| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Neuroradiology |
1 From the Departments of Radiology (S.C., D.C.S., A.D.A.) and Neurology (D.J.L., L.P.R.), Neurological Institute of New York, Columbia University, Milstein Hospital Bldg, 3rd Floor, 177 Fort Washington Ave, New York, NY 10032. Received July 16, 1998; revision requested August 12; revision received November 17; accepted January 27, 1999. S.C. supported by an RSNA Scholars Award and a General Electric-AUR Radiology Research Academic Fellowship. D.C.S. supported by an R-29 award from the National Institute of Neurological Disorders and Stroke (National Institutes of Health). Address reprint requests to S.C.
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: In 43 of 50 patients with motor neuron disease and 14 of 14 control subjects, matching sets of MR spectra were obtained in the left and right motor cortex. The ratio of N-acetylaspartate (NAA) to creatine (Cr) was derived from peak area measurements. Mean ratios were calculated for control subjects and several patient groups, including patients with amyotrophic lateral sclerosis (ALS) or primary lateral sclerosis (PLS). MR images were evaluated for corticospinal tract hyperintensity and central sulcus dilatation.
RESULTS: Mean NAA/Cr values were significantly different between control subjects and the ALS or PLS groups (P < .05). With an optimal cutoff of 2.5, NAA/Cr values were abnormal in 15 (79%) of 19 patients with ALS, 12 (67%) of 18 patients with PLS, and one (7%) of 14 control subjects. Corticospinal tract hyperintensity, central sulcus enlargement, or both were found in 43% of the ALS group, 24% of the PLS group, and 7% of the control group.
CONCLUSION: NAA/Cr values determined at single-voxel proton MR spectroscopy are more sensitive than are standard findings at MR imaging in the detection of upper motor neuron disease.
Index terms: Amyotrophic lateral sclerosis, 13.89 • Brain, cortex, 13.89 • Brain, diseases, 13.89 • Brain, MR, 13.121411, 13.121412, 13.121416, 13.12145 • Magnetic resonance (MR), comparative studies, 13.121411, 13.121412, 13.121416, 13.12145 • Magnetic resonance (MR), spectroscopy, 13.12145
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
In single-voxel MR spectroscopy, a specific region of the brain must be targeted for study. The motor cortex is usually affected in ALS, so targeting of the motor cortex at single-voxel MR spectroscopy should yield relevant clinical information. Although single-voxel MR spectroscopy does not have the spatial resolution of MR spectroscopic imaging, it is now widely available on commercial MR imagers, is fully automated after voxel placement, and requires only 7–15 minutes of additional imaging time. Therefore, single-voxel MR spectroscopy may be implemented as part of a clinical MR imaging study with no increase in patient discomfort or cost. In this study, we evaluated use of single-voxel MR spectroscopy to help distinguish patients with disorders involving the upper motor neuron, such as ALS, from those with non–upper motor neuron disorders.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The patient population comprised 28 men and 24 women with motor neuron disease (age range, 31–78 years; mean age, 60.5 years; SD, 10.7 years). The clinical diagnoses were based on institutional criteria for four broad groups: (a) ALS; (b) ALS with probable upper motor neuron signs; (c) primary lateral sclerosis (PLS); and (d) non–upper motor neuron disorders, which included pure lower motor neuron disorders (such as progressive spinal muscular atrophy or multifocal motor neuropathy) and focal spinal cord disorders, including the myelopathy of multiple sclerosis or spinal cord trauma.
Control subjects comprised eight men and six women (age range, 45–77 years; mean age, 56.8 years; SD, 10.7 years) who were recruited by means of an advertisement. People were excluded who were younger than 45 years or who had a known neurologic disease.
Institutional Criteria
Patients with motor neuron disease were divided into five diagnostic categories on the basis of the presence or absence of lower motor neuron signs or upper motor neuron signs, respectively. Lower motor neuron signs include focal weakness and wasting and is often accompanied by overt fasciculation. Upper motor neuron signs include Hoffmann sign, Babinski sign, clonus, snout reflex, or jaw jerk, any one of which was considered an unequivocal sign of upper motor neuron disease. The presence of brisk tendon reflexes (hyperreflexia) was considered a "probable" sign of upper motor neuron disease.
ALS is a progressive disorder denoted by the presence of both lower and upper motor neuron signs.
