A new study may provide a clue about what caused the inflammation -- a discovery researchers hope will enable them to figure out how to avoid it.
In the current study, researchers inoculated mice with the same vaccine that had shown such promise in earlier studies. But this time, they used a different mouse model. They key difference was that the brains of the new mice had a condition that more closely mimicked Alzheimer's in humans than the previous mouse model.
When these new mice were inoculated, they developed serious brain hemorrhaging, said Peter M. Mathews, one of the study's investigators and an assistant professor of psychiatry at New York University School of Medicine.
"All of the mouse studies have looked incredibly positive, but when we went to humans it was disastrous," Mathews said. "What this study adds is now we've gone to an animal model that looks more like the human disease and we found this really severe side effect."
The study appears in tomorrow's issue of Science.
A hallmark of Alzheimer's disease is clumps of peptides in the brain. These peptide clumps, called amyloid-beta plaques, are believed to play a key role in causing Alzheimer's, although doctors don't know exactly how they do their damage.
Previous research showed that inoculating mice who had these peptide deposits in their brain jump-started the immune system to produce antibodies, Mathews said.
The antibodies cleared the plaques from the brain, suggesting that the vaccine could slow or event prevent dementia. Further research showed the mice showed improvement in their memory.
"This added a lot of impetus to the idea of using amyloid-beta inoculations as a potential treatment," Mathews said. "Not only did you have a reduction of amyloid-beta plaque, which we think is related to the disease, but we had improved memory, which is clearly related to the disease."
The next step was trials in humans. About 300 people in Europe and the United States received the vaccine. About 17 of them developed the potentially-dangerous brain inflammation.
"It was a real setback," Mathews said. "All of the animals studies had been incredibly successful up until this point."
Researchers immediately stopped giving any further doses of the vaccine in humans, said Bill Thies, vice president for medical and scientific affairs for the Alzheimer's Association in Chicago.
However, the patients who received the vaccine are still being observed by researchers, he said. Some early published reports have shown that the vaccine is continuing to work at least in some patients and that the brain swelling subsided after the doses were stopped.
In this latest study, researchers tried the vaccine on a different mouse model. Mice used in previous experiments had amyloid-beta deposits only around the neurons of their brains. The new mice also had amyloid-beta deposits in the blood vessels in the brain -- a condition that more closely resembles that of human Alzheimer's patients.
"Most Alzheimer's patients have amyloid deposits around the blood vessels, but the mice used in the earlier research didn't take that into account," Mathews said. Amyloid deposits can cause a weakening of the blood vessels, he added.
When the new mice were inoculated with the vaccine, they developed brain hemorrhaging.
Researchers don't yet know exactly what happened in the brains of the 17 people who developed the inflammation after the vaccine. But hemorrhaging could have very likely been a part of the problem, he said.
"Now that we've identified this side effect of the vaccine, we can take a step back, go back to animal studies and figure out how to avoid it," he said.
Thies said the study is interesting, but by no means answers all the questions about what went wrong with the vaccine. This mouse model may be more realistic that the previous mouse model, but it's not perfect, he said.
"There are multiple mouse models for Alzheimer's disease, and none of them are identical to the condition in humans," he said. "While this model may have advantages in some areas, it may have disadvantages in others. We'll probably need multiple mouse models to answer all the questions."
What To Do
For more information, visit the Alzheimers Assocation or the National Institute of Neurological Disorders and Stroke.