Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print November 21, 2007, 10.1148/radiol.2461062075
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Supplemental Tables
Right arrow All Versions of this Article:
2461062075v1
246/1/222    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yu, C.
Right arrow Articles by Chan, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yu, C.
Right arrow Articles by Chan, P.
(Radiology 2008;246:222-228.)
© RSNA, 2007

Neuroradiology

Pathogenesis of Normal-appearing White Matter Damage in Neuromyelitis Optica: Diffusion-Tensor MR Imaging1

Chunshui Yu, MD, Fuchun Lin, PhD, Kuncheng Li, MD, Tianzi Jiang, PhD, Wen Qin, MS, Hong Sun, MD, and Piu Chan, MD

1 From the Departments of Radiology (C.Y., K.L., W.Q.) and Neurology (H.S., P.C.), Xuanwu Hospital, Capital University of Medical Sciences, 45 Chang-Chun St, Xuanwu District, Beijing 100053, People's Republic of China; and National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing, People's Republic of China (F.L., T.J.). Received December 6, 2006; revision requested February 15, 2007; revision received March 21; final version accepted May 9. Supported in part by the Natural Science Foundation of China (nos. 30670601, 30470519, 30425004 and 30570509), Beijing Scientific and Technological New Star Program (2005B21), and the National Key Basic Research and Development Program (2004CB318107). Address correspondence to K.L. (e-mail: kunchengli1955{at}yahoo.com.cn).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCE IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE...
 References
 
Purpose: To prospectively evaluate diffusion indexes of the corticospinal tract (CST), corpus callosum (CC), optic radiation (OR), and cingulum in patients with neuromyelitis optica (NMO) without visible lesions in the brain.

Materials and Methods: All participants provided informed consent, and the study was approved by the institutional review board. Nineteen patients with NMO (one man, 18 women; mean age, 35.1 years; range, 19–55 years) with normal brain magnetic resonance (MR) imaging findings and 19 sex- and age-matched healthy control subjects were examined with diffusion-tensor MR imaging. The CST, CC, OR, and cingulum were globally and regionally analyzed by using mean diffusivity, fractional anisotropy, and primary ({lambda}1) and transverse ({lambda}23) eigenvalues. Correlations of diffusion indexes of the CST and OR with the pyramidal and visual components of the Kurtzke Functional Systems (KFS) and Expanded Disability Status Scale scores were also investigated. Student t testing and Pearson correlation were performed.

Results: As compared with values in control subjects, both global and regional analyses showed significant (P < .01) increases in mean diffusivity and {lambda}23 of the CST and OR but not in any of the diffusion indexes of the CC and cingulum in patients with NMO. In patients with NMO, mean diffusivity (r = 0.556, P = .013) and {lambda}1 (r = 0.556, P = .013) of the CST were correlated with pyramidal KFS scores, and mean diffusivity (r = 0.523, P = .022) and {lambda}1 (r = 0.504, P = .027) of the OR were correlated with visual KFS scores.

Conclusion: Axonal degeneration secondary to lesions in the spinal cord and optic nerves is a cause of normal-appearing white matter damage in NMO.

Supplemental material: http://radiology.rsnajnls.org/cgi/content/full/2461062075/DC1

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCE IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE...
 References
 
Neuromyelitis optica (NMO) is a severe demyelinating disease characterized by selective involvement of the spinal cord and optic nerves and is considered to frequently spare the brain (1–3). Two recent studies (4,5) have revealed abnormal changes in normal-appearing gray matter and normal-appearing white matter (NAWM) in the brain in this disorder. However, the pathogenesis underlying such abnormalities has not been fully elucidated. We speculate that tract-based quantitative analyses of diffusion indexes may help to explain these findings.

To explain the diffusion abnormalities of brain NAWM in NMO, we hypothesize that they are caused by (a) secondary axonal degeneration, (b) small discrete lesions beyond the spatial resolution of conventional magnetic resonance (MR) imaging, or (c) both a and b. Hence, the purpose of our study was to prospectively evaluate diffusion indexes of the corticospinal tract (CST), corpus callosum (CC), optic radiation (OR), and cingulum in patients with NMO without visible lesions in the brain.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCE IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE...
 References
 
