There are no presymptomatic diagnostic tests for Alzheimer's disease (AD) thus limiting experimentation with any potential disease prevention strategy. The goal of this 5-year longitudinal project is to evaluate whether brain amyloid imaging using positron emission tomography (PIB-PET) and CSF tau biomarkers that are biologically valid for the brain changes of AD: a) predict cognitive decline in cognitively normal (NL) subjects;and b) confirm and improve predictions of cognitive decline over hippocampus (HIP) glucose metabolism (MRglc) scanning with FDG-PET. In the current cycle we created and validated the automated HipMask image analysis procedure to sample HIP and other regional MRglc on FDG-PET. HIP MRglc sampling accurately predicted the conversion of NL to MCI and to autopsy confirmed AD. However, MRglc measures lack pathological specificity for AD. The definite diagnosis of """"""""brain AD"""""""" requires detection of neurofibrillary tangles (tauopathy) and amyloid beta (A2) plaques. At post mortem, AD lesions in NL subjects are associated with mild memory impairments, but the risk for the future cognitive symptoms cannot be known. We also initiated longitudinal CSF AD-biomarker studies and extended HipMask to sampling the PET PIB scan used for imaging amyloid beta (A2). Our preliminary studies suggest that CSF biomarkers and PIB-PET imaging can bridge the gap between preclinical data and pathological confirmation. Our data show that elevated levels of CSF P-tau231 (reflecting tauopathy) are: 1) associated with reduced entorhinal cortex MRglc in NL ApoE 54 carriers with subjective memory complaints;2) predict decline from MCI to AD;3) are exclusively brain derived;and 4) are diagnostically specific for AD. Our pilot PIB-PET studies show accentuated prefrontal cortex A2 pathology in both MCI and AD relative to control. Our plan is to test whether the biologically specific CSF P-tau231 and PIB-PET biomarkers contribute to the non-specific HIP MRglc in the prediction of cognitive decline in NL. We will study 110 NL subjects (65-80 yrs). To observe adequate numbers showing declining memory and MCI, 80 individuals with subjective memory complaints will also have a history of a first-degree relative with AD after age 65. This confers a 30% aggregate risk of decline over 3-years. We will conduct three clinical exams at 18-month intervals. FDG and PIB-PET scans and CSF will be obtained twice, at baseline and after 3-years. We will test three major hypotheses: 1) among the imaging and CSF measures, the superior predictors are: HIP MRglc (FDG-PET), P-tau231 (CSF), and prefrontal PIB uptake (PIB-PET);2) P-tau231 (tauopathy) and prefrontal PIB (A2) measures both contribute to the HIP MRglc prediction of outcome;3) CSF and imaging biomarkers can be staged as having diagnostic value in early (presymptomatic) or in later (symptomatic) clinical stages. We will use standardized and quality controlled protocols and there is adequate power for hypothesis testing.
Future prevention trials for Alzheimer's disease will require accurate identification of normal individuals at increased risk for cognitive decline. Currently, FDG-PET is the best tool for the prediction of AD in presymptomatic subjects;however, this modality is not specific for """"""""brain AD"""""""". This project will test the hypothesis that validated biological markers for neurofibrillary tangle and amyloid plaque pathology contribute to FDG-PET in predicting future cognitive decline in cognitively normal subjects.
Showing the most recent 10 out of 109 publications
