Iron and volume in the deep gray matter: association with cognitive impairment in multiple sclerosis

Abstract

Background and purpose: There is a well-established correlation between deep gray matter atrophy and cognitive dysfunction in MS. However, the cause of these signs of neurodegeneration is poorly understood. Iron accumulation in the deep gray matter is higher in patients with MS compared with age- and sex-matched healthy controls, and could contribute to disease progression. Our objective was to evaluate the relationship between iron and cognition in several deep gray matter structures while accounting for the influence of volume loss.

Materials and methods: Eighty-five patients with MS and 27 healthy volunteers underwent 3T MR imaging and neuropsychological examination. We used SWI filtered phase to analyze the mean phase of low-phase voxels, indicative of abnormal iron accumulation.

Results: Correlations between mean phase of low-phase voxels and cognitive tests were found in the caudate nucleus (r = 0.240 and 0.232), putamen (r = 0.368, 0.252, and 0.238), globus pallidus (r = 0.235), and pulvinar nucleus of thalamus (r = 0.244, 0.255, and 0.251) (P < .05). However, correlations between structure volume and cognition were more robust. Furthermore, the introduction of structure volume into hierarchical regression analyses after iron metrics significantly improved most models, and mean phase of low-phase voxels did not account for significant variance after volume.

Conclusions: These findings suggest that iron accumulation plays a significant, if minor, role in MS cognitive decline.

Fig 1.

Scan demonstrating deep gray matter structure segmentation. FMRIB's Integrated Registration and Segmentation Tool was applied to the 3D high-resolution T1WI to segment the DGM. Pulvinar (yellow and red) was segmented using a semiautomated contouring technique.