Participants
The institutional review board of Xuanwu Hospital approved the study, and written informed consent was obtained from each participant. From April 2003 to July 2005, we prospectively examined 120 patients with demyelinating diseases. Among these patients, 26 satisfied the diagnostic criteria for NMO (2). The diagnosis of NMO required all absolute criteria (optic neuritis, acute myelitis, and no evidence of clinical disease outside the optic nerves or spinal cord) and one major supportive criterion (negative brain MR imaging findings at onset, spinal cord MR imaging with signal intensity abnormality extending over three or more vertebral segments, cerebrospinal fluid pleocytosis of more than 50 white blood cells or more than five neutrophils per cubic millimeter) or two minor supportive criteria (bilateral optic neuritis; severe optic neuritis with fixed visual acuity worse than 20/200 in at least one eye; severe, fixed, attack-related weakness [Medical Research Council grade ≤ 2] in one or more limbs) (2). Two neuroradiologists (K.L., with 13 years of experience in MR image evaluation, and C.Y., with 10 years of experience) and two neurologists (H.S., with 11 years of experience in neurology, and P.C., with 18 years of experience) assigned final clinical diagnoses for all patients in consensus by reviewing all case histories, physical examination results, MR images, and laboratory data.

Nineteen of the 26 patients with NMO had normal brain MR imaging findings, while seven had abnormalities at MR imaging. The mean age of these seven patients was 53.4 years ± 4.9 (standard deviation) (range, 46–59 years); five of the patients had multiple age-related lesions in the white matter around the lateral ventricle, one had lacunar infarction in the left internal capsule, and one had a small lesion in the deep white matter of the right frontal lobe. Because brain lesions may create uncertainty in finding the principal diffusion direction when tractography is used, we included only the 19 patients (one man, 18 women; mean age, 35.1 years ± 10.9; range, 19–55 years) with normal brain MR imaging findings in our study. The mean duration of the disease was 4.3 years ± 2.5 (range, 1–8 years), the mean number of attacks was 4.5 ± 1.6 (range, two to eight), the mean Kurtzke Expanded Disability Status Scale (EDSS) score was 3.2 ± 1.4 (range, 0–5), the mean score for the pyramidal component of the Kurtzke Functional Systems (KFS) test was 1.6 ± 1.1 (range, 0–3), and the mean score for the visual component of the KFS test was 3.3 ± 2.0 (range, 0–6).

All patients with NMO had experienced optic neuritis, and in 12 of them it involved both optic nerves. In the 19 patients with NMO, the mean duration of optic neuritis was 3.8 years ± 2.7 (range, 1–8 years), and the mean number of optic neuritis attacks was 2.7 ± 1.8 (range, one to eight). Sixteen patients had attack-related weakness; the mean duration of weakness was 2.2 years ± 2.1 (range, 0–8 years), and the mean number of attacks was 2.9 ± 2.8 (range, zero to eight). Spinal MR images showed that 17 patients had cord lesions; 13 of these patients had a cord lesion in more than three vertebral segments.

Nineteen sex-and age-matched healthy volunteer control subjects (one man and 18 women; mean age, 35.3 years ± 10.9; range, 19–55 years) with no history of neurologic disorders and normal neurologic examination findings were recruited to serve as control subjects. None of the participating patients with NMO had been treated with related medications (eg, corticosteroids and immunosuppressants) within 3 months of the MR images being obtained. One neurologist (H.S.) performed the neurologic examinations in the patients and volunteers and generated the EDSS and KFS scores for each subject.

MR Imaging Examination
Brain and spinal cord imaging was performed in all patients by using a 1.5-T MR imaging unit (Sonata; Siemens Medical Systems, Erlangen, Germany). Diffusion-tensor MR imaging, turbo spin-echo T2-weighted, spin-echo T1-weighted, and fluid-attenuated inversion recovery sequences were used to image the brain of each subject. All sections of each sequence were consistently positioned to parallel a line that joined the most inferoanterior and inferoposterior parts of the CC. All sequences were performed with identical field of view (240 x 210 mm), number of sections (ie, 30), section thickness (4 mm), and intersection gap (0.4 mm). Detailed imaging parameters were as follows: Diffusion-tensor MR imaging was performed by using a spin-echo single-shot echo-planar imaging sequence with a repetition time msec/echo time msec of 5000/100, 10 signals acquired, and a matrix of 128 x 112. A total of seven image sets were acquired: six with noncollinear diffusion-weighting gradients and with a b value of 1000 sec/mm2, and one without diffusion weighting (b value, 0 sec/mm2). Turbo spin-echo T2-weighted imaging was performed with 5500/94, an echo-train length of 11, three signals acquired, and a matrix of 256 x 224. Spin-echo T1-weighted imaging was performed with 650/6, three signals acquired, and a matrix of 256 x 224. The fluid-attenuated inversion recovery sequence was performed with 8500/150, an inversion time of 2200 msec, an echo train length of eight, one signal acquired, and a matrix of 256 x 224.

