Lesion Location and Functional Connections Reveal Cognitive Impairment Networks in Multiple Sclerosis.

Objective: Cognitive impairment, fatigue, and depression are common in multiple sclerosis (MS), potentially due to disruption of regional functional connectivity caused by white matter (WM) lesions. We explored whether WM lesions functionally connected to specific brain regions contribute to these MS-related manifestations.

Methods: A total of 596 MS patients underwent 3T brain MRI acquisition, neurologic assessment, and neuropsychological evaluation (Brief Repeatable Battery, Modified Fatigue Impact Scale [MFIS], and Montgomery-Åsberg Depression Rating Scale [MADRS]). Voxel-wise lesion probability maps were compared between subgroups based on cognition, fatigue, or depression. Lesion distributions were linked to a brain functional connectivity atlas to map lesion network associations. Lesion network maps (LNMs) were then compared among subgroups (p < 0.05, FWE-corrected).

Results: One hundred twenty-six (27.2%) MS patients were cognitively impaired and showed significantly more widespread WM lesions, more strongly functionally connected to bilateral hippocampi, thalami, cerebellum, and occipital cortices (corrected-p < 0.05) than cognitively preserved patients. Lesion networks were similar for impaired processing speed/attention. Verbal memory deficits were associated with WM lesions connected to parahippocampi, temporal pole, and cerebellum (corrected-p ≤ 0.05), while verbal fluency deficits involved connections to thalami, putamen, caudate nuclei, anterior cingulate cortex, and cerebellum (corrected-p ≤ 0.05). No significant lesion distribution or network connectivity differences were found in patients with visual memory deficits, fatigue (MFIS ≥ 38, 184/493 [37.3%]) or depression (MADRS > 9, 192/495 [38.8%]).

Interpretation: Regional WM lesions disrupting connections to the hippocampus, thalamus, cerebellum, and temporo-occipital cortices contribute to cognitive impairment, but not fatigue or depression. LNM may clarify mechanisms underlying cognitive deficits in MS.