Abstract
Objective: To assess corticospinal tract involvement in patients with amyotrophic lateral sclerosis (ALS) by correlating diffusion tensor imaging (DTI) measures with intra- and extracranial central motor conduction time (CMCT) and clinical features of the patients.
Methods: We investigated 31 patients with ALS and 31 normal volunteers by DTI and measured fractional anisotropy (FA) within the corticospinal tracts and in the extramotor white matter. We measured CMCT for the first dorsal interosseous muscle and segmented it into cortical-brainstem (CTX-BS CT) and brainstem-cervical root (BS-CV CT) conduction times by magnetic brainstem stimulation at the foramen magnum level. Clinical status of each patient was evaluated with the ALS Functional Rating Scale–Revised (ALSFRS-R) and upper motor neuron (UMN) score devised for this study.
Results: We found a significant decrease of mean FA in all regions of the corticospinal tracts in patients with ALS as compared with controls. We found that FA along the corticospinal tract decreased significantly with higher UMN scores. There was no significant correlation between FA and ALSFRS-R, to which both upper and lower motoneuron involvements contribute. FA showed a significant correlation with the intracranial part of the central motor conduction (CTX-BS CT) but not with the extracranial conduction time.
Conclusions: Fractional anisotropy reflects functional abnormality of intracranial corticospinal tracts and can be used for objective evaluation of upper motor neuron impairment in amyotrophic lateral sclerosis.
GLOSSARY: ALS = amyotrophic lateral sclerosis; ALSFRS-R = ALS Functional Rating Scale–Revised; BS-CV CT = brainstem-cervical root conduction time; CMCT = central motor conduction time; CTX-BS CT = cortical-brainstem conduction time; FA = fractional anisotropy; FDI = first dorsal interosseous; LMN = lower motor neuron; ROI = region of interest; UMN = upper motor neuron.
Amyotrophic lateral sclerosis (ALS) is clinically diagnosed by lower motor neuron (LMN) signs of limb and bulbar muscles associated with upper motor neuron (UMN) signs. Subclinical LMN involvement is detectable by needle electromyographic findings of denervation, which are incorporated in revised El Escorial criteria.1 UMN involvement can be evaluated by physiologic measures or neuroimaging techniques,2 although these have not been sufficiently well established to be incorporated into diagnostic criteria. Diffusion tensor MRI visualizes the overall orientation of the fiber tracts and their integrity in the white matter by measuring anisotropic water diffusion.3 Decreased fractional anisotropy (FA) along the corticospinal tract has recently been reported in patients with ALS.4,5 However, the pathophysiology of such reduced FA remains unclear. The present investigation was undertaken with the intent of clarifying the mechanism for reduced FA in ALS by studying correlations of the FA value with central motor conduction time segmented into intracranial and extracranial conduction times using brainstem stimulation.
METHODS
Subjects.
We recruited 31 patients with ALS and 31 age-matched normal subjects (ALS 60.7 ± 12.9 years, normal 57.1 ± 13.0 years, p = 0.166). The Ethical Review Committee of the University of Tokyo approved this study. All subjects gave their written informed consent to participate in the study. All patients were enrolled if they met definite, probable, or possible categories of revised El Escorial criteria.1 The degree of abnormality was quantified using the ALS Functional Rating Scale-Revised (ALSFRS-R). The UMN score was designed to assess UMN impairment. The following neurologic signs were rated on a 0 to 2 scale according to their severity (0 = absent or normal, 1 = moderately impaired, and 2 = greatly impaired): jaw jerk, other pathologic reflexes of the cranial regions, overactive tendon reflexes in upper limbs, overactive finger flexor reflexes, overactive tendon reflexes in lower limbs, pathologic reflexes in lower limbs, spasticity, and presence of clonus. The scale generates a score from 0 to 16.
Transcranial magnetic stimulation of corticospinal pathways.
Central motor conduction time (CMCT) was measured with methods described previously,6 recorded from the first dorsal interosseous (FDI) muscles. A round coil was used for motor cortical and spinal motor root stimulation, and a double cone coil was used for brainstem stimulation. CMCT (motor-evoked potential latency difference between motor cortical and cervical root simulation), cortical-brainstem conduction time (CTX-BS CT, latency difference between motor cortical and brainstem stimulation), and brainstem-cervical root conduction time (BS-CV CT, latency difference between brainstem and cervical root stimulation) were calculated and evaluated by neurophysiologists blinded to MRI results.
Diffusion tensor MRI scanning protocol.
