Abstract
1458
Objectives: To evaluate relationship between cerebral amyloid deposition, glucose metabolism and neurocognitive function in elderly with normal cognition (healthy control, HC), mild cognitive impairment (MCI) and Alzheimer’s disease (AD) using observer-independent 3D-SSP analysis of [F-18]fluorodeoxyglucose (FDG) and [F-18] florbetapir brain PET using Thai population normal database (Thai NDB).
Methods: [F-18]FDG (FDG) and [F-18]florbetapir (AMY) PET were performed in 78 subjects (20 HCs, 27 MCIs and 31 ADs) within 6-week from neurocognitive assessment. AMY and FDG PET images were obtained with the ADNI2 protocol. FDG and AMY PET images were coregistered and transformed stereotactically using NEUROSTAT (University of Utah). Brain PET image sets from HCs were used to create Thai NDB. 3D-SSP extracted cortical activities were compared to the Thai NDB by generating z-score maps. Global and regional cerebral AMY and FDG uptake were normalized by cerebellar and pontine activities, respectively. Regional and global correlations between image and cognitive data were examined. Chi-square tests and one-way ANOVA followed by the post hoc test were applied for statistical comparison between groups. Spearman’s correlation between cerebral amyloid deposition, glucose metabolism and MMSE scores were assessed. Statistical significance was defined as p-value of below 0.05.
Results: The automatic image processing of AMY and FDG PET datasets were successful in 77 subjects (98.7%). Significant negative correlation was observed between AMY and FDG uptake at all cerebral cortices (frontal>temporal>parietal) except sensorimotor and primary visual cortices. There was statistically significant difference in AMY SUVr between HC vs AD and MCI vs AD for all cerebral cortices (p<0.05 adjusted for multiple comparisons), but only at anterior cingulate cortex between HC vs MCI. For FDG, there was significant difference in SUVr between HC and AD, MCI and AD and HC vs MCI for all regions except anterior cingulate, occipital and primary visual cortices where significant difference was only observed between HC vs AD and MCI vs AD. Significant negative correlation between AMY uptake and MMSE score and significant positive correlation between FDG uptake and MMSE score were observed (p<0.05). In MCI, the correlation between AMY and MMSE score was found only at anterior cingulate cortex, while correlation between FDG and MMSE score was found at frontal, temporal, anterior cingulate, posterior cingulate and primary visual cortices. In ADs, the correlation between AMY and MMSE score was noted at occipital and primary visual cortices while FDG uptake was correlated with MMSE score at all regions except sensorimotor cortex. In 3D-SSP analysis, AMY showed stronger and more extensive regional correlation with MMSE scores than FDG, particularly at sensorimotor and medial occipital cortices where no correlation between FDG uptake and MMSE scores was noted. In MCIs, AMY showed weak correlation with MMSE scores at all cerebral cortical regions, while FDG showed no correlation. However, FDG showed stronger correlation with MMSE scores than AMY at precuneus, posterior cingulate, lateral parietal, temporal and lateral occipital cortices in ADs. AMY and FDG uptake had no correlation with MMSE scores in HCs.
Conclusions: Amyloid deposition had negative correlation with glucose metabolism in most regions of cerebral cortex except sensorimotor cortex. Amyloid burden and cerebral hypometabolism were significantly higher in AD group than those in other groups. Amyloid deposition was strongly correlated with cognitive function at relatively early stage of AD, while cerebral glucose metabolism better described changes in cognitive function at relatively later stage of AD.