Dorsal medulla surface texture: Differentiating neuromyelitis optica spectrum disorder from multiple sclerosis

Background and Purpose The timely and accurate diagnosis of neuromyelitis optica spectrum disorder (NMOSD) is essential and exposure to multiple sclerosis (MS) disease-modifying therapies may result in permanent neurological disability. Methods Standardized 3-Tesla 3-dimensional brain MRI studies were retrospectively studied from people with NMOSD, MS, other CNS neurological diseases, and healthy control subjects. Comparisons of surface texture characteristics at the area postrema involving absolute introverted planar triangle counts, representing more complex and concave tissue topography, along with the spatial dissemination pattern of these triangles were performed cross-sectionally and longitudinally. An ideal introverted planar triangle threshold separating groups with NMOSD and MS was accomplished using the highest Youden's J statistic. For the classification of NMOSD, out-of-sample and in-sample measurements of the following were acquired: sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Results The study cohort included 60 people with NMOSD, 100 people with MS, 12 with other neurological diseases, and five healthy controls. Significantly higher cross-sectional median introverted triangle counts were observed when the NMOSD (median [interquartile range]: 100 [23.5]) group was compared to MS (65 [20.25]; p < .0001) and other neurological diseases (66 [13.75]; p < .0001). Distinct spatial dissemination patterns of triangles extending craniocaudally at the region of interest within the dorsal medulla was also seen between groups with NMOSD and MS (p < .0001). For the identification of NMOSD, out-of-sample sensitivity (83%), specificity (100%), PPV (100%), and NPV (60%) were achieved. Conclusions Cross-sectional and longitudinal dorsal medulla surface texture differences within selective regions of vulnerability differentiate NMOSD from MS.