Diffusion tensor images were acquired with 1.5-tesla Signa Horizon LX MRI system (GE Medical Systems), using single-shot spin-echo echoplanar sequences (repeat time 6,000 msec, echo time 78 msec, field of view 24 cm, NEX 4, 128 × 128-pixel matrix, diffusion gradients [b-value of 1,000 sec/mm2], 3-mm slice thickness). Diffusion properties were measured along 13 noncollinear directions. FA was measured using a region-of-interest (ROI) method. Elliptical ROIs were placed along bilateral corticospinal tracts (corona radiata, internal capsule, cerebral peduncle, basis pontis, and medulla oblongata) and extramotor white matters (genu and splenium of the corpus callosum, superior, middle, and inferior cerebellar peduncle, and cerebellar white matter) on FA maps by one author blinded to subject clinical status, based on empirical anatomic knowledge and reference to pertinent literature.
Statistical analyses.
We used a two-way analysis of variance (ANOVA) (factors of subject group and region). We used Scheffe analysis as post hoc multiple comparisons (significance level 0.05). Linear regression analyses were applied for all correlations (significance level 0.05) using StatView software (version 5; SAS Institute). Because FA is reported to decline with advancing age,7 correlations between FA and clinical or physiologic measures are examined by using ratio of FA at each ROI to that of the splenium of the corpus callosum, to compensate for interindividual variability of absolute FA values. In evaluation of correlations between FA and CMCT as well as CTX-BS CT and BS-CV CT, we used all FA values and compatible physiologic measures, such as a FA on one side and a physiologic measure for the contralateral FDI. When a correlation between anisotropy data and ALSFRS-R or UMN score was analyzed, we averaged values from right and left sides to provide a single mean FA at a site for each individual.
RESULTS
Fractional anisotropy.
Individual plots of FA at each ROI for patients with ALS and controls are shown in figure 1. Two-way ANOVA showed an effect of the subject group (patient and control) and region (effect of subject group: F = 205.763, p < 0.0001; effect of region: F = 395.421, p < 0.0001). It also showed an interaction between the subject group and region (F = 25.457, p < 0.0001). Post hoc analyses showed that the mean FA was lower in patients with ALS than in controls in all ROI within the corticospinal tracts (the corona radiata, posterior limb of the internal capsule, cerebral peduncle, basis pontis, pyramid of the medulla oblongata) (p < 0.0005). No significant differences were found within extramotor white matter. FA decreased significantly with higher UMN scores at corona radiata, internal capsule, and pyramids of medulla oblongata. No correlation was apparent between UMN scores and FA in extramotor white matter. FA showed no significant correlation with ALSFRS-R in any ROI (table 1).
Figure 1 Individual plots of fractional anisotropy (FA) at each region of interest (ROI) for patients with amyotrophic lateral sclerosis (filled circles) and controls (open circles) (mean ± SD)
Table 1 Correlation between fractional anisotropy (FA) and clinical/physiologic measures
Transcranial magnetic stimulation of corticospinal pathways.
We examined 47 limbs of 25 patients. In 16 limbs of 10 patients, no responses were obtained with motor cortical, brainstem, or motor root stimulation. The averaged ALSFRS-R of these patients was 29.7 ± 11.7, which was worse than that of the rest of the patients (36.7 ± 8.2, p < 0.05). There was no difference in the averaged UMN scores between the two patient groups (6.3 ± 3.8 and 6.2 ± 4.1, p > 0.05). Theoretically, unobtainable responses are attributable to cortical inexcitability resulting from motor cortical cell loss, severe peripheral involvement, or a combination of both. However, in the patients studied here, based on the above results of correlations, we can infer that dysfunction of LMN contributes more than that of the UMN to the lack of responses. These absent responses were excluded from the following correlation analyses because there were no measurable latencies.
In all, we obtained 43 CMCTs and 31 CTX-BS CTs as well as BS-CV CTs. The averaged CMCT of FDI of the patients was 8.5 ± 3.4 msec (the average ± SD of the normal subjects at our facility was 7.0 ± 0.4 msec). Seventeen of 43 CMCTs were abnormally delayed (above the average + 2SD of the normal values). The average of CTX-BS CTs from the patients was 4.4 ± 3.0 msec (the normal average was 3.3 ± 0.3 msec). Twelve of 31 CTX-BS CTs were delayed. The average of BS-CV CTs of the patients was 4.3 ± 2.9 (the normal average was 3.7 ± 0.5), and 12 of 31 BS-CV CTs were abnormally prolonged. We found overall abnormal results including absent responses to either cortical or root stimulation, and delayed responses, in 44.7% of all the limbs studied. Delayed CMCT, CTX-BS CT, and BS-CV CT were found in 39.5%, 38.7%, and 38.7% of recorded responses. CMCT and CTX-BS CT correlated significantly with both ALSFRS-R and UMN scores, but BS-CV CT correlated only with ALSFRS-R (table 2).
