Diffusion imaging of fornix and interconnected limbic deep grey matter is linked to cognitive impairment in multiple sclerosis

Abstract

Diffusion tensor imaging (DTI) and volumetric magnetic resonance imaging (MRI) have shown white matter (WM) and deep grey matter (GM) abnormalities in the limbic system of multiple sclerosis (MS) participants. Structures like the fornix have been associated with cognitive impairment (CI) in MS, but the diffusion metrics are often biased by partial volume effects from cerebrospinal fluid (CSF) due to its small bundle size and intraventricular location. These errors in DTI parameter estimation worsen with atrophy in MS. The goal here was to evaluate DTI parameters and volumes of the fornix, as well as associated deep GM structures like the thalamus and hippocampus, with high‐resolution fluid‐attenuated inversion recovery (FLAIR)‐DTI at 3T in 43 MS patients, with and without CI, versus 43 controls. The fornix, thalamus and hippocampus displayed atrophy and/or abnormal diffusion metrics, with the fornix showing the most extensive changes within the structures studied here, mainly in CI MS. The affected fornix volumes and diffusion metrics were associated with thalamic atrophy and atypical diffusion metrics in interconnected limbic GM, larger total lesion volume and global brain atrophy. Lower fractional anisotropy (FA) and higher mean and radial diffusivity in the fornix, lower hippocampus FA and lower thalamus volume were strongly correlated with CI in MS. Hippocampus FA and thalamus atrophy were negatively correlated with fatigue and longer time since MS symptoms onset, respectively. FLAIR‐DTI and volumetric analyses provided methodologically superior evidence for microstructural abnormalities and extensive atrophy of the fornix and interconnected deep GM in MS that were associated with cognitive deficits.

High‐resolution diffusion tensor imaging and T2 mapping detect regional changes within the hippocampus in multiple sclerosis

“…Third, left and right hippocampal metrics were averaged together to minimize the already large number of comparisons. Separate left and right analyses showed similar findings (data not shown) and most previous studies have shown similar left and right atrophy in MS 13,16,18 or have analyzed left and right metrics together, 10,11,15,17,20 including DTI studies 25–27,60 . Fourth, MRI metrics such as MD and T2 are sensitive to increased water content, which could occur from partial volume effects with CSF and be exacerbated as a result of greater atrophy, even at this improved high resolution.…”

“…Neuronal degeneration caused by axonal transection due to lesions in the fornix, which is the major efferent WM pathway from the hippocampus, could also trigger diffuse neuronal loss in the hippocampus that could happen anywhere along the anterior posterior long axis of the hippocampus. A previous CSF suppressed DTI study in the same MS cohort reported marked fornix atrophy (~26%) and microstructural diffusion abnormalities (6%–12%) in MS that were greater in the CI group 25 . As the extent of fornix atrophy and diffusion abnormalities are greater than hippocampus atrophy, diffusion, and T2 relaxometry differences when compared with controls in this MS cohort, we hypothesize that fornix injury could be inducing hippocampus damage here.…”

“…The slices were aligned to the long axis of the hippocampus (see the red rectangle in Figure 1A) using a whole brain high‐resolution 3D T1‐weighted magnetization‐prepared rapid acquisition with gradient echo (MPRAGE) for reference (0.85 mm isotropic; 3 min 39 s). Whole brain 3D sampling perfection with application optimized contrasts using different flip angle evolution (SPACE) FLAIR was also performed (1.2 mm isotropic, 3 min 7 s) in the MS cohort to yield total WM lesion volumes (TLVs) 25 . T2 relaxometry and higher in‐plane resolution T2‐weighted imaging were acquired for the same 20 1 mm axial‐oblique slices as the DTI and included: (a) 2D T2 multiecho spin echo with 16 echoes (TE = 10.7–171.2 ms with 10.7 ms interecho spacing), TR = 3560 ms, 1.1 × 1.1 × 1 mm 3 , 20 slices in 5 min 47 s, and (b) 0.5 × 0.5 × 1 mm 3 T2‐weighted images with TE = 52 ms, TR = 5440 ms, 20 slices in 4 min 39 s (Figure 1C).…”

“…However, the poor spatial resolution of DTI in vivo (i.e., ≥ 2 mm isotropic typically) has led to limited application to the study of the hippocampus in MS. The majority of these few DTI studies have thus evaluated diffusion metrics averaged over the entire hippocampus, demonstrating group differences of higher MD 16,[25][26][27][28] and lower FA [25][26][27] in MS versus healthy cohorts, as well as correlations of hippocampal diffusion metrics with memory impairment, 26 visuospatial skills and information processing speed, 29 and fatigue. 25 The development of a 6-min clinically applicable 1 mm isotropic resolution DTI protocol of the hippocampus has shown excellent visualization of known internal anatomy and the ability to make regional diffusion measurements.…”

Analysis of the extent of limbic system changes in multiple sclerosis using FreeSurfer and voxel-based morphometry approaches

“…VBM studies have shown changes in the insula, orbitofrontal cortices, temporal lobes, thalamus, and cingulate cortex [ 35 ]. Using diffusion tensor and tractography, the hippocampus, fornix, and thalamus show significant difference from controls [ 36 , 37 ]. Functional MRIs studies highlighted alterations in hippocampal and amygdaloid regions [ 38 ].…”