A comparative study of 18F‐FDG PET brain metabolism and cognitive performance between normal controls and amnestic Alzheimer’s disease patients

Abstract

AbstractBackground18F‐fluorodeoxyglucose Positron Emission Tomography (18F‐FDG PET) allows the measurement of brain metabolism in vivo. In neurodegenerative diseases, hypometabolism´s topography patterns usually correlate with certain types of cognitive dysfunction. In the amnestic presentation of Alzheimer’s Disease (AD), the temporoparietal hypometabolism is the typical 18F‐FDG PET signature. However, much less is known about its spatial correlation with other cognitive manifestations typically seen in the clinical setting. Objectives: To compare the distribution of 18F‐FDG PET between Normal Controls (NC), patients with Mild Cognitive impairment due to AD (MCI‐AD), and patients with AD Dementia (ADD) and to analyze each cognitive test as a covariate.MethodCross‐sectional study, including 14 NC (13 PET‐PIB ‐, 1 PET‐PIB +), 25 MCI, and 28 ADD patients, (all PET‐PIB +) matched by age and education. Images were normalized to a template using non‐linear methods. The metabolic differences between the 3 groups were evaluated by a voxel‐wise analysis based on Statistical Parametric Mapping. Statistical models were applied using neuropsychological tests as covariates, and VOIS statistics were generated using the Boogle‐DKT atlas.ResultDifferences were observed between the groups analyzed in middle and inferior temporal, posterior parietal, posterior cingulate, and precuneus cortical regions, among other areas of interest (p <0.005). Some of these areas correlated with a worse performance in verbal episodic memory tests (left temporoparietal regions), visuospatial tests (right parietal regions) and phonological fluency tests (left frontal and temporoparietal regions), among other results (p<0.005).ConclusionIn this study, 18F‐FDG PET voxel‐wise analysis showed metabolic differences between the groups included, following the severity spectrum of each amnestic AD stage considered. Although these differences correlated with some cortical areas usually reported as the “AD metabolic signature”, some cortical regions were associated with cognitive functions that weren´t previously reported, allowing a better understanding of the cortical networks involved in the clinical manifestations of AD.