Damage to the optic radiation in multiple sclerosis is associated with retinal injury and visual disability

Abstract

Objective: To determine whether damage to the optic radiation (OR) in multiple sclerosis (MS) is associated with optic nerve injury and visual dysfunction.

Design: Case-control study.

Setting: Referral center.

Participants: Ninety referred patients with MS and 29 healthy volunteers.

Main outcome measures: Magnetic resonance imaging indices along the OR were reconstructed with diffusion tensor tractography. Retinal nerve fiber layer thickness and visual acuity at high and low contrast were measured in a subset of the MS group (n = 36).

Results: All tested magnetic resonance imaging indices (fractional anisotropy [FA]; mean, parallel, and perpendicular [lambda( perpendicular)] diffusivity; T2 relaxation time; and magnetization transfer ratio) were significantly abnormal in patients with MS. Mean retinal nerve fiber layer thickness was significantly correlated with FA (r = 0.55; P < .001) and lambda( perpendicular) (r = -0.37; P = .001). The retinal nerve fiber layer thickness in the nasal retinal quadrant was also specifically correlated with FA and lambda( perpendicular) in the synaptically connected contralateral OR. In individuals with less severely damaged optic nerves (mean retinal nerve fiber layer thickness >80 mum), letter acuity scores at 2.5% contrast were correlated with OR-specific FA (r = 0.55; P = .004), lambda( perpendicular) (r = -0.40; P = .04), and magnetization transfer ratio (r = 0.54; P = .01), as well as the fraction of OR volume made up of lesions (r = -0.69; P < .001).

Conclusions: Fractional anisotropy and lambda( perpendicular) are potentially useful quantitative magnetic resonance imaging biomarkers of OR-specific damage in MS. Such damage is associated with retinal injury and visual disability.

Figure 1

Regions of interest used for diffusion tensor imaging-based tractography of the optic radiation. Images are from a healthy person. On coronal reformations of the axially acquired data, regions of interest were drawn around the expected locations of the anterior optic radiation, near the lateral geniculate nuclei (panel A) and around the posterior portion of the tract, just anterior to its termination in the primary visual cortex (panel B). Only the portions of the fibers that ran between both regions of interest were further analyzed. Panel C shows a three dimensional reconstruction of the reconstructed right and left optic radiation.

Figure 2

Tract profiles of the optic radiation spanning the lateral geniculate nucleus and occipital lobe for a variety of MRI indices. The horizontal axis uses normalized distance units. Data are averaged across 90 study participants with MS (red) and 29 healthy controls (black). Error bars denote standard errors of the mean. Additional curves correspond to subgroups of MS participants: relapsing remitting (green), secondary progressive (blue), and primary progressive (cyan). MD, mean diffusivity; λ_∥, parallel diffusivity; λ_⊥, perpendicular diffusivity; FA, fractional anisotropy; T2, absolute T2 relaxation time; MTR, magnetization transfer ratio.

Figure 3

Tract profiles of the optic radiation for the unsegmented tract (red), lesional portions only (green), and portions that appear normal on conventional MRI (blue). Profiles from healthy controls (black) are shown for comparison. Abbreviations are as in Figure 2.

Figure 4

Association of optic radiation MRI indices with retinal nerve fiber layer thickness (A-D) and visual acuity at 2.5% contrast (E-H). Only MS data are included, and marker size is proportional to the number of scans obtained for each participant within 30 days of acuity or retinal testing. The left column (A, C, E, G) shows raw data, whereas the right column (B, D, F, H) shows data adjusted for age, sex, corticospinal tract MRI indices, and, for plots of visual acuity, mean retinal nerve fiber layer thickness. Each data point represents a single participant, and marker size is proportional to the number of observations for that participant. Weighted multiple linear regression analysis was used to derive the values in the right column. RNFL, retinal nerve fiber layer; FA, fractional anisotropy; MTR, magnetization transfer ratio.