The correlation of conventional and advanced MRI findings with cognitive function in multiple sclerosis: A systematic review and meta-analysis

Abstract

There are limited data on the possible association between conventional and advanced magnetic resonance imaging (MRI) findings and cognitive function in patients with multiple sclerosis (MS). Therefore, this systematic review and meta-analysis aimed to explore the correlation between MRI-derived metrics and cognitive tests in patients with MS. An electronic literature search of the PubMed, Web of Sciences, Embase, and Scopus databases was performed to identify related studies. The correlation coefficients of the MRI indices and cognitive tests were pooled. Thirteen studies were selected for inclusion of 824 patients diagnosed with MS. Most evaluated patients (60.44%) had relapsing-remitting MS (RRMS). The Paced Auditory Serial Addition Test (PASAT-3), Brief Visuospatial Memory Test (BVMT), and Symbol Digit Modalities Test (SDMT) were inversely correlated with the mean diffusivity (MD) of the brain with pooled correlation coefficient of -0.225, -0.361, and -0.438, respectively (P<0.0001). The SDMT test positively correlated with fractional anisotropy (FA) with a correlation coefficient of 0.351 (P<0.0001) and inversely correlated with T2 lesion volume with a correlation coefficient of -0.367 (P<0.0001). In the case of other tests, there was low number of studies with significant correlations being reported. We found significant correlations between some neuropsychological tests and MRI findings in patients with MS. Brain atrophy might disrupt the process of correct registration between anatomical and MRI diffusion scans. However, we did not have enough studies with exactly matched anatomical areas to evaluate correlations and we recommend that histological validation of diffusion tensor imaging (DTI) findings for brain atrophy is needed as a basis for picture processing procedures and correlation with cognition status.

Keywords: Cognition, Diffusion tensor imaging, Magnetic resonance imaging, Meta-analysis, Multiple sclerosis