The following pulse sequences were used to image the cervical and thoracic portions of the spinal cord, respectively: (a) turbo spin-echo T2-weighted imaging with 3000/120, three signals acquired, an echo train length of 20, a matrix of 512 x 256, and a field of view of 280 x 280 mm; and (b) spin-echo T1-weighted imaging with 500/12, three signals acquired, a matrix of 256 x 256, and a field of view of 280 x 280 mm. T2- and T1-weighted sequences were each performed to obtain seven 3-mm-thick sagittal sections with an intersection gap of 0.3 mm; T2-weighted sequences were also performed to obtain 20 5-mm-thick transverse sections with an intersection gap of 5 mm.

Preprocessing of Diffusion-Tensor MR Imaging Data
Distortion induced by eddy currents was first corrected by using free software (Diffusion Toolbox; Oxford Centre for Functional MRI of the Brain, Oxford, England, available at http://www.fmrib.ox.ac.uk/fsl/). The diffusion tensor of each voxel was calculated by using the linear least-square fitting algorithm (6). After the diffusion tensor was diagonalized, diffusion-tensor eigenvalues were obtained. The {lambda}1 eigenvalue measures the diffusion coefficient along the direction of maximum diffusivity. The {lambda}3 eigenvalue measures the diffusion coefficient along the direction of minimum diffusivity. The {lambda}2 eigenvalue measures the diffusivity along the third direction, which is orthogonal to the other two directions. The {lambda}23 eigenvalue, generated by averaging the {lambda}2 and {lambda}3 eigenvalues to avoid sorting bias, reflects the average diffusion coefficient perpendicular to the direction of maximum diffusivity (7). Mean diffusivity, or MD, and fractional anisotropy (FA) were derived for each pixel according to the following equations:

Formula
and

Formula

Here, mean diffusivity measures the average magnitude of molecular motion and FA measures its directionality.

Tractography
The 30 transverse sections were interpolated into 136 sections to obtain a voxel size of 0.94 x 0.94 x 0.94 mm. Then, tractography was performed according to the streamline method (8–10). All seed regions of interest (ROIs) were manually drawn by one author (C.Y.). We used two ROIs to reconstruct the CST. One was placed in the middle third of the cerebral peduncle, and another was placed in the posterior limb of the internal capsule (11,12). The first ROI was used as a starting region, and the second ROI was used as a filtering region to exclude non-CST fibers. We used two ROIs to reconstruct the OR. The first ROI was placed at the level of the trigone in the coronal plane, and the second ROI was located in the white matter of the occipital lobe in the parasagittal plane near the midline. The second ROI was set to filter out the fibers that did not terminate at the primary visual cortex. The seed ROI for tracing the CC was placed on the midsagittal FA image so that it encompassed the whole tract (13). Two seed ROIs in the coronal planes were used to reconstruct the cingulum: one passed through the genu-trunk junction of the CC, and the other was placed through the trunk-splenium junction of the CC. All of the ROIs were drawn on the FA maps. To reduce the partial volume effect, each voxel of the seed ROI was divided into 16 equally spaced starting seeds. Then the Frenet equation, which describes the evolution of a three-dimensional fiber tract, was solved numerically by using the fourth-order Runge-Kutta method within the imaging volume (14). Fiber tracking was terminated when a fiber made a turn of greater than 45° between two successive eigenvectors or when the FA value of a voxel fell below 0.2 (10,15). Reconstruction of the CSTs was terminated when the lowest section of the midbrain was reached because the fibers of the CST are crossed with transverse fibers in the pons, and this might result in artificial trajectories in the CST. Additional restrictions were not applied to the reconstruction of other tracts.

Tract-based Quantitative Analysis
After we tested for and did not find side differences (differences between left and right sides) of diffusion indexes of the CST, OR, and cingulum in both patients and control subjects, the region of each of the above tracts was defined by collecting all voxels traversed by the right and left sides of the tract. For example, the region of the CC was defined as all voxels traversed by the tract. Diffusion indexes of these tracts were then calculated for patients with NMO and healthy control subjects.

