Imaging Uncovers Brain Changes Long Before Alzheimer's Diagnosis

The conversion to t he Alzheimer's disease process is known to involve a decline in the level of beta amyloid in cerebrospinal fluid (which corresponds to a rising level of beta amyloid in the brain), a decline in brain glucose metabolism, and shrinkage of the brain's memory center, the hippocampus.

However, few studies have examined the dynamic change of two or three of these biomarkers simultaneously. Yet such a comparison is crucial to understanding their roles in the Alzheimer's conversion process.

Raymond Lo, M.D., a neurologist and doctoral candidate in epidemiology at the University of California, Berkeley, and his team decided to undertake such a comparison in a large study called the Alzheimer's Disease Neuroimaging Initiative. It took place at 59 study sites and included 818 subjects aged 55 to 90 with either normal cognition, mild cognitive impairment, or Alzheimer's.

The subjects were followed for up to three years to track longitudinal changes in the three biomarkers of interest—beta amyloid in cerebrospinal fluid, brain glucose metabolism, and the hippocampus. After the baseline visit, subsequent visits occurred at six- or 12-month intervals. The relationship between the biomarkers and cognitive change was evaluated. At baseline, subjects were genotyped to see whether they carried the Alzheimer's risk gene variant APOE-e4, so that the researchers could see whether having the APOE-e4 variant influenced the rate of change for each of the three biomarkers.

Cerebrospinal fluid levels of beta amyloid appeared to decrease a little faster in subjects with normal cognition than in subjects with mild cognitive impairment or Alzheimer's, Lo and his team reported online June 13 in the Archives of Neurology. This was somewhat surprising to them, Lo told Psychiatric News. "We didn't know why this happened," he said. "It could be part of the aging or disease process."

In contrast, brain glucose metabolism appeared to decrease least in subjects with normal cognition, somewhat more in subjects with mild cognitive impairment, and most in subjects with Alzheimer's, and the hippocampus seemed to shrink least in subjects with normal cognition, somewhat more in subjects with mild cognitive impairment, and most in subjects with Alzheimer's.

Furthermore, for subjects with mild cognitive decline, change in all three biomarkers was related to cognitive decline. And as for subjects diagnosed with Alzheimer's, change in glucose metabolism and the hippocampus, but not in beta amyloid levels, was related to cognitive decline.

"These findings imply that the cerebrospinal fluid level of beta amyloid declines prior to the onset of cognitive impairment in relation to aging or preclinical Alzheimer's, whereas ... glucose metabolism and hippocampal atrophy change with disease severity and stage," Lo and his team concluded.

Commenting on the study in an accompanying editorial, Roger Rosenberg, M.D., editor of the Archives of Neurology, said, "Their data provide compelling evidence that amyloid deposition [in the brain as reflected by low beta amyloid levels in cerebrospinal fluid] occurs early in the disease process, before hypometabolism or hippocampal atrophy."

The study also shed light on how the APOE-e4 gene variant contributes to the Alzheimer's process. Having that variant did not seem to modify the rate of change in cerebrospinal fluid levels of beta amyloid or glucose metabolism in any of the three groups of subjects. But it did appear to accelerate hippocampal atrophy in subjects with mild cognitive impairment and in subjects with Alzheimer's.

These findings have important clinical implications, Rosenberg said in his editorial. "They emphasize that a narrow but vital window of opportunity exists for preventing Alzheimer's," and that is when individuals are still cognitively normal, but where beta amyloid is already starting to build up in their brains. The many compounds being developed to prevent Alzheimer's should be tested at this stage, he said.

Rosenberg also suggested that the reason why the Elan Pharmaceutical trial between 2000 and 2002 to see whether immunization with a vaccine composed of beta amyloid could successfully treat Alzheimer's failed may have been because the vaccine was given too late in the disease process. Even though the vaccine was shown to reduce brain levels of beta amyloid in subjects, he said, the subjects still progressed to "cognitive loss, and all developed severe and unremitting dementia."

The study was funded by the National Institutes of Health, several pharmaceutical companies, the Alzheimer's Association, the Alzheimer's Drug Discovery Foundation, and the Dana Foundation.