White matter alterations differ in primary lateral sclerosis and amyotrophic lateral sclerosis

Abstract

Primary lateral sclerosis is a sporadic disorder characterized by slowly progressive corticospinal dysfunction. Primary lateral sclerosis differs from amyotrophic lateral sclerosis by its lack of lower motor neuron signs and long survival. Few pathological studies have been carried out on patients with primary lateral sclerosis, and the relationship between primary lateral sclerosis and amyotrophic lateral sclerosis remains uncertain. To detect in vivo structural differences between the two disorders, diffusion tensor imaging of white matter tracts was carried out in 19 patients with primary lateral sclerosis, 18 patients with amyotrophic lateral sclerosis and 19 age-matched controls. Fibre tracking was used to reconstruct the intracranial portion of the corticospinal tract and three regions of the corpus callosum: the genu, splenium and callosal fibres connecting the motor cortices. Both patient groups had reduced fractional anisotropy, a measure associated with axonal organization, and increased mean diffusivity of the reconstructed corticospinal and callosal motor fibres compared with controls, without changes in the genu or splenium. Voxelwise comparison of the whole brain white matter using tract-based spatial statistics confirmed the differences between patients and controls in the diffusion properties of the corticospinal tracts and motor fibres of the callosum. This analysis further revealed differences in the regional distribution of white matter alterations between the patient groups. In patients with amyotrophic lateral sclerosis, the greatest reduction in fractional anisotropy occurred in the distal portions of the intracranial corticospinal tract, consistent with a distal axonal degeneration. In patients with primary lateral sclerosis, the greatest loss of fractional anisotropy and mean diffusivity occurred in the subcortical white matter underlying the motor cortex, with reduced volume, suggesting tissue loss. Clinical measures of upper motor neuron dysfunction correlated with reductions in fractional anisotropy in the corticospinal tract in patients with amyotrophic lateral sclerosis and increased mean diffusivity and volume loss of the corticospinal tract in patients with primary lateral sclerosis. Changes in the diffusion properties of the motor fibres of the corpus callosum were strongly correlated with changes in corticospinal fibres in patients, but not in controls. These findings indicate that degeneration is not selective for corticospinal neurons, but affects callosal neurons within the motor cortex in motor neuron disorders.

Trial registration: ClinicalTrials.gov NCT00334516.

Figures

Figure 1

White matter tracts defined by diffusion tensor tractography using the FACT fibre tracking algorithm. The corticospinal tract (red), and the genu (green), motor fibres (yellow), and splenium (blue) of the corpus callosum are shown in one healthy control.

Figure 2

Profiles of group mean diffusion tensor tractography measures along the rostrocaudal extent of the corticospinal tract, from mid-pons to motor cortex, for controls (black lines), patients with primary lateral sclerosis (blue lines) and patients with ALS (red lines) for (A) fractional anisotropy (FA), (B) mean diffusivity (MD) and (C) cross-sectional area. The right and left corticospinal tract were averaged for each subject. Comparing anatomical levels where the corticospinal tract is least variable (dashed lines from left to right: peduncle, internal capsule, subcortical white matter), corticospinal tract fractional anisotropy of patients with ALS was significantly reduced and mean diffusivity significantly increased from the controls. Corticospinal tract cross-sectional area profiles in patients with primary lateral sclerosis was reduced compared with controls. There were no significant differences between the three levels.

Figure 3

Correlations between corticospinal tract fractional anisotropy (FA) and upper motor neuron (UMN) rapidity index (A and C) and disease duration (B and D) for patients with ALS (A and B) and primary lateral sclerosis (C and D). Each point represents one patient with right and left side corticospinal tract averaged. Significant Pearson's correlation r-values are shown (P < 0.001). Dashed lines represent corticospinal tract fractional anisotropy range of control subjects (mean ± 2 SD). Upper motor neuron rapidity index = global upper motor neuron-impairment score/disease duration in years. Disease duration is counted from symptom onset.

Figure 4

TBSS analysis showing clusters of voxels with significantly lower fractional anisotropy (FA) (red) in (A) patients with ALS and (B) patients with primary lateral sclerosis (PLS) as compared with healthy controls (HC). Clusters of voxels showing significantly greater mean diffusivity (MD) (blue) in (C) patients with primary lateral sclerosis as compared with healthy controls (P < 0.05), corrected for multiple comparisons across space (FWE) using threshold-free cluster enhancement. Patients with ALS did not have significant differences in mean diffusivity compared with healthy controls (not illustrated). Each panel shows the significant voxel clusters superimposed on the mean fractional anisotropy map in axial (upper row), coronal (middle row) and sagittal (bottom row) slices. Images are displayed in radiological convention (right brain on left side).

Figure 5

TBSS analysis of clusters of voxels with significantly lower fractional anisotropy (FA) (red) in (A) patients with ALS compared with patients with primary lateral sclerosis (PLS), and significantly greater mean diffusivity (MD) (blue) in (B) patients with primary lateral sclerosis compared with patients with ALS in selected axial, coronal and sagittal sections (P < 0.05), corrected for multiple comparisons across space (FWE) using threshold-free cluster enhancement. MNI coordinates are shown. Images are displayed in radiological convention (right brain on left side).

Figure 6

TBSS analysis showing clusters of voxels with significant correlation between the lateralized upper motor neuron impairment score and fractional anisotropy (FA) (red) in (A) patients with ALS and (B) patients with primary lateral sclerosis (PLS), and between the lateralized upper motor neuron impairment score and mean diffusivity (MD) (blue) in (C) patients with primary lateral sclerosis (P < 0.05, corrected for multiple comparisons across space (FWE) using threshold-free cluster enhancement). Each panel shows the significant voxel clusters superimposed on the mean fractional anisotropy map in axial (upper row), coronal (middle row) and sagittal (bottom row) slices. Images are displayed in radiological convention (right brain on left side). Patients with ALS did not exhibit correlations between the upper motor neuron impairment score and mean diffusivity (not illustrated).