Objectives: Many multiple sclerosis (MS) lesions develop around small veins that are surrounded by perivenular inflammatory cells, but whether veins in the brains of people with MS are smaller or larger than similar veins in healthy volunteers or people with other neurologic diseases remains unknown. This question can be addressed by high-resolution, high-field-strength MRI.Methods: In a cross-sectional study performed on a standard 3 T clinical scanner, we acquired wholebrain T2*-weighted images with 0.55 mm isotropic voxels and reconstructed the courses of deep and superficial veins within the white matter. We compared the apparent diameters of intralesional and perilesional veins to those of extralesional MS veins, veins in healthy volunteers, and veins in individuals with other neurologic diseases.Results: We studied veins in 19 MS cases, 9 healthy volunteers, and 8 individuals with other neurologic diseases, analyzing a total of 349 veins. The mean diameter of intralesional veins (0.76 6 0.14 mm) was smaller than that of perilesional (1.18 6 0.13 mm; p , 0.001) and extralesional (1.13 6 0.14 mm; p , 0.001) veins, regardless of lesion size and location. Perilesional and extralesional MS veins were larger than non-MS veins (0.94 6 0.14 mm; p , 0.001), and intralesional MS veins were smaller (p , 0.001). Most multiple sclerosis (MS) lesions in the cerebral white matter expand outward from small, inflamed veins, 1,2 although demyelinated lesions not centered by small veins have been reported.
Conclusions:
3The walls of veins in MS plaques are denser and heavily cellular, and perivenular fibrosis can be seen in the lesion center. 1 In chronic lesions, due to fibrosis and collagen deposition in the vein wall and perivascular space, venous caliber is significantly narrowed. 1,4 MRI venography at standard clinical field strength can demonstrate the perivenous distribution of MS lesions in vivo. 5 The ability to detect veins improves greatly as field strength increases, 6 because the degree to which deoxyhemoglobin in venous blood reduces signal intensity on T2*-weighted images becomes magnified.7 At 7 T, a central vessel can be identified in 87% of visible white matter lesions.
8An important open question is whether veins centering lesions are macroscopically different from veins outside lesions and from similar veins in healthy people and individuals with other neurologic diseases. One possibility is that the histopathologically observed perivenular cuffing and wall thickening induce macroscopic changes in intralesional veins. It is also possible, however, that such changes on MRI occur diffusely, affecting many if not all parenchymal veins in the MS brain; in this case, the changes would not be directly related to lesions, and their pathophysiologic significance would be less clear.