Region specific fractional anisotropy in diffusion tensor MR imaging reveals EDSS correlation in patients with Multiple Sclerosis

Background: Assessment of stage of disease in patients with Multiple Sclerosis (MS) in the setting of clinical studies is usually performed using categorical clinical scales such as the EDSS. More objective measures are desirable to quantitatively and sensitively assess the stage and progression of MS. Recently the non-invasive assessment of white matter lesions using diffusion tensor MR imaging (DTI) and analyzing fractional anisotropy (FA) in regions of interest (ROIs) became available. So far, no conclusive results were obtained when evaluating FA as a measure of disease severity in MS.

Objective: To investigate whether FA as assessed by DTI in white matter ROIs correlates to disease severity in patients with MS as assessed clinically by the EDSS.

Subjects and methods: DTI scans were obtained from 29 patients with MS using a 3,0 Tesla scanner. FA was calculated from DTIs. Mean FA was calculated for different white matter ROIs: frontal lobe, temporal lobe, occipital lobe, capsula interna, corona radiata, corpus callosum, genu, splenium, pyramidal tract, corticospinal tract, brain stem, and cerebellum. All subjects were clinically assessed using the EDSS by a physician experienced in the care of patients with MS. EDSS scores and FA were assessed in a double-blinded manner. Statistics were performed using Pearson correlation analysis (SPSS 13.0). Results were considered significant at p<0.01.

Results: Significant correlations of FA with the severity of clinical symptoms as assessed by the EDSS score were obtained for the following ROIs: corona radiata (r=-0.515, p=0.005), pyramidal tract (r=-0.526, p=0.004), corticospinal tract (r=-0.527, p=0.004), and brain stem (r=-0.558, p=0.002). Mean FA from these regions also correlated significantly to the EDSS (r=-0.563, p=0.0017). Two other regions were closed to a significant correlation: capsula interna (r=-0.444, p=0.018), and splenium (r=-0.479, p=0.011).

Conclusion: The data suggests that region specific analysis of FA may be a useful imaging surrogate marker in assessing disease severity in patients with MS. In contrast to some previous studies, region specific FA did show a correlation to clinical disease phenotype expression as assessed in the EDSS. FA may be an objective and quantitative non-invasive outcome-measure for clinical studies in MS and warrants further exploration, e.g., in a blinded multicenter follow-up study.