Regionally distinct white matter lesions do not contribute to regional gray matter atrophy in patients with multiple sclerosis.

PURPOSE: To determine to what extent T1- and T2-regional lesion volumes (RLVs) contribute to total and/or regional gray matter (GM) atrophy in multiple sclerosis (MS). METHODS: We studied 110 (67 relapsing-remitting and 43 secondary-progressive) MS patients. SABRE program was used to parcel the brain into 13 regions per hemisphere. Total and regional GM fractions (GMFs) were determined in each region to correct for intraregional size variability. Partial correlations were used to determine associations (holding the converse constant) between RLVs, GMF, and regional GMFs (P < .001 to avoid Type 1 error). RESULTS: Partial correlations between RLVs and regional GMFs (controlling for total GMF) for the total MS group were not significant for any of the 26 regions for T2, whereas they were significant for two of the 26 regions for T1. Partial correlations between RLVs and total GMF (controlling for regional GMF) for the total MS group were significant in 9 of 26 regions for T2 (largest r = right lateral inferior frontal, -.45) and 5 of 26 regions for T1 (largest r = right inferior parietal, -.45). CONCLUSIONS: Results suggest a model whereby a distinct generalized disease process accounts for GM atrophy better than regionally distinct Wallerian degeneration.

Independent contributions of cortical gray matter atrophy and ventricle enlargement for predicting neuropsychological impairment in multiple sclerosis.

The primary goal of this study was to investigate associations between regional gray matter (GM) atrophy and neuropsychological function in multiple sclerosis (MS), while accounting for the influence of central brain atrophy (i.e. third ventricle enlargement). Using a cross-sectional design, we studied 59 MS patients with brain MRI and neuropsychological testing. Regional gray matter fractions (rGMFs) were calculated from MRI images for 11 homologous brain areas using the semiautomatic brain region extraction (SABRE) technique. Neuropsychological testing followed consensus panel guidelines and included tests emphasizing episodic memory, working memory and processing speed. The analytic approach was stepwise linear regression, with forward selection and p<0.05 threshold for significance. Consistent with previous research, there were significant correlations between third ventricle width and neuropsychological tests. Stepwise linear regression analyses controlling for third ventricle width retained rGMFs obtained from specific regions within the prefrontal cortex. Left frontal atrophy was associated with tests emphasizing auditory/verbal memory. Right frontal atrophy was associated with impairment in visual episodic and working memory. For the first time, we show an independent relationship between cortical atrophy and cognitive impairment after accounting for the effects of central atrophy.