Assessment of White Matter Tract Damage in Patients with Amyotrophic Lateral Sclerosis: A Diffusion Tensor MR Imaging Tractography Study

BACKGROUND AND PURPOSE: Most DTI studies in ALS have been limited to the assessment of the CST damage. In this study, we used DTI tractography to investigate whether microstructural abnormalities occur in the major motor and extramotor WM tracts in mildly disabled patients with ALS. MATERIALS AND METHODS: Brain conventional MR imaging and DTI were performed in 24 patients with probable or definite ALS and mild disability (ALSFRS score, ≥20) and 20 healthy controls. The mean disease progression rate was 0.62 (range = 0.08–2.50). DTI tractography was used to segment the CST, the corpus callosum, and the major WM association tracts (ie, cingulum, uncinate fasciculus, inferior fronto-occipital, inferior longitudinal, and superior longitudinal fasciculi). RESULTS: Compared with healthy controls, patients with ALS showed significantly decreased FA and significantly increased MD and radial D of the CST bilaterally (P values from .005 to .01). Patients with ALS also had a significantly increased axial D of the right uncinate fasciculus relative to controls (P = .04). CST FA significantly correlated with the rate of disease progression (right CST: r = −0.50, P = .02; left CST: r = −0.41, P = .05). CONCLUSIONS: Patients with ALS and mild disability have preferential damage to the CST. The association of CST damage with the rate of disease progression suggests that DTI has the potential to provide in vivo markers of ALS evolution. The subtle involvement of the uncinate fasciculus may precede the appearance of behavioral symptoms in patients with ALS.

A LS is the most common adult-onset motor neuron disease. 1 The pathologic hallmark of sporadic ALS is the loss of upper motor neurons in the motor and premotor cortices and lower motor neurons in the brain stem and spinal cord. 2 ALS is characterized by an extramotor cerebral pathology that, to a variable extent, overlaps the clinicopathologic features of frontotemporal lobar degeneration. 3 To date, the only specific marker of sporadic ALS is the presence of inclusions staining positively for ubiquitin and TAR deoxyribonucleic acidϪbinding protein 43 (TDP-43) in degenerating neurons. 4 By quantifying the magnitude and directionality of water diffusion within a tissue, DTI allows inferences about WM microstructure in vivo. 5,6 Most DTI studies in ALS have been limited to the assessment of the CST damage. 7 Whether on the basis of regions of interest, [8][9][10][11][12][13][14][15][16] voxel-based analysis, [17][18][19][20] or tractography, [21][22][23] DTI studies in ALS have shown reduced FA along the CST.
More recently, voxel-based DTI studies have assessed the integrity of extramotor brain regions in patients with ALS. [17][18][19][20] In agreement with pathologic findings, 24,25 FA decrease was found in the anterior corpus callosum [18][19][20] and in the prefrontal 18,20 and temporal 20 WM regions. Because of the intrinsic limitations of region of interestϪ and voxel-based analyses, however, such studies could only infer the specific WM tracts involved. On the other hand, DTI tractography allows one to gain quantitative information on the localization of damage to specific neuronal pathways. 26 To the best of our knowledge, no studies have yet assessed the integrity of WM tracts, other than the CST, in patients with ALS.
In this study, we used DTI tractography to assess the structural integrity of the CST, the corpus callosum, and the major WM association tracts (ie, cingulum, uncinate fasciculus, inferior fronto-occipital, inferior longitudinal, and superior longitudinal fasciculi) in patients with ALS with mild disability, relative to age-and sex-matched healthy controls.

Materials and Methods
The study was conducted with institutional review board approval. Written informed consent was obtained from each participant.

Patients
From June 2005 to March 2006, we recruited 24 patients (13 men and 11 women; mean age ϭ 55 years, range ϭ 27-75 years; mean disease duration ϭ 35 months, range ϭ 6 -58 months) with probable or definite ALS, 27 and mild disability, defined as a score Ն20 on the ALSFRS. 28 Six patients had a bulbar-onset and 18 patients had a limbonset disease. Within 48 hours of study entry, patients were assessed clinically by a single physician who was unaware of the MR imaging results. Disease severity was assessed using ALSFRS. 28 The mean ALSFRS score was 29 (range ϭ 21-38). Patients with evidence of cognitive deficit or overt dementia were excluded. The rate of disease progression was calculated by using the following formula: 40 Ϫ ALSFRS score) / disease duration. 23 The mean disease progression rate was 0.62 (range ϭ 0.08 -2.50). Twenty sex-and age-matched healthy subjects (11 men and 9 women, mean age ϭ 53 years, range ϭ 28 -73 years) served as controls.

