Longitudinal imaging of gray matter lesions in multiple sclerosis and the relation with neurocognition

• Published in: Multiple sclerosis Vol. 14; p. S26
• Main Authors: Roosendaal, S D, Moraal, B, Vrenken, H, Pouwels, P J, Castelijns, JA, Barkhof, F, Geurts, J J
• Format: Journal Article
• Language: English
• Published: 01.09.2008
• ISSN: 1352-4585

Background: New MRI sequences, such as double-inversion recovery (DIR), have been shown to substantially improve the detection of gray matter (GM) lesions in multiple sclerosis (MS). GM lesions were found to be abundant and occur early in the disease course. However, so far little is known about the progression of GM lesions over time. Objective: To study the development of GM lesions within three-year follow-up, using 3D-DIR images and to explore whether the number of GM lesions is associated with cognitive decline. Methods: 3DDIR images (TR/TE/T11/TI2:6500/349/2350/350; 1.2 x 1.2 x 1.3 mm super(3)) were acquired of 13 MS patients (mean disease duration 15.9 years, standard deviation (SD) 5.8) and seven healthy controls on two time points (mean interval 2.8 years, SD 0.2) on the same 1.5 T scanner (Siemens Sonata, Erlangen, Germany). At follow-up, subjects underwent the Letter Digit Substitution Test (LDST) to assess processing speed and to the Location Learning Test (LLT) to assess visuospatial memory. Lesions were classified by the same rater as white matter (WM; that is, total of periventricular and deep WM) or cortical (total of intracortical, juxtacortical and mixed WM-GM lesions). Finally, hippocampal lesions were also scored. Results: Mean cortical lesion number increased from 29.3 to 35.6 (Wilcoxon's Z: 2.5; p<0.05), mean WM lesion number increased from 55.0 to 66.3 (Z: 2.4; p<0.05), no significant increase was found for hippocampal lesions. Cortical lesions had increased more in patients who already showed high cortical lesion numbers at baseline. Significant relations were found between cortical lesion number at follow-up and both LLT delay score (Spearman's rho: 0.6; p<0.05) and LDST (Spearman's rho: -0.6; p=0.05). WM lesion number was strongly related to cortical lesion number (rho: 0.8; p<0.01) and thus showed the same relations to cognition. Hippocampal lesion number was related to the LLT delay score (Spearman's rho: 0.7; p<0.01). Conclusions: In conclusion, cortical and WM lesions increase substantially over a relatively short period of time. Patients with high cortical lesion number at baseline showed a higher cortical lesion increase over time. Cortical and hippocampal lesions were related to cognitive deficits.