Fig. 5: Characterization of composite cognitive and physical disability outcomes in relation to traditional outcomes used in multiple sclerosis (MS) clinical trials.
a The brain damage outcome was generated as the first component of a principal component analysis (PCA) of brain parenchymal fraction (BPFr; x-axis, volumetric magnetic resonance imaging [MRI] ratio of brain tissue to intracranial volume) and Symbol-digit modalities test (SDMT; y-axis). Brain damage correlated strongly with both measures (plots on right). b The Combinatorial Weight-adjusted Disability Scale (CombiWISE), a linear combination of four disability scores (weights decreasing as shown by the red triangle), correlated strongly with traditional MS disability scales, the Expanded Disability Status Scale (EDSS) and Timed 25-foot walk (T25FW) (plots on right). c The spinal cord (SC) disability outcome was calculated as residuals of a linear regression model of CombiWISE (y-axis) versus brain damage (x-axis). For similar brain damage, positive residuals (orange) indicated proportionally more physical disability, while negative residuals (blue) indicated less. MRI examples show patients with (1) high SC disability (low brain atrophy, high SC atrophy) and (2) low SC disability (marked brain atrophy, minimal SC atrophy). d Quantitative spinal cord atrophy, measured as C1–C2 SC area (x-axis), correlated strongly with SC disability (y-axis) in untreated (left) and all samples (right; treated samples in red). Scatter plot associations are shown as regression lines with 95% confidence intervals (black line, gray shading) and characterized by Pearson correlation coefficient (R), coefficient of determination (R²), and unadjusted p-values. ns = number of samples; np = number of patients. All regression coefficients were tested using two-sided p values. Created in BioRender. Kosa, P. (2025) https://BioRender.com/l8nkay6.
