Clinical and MRI predictors of cognitive decline in patients with relapsing-remitting multiple sclerosis: A 2-year longitudinal study

Background In this study, we conducted a prospective 2-year cohort of patients with RRMS and healthy controls (HCs) to investigate the rate and clinical/imaging predictors of cognitive decline in relapsing-remitting multiple sclerosis (RRMS). Methods A total of 107 patients with clinically definite RRMS and 74 HCs were recruited at Hebei General Hospital, Shijiazhuang, Hebei. Patients were assessed with the Minimal Assessment of Cognitive Function in MS (MACFIMS) at baseline and 2-year follow-up visits and were classified into cognitively-declining and cognitively-stable RRMS. Baseline demographic, clinical, and imaging parameters were inserted in separate multivariate regression models to investigate the predictive power of these factors for future cognitive decline. Results Based on the classification protocol and the data from HCs, 35.5% of RRMS patients were categorized as cognitively-declining. The multivariate logistic regression analyses demonstrated that disease duration, EDSS, and average disease attack/year were the clinical parameters with significant predictive value for future cognitive decline (R2=0.344). Within whole-brain MRI measures, total brain, cortical grey matter (GM), and subcortical GM volumes could significantly predict cognitive decline (R2=0.566). WM lesion volume could also significantly predict cognitive decline (R2=0.645). Within lobar brain measures, frontal and temporal lobe volumes (R2=0.666), and finally within subcortical GM volumes, hippocampus, pallidum, putamen, and thalamus were predictive of future cognitive decline (R2=0.732). Conclusions Our findings suggest that RRMS patients with more disease severity, higher GM atrophy, and increased WM lesion volume are more susceptible to cognitive decline. Further studies could assess the underpinnings of cortical and subcortical atrophy that lead to cognitive decline in RRMS patients.