ROI Analysis
ROI analysis based on results of tractography was then performed by one author (C.Y.) in each tract to validate the findings from the entire tract analysis. For each tract, we selected one ROI in which the fibers were coherently arranged. The ROI of the CST was located in the middle transverse section of the cerebral peduncle (after interpolation, the cerebral peduncle could be observed in about 15 transverse sections; the middle transverse section was selected for analysis), the ROI of the CC was located in the midsagittal plane, the ROI of the OR was located in the coronal plane through the most posterior margin of the CC, and the ROI of the cingulum was located in the coronal plane through the most superior margin of the CC. After we tested for and did not find side differences of diffusion indexes of the ROIs of the CST, OR, and cingulum in both patients and control subjects, all pixels traversed by the bilateral fibers of a tract in an anatomically specified section were collected as pixels of the ROI in this section (Fig 1). Then the mean diffusivity, FA, {lambda}1, and {lambda}23 of each ROI were calculated and analyzed.


Figure 1A
View larger version (74K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: Placement of ROIs (in red) in different brain white matter tracts on FA images. ROIs were lo-cated (a) in the middle transverse section of the cerebral peduncle for the CST, (b) in the midsagittal plane for the CC, (c) in the coronal plane through the most posterior margin of the CC for the OR, and (d) in the coronal plane through the most superior margin of the CC for the cingulum.

 

Figure 1B
View larger version (79K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: Placement of ROIs (in red) in different brain white matter tracts on FA images. ROIs were lo-cated (a) in the middle transverse section of the cerebral peduncle for the CST, (b) in the midsagittal plane for the CC, (c) in the coronal plane through the most posterior margin of the CC for the OR, and (d) in the coronal plane through the most superior margin of the CC for the cingulum.

 

Figure 1C
View larger version (84K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1c: Placement of ROIs (in red) in different brain white matter tracts on FA images. ROIs were lo-cated (a) in the middle transverse section of the cerebral peduncle for the CST, (b) in the midsagittal plane for the CC, (c) in the coronal plane through the most posterior margin of the CC for the OR, and (d) in the coronal plane through the most superior margin of the CC for the cingulum.

 

Figure 1D
View larger version (89K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1d: Placement of ROIs (in red) in different brain white matter tracts on FA images. ROIs were lo-cated (a) in the middle transverse section of the cerebral peduncle for the CST, (b) in the midsagittal plane for the CC, (c) in the coronal plane through the most posterior margin of the CC for the OR, and (d) in the coronal plane through the most superior margin of the CC for the cingulum.

 
Reproducibility of Defining Seed ROIs for Tractography
Because the seed ROIs for tractography were manually defined by a single rater (C.Y.), the intrarater reproducibility of the positioning of the seed ROIs was determined by redefining these seed ROIs, reconstructing the tracts, and remeasuring the {lambda}1 and {lambda}23 of the four tracts in all 38 participants on two separate occasions (separated by at least 3 months).

Statistical Analysis
A Student t test was used to test the side differences in diffusion indexes of the CST, OR, and cingulum and to test the differences in diffusion indexes of each region between patients with NMO and control subjects because these diffusion indexes satisfied a normal distribution. To correct for multiple comparisons to reduce type I errors and to simultaneously reduce type II errors, P ≤ .01 was considered to indicate a statistically significant difference, as in previous studies (16,17). After we tested and confirmed the normality of clinical scores (EDSS, visual KFS, and pyramidal KFS), Pearson correlation coefficients were used to test the relationship of diffusion indexes in the ROI of the CST at the level of the cerebral peduncle with the EDSS and pyramidal KFS scores and the relationship of diffusion indexes in the ROI of the OR at the level of the most posterior margin of the CC with the EDSS and visual KFS scores. For these tests, P < .05 was considered to indicate a statistically significant difference. Intraclass correlation coefficients calculated with a one-way random effects model were used to test intrarater reproducibility (18). All of the statistical evaluations were performed with statistical software (SPSS for Windows, version 11.5; SPSS, Chicago, Ill). A retrospective power analysis was also performed for all significant results by using software (Power and Precision 2, available at http://www.power-analysis.com).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCE IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE...
 References
 
Reproducibility of Tractography
We were able to reconstruct the CST, CC, OR, and cingulum in all participants (Fig 2) in a manner consistent with the descriptions of known anatomy; thus, to some extent, we confirmed the reproducibility of the tractography method used. The intraclass correlation coefficient of the {lambda}1 was 98.52% for the CST, 97.26% for the OR, 98.23% for the CC, and 97.98% for the cingulum, and the intraclass correlation coefficient of the {lambda}23 was 98.78% for the CST, 97.47% for the OR, 98.50% for the CC, and 98.12% for the cingulum. These findings indicated an acceptable reproducibility in defining seed ROIs.