ALS with probable upper motor neuron signs is a progressive disorder denoted by the presence of brisk tendon reflexes in limbs with unequivocal lower motor neuron signs and the absence of unequivocal upper motor neuron signs.
Adult-onset progressive spinal muscular atrophy is a progressive disorder characterized by the presence of lower motor neuron signs and the absence of tendon reflexes and upper motor neuron signs. Electromyography reveals no conduction abnormality.
Multifocal motor neuropathy is a clinical syndrome similar to progressive spinal muscular atrophy except for the presence of conduction block or slowing at electromyography.
PLS is a syndrome with slowly progressive gait disturbance and only upper motor neuron signs. The diagnosis is one of exclusion after consideration of multiple sclerosis, spinal cord trauma, compression of the cervical spinal cord, syringomyelia, bilateral strokes, and other conditions. Electromyography reveals no conduction abnormality.
MR Studies
All MR studies were conducted with a 1.5-T MR imager (Signa; GE Medical Systems, Milwaukee, Wis). For all patients, a standard brain MR imaging protocol included the following: (a) scout T1-weighted spin-echo (SE) sagittal imaging (repetition time msec/echo time msec = 500/14–16 with one signal acquired); (b) high-spatial-resolution, intermediate-weighted, fast SE axial imaging (2,500/21; echo train length, four; field of view, 20 cm; two signals acquired; imaging matrix, 256 x 256; section thickness, 5 mm with 1.5-mm gap; flow compensation; imaging time, 5.5 minutes); and (c) standard T2-weighted fast SE axial and coronal imaging (4,000/96; echo train length, eight; field of view, 22 cm at 75% field of view; one or two signals acquired; imaging matrix, 192 x 256; section thickness, 5 mm with 1.5-mm gap; flow compensation; imaging time, 2–3 minutes). The brain MR imaging study was followed by the brain MR spectroscopic study. Finally, in 35 of the 52 patients, an adequate MR imaging study of the cervical spine was obtained, which included high-spatial-resolution axial imaging of the cervical cord (4,000/92; echo train length, eight; field of view, 16 cm; four signals acquired; imaging matrix, 256 x 256; section thickness, 5 mm with 1.5-mm gap; imaging time, 8.5 minutes).
Because of limited availability of the MR imager, the control subjects underwent abbreviated MR imaging and MR spectroscopic studies of the brain. These consisted mainly of T1-weighted SE sagittal imaging (500/14–16 with one signal acquired) and single-echo T2-weighted fast SE axial imaging (4,000/96 with one or two signals acquired) followed by proton MR spectroscopy. Intermediate-weighted axial imaging was omitted, as was the MR imaging study of the cervical spine.
The single-voxel proton MR spectroscopic studies were performed with a point-resolved spectroscopic sequence, or PRESS, supplied by the manufacturer (2,000/272 [SE delay] with 96 signals acquired). A single 8-cm3 voxel (2 x 2 x 2 cm) was placed over the motor cortex (Fig 1a, 1b), first on the right side and then on the left side. In all cases, correct and reproducible voxel placement was achieved by first locating the central sulcus in the most superior sections on the T2-weighted axial images (9) and then tracing it down to the axial section located about 1 cm above the top of the lateral ventricles (defined as the "plan section"). The voxel was then centered on the precentral gyrus. To ensure maximal sampling of brain parenchyma within the boundaries of the voxel, the location of the voxel was adjusted to fulfill the following three conditions: (a) the posterior edge of the voxel intersected or came within 2 mm of the obliquely oriented central sulcus at some point in the plan section, (b) the medial edge of the voxel did not include the lateral ventricle in the section 1 cm inferior to the plan section (the section that included the top of the lateral ventricles), and (c) the lateral edge of the voxel did not overlap the bone marrow of the calvarium in the axial section 1 cm superior to the plan section. Preimaging required about 3 minutes, and imaging required 4 minutes 24 seconds, or about 7.5 minutes to obtain each MR spectrum of the motor cortex. The entire MR spectroscopic examination required about 15 minutes.