Table 2 Correlations between central motor conduction times and clinical indices
Correlation between FA and CMCTs.
FAs at most regions along the corticospinal tract decreased significantly with delayed CMCT and CTX-BS CT (table 1, figure 2A). However, BS-CV CT did not correlate with FA in any ROI (figure 2B). No significant correlation was found between FA of extramotor regions and CMCT or CTX-BS CT.
Figure 2 Regression plots of fractional anisotropy at the internal capsule as a function of cortical-brainstem conduction time (CTX-BS CT) (A) and brainstem-cervical root conduction time (BS-CV CT) (B)
Correlation was only apparent between FA and cortical-brainstem conduction time (|r| = 0.535, p = 0.002). No correlation was observed between FA and the extracranial conduction time (brainstem-cervical root conduction time) (|r| = 0.249, p = 0.176).
DISCUSSION
In ALS, we have demonstrated reduced FA restricted to the corticospinal tracts. We also found that FA along the corticospinal tract decreased with higher UMN scores or delayed CTX-BS CT. The brainstem stimulation is considered to differentiate the CMCT delay due to an intracranial lesion from that due to an extracranial spinal lesion.8 The CTX-BS CT must purely reflect UMN function, whereas the BS-CV CT must mostly reflect LMN function, and CMCT both LMN and UMN functions. Based on this theory, our present results suggest that FA measurement can evaluate UMN function in patients.
We showed that CTX-BS CT, but not BS-CV CT, delayed significantly with decreased FA at sites of the corticospinal tracts other than the cerebral peduncle where CSF has a greater partial volume effect on images. Decreased FA suggests tract degeneration that engenders the loss of organized coherent structures. FA changes are explainable by both intracellular water diffusion changes and extracellular matrix changes. Previous histopathologic studies of ALS suggest that the former corresponds to degeneration of the corticospinal tract axon itself with associated astrocytosis and accumulation of axonal spheroids, whereas the latter corresponds to extracellular matrix expansion and astrocytosis within interaxonal spaces. Both processes can cause reduced anisotropy of water diffusion. Meanwhile, slowing of the conduction time is considered to result from the loss of larger and faster conducting neurons and reduction of functioning rapidly conducting axons following corticospinal cell loss.9 From animal experiments, the magnitude of latency delay by failed firing attributable to functioning fiber loss is estimated to be a few milliseconds at maximum from the cortex to cervical spinal cord. A striking increase of the conduction time in excess of this range would suggest slowing of conduction itself, which might be attributable to conduction through slowly conducting fibers due to degeneration of rapidly conducting fibers, or secondary demyelination when cortical neuronal loss is severe. These inferences are supported by white matter histopathology of the corticospinal tract: Myelin loss is commonly observed, especially in advanced patients, and the severity of that loss is generally related to neuronal loss of the motor cortex.10 With the finding that the FA along the corticospinal tract decreased with intracranial motor conduction delay in ALS, we can infer that impaired axonal function, rather than extracellular factors, mainly contributes to the reduced FA. Demyelination secondary to motor axonal loss may add water diffusion changes in patients with excessive delayed motor conduction. This idea is consistent with the current view of determinants of anisotropy that the primary contributor is axonal membrane function, whereas other microstructures such as the myelin sheath, the neurofibrils (microtubules, neurofilaments), and axonal transport can play a secondary modulating role.11 For more precise elucidation on potential determinants of anisotropic changes in ALS, thorough comparative studies of diffusion tensor imaging and postmortem specimens are necessary.
We demonstrated a significant correlation of FA with other clinical or physiologic indices. Potential applications of this method for patients with ALS include its use as an objective marker in following the natural course of the disease or modified course in therapeutic trials, or detecting a mild lesion of the corticospinal tracts at early stages.
ACKNOWLEDGMENT
The authors thank Dr. Peter T. Lin for helpful comments.
Footnotes
-
Supported by Research Project Grant-in-Aid for Scientific Research 16500194 from the Ministry of Education, Culture, Sports, Science, and Technology of Japan; Research Grant 15B-2 for Nervous and Mental Disorders from the Ministry of Health, Labor, and Welfare of Japan; a grant from the Committee of the Study of Human Exposure to EMF; the Ministry of Internal Affairs and Communications; grants from the Life Science Foundation of Japan and the Association of Radio-industry and Business; and the Nakabayashi Trust for ALS Research.
Disclosure: The authors report no conflicts of interest.
Received September 21, 2005. Accepted in final form August 8, 2007.
Disputes & Debates: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.