Figure 2A
View larger version (81K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: Reconstructed tracts of the (a) CST, (b) CC, (c) OR, and (d) cingulum were overlaid on coronal, sagittal, transverse, and sagittal T2-weighted MR images (5500/94), respectively. The fibers in red are consistent with known anatomy.

 

Figure 2B
View larger version (102K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: Reconstructed tracts of the (a) CST, (b) CC, (c) OR, and (d) cingulum were overlaid on coronal, sagittal, transverse, and sagittal T2-weighted MR images (5500/94), respectively. The fibers in red are consistent with known anatomy.

 

Figure 2C
View larger version (69K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: Reconstructed tracts of the (a) CST, (b) CC, (c) OR, and (d) cingulum were overlaid on coronal, sagittal, transverse, and sagittal T2-weighted MR images (5500/94), respectively. The fibers in red are consistent with known anatomy.

 

Figure 2D
View larger version (76K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d: Reconstructed tracts of the (a) CST, (b) CC, (c) OR, and (d) cingulum were overlaid on coronal, sagittal, transverse, and sagittal T2-weighted MR images (5500/94), respectively. The fibers in red are consistent with known anatomy.

 
Global Analyses of Tracts
We did not find side differences of the mean diffusivity, FA, {lambda}1, and {lambda}23 of the entire CST, OR, and cingulum (P > .05 for each index of each tract) in both patients and control subjects. In the global analysis of brain white matter tracts, the diffusion indexes of a tract on the left and right sides were then averaged to obtain one measure per tract. The mean diffusivity and {lambda}23 of the CST in patients with NMO were significantly higher (P < .001 for mean diffusivity and P = .003 for {lambda}23) than those in healthy control subjects, while the FA and {lambda}1 of the CST failed to show any significant changes. The mean diffusivity and {lambda}23 of the OR in patients with NMO were also significantly higher (P = .003 for mean diffusivity and P = .007 for {lambda}23) than those in healthy control subjects, while the FA and {lambda}1 of the OR failed to show any significant changes. Compared with healthy control subjects, patients with NMO did not show any significant differences in the mean diffusivity, FA, {lambda}1, and {lambda}23 of the CC and cingulum (Table E1, http://radiology.rsnajnls.org/cgi/content/full/2461062075/DC1).

ROI Analyses
We did not find side differences of the mean diffusivity, FA, {lambda}1, and {lambda}23 in the ROIs of the CST, OR, and cingulum (P > .05 for each index of each tract) in both patients and control subjects. In the ROI analysis of brain white matter tracts, the pixels traversed by the bilateral fibers of a tract in an anatomically specified section were collected as pixels of the ROI in that section. Compared with the values in healthy control subjects, the mean diffusivity and {lambda}23 of the CST at the middle transverse section of the cerebral peduncle in patients with NMO were significantly higher (P = .003 for mean diffusivity, P < .001 for {lambda}23), while the FA was lower (P = .007) and the {lambda}1 remained unchanged. The mean diffusivity and {lambda}23 of the OR at the coronal plane through the posterior margin of the CC were significantly increased (P = .007 for mean diffusivity and P = .006 for {lambda}23) in patients with NMO compared with these values in healthy control participants. However, none of the diffusion indexes showed any significant changes in the ROIs of the CC and cingulum in patients with NMO as compared with healthy control participants (Table E2, http://radiology.rsnajnls.org/cgi/content/full/2461062075/DC1).

Correlation Analysis
Correlations of diffusion indexes of the CST and OR with EDSS, pyramidal KFS, and visual KFS scores in patients with NMO are reported in the Table. In patients with NMO, none of the diffusion indexes of the CST or the OR were correlated with EDSS scores. However, the mean diffusivity (r = 0.556, P = .013) and {lambda}1 (r = 0.556, P = .013) of the CST were correlated with pyramidal KFS scores, and the mean diffusivity (r = 0.523, P = .022) and {lambda}1 (r = 0.504, P = .027) of the OR were correlated with visual KFS scores.