|
|
|
|
Sensitivity and specificity analyses were performed for a range of NAA/Cr values to generate a receiver operating characteristic curve (12) and to determine the optimal cutoff of NAA/Cr for the detection of upper motor neuron abnormality. The optimal cutoff point was defined as the NAA/Cr value at which the sum of the sensitivity and the specificity was maximized (13). An MR spectroscopic study was considered abnormal if the NAA/Cr value was lower than the cutoff in the right motor cortex, the left motor cortex, or both. For the purpose of identifying an optimal cutoff point for each patient group, we defined the disease groups as the combined group, with ALS alone and ALS with probable upper motor neuron signs, or the PLS group, and the nondisease group as the control group. Spectroscopic data from the groups with PLS or non–upper motor neuron disorders were analyzed separately because, in comparison to ALS alone and ALS with probable upper motor neuron signs, less is known about the severity and pathologic extent of upper motor neuron involvement in PLS or lower motor neuron disorders.
The high-spatial-resolution MR images were interpreted prospectively by one neuroradiologist (S.C.), who was blinded to clinical diagnosis. He evaluated the MR images for the following signs: (a) focal increased signal intensity on intermediate-weighted axial images within the corticospinal tracts in the posterior limb of the internal capsule (Fig 2) (14), (b) disproportionate enlargement of the central sulcus on T1-weighted sagittal images (Fig 3) (15), and (c) focally increased signal intensity in the lateral columns of the cervical cord on T2-weighted axial images (16). For purposes of reliability, a second blinded neuroradiologist (A.D.A.) retrospectively reviewed the 52 MR studies of the brain to determine the presence or absence of each of the first two findings. The retrospective readings were then compared with the prospective readings; disagreement among readings was resolved by means of consensus review. Interrater reliability was determined with the 
statistic (17). Sensitivity and specificity analyses were performed of the cerebral MR imaging findings; comparative receiver operating characteristic data were generated by using the corticospinal tract finding alone or the combination of the corticospinal tract and the central sulcus findings. No retrospective reading was performed of the 35 MR studies of the cervical spine because (a) the initial prospective readings indicated a low sensitivity for detection of corticospinal tract abnormalities within the cord (n = 2) and (b) the main focus of the study was the comparison of MR spectroscopy of the motor cortex with MR imaging of the brain.
|
|
|
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
|
|
The sensitivity of specific MR imaging abnormalities for upper motor neuron disease was compared with that of abnormal NAA/Cr at single-voxel MR spectroscopy (Table 2). Interrater reliability measurements for brain MR image readings yielded the following: corticospinal tract hyperintensity, = 0.66; central sulcus enlargement, = 0.57; overall, = 0.62. A value of 0.61–0.81 is taken as good interrater agreement and of 0.41–0.60 as moderate agreement (17).
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The MR spectroscopic decision criterion involved detection of an abnormality within the motor cortex of either one or both cerebral hemispheres. With this criterion, the MR spectroscopic findings in 79% (15 of 19) of patients with ALS alone and ALS with probable upper motor neuron signs and 67% (12 of 18) of patients with PLS were deemed abnormal. These percentages compare favorably to those associated with both MR imaging abnormalities, either corticospinal tract hyperintensity or central sulcus enlargement: 43% (nine of 21) for ALS alone and ALS with probable upper motor neuron signs and 24% (five of 21) for PLS. With the same criterion, a false-positive MR spectroscopic diagnosis was seen in 7% (one of 14) of control subjects and 17% (one of six) of patients with non–upper motor neuron disorders. We note that it is likely that some false-positive findings of non–upper motor neuron disorders were clinically occult upper motor neuron disease; in autopsy examination of patients with motor neuron disease without upper motor neuron signs (including adult-onset progressive spinal muscular atrophy), more than half (17 of 25 patients) had degeneration of the corticospinal tracts (20,21). In fact, postmortem examination of one of our patients with non–upper motor neuron disorders (who had clinically diagnosed spinal muscular atrophy) revealed degeneration of the corticospinal tracts. This pathologic result correlates well with the antemortem findings at MR spectroscopy (NAA/Cr = 1.7) and MR imaging that were indicative of upper motor neuron disease.
The spectroscopic abnormality was unilateral in five of 15 patients with ALS alone or ALS with probable upper motor neuron signs and in seven of 12 PLS patients. This could arise from asymmetric loss of motor neurons in the two cerebral hemispheres. In one patient with ALS who had both left-sided corticospinal tract involvement and left central sulcus enlargement, NAA/Cr was significantly lower in the left (1.8) than in the right (2.4) motor cortex. However, as the full extent of variation in NAA/Cr in the motor cortex is not known, more data are needed to ascertain the importance of cerebral asymmetries seen at MR imaging and MR spectroscopy.