MR Imaging Analysis
All MR imaging analysis was performed by an experienced observer, unaware of subject's identity. Axial DE and coronal T2-weighted images were analyzed to assess the presence and location of areas with increased signal intensity. DTI analysis was performed by using in-house software. Pulsed gradient spin-echo single-shot echo-planar images were first corrected for distortion induced by eddy currents, 29 then the diffusion tensor was estimated by linear regression 5 and MD and FA maps were computed. 30 Maps of axial D, which is equivalent to the magnitude of the largest eigenvalue of the tensor, and radial D, which is the average of the 2 smallest eigenvalues of the tensor, were also calculated. 30 An FA atlas was obtained by using DTIs from healthy subjects ranging from 21 to 40 age years of age with no history of neurologic or psychiatric disorders (reference group), as previously described. 31 Briefly, their DE scans were registered to the standard Montreal Neurologic Institute space 32 with affine transformation by using the VTK CISG Registration Toolkit (http://vtk.org). 33 This transformation was then applied to FA images to correct for differences in head size between controls. FA maps were then nonlinearly transformed with an iterative procedure to produce an average shape and intensity image atlas. 31 On reference FA maps, fiber tracking 34 was performed to segment the major brain WM tracts, bilaterally. 35 These included the following: the CST, the corpus callosum, the cingulum, the uncinate fasciculus, the inferior fronto-occipital fasciculus, the inferior longitudinal fasciculus, and the superior longitudinal fasciculus. A single subject's WM tracts were then registered to the standard space by using the transformation matrices previously computed (see above) and were averaged to produce WM tract probability maps. These maps were thresholded at 40%. WM tract probability maps are shown in Fig 1. Finally, the nonlinear transformation between the FA atlas and the FA maps of each subject was estimated 36 and applied to each subject's MD, FA, axial D, and radial D maps. WM tract probability maps were used as masks, and average MD, FA, axial D, and radial D values of each tract were measured.

Statistical Analysis
Statistical analysis was performed with the Statistical Package for the Social Sciences, Version 13.0 (SPSS, Chicago, Illinois). DTI variables were compared between patients with ALS and healthy controls by using a univariate analysis of variance. The subject's age was entered into the statistical analysis as a covariate. The same model was used to compare DTI changes between patients with and without hyperintensities along the brain CST. Univariate correlations were assessed by using the Spearman rank correlation coefficient. The significance threshold was set at P Ͻ .05.

Results
T2 hyperintensities along the CST were detected bilaterally in 13/24 patients with ALS (54%). 17 No T2 hyperintensities were detected beyond the CST. On-line Tables 1 and 2 report the DTI metrics from patients with ALS and healthy controls. Compared with healthy controls, patients with ALS showed significantly increased MD and radial D values and significantly decreased FA of the CST bilaterally (P values ranging from .01 to .03). CST DTI metrics did not differ between patients with ALS with and without hyperintensities along the CST (data not shown). Patients with ALS also had a significant increased axial D of the right uncinate fasciculus relative to healthy controls (P ϭ .04). No significant between-group difference was found in the other WM tracts (On-line Tables 1  and 2).
CST FA values significantly correlated with the rate of disease progression (right CST: r ϭ Ϫ0.50, P ϭ .02; left CST: r ϭ Ϫ0.41, P ϭ .05). No correlation was found between DTI metrics and the ALSFRS score.

