Abstract Background Automated cross-sectional magnetic resonance imaging (MRI) measures of brain volume are becoming available in clinical practice; however, their prognostic utility in multiple sclerosis remains unclear. Objectives We investigated associations between cross-sectional brain and brain lesion volume measures and subsequent disability outcomes in relapsing–remitting multiple sclerosis (RRMS). Methods We conducted a multicentre international cohort study using data collected prospectively in the MSBase Registry. MRI analysis was performed using icobrain ms. Associations between MRI volumetric measurements and confirmed disability worsening (CDW), progression independent of relapse activity (PIRA), secondary progressive multiple sclerosis (SPMS) and confirmed disability improvement (CDI) were examined using multivariable Cox proportional hazards models. Results A total of 1598 patients were included and 68.6% received high-efficacy disease-modifying therapy during follow-up. Normalised lateral ventricular volume (mL) was significantly associated with all disability outcomes, including PIRA (hazard ratio 1.007; 95% confidence interval 1.002 − 1.012; p = 0.009). Higher normalised whole brain, grey matter and white matter volumes were all significantly associated with a lower hazard of CDW and higher hazard of achieving CDI. Higher normalised white matter volume was also significantly associated with a lower hazard of PIRA. Higher T1-weighted black hole lesion volume was significantly associated with an increased hazard of conversion to SPMS; lesion volume measures were otherwise not significantly associated with disability outcomes. Conclusions Automated MRI measures of brain volume from a single time point demonstrated prognostic utility in a highly treated cohort of people with RRMS. The results suggest that normalised lateral ventricular volume is a particularly informative prognostic marker across multiple disability outcomes.
Hayes et al. (Wed,) studied this question.