Abstract
1569
Objectives To evaluate the diagnostic potential of glucose metabolism, amyloid deposition measured by PET and CSF biomarkers in monitoring the progression from mild cognitive impairment (MCI) to Alzheimer’s disease (AD) and cognitively normal (NC) to MCI in a longitudinal study.
Methods We investigated 82 Alzheimer’s Disease Neuroimaging Initiative (ADNI) subjects [48 MCI (24 converted to AD) and 34 NC (10 converted to MCI or AD)] for up to 96 months (median= 72 months). While FDG PET was used to measure glucose metabolism, 18F-florbetapir (AV45) and 11C-Pittsburgh Compound-B (PiB) PET focused on amyloid deposition. CSF β-amyloid (ABETA), phosphorylated tau (PTAU), and total tau (TAU) were collected. All preprocessed PET images with structural MRIs were downloaded from ADNI database. All PET scans were spatially normalized to MNI space using MRI and SPM8. 35 regions of interest (ROI) were manually drawing in a high resolution MRI template from VBM8. Standard uptake values ratios (SUVR) to cerebellum were calculated. Statistical ROC analysis was applied to ROI SUVRs and CSF biomarkers with clinical diagnosis.
Results In MCI group, the AUC values for posterior cingulate and posterior precuneus were 0.75 in FDG scans, which is comparable to the CSF Tau and orbital frontal cortex in PIB. None of regions in AV45 scans showed significant difference between converters and non-converters. In NC group, parietal (AUC=0.77) in FDG scans was significantly different between converters and non-converters; AUC values for laterial temporal cortex, posterior precuneus and anterior cingulate were over 0.76 in AV45 scans, while CSF abeta AUC was 0.87.
Conclusions 18F-AV45 and FDG scans were sensitive to detect early stage of disease progression (NC group), and 11C-PiB scans act effectively in distinguishing MCI converters from non-converters. FDG PET and CSF Tau measurements are sensitive to detect the progression from MCI to AD.