Fatigue and cognitive impairment are common in multiple sclerosis (MS) affecting up to 95% and 65% of the patients, respectively. Only little is known about the interplay of different volumetric magnetic resonance imaging (MRI) parameters, the serum biomarkers neurofilament light chain (sNfL) and glial fibrillary acidic protein (sGFAP), cognition, and fatigue in MS patients without signs of progression independent of relapse activity (PIRA; i.e. clinically stable). This study aimed to evaluate whether volumetric MRI and serum NfL and GFAP are associated with cognitive impairment and fatigue in clinically stable relapsing-remitting multiple sclerosis (RRMS) and to examine whether the Expanded Disability Status Scale (EDSS) has an added value. A retrospective mono-centric cohort of 54 clinically stable (no disability worsening/relapse-free) RRMS patients underwent comprehensive clinical assessment with the EDSS, Brief International Cognitive Assessment for Multiple Sclerosis (BICAMS-M), Fatigue Scale for Motor and Cognitive Functions (FSMC), in addition to established biomarkers of MS pathology, including sNfL, sGFAP, and magnetic resonance imaging (MRI: thalamic/hippocampal/whole brain parenchymal volumes/annualized brain volume loss (BVL/year)). Associations were examined using regressions, correlations, and group comparisons. Thalamic and hippocampal volumes were significantly lower in cognitively impaired than non-impaired patients (both p.35; p>.09). Fatigue was not associated with MRI or serum markers (p>.09). BICAMS-M correlated with EDSS (r=–.56; p<.001) to a similar extent as with thalamic or hippocampal volumes (r=.55; p<.001). BVL/year was highly associated with cognitive impairment, explaining 63.2% (BICAMS-M) of the variance (p<.001). In our clinically stable, small RRMS cohort, BVL/year, thalamic and hippocampal volumes are associated with cognition in contrast to sNfL and sGFAP. EDSS as a global disability score correlates with cognition but adds little beyond MRI. These results require further validation using a prospective design and broader cohort.
Erhart et al. (Fri,) studied this question.