View this table:
[in this window]
[in a new window]

  		Correlations of Diffusion Indexes of CST and OR with Clinical Measures in Patients with NMO

 
Retrospective Power Analysis
Retrospectively, the powers of all significant results were calculated with {alpha} = .05 and a sample size of 19 for each group. These powers ranged from .81 to .97, with the exception of one outlier, which had a power of .79.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCE IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE...
 References
 
In previous MR imaging studies of NMO, attention was focused on the changes in the spinal cord and optic nerves, but not in the brain (1–3). More recent investigations in patients with NMO revealed occult damage in the normal-appearing gray matter and NAWM (4,5,19), although a study involving magnetization transfer imaging failed to reveal any abnormality in the brain in eight patients with NMO (20). One study by Rocca and colleagues (4) found abnormalities in the normal-appearing gray matter in 10 patients with NMO at diffusion-tensor and magnetization transfer MR imaging. Another study by this group (19) also showed an abnormal pattern of movement-associated cortical activations at functional MR imaging. Additionally, another diffusion-tensor MR imaging study showed that the abnormal changes in mean diffusivity and FA were present in both normal-appearing gray matter and NAWM (5). However, the above articles either did not report changes in diffusion-tensor eigenvalues or did not explain the pathogenesis of the abnormalities in detail.

In our study, we found that the mean diffusivity and {lambda}23 of the CST and OR in patients with NMO were significantly higher than those in healthy control subjects, while none of the mean diffusivity, FA, {lambda}1, and {lambda}23 values of the CC and cingulum showed any significant changes in patients with NMO as compared with these values in control subjects. We also validated these findings by selecting one ROI in each tract in which the fibers were coherently arranged. These results indicated that the abnormal diffusion was limited to the regions with connections to the spinal cord and optic nerves in patients with NMO and argue against the presence of small discrete lesions beyond the spatial resolution of conventional MR imaging in NAWM in NMO. Thus, the most likely pathogenesis of abnormal diffusion in brain white matter would appear to be axonal degeneration secondary to lesions in the spinal cord and optic nerves in patients with NMO.

The changes in diffusion-tensor eigenvalues may provide further information about disease (21) because they are the source indexes for calculating the mean diffusivity and FA. The {lambda}1 measures the diffusion coefficient along the direction of maximum diffusivity and reflects changes in restrictive barriers along the direction of a tract. The {lambda}23 measures the diffusion coefficient perpendicular to the direction of maximum diffusivity and mainly reflects changes in the axonal membrane and myelin sheath (22,23). In our study, we found that the abnormal changes in diffusion-tensor eigenvalues in the CST and OR were restricted to the directions transverse to the tracts, in line with characteristic findings at the chronic stage of axonal degeneration (23–25). A decrease in {lambda}1 was found at the acute stage of axonal degeneration that was caused by fragmentation of axons creating barriers to the longitudinal displacement of water molecules (26–28). At the chronic stage of axonal degeneration, {lambda}1 returns to normal because the axonal fragments are cleared, while {lambda}23 increases because of the degradation of axons and myelin sheaths (28,29). These findings also indicate that axonal degeneration secondary to lesions in the spinal cord and optic nerves may account for the abnormal diffusion in brain white matter in patients with NMO.

The pyramidal and visual KFS scores reflect the severity of the damage to motor and visual pathways, which mainly depends on lesions in the spinal cord and optic nerves in patients with NMO. The findings of the correlations of diffusion indexes of the CST and OR with the corresponding KFS scores indirectly indicate that the damage to the CST and OR is related to lesions in the spinal cord and optic nerves. That is to say, a patient with NMO with more severe damage in the motor pathway of the spinal cord often had a higher pyramidal KFS score and more severe damage to the CST. In the same way, a patient with NMO with more severe damage in the optic nerves often had a higher visual KFS score and more severe damage to the OR. Thus, these findings support the thinking that axonal degeneration secondary to lesions in the spinal cord and optic nerves may lead to the abnormal diffusion in brain white matter in patients with NMO. Additionally, we also partly excluded the presence of small discrete lesions in the brain on the basis of our finding of no significant correlations between diffusion indexes of the CST and OR and EDSS scores. Our study shows the potential of tract-based analysis of diffusion indexes to match abnormalities in appropriate anatomic fiber tracts to results with specific clinical scoring systems in patients with NMO and to help monitor the progress of the disease and assess the effectiveness of new therapeutic methods by measuring the mean diffusivity and {lambda}1 of the CST and OR.