Discussion
In this study, we investigated the structural integrity of the major cerebral WM tracts in mildly disabled patients with ALS. We found that patients with ALS with mild disability had preferential damage to the CST, which is associated with the rate of disease progression. In addition, the assessment of DTI eigenvalues in the investigation of WM damage revealed a subtle involvement of the right uncinate fasciculus, which may precede the appearance of behavioral symptoms in patients with ALS.
The finding of increased MD and radial D and reduced FA in the CST of patients with ALS is in agreement with previous studies, [8][9][10][11][12][13][14][15]23 confirming that DTI is a valuable tool for assessing CST changes in ALS. An increased average MD can result from the enlargement of extracellular spaces, which indicates a breakdown of the barriers that restrict free water movement. 6 The reduction of average FA might reflect both intracellular abnormalities of surviving axons and formation of "new" isotropic barriers, due to the presence of cell debris resulting from partially degenerated or disintegrated CST fibers, inflammatory infiltrates, and astrocytosis. 6 From the evaluation of the results of the diffusion tensor eigenvalue analysis, it appears that MD and FA changes are secondary to an increased diffusion perpendicular to WM fibers (as indicated by an increased radial D) rather than decreased diffusion along the fiber bundles (as indicated by unchanged axial D values). It is usually assumed that radial D, which is the average diffusion perpendicular to fiber bundles, is modulated by the extracellular distance between membranes, axon diameter, and degree of myelination. 37,38 As a consequence, it is thought that loss of fiber tracts and disruption of myelin sheaths in the course of Wallerian degeneration could explain enhanced transverse diffusion. 37,38 The main pathologic changes in patients with ALS include loss of pyramidal motor neurons in the primary motor cortex and axonal degeneration of the CST. 2 These abnormalities, together with the proliferation of glial cells, the extracellular matrix expansion, and the intraneuron abnormalities, 2 may contribute to the observed CST changes in patients with ALS. Most interesting, no difference in CST DTI features was found between patients with ALS with and without CST hy-perintensities on conventional MR images. These findings are in line with previous conventional MR imaging studies in ALS showing the low sensitivity and specificity of such abnormalities. [39][40][41][42][43][44][45][46] In patients with ALS, the severity of CST damage correlated with the rate of disease progression, thus suggesting that the pathologic changes of the CST detected by DTI are likely to contribute to the rapidity of disease evolution in patients with ALS. In a previous tractography study, 23 a strong association was found between disease-progression rate and left CST structural connectivity measures, while no association was found with CST FA. Differences in image preprocessing and tractography analysis may help explain the different results of the 2 studies and, in turn, do not allow a direct comparison of the results. In addition, Ciccarelli et al 23 investigated DTI changes only in the CST above the internal capsule, while previous studies have reported that the most significant changes in FA are evident at the level of the cerebral peduncles 47 and the internal capsule. 9,10 Our failure to identify a correlation between CST damage and the ALSFRS score is in keeping with previous data obtained from patients with ALS. 10,14,23,48 In addition, in our study, patients had a relatively small spread of ALSFRS scores, and this might have worked against the ability to detect a correlation between structural CST changes and disability.
Another intriguing finding is that patients with ALS showed a subtle change (ie, increased axial D) of the right uncinate fasciculus compared with healthy controls. This finding is in agreement with pathologic, 24,25 DTI,18,20 and voxel-based morphometry 49 studies showing the involvement of the frontal lobes and the temporal and limbic areas in ALS. Regions of significant increase of axial D in the major corticocortical association tracts have been recently observed with normal aging 50 and in other neurodegenerative diseases, such as Alzheimer disease. 51,52 In a previous voxel-based DTI study of the same patients with ALS, 17 we found clusters of extramotor increase of MD in the frontal, temporal, and occipital lobes, which only partially overlapped the uncinate fasciculus. One explanation for this would be the different spatial resolution (tract-versus voxel-based), because the presence of microstructural damage in a small region may not have affected the average DTI metrics of the WM tract. Although the functional role of the uncinate fasciculus is still debated, its termination in the anterior temporal lobe and in the ventral and orbital frontal regions 35 suggests that its involvement is likely to contribute to the behavioral symptoms of patients with ALS.
ALS patients included in our study did not show clinical evidence of a frontal cognitive or behavioral syndrome. 17 Thus, one may speculate that damage to the right uncinate may precede the appearance of such symptoms. According to Lomen-Horeth et al, 53 approximately 25% of patients with ALS meet the Neary criteria for FTD and approximately another 25% manifest executive or behavioral dysfunction compatible with early FTD. This hypothesis is also supported by recent findings of a similarly damaged uncinate fasciculus in patients with behavioral FTD 54,55 and semantic dementia 54,56 and by the association between medial and ventral frontal WM damage and behavioral symptoms in patients with behavioral FTD. 57 Longitudinal studies are now warranted to assess whether the extent and severity of the MR imagingϪ detectable extramotor damage is predictive of subsequent development of cognitive impairment in patients with ALS.
We should mention some limitations of our study: first, the relatively small sample size, which was nevertheless comparable with previous reports. 7 Due to the exploratory nature of our study, we did not correct for multiple comparisons; therefore, some of the significances might have been overestimated. However, we observed DTI tractography changes only in brain areas that are known to be the preferential sites of ALS damage from neuropathologic studies. 2 Finally, some methodologic issues related to DTI should be addressed. Regions of crossing fibers are characterized by low anisotropy and, thus, the determination of fiber orientation may be difficult in these voxels, 58 leading to unreliable tensor estimates. However, the fasciculi that we investigated are major WM tracts associated with high anisotropy; therefore, the effect of crossing fibers should not have affected our results a great deal, especially in the case of CST. 58 Despite such limitations, this study highlights the potential of DTI of the brain to provide in vivo markers of cerebral involvement in ALS. 7