Several types of MR abnormalities have been found in patients with ALS or PLS. First, the corticospinal tract can be identified readily on intermediate- and T2-weighted images in the posterior third of the posterior limb of the internal capsule (22). In one MR imaging study in patients with ALS (14), hyperintensity of the corticospinal tract was identified primarily at the level of the posterior limb of the internal capsule, but, in some cases, high signal intensity extended up to the cortex or down through the cerebral peduncle and brainstem. Findings in another study suggested that this abnormality, when present, is characteristic of ALS but is not sensitive for the diagnosis of ALS (18). This abnormality was also reported in one case of PLS (23). Second, atrophy of the precentral gyrus may be identified indirectly as enlargement of the adjacent central sulcus (15). Although this was initially described in cases of PLS, we also found the abnormality in patients with either ALS or lower motor neuron disease. Third, cortical signal intensity may be decreased in the precentral gyrus on T2-weighted conventional SE axial images (24,25). We did not evaluate this finding because we used fast SE imaging, which is known to be less sensitive to magnetic susceptibility effects than is conventional SE imaging (26). The use of multiple "refocusing" 180° pulses in fast SE pulse sequences results in attenuation of the effects of T2*-weighted signal decay and, therefore, of magnetic susceptibility effects. This abnormality has also been seen in the motor cortex of healthy elderly subjects (27), which suggests that this finding may be less characteristic of motor neuron disease than was previously believed.
Our MR imaging data included two noteworthy findings. First, we did not confirm the findings of Pringle et al (15) that enlargement of the central sulcus is seen in cases of PLS but not ALS. By using qualitative visual assessments of the central sulcus, we may have failed to detect central sulcus enlargement in some cases of PLS in which an abnormality might have been identified with the quantitative approach of Pringle et al. However, our qualitative approach sufficed to help identify central sulcus enlargement in cases of ALS. Second, we found marked corticospinal tract hyperintensity in two patients with clinically diagnosed PLS, similar to that reported in another case of PLS (23). However, in both of our PLS cases, MR spectroscopy of the motor cortex indicated a nonsignificant decrease in NAA/Cr (2.8 in both cases) compared with the decrease in the control group (mean, 3.1). (With our cutoff criterion of NAA/Cr of 2.5 or less, these two cases would be considered normal.) This suggests that the upper motor neuron disorder in some patients with PLS may arise in the corticospinal tracts or in subcortical neurons.
Single-voxel MR spectroscopy of the motor cortex will be made even easier by the introduction of automated spectral quantitation for rapid calculation of NAA, choline, and myo-inositol levels (relative to Cr) during diagnostic MR examinations. These techniques became available to us only after we concluded the study. After 1996, preliminary data implied good correlation between the NAA/Cr values generated with autoquantitation and by our spectroscopist; the largest discrepancy was about 10% (Shungu DC, unpublished data). We are currently using the autoquantitative technique to perform single-voxel MR spectroscopy of the motor cortex, and we are further delineating the sensitivity and specificity of this refinement of the method.
In conclusion, our data indicate that MR spectroscopy of the motor cortex is useful in the detection of abnormality of the upper motor neuron. MR spectroscopy was more sensitive than MR imaging in the detection of an upper motor neuron abnormality in ALS and PLS. We recommend further exploration of this technique in evaluation of motor neuron disease. A reliable measure of upper motor neuron dysfunction could offer the following clinical advances: (a) provide a useful diagnostic tool for monitoring therapy of ALS and other motor neuron diseases; (b) help identify cases of adult-onset progressive spinal muscular atrophy that are, in fact, ALS; and (c) help ascertain the clinical diagnosis of PLS.
