused this serious side effect. Despite the setback, proponents of the vaccine still believe that the immune system can be taught to fight Alzheimer's--even if they aren't certain how the vaccine works.
The Elan vaccine, referred to as AN-1792, is a synthetic version of the beta amyloid protein. In Alzheimer's the protein becomes insoluble and accumulates as whitish plaques in the brain, thereby leading to nerve cell damage and dementia.
In 1999 Dale B. Schenk, now a vice president at Elan, announced that experiments in mice with AN-1792 suggested that it could halt and perhaps even cure Alzheimer's. Since then, it has been shown that vaccinated mice genetically engineered to develop the human disease make antibodies that not only prevent the sticky protein from accumulating in the brain but also help to clear existing amyloid plaques.
Just how antibodies remove plaques remains unknown. The prevalent theory has been that antibodies against beta-amyloid are able to cross the blood-brain barrier. Then the antibody forms a complex with beta-amyloid that triggers microglia--support cells in the brain--to destroy the plaques.
The problem with this explanation is that antibodies are typically too large to cross the blood-brain barrier. So it is unlikely that this is the main mechanism by which the vaccine reduces amyloid pathology, notes David M. Holtzman, a neurologist at the Washington University School of Medicine. In his studies with mice, he found that only about 0.05 percent of all the antibodies circulating in the blood are found in the cerebrospinal fluid. "There aren't enough antibodies crossing the blood-brain barrier to activate microglia," Holtzman says.
New research indicates that antibodies don't need to go into the brain to clear amyloid plaques. Holtzman and others have strong evidence for the "sink" hypothesis, whereby antibodies bind beta-amyloid in the blood, outside the central nervous system. The blood then serves as a sink for beta-amyloid, pulling it from the brain and shifting the equilibrium in favor of the soluble protein over the plaque-associated insoluble form.
Holtzman's group found that after immunization, Alzheimer's mice had 1,000-fold more beta-amyloid in their blood compared with those that did not receive the vaccine. This surge in amyloid appears to be coming from the brain, Holtzman explains. And treated mice had far fewer amyloid plaques after five months than the control animals.
Cynthia A. Lemere, a neuropathologist at the Center for Neurologic Diseases at Brigham and Women's Hospital in Boston, has seen a similar sink effect in mice immunized with her beta-amyloid nasal vaccine, which is set to go into clinical trials at the end of this year. Most recently, Lemere discovered antibody-amyloid complexes in the spleen of vaccinated animals, indicating that antibodies bound to amyloid in the blood are being processed in the same way as any other antibody-bound protein. "This gives the sink phenomenon a big boost, but I'm not convinced it's the only mechanism," she says.
Indeed, the side effect seen in the Elan trial supports the idea that microglia are activated by some antibodies crossing into the brain. "The first reaction I always see after I vaccinate the mice is inflammation in the brain," notes neuroscientist David Morgan of the University of South Florida. "Patients in the Elan trial probably had a similar reaction." The inflammation does not kill the mice and eventually subsides as the microglia stop reacting to continued exposure to the vaccine. The Elan patients may have been fine had they been given enough time to let the swelling subside, he speculates.
Morgan's theory, though, won't be put to the test now that Elan has abandoned the trial. Patients who received AN-1792 will continue to be monitored to watch for additional side effects and to see if the injections had any benefit against Alzheimer's. The company has several other variations of the AN-1792 vaccine in the pipeline, and it hopes to test them in patients soon. Still, even if these vaccines don't cause side effects the way AN-1792 has, many questions will have to be answered before an Alzheimer's vaccine is approved. The most alarming is whether the vaccine itself might cause neurodegeneration; beta-amyloid tends to stick together and might cause a "seeding" effect. "We haven't seen this side effect in the mice," Holtzman says, "but no one knows for sure."