Gray matter atrophy in multiple sclerosis despite clinical and lesion stability during natalizumab treatment

Background Brain volume loss is an important surrogate marker for assessing disability in MS; however, contribution of gray and white matter to the whole brain volume loss needs further examination in the context of specific MS treatment. Objectives To examine whole and segmented gray, white, thalamic, and corpus callosum volume loss in stable patients receiving natalizumab for 2-5 years. Methods This was a retrospective study of 20 patients undergoing treatment with natalizumab for 24-68 months. Whole brain volume loss was determined with SIENA. Gray and white matter segmentation was done using FAST. Thalamic and corpus callosum volumes were determined using Freesurfer. T1 relaxation values of chronic hypointense lesions (black holes) were determined using a quantitative, in-house developed method to assess lesion evolution. Results Over a mean of 36.6 months, median percent brain volume change (PBVC) was -2.0% (IQR 0.99-2.99). There was decline in gray (p = 0.001) but not white matter (p = 0.6), and thalamic (p = 0.01) but not corpus callosum volume (p = 0.09). Gray matter loss correlated with PBVC (Spearman's r = 0.64, p = 0.003) but not white matter (Spearman's r = 0.42, p = 0.07). Age significantly influenced whole brain volume loss (p = 0.010, multivariate regression), but disease duration and baseline T2 lesion volume did not. There was no change in T1 relaxation values of lesions or T2 lesion volume over time. All patients remained clinically stable. Conclusions These results demonstrate that brain volume loss in MS is primarily driven by gray matter changes and may be independent of clinically effective treatment.