Volumetric Imaging of Gray and White Matter in Corticobasal Syndrome (P05.035)

Abstract

Objective: Demonstrate the ability of volumetric MRI to distinguish patients with corticobasal syndrome (CBS) who have underlying Alzheimer9s disease (AD) from those with non-AD pathology. Background CBS is associated with various pathologies including AD and frontotemporal lobar degeneration (e.g., corticobasal degeneration and progressive supranuclear palsy). A ratio of cerebrospinal fluid (CSF) tau to amyloid beta 1-42 ≥ 0.34 has been associated with underlying AD in autopsy-validated studies. We compared gray matter (GM) and white matter (WM) imaging in CBS patients with CSF suggestive of AD versus non-AD pathology. Design/Methods: High-resolution T1 and diffusion-weighted images were available for 18 CBS patients (CBS-AD=8, CBS-nonAD=10) and 35 matched controls. For T1 images, we performed a probabilistic segmentation to compute GM density. We extracted fractional anisotropy (FA) along WM tracts from 30-direction diffusion tensor images. GM density and FA for the CBS-AD and CBS-nonAD groups were compared to controls using two-sample t-tests. We identified GM and WM regions that differed between patients and controls (p Results: All CBS patients showed GM atrophy in bilateral parietal and temporal cortex relative to controls. GM atrophy was also present in bilateral inferior and dorsolateral prefrontal cortex in CBS-nonAD patients. FA was reduced in posterior corpus callosum and superior longitudinal fasciculus in both groups. CBS-nonAD patients also showed reduced FA in anterior corpus callosum and uncinate. In the ROC curve analysis, three GM regions (right intraparietal sulcus, right inferior frontal cortex, anterior cingulate) and one WM region (anterior corpus callosum) distinguished between CBS-AD and CBS-nonAD patients (p Conclusions: Volumetric MRI can distinguish CBS patients likely to have underlying AD from those with non-AD pathology. The ability to predict underlying pathology in CBS non-invasively is essential as disease-specific therapies become available. Supported by: NIH grants NS44266, NS53488, AG15116, AG17586, AG32953, HD060406, NS065347, and NS045839, as well as the Wyncote Foundation. Disclosure: Dr. Goldmann Gross has nothing to disclose. Dr. McMillan has nothing to disclose. Dr. Cook has nothing to disclose. Dr. Grossman has nothing to disclose.