Refining Models of Amyloid Accumulation in Alzheimer’s Disease

Changes in MMSE and PET uptake over time
Changes in MMSE and PET uptake over time

A new study published in Alzheimer’s and Dementia, the Journal of the Alzheimer’s Association, proposes to stage amyloid PET images using regional information. The research was conducted by Dr. Bernard Hanseeuw, instructor at MGH Gordon Center, and Dr. Keith Johnson, leader of the Aging NeuroImaging Program at the MGH Gordon Center, and the Harvard Aging Brain Study.

Using longitudinal amyloid PET imaging data collected over three years in more than 1,400 participants including clinically normal (CN) older adults and patients with mild cognitive impairment (MCI) or Alzheimer’s dementia (AD), the authors provided in-vivo evidence that amyloid deposits first in neocortex and then in striatum, a subcortical brain structure. This progressive regional involvement from neocortex to striatum had been suspected for long from autopsy data (referred to in the literature as “Thal phases”), but it had never been demonstrated in living humans.

The results of the study showed that regional expansion of amyloid pathology in striatum was predictive of subsequent cognitive decline and progression to Alzheimer’s dementia. Participants with striatal amyloid declined faster than those who only had cortical amyloid. Higher levels of striatal amyloid were also associated with higher levels of tau pathology and hippocampal atrophy, confirming that striatal amyloid was indicative of disease progression.

Amyloid-PET is commonly expressed as a binary measure of cortical deposition (low/high). However, not all individuals with high-cortical amyloid experience rapid cognitive decline.
Using a three-stage PET classification (low cortex/high cortex, low striatum/high striatum) allow a better identification of the most at-risk individuals. Such a staging system could also help preventive trials for selecting normal participants based on their risk of developing the disease in the following years.

View full paper