Our study had limitations. One limitation was that this cohort of patients with NMO may have suffered from selection bias because we excluded patients with nonspecific brain MR imaging abnormalities because of a technical reason. A major limitation of the streamline method for tractography is the greater uncertainty in finding the principal diffusion direction when FA decreases (such as in brain lesions), which may lead to incorrect results. Nonspecific brain MR imaging abnormalities may result in abnormal diffusion in NAWM, but such changes cannot be attributed to NMO. Thus, we only recruited 19 patients with NMO with normal brain MR imaging findings for studying the pathogenesis of NAWM damage. Another limitation was that only six directions for diffusion encoding were used in our study, which may affect the accuracy of the estimation of diffusion tensors. The main reason for using six directions was to reduce the total image acquisition time so that it would be tolerable by most of the patients. An additional limitation was that the pathogenesis of white matter tract damage in NMO, albeit indirectly supported by our diffusion-tensor MR imaging findings, needs further confirmation in postmortem studies.

In conclusion, we found that (a) the abnormal diffusion in brain white matter was restricted to the tracts (CST and OR) with connections to the spinal cord and optic nerves, (b) diffusion indexes of CST and OR were correlated with clinical measures reflecting the involvement of these pathways, and (c) the abnormalities in diffusion-tensor eigenvalues of these tracts were restricted in the directions transverse to the tracts but did not appear in the direction parallel to the tracts. Our study results suggest that axonal degeneration secondary to lesions in the spinal cord and optic nerves might be a cause of the abnormal diffusion in NAWM in patients with NMO.


   ADVANCE IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCE IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE...
 References
 


   IMPLICATIONS FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCE IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE...
References
 


   FOOTNOTES
 

Abbreviations: CC = corpus callosum • CST = corticospinal tract • EDSS = Expanded Disability Status Scale • FA = fractional anisotropy • KFS = Kurtzke Functional Systems • NAWM = normal-appearing white matter • NMO = neuromyelitis optica • OR = optic radiation • ROI = region of interest

Guarantors of integrity of entire study, C.Y., F.L., K.L.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, C.Y., F.L., K.L., T.J., W.Q.; clinical studies, C.Y., W.Q., H.S., P.C.; statistical analysis, C.Y., F.L., H.S.; and manuscript editing, all authors

Authors stated no financial relationship to disclose.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCE IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE...
 References
 

