Abstract
598
Objectives: Huntington’s disease (HD) is a hereditary movement disorder with progressive neurodegeneration and functional decline. Using brain network analysis of FDG PET data we have shown the existence of a HD related covariance pattern (HDRP) to describe altered glucose metabolism in presymptomatic (pHD) gene carriers. HDRP network activity is sensitive in revealing the extent of metabolic abnormality in both preclinical and early onset HD subjects. Here we evaluate the temporal evolution of HDRP by identifying a specific metabolic topography in a pHD cohort in a longitudinal study.
Methods: We used PET scans of glucose metabolism and RAC binding from 12 pHD gene-carriers (CAG repeat 42 ± 2; age 47 ± 11 years) at baseline, 18 mo and 44 mo and 11 age-matched controls (age 41 ± 15 years). All images from each subject were spatially realigned and normalized. pHD carriers at each time point were then separately compared with controls using network analysis on a voxel basis (1). Subject scores of each abnormal brain network were computed prospectively at two other time points in the same HD cohort to assess its change over time. Striatal RAC binding was estimated by DVR relative to the occipital cortex.
Results: Network analysis replicated the HDRP dominated by striatal hypometabolism and hypermetabolism in the cerebellum, thalamus, occipital and parietal cortices (2). Each time point had a slightly different HDRP. Subject scores in the pHD cohort discriminated between 3 time points (RMANOVA, p < 0.01); scores increased at 18 mo (p < 0.05) relative to baseline but decreased at 44 mo. Scores at the third time point were still above the baseline. HDRP scores correlated negatively (p < 0.01) with caudate RAC binding (r > 0.79) and estimated years to onset (r > 0.81) at each time point. All subject scores in the pHD group were higher than the controls (p < 0.05).
Conclusions: Abnormal patterns of glucose metabolism evolve with time in pHD subjects even before reaching early symptomatic phases. Network method reveals a unique metabolic topography associated with compensatory processes in functional brain circuits (3). HDRP activity is correlated with loss of striatal dopamine D2 receptor and can serve as a predictor of years to onset in pHD carriers. It allows us to more accurately delineate deteriorating metabolic substrates as the disease progresses from presymptomatic to symptomatic stages. This may shed important light on clinical and behavior correlates of an imaging marker based on spatial covariance.
Research Support (if any): Supported by NIH grant RO1 NS-37564
- Society of Nuclear Medicine, Inc.