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, S.C., L.P.R.; study concepts, S.C., D.C.S., L.P.R.; study design, S.C., L.P.R.; definition of intellectual content, S.C., D.C.S., L.P.R.; literature research, S.C., D.C.S., A.D.A., L.P.R.; clinical studies, S.C., D.J.L., L.P.R.; data acquisition, S.C., D.C.S., A.D.A.; data analysis, S.C., D.C.S.; statistical analysis, S.C., D.C.S.; manuscript preparation, S.C.; manuscript editing and review, all authors.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Mitsumoto, A. M. Ulug, S. L. Pullman, C. L. Gooch, S. Chan, M. -X. Tang, X. Mao, A. P. Hays, A. G. Floyd, V. Battista, et al. Quantitative objective markers for upper and lower motor neuron dysfunction in ALS Neurology, April 24, 2007; 68(17): 1402 - 1410. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.A. Butman and M.K. Floeter Decreased Thickness of Primary Motor Cortex in Primary Lateral Sclerosis AJNR Am. J. Neuroradiol., January 1, 2007; 28(1): 87 - 91. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Wang, H. Poptani, J. H. Woo, L. M. Desiderio, L. B. Elman, L. F. McCluskey, J. Krejza, and E. R. Melhem Amyotrophic Lateral Sclerosis: Diffusion-Tensor and Chemical Shift MR Imaging at 3.0 T Radiology, June 1, 2006; 239(3): 831 - 838. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Kaufmann, S. L. Pullman, D. C. Shungu, S. Chan, A. P. Hays, M. L. Del Bene, M. A. Dover, M. Vukic, L. P. Rowland, and H. Mitsumoto Objective tests for upper motor neuron involvement in amyotrophic lateral sclerosis (ALS) Neurology, May 25, 2004; 62(10): 1753 - 1757. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Zhai, F. Pagan, J. Statland, J. A. Butman, and M. K. Floeter Primary lateral sclerosis: A heterogeneous disorder composed of different subtypes? Neurology, April 22, 2003; 60(8): 1258 - 1265. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kalra, N. R. Cashman, Z. Caramanos, A. Genge, and D. L. Arnold Gabapentin Therapy for Amyotrophic Lateral Sclerosis: Lack of Improvement in Neuronal Integrity Shown by MR Spectroscopy AJNR Am. J. Neuroradiol., March 1, 2003; 24(3): 476 - 480. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Block, F. Jessen, F. Traber, S. Flacke, C. Manka, R. Lamerichs, E. Keller, R. Heun, and H. Schild Regional N-Acetylaspartate Reduction in the Hippocampus Detected With Fast Proton Magnetic Resonance Spectroscopic Imaging in Patients With Alzheimer Disease Arch Neurol, May 1, 2002; 59(5): 828 - 834. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Suhy, R.G. Miller, R. Rule, N. Schuff, J. Licht, V. Dronsky, D. Gelinas, A.A. Maudsley, and M.W. Weiner Early detection and longitudinal changes in amyotrophic lateral sclerosis by 1H MRSI Neurology, March 12, 2002; 58(5): 773 - 779. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Ellis, J. Suckling, E. Amaro Jr., E. T. Bullmore, A. Simmons, S. C.R. Williams, and P. N. Leigh Volumetric analysis reveals corticospinal tract degeneration and extramotor involvement in ALS Neurology, November 13, 2001; 57(9): 1571 - 1578. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. P. Rowland and N. A. Shneider Amyotrophic Lateral Sclerosis N. Engl. J. Med., May 31, 2001; 344(22): 1688 - 1700. [Full Text] [PDF]
|
|
|
|
|
|
B. C. Bowen and W. G. Bradley Amyotrophic Lateral Sclerosis: The Search for a Spectroscopic Marker of Upper Motoneuron Involvement Arch Neurol, May 1, 2001; 58(5): 714 - 716. [Full Text] [PDF]
|
|
|
|
|
|
C. Pohl, W. Block, J. Karitzky, F. Traber, S. Schmidt, C. Grothe, R. Lamerichs, H. Schild, and T. Klockgether Proton Magnetic Resonance Spectroscopy of the Motor Cortex in 70 Patients With Amyotrophic Lateral Sclerosis Arch Neurol, May 1, 2001; 58(5): 729 - 735. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Mascalchi, M. Cosottini, F. Lolli, F. Salvi, C. Tessa, M. Macucci, M. Tosetti, R. Plasmati, A. Ferlini, C. A. Tassinari, et al. Proton MR Spectroscopy of the Cerebellum and Pons in Patients with Degenerative Ataxia Radiology, May 1, 2002; 223(2): 371 - 378. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