  1. O'Riordan JI, Thompson AJ, Kingsley DP, et al. Clinical, CSF, and MRI findings in Devic's neuromyelitis optica. J Neurol Neurosurg Psychiatry 1996;60:382–387.
  2. Wingerchuk DM, Hogancamp WF, O'Brien PC, Weinshenker BG. The clinical course of neuromyelitis optica (Devic's syndrome). Neurology 1999;53:1107–1114. [Abstract/Free Full Text]
  3. Ghezzi A, Bergamaschi R, Martinelli V, et al. Clinical characteristics, course and prognosis of relapsing Devic's Neuromyelitis Optica. J Neurol 2004;251:47–52. [CrossRef][Medline]
  4. Rocca MA, Agosta F, Mezzapesa DM, et al. Magnetization transfer and diffusion tensor MRI show gray matter damage in neuromyelitis optica. Neurology 2004;62:476–478. [Abstract/Free Full Text]
  5. Yu CS, Lin FC, Li KC, et al. Diffusion tensor imaging in the assessment of normal-appearing brain tissue damage in relapsing neuromyelitis optica. AJNR Am J Neuroradiol 2006;27:1009–1015. [Abstract/Free Full Text]
  6. Basser PJ, Mattiello J, LeBihan D. Estimation of the effective self-diffusion tensor from the NMR spin echo. J Magn Reson B 1994;103:247–254. [CrossRef][Medline]
  7. Basser PJ, Pajevic S. Dealing with uncertainty in diffusion tensor MR data. Isr J Chem 2003;43:129–144. [CrossRef]
  8. Conturo TE, Lori NF, Cull TS, et al. Tracking neuronal fiber pathways in the living human brain. Proc Natl Acad Sci U S A 1999;96:10422–10427. [Abstract/Free Full Text]
  9. Basser PJ, Pajevic S, Pierpaoli C, Duda J, Aldroubi A. In vivo fiber tractography using DT-MRI data. Magn Reson Med 2000;44:625–632. [CrossRef][Medline]
  10. Catani M, Jones DK, Donato R, Ffytche DH. Occipito-temporal connections in the human brain. Brain 2003;126:2093–2107. [Abstract/Free Full Text]
  11. Glenn OA, Henry RG, Berman JI, et al. DT-MRI-based three-dimensional tractography detects differences in the pyramidal tracts of infants and children with congenital hemiparesis. J Magn Reson Imaging 2003;18:641–648. [CrossRef][Medline]
  12. Lee JS, Han MK, Kim SH, Kwon OK, Kim JH. Fiber tracking by diffusion tensor imaging in corticospinal tract stroke: topographical correlation with clinical symptoms. Neuroimage 2005;26:771–776. [CrossRef][Medline]
  13. Hagmann P, Thiran JP, Jonasson L, et al. DTI mapping of human brain connectivity: statistical fibre tracking and virtual dissection. Neuroimage 2003;19:545–554. [CrossRef][Medline]
  14. Press WH, Teukolsky SA, Vetterling WT, Flannery BP. Numerical recipes in C: the art of scientific computing. Cambridge, England: Cambridge University Press, 1992; 711.
  15. Kunimatsu A, Aoki S, Masutani Y, et al. The optimal trackability threshold of fractional anisotropy for diffusion tensor tractography of the corticospinal tract. Magn Reson Med Sci 2004;3:11–17. [CrossRef][Medline]
  16. Filippi M, Cercignani M, Inglese M, Horsfield MA, Comi G. Diffusion tensor magnetic resonance imaging in multiple sclerosis. Neurology 2001;56:304–311. [Abstract/Free Full Text]
  17. Bozzali M, Falini A, Cercignani M, et al. Brain tissue damage in dementia with Lewy bodies: an in vivo diffusion tensor MRI study. Brain 2005;128:1595–1604. [Abstract/Free Full Text]
  18. McGraw KO, Wong SP. Forming inferences about some intraclass correlations. Psychol Methods 1996;1:30–46. [CrossRef]
  19. Rocca MA, Agosta F, Mezzapesa DM, et al. A functional MRI study of movement-associated cortical changes in patients with Devic's neuromyelitis optica. Neuroimage 2004;21:1061–1068. [CrossRef][Medline]
  20. Filippi M, Rocca M, Moiola L, et al. MRI and magnetization transfer imaging changes in the brain and cervical cord of patients with Devic's neuromyelitis optica. Neurology 1999;53:1705–1710. [Abstract/Free Full Text]
  21. Hasan KM. Diffusion tensor eigenvalues or both mean diffusivity and fractional anisotropy are required in quantitative clinical diffusion tensor MR reports: fractional anisotropy alone is not sufficient. Radiology 2006;239:611–612. [Free Full Text]
  22. Le Bihan D. Looking into the functional architecture of the brain with diffusion MRI. Nat Rev Neurosci 2003;4:469–480.
  23. Pierpaoli C, Barnett A, Pajevic S, et al. Water diffusion changes in Wallerian degeneration and their dependence on white matter architecture. Neuroimage 2001;13:1174–1185. [Medline]
  24. Thomalla G, Glauche V, Koch MA, Beaulieu C, Weiller C, Rother J. Diffusion tensor imaging detects early Wallerian degeneration of the pyramidal tract after ischemic stroke. Neuroimage 2004;22:1767–1774. [CrossRef][Medline]
  25. Henry RG, Oh J, Nelson SJ, Pelletier D. Directional diffusion in relapsing-remitting multiple sclerosis: a possible in vivo signature of Wallerian degeneration. J Magn Reson Imaging 2003;18:420–426. [CrossRef][Medline]
  26. George R, Griffin JW. The proximo-distal spread of axonal degeneration in the dorsal columns of the rat. J Neurocytol 1994;23:657–667. [CrossRef][Medline]
  27. Kerschensteiner M, Schwab ME, Lichtman JW, Misgeld T. In vivo imaging of axonal degeneration and regeneration in the injured spinal cord. Nat Med 2005;11:572–577. [CrossRef][Medline]
  28. Concha L, Gross DW, Wheatley BM, Beaulieu C. Diffusion tensor imaging of time-dependent axonal and myelin degradation after corpus callosotomy in epilepsy patients. Neuroimage 2006;32:1090–1099. [CrossRef][Medline]
  29. George R, Griffin JW. Delayed macrophage responses and myelin clearance during Wallerian degeneration in the central nervous system: the dorsal radiculotomy model. Exp Neurol 1994;129:225–236. [CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Supplemental Tables
Right arrow All Versions of this Article:
2461062075v1
246/1/222    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yu, C.
Right arrow Articles by Chan, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yu, C.
Right arrow Articles by Chan, P.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE