Study of the features of proton MR spectroscopy (1H-MRS) on amyotrophic lateral sclerosis




To study the features of proton magnetic resonance spectroscopy (1H-MRS) on amyotrophic lateral sclerosis (ALS) and its relation with clinical scale.

Materials and Methods:

Fifteen patients with definite or probable ALS and 15 age- and gender-matched normal controls were enrolled. 1H-MRS was performed on a 3.0 Tesla GE imaging system (GE Healthcare, Milwaukee, WI). TE-averaged Point Resolved Selective Spectroscopy was used. N-acetylaspartate (NAA), creatine (Cr), Glu, and Glx (glutamate + glutamine) values of the motor cortex and posterior limb of internal capsule were acquired. The t-test was used to compare differences between groups, and the correlations between the above values and clinical scale were analyzed.


The motor area and posterior limb of the internal capsule of ALS patients had lower NAA/Cr (1.91 ± 0.34, 1.53 ± 0.17) compared with normal subjects (2.23 ± 0.33, 1.66 ± 0.07), and the differences between groups were statistically significant (P < 0.01, 0.01). ALS patients had higher Glu/Cr (0.34 ± 0.05, 0.29 ± 0.06) and Glx/Cr (0.40 ± 0.04, 0.33 ± 0.06) compared with normal subjects (0.30 ± 0.03, 0.25 ± 0.04) and (0.32 ± 0.05, 0.26 ± 0.03), and the differences between groups were statistically significant (P < 0.01, 0.01). The Norris scale was negatively correlated with Glx/Cr of primary motor cortex by lineal correlation analysis (r = −0.75), and this correlation had statistical significance (F = 16.60; P = 0.001).


Neuronal loss and Glu+Gln increase can be detected by using proton MRS in ALS patients. 1H-MRS is an useful tool in reflecting the characteristic changes of metabolite in ALS. J. Magn. Reson. Imaging 2010; 31: 457–465. © 2010 Wiley-Liss, Inc.

AMYOTROPHIC LATERAL SCLEROSIS (ALS) is a neurodegenerative disorder that involves upper motor neuron (UMN) and lower motor neuron (LMN) simultaneously, and results in neuron degeneration. The real etiology remains unknown. With the development of neuroimaging techniques, MR is applied more and more in the diagnosis, differential diagnosis, pathologic studies, and therapeutic effect evaluation of ALS. Many MR studies focus on regular imaging features and its diagnostic values in ALS. Regular MR imaging with high soft tissue resolution can show characteristic abnormal signals in the brain of ALS patients, but the sensitivity and specificity are low. Proton MR spectroscopy (1H-MRS) has been suggested as a noninvasive method of detecting brain metabolites, such as N-acetylaspartate (NAA), creatine (Cr), glutamate (Glu), and glutamate + glutamine (Glx). There is increasing evidence that Glu excitotoxicity may play an important role in the pathogenesis of ALS, and Glu has been found to be elevated in the plasma and CSF of patients with ALS (1, 2). Because 1H-MRS can reveal and quantitatively analyze the changes of brain metabolites including Glu and Glx in ALS patients, it may provide useful information in the disease diagnosis, staging, and therapeutic effect evaluation of ALS.


A total of 15 patients (6 males, 9 females; age range, 30–62 years; mean, 51.33 ± 10.12 years) with clinically definite (10 cases) or probable (5 cases) ALS as defined by the El Escorial criteria were enrolled. The clinical course was from 6 to 48 months (mean: 13.10 ± 9.44 months). No patients had other nervous system disease or took any medicine within the last 3 months. The 15 healthy volunteers (7 males, 8 females; age range, 37–60 years; mean, 49.47 ± 7.37 years) were enrolled as control. All the patients were scored by modified Norris score and ALS functional rating score (ALSFRS).

All subjects underwent a scanning session with the same 3.0 Tesla (T) MRI system (GE Healthcare, Milwaukee, WI) and the same protocol with the gradient field of 23 mT/m and slew rate of 150 mT/m/ms. A standard orthogonal head coil was used.

The MRI regular scan was performed using fast spin echo (FSE) T2 weighted imaging (repetition time/echo time [TR/TE] = 4,000/102 ms) to obtain images in coronal and axial planes (slice thickness = 5 mm, interval = 0 mm).

The MRS scan was performed using TE-averaged PRESS (Point Resolved Selective Spectroscopy) described previously (3), with TR = 1500 ms and TE base value = 25 ms. Data were collected by averaging 18 steps with a 5-ms increment time to 115 ms. Under this TE set (25–115 ms), the effective TE was 67 ms (calculated according to gray matter Cr with T2 relaxation time of 143 ms). Field of view = 24 cm × 24 cm, NEX = 4, voxel = 2.0 cm × 2.0 cm × 2.0 cm, scan time = 4 min 12 s.

The study was approved by the local Ethics Committee of the PLA General Hospital. All participants signed the informed consent form before their inclusion in the study.

The participants were asked to keep calm during the MR scan process. The voxels were placed bilaterally at the motor area and posterior limb of the internal capsule (Fig. 1), and MRS values were acquired. Auto shimming and water suppression were performed before 1H-MRS. A spectral line was acquired with the line width less than 7 Hz.

Figure 1.

MRS data acquired from the bilateral motor cortex (a) and posterior limb of the internal capsule (b).

Data processing of the 1H MRS was performed on a SUN imaging workstation (Advantage Windows TM 4.2) using SAGE (Spectroscopic Analysis, GE). The spectra of NAA, Glu, Glx, and Cr were collected, and values of peak height and peak area were calculated. Then the relative magnitude values of peak height and peak area of NAA/Cr, Glu/Cr, and Glx/Cr were calculated.

The process was done by two radiologists who knew nothing about the study design.

All statistical tests were performed using SPSS11.5. Mann-Whitney U-test and lineal correlation analysis were used, and the difference was considered to be statistically significant at P < 0.05.


Spectral Line Acquisition

Since the TE-averaged PRESS sequence was introduced in the current study, Glu peak at 2.35 ppm and Glx (Glu + Gln) peak at 3.75 ppm were isolated separately, and NAA at 2.0 ppm and Cr at 3.0 ppm were also identified (Fig. 2). MR data from a total of 120 voxels were acquired in 30 participants, and the data from only 8 voxels showed obscure Glu peak and 8 with obscure Glx peaks, and the rate of demonstration for Glu and Glx peak was 93%.

Figure 2.

Spectrum of ALS (a) and control (b).

Spectral Line Analysis

Tables 1 and 2 showed that NAA/Cr values of bilateral motor cortex (BMC) and posterior limb of the internal capsule (PLIC) in ALS patients (1.91 ± 0.34 [BMC] and 1.53 ± 0.17 [PLIC]) were lower than that in controls (2.23 ± 0.33 [BMC] and 1.66 ± 0.07 [PLIC]), and the difference had statistical significance (P < 0.01, 0.01). Glu/Cr (0.34 ± 0.05 [BMC] and 0.29 ± 0.06 [PLIC]) and Glx/Cr (0.40 ± 0.04 [BMC] and 0.33 ± 0.06 [PLIC]) in the ALS group were higher than those in controls: Glu/Cr (0.30 ± 0.03 [BMC] and 0.25 ± 0.04 [PLIC]), Glx/Cr (0.32 ± 0.05 [BMC] and 0.26 ± 0.03[PLIC]), and the differences between the two groups had statistical significance (P < 0.01, 0.01).

Table 1. NAA/Cr, Glx/Cr, and Glu/Cr Values in BMC and Statistical Analysis
MetabolitesNAA/Cr (control = 15) (ALS = 15)T'PGlx/Cr (control = 12) (ALS = 15)T'PGlu/Cr (control = 13) (ALS = 15)T'P
Control2.23 ± 0.3345.5(T') (P < 0.01)0.32 ± 0.05102.5(T') (P < 0.01)0.30 ± 0.03154.2(T') (P < 0.01)
ALS1.91 ± 0.340.40 ± 0.040.34 ± 0.05
Table 2. NAA/Cr, Glx/Cr, and Glu/Cr Values in PLIC and Statistical Analysis
MetabolitesNAA/Cr (control = 15) (ALS = 15)T'PGlx/Cr (control = 10) (ALS = 13)T'PGlu/Cr (control = 9) (ALS = 13)T'P
Control1.66 ± 0.07130.4(T') (P < 0.01)0.26 ± 0.0344.5(T') (P < 0.01)0.25 ± 0.0461.8(T') (P < 0.01)
ALS1.53 ± 0.170.33 ± 0.060.29 ± 0.06

Correlation Between 1H-MRS and Clinical Score

In linear correlation analysis, NAA/Cr values showed positive correlation with modified Norris score and ALSFRS, but without statistical significance (r = 0.55, P = 0.06; and r = 0.53, P = 0.07). The motor cortex Glx/Cr values had negative correlation with modified Norris score (r = −0.75) and the correlation had statistical significance (F = 16.60; P = 0.001) (Fig. 3). The BMC Glx/Cr value had negative correlation with ALSFRS (r = −0.47), but without statistical significance (F = 3.61; P = 0.08). The PLIC Glx/Cr value showed negative correlation with ALSFRS and modified Norris score, but without statistical significance (r = −0.25, F = 2.49; P = 0.08; r = −0.42; F = 4.51; P = 0.08).

Figure 3.

Scatter plot of Glx/Cr values and Norris scores.


In 1995, Jones et al (4) first applied MRS on ALS and found the reduction of NAA and NAA/Cho in motor cortex and adjacent cortex (59). Rule et al (10) reported that NAA/Cho and NAA/Cr+Cho were reduced not only in the motor cortex and adjacent cortex but also in subcortex white matter. Yin et al (11) observed and acquired NAA/Cr values in different levels of corticospinal tract (CST), including subcortex white matter, paraventricular white matter and posterior limb of the internal capsule. The NAA/Cr values dropped more in subcortex white matter and paraventricular white matter in ALS. By using 3.0T scanner, we acquired high signal-to-noise spectra in our study. The NAA/Cr values of the motor cortex and posterior limb of the internal capsule in ALS patients were much lower than those in the control group (P < 0.01, 0.01).

The cause of NAA, NAA/Cho, and NAA/Cr reduction in ALS may be the neuronal loss. Gredal et al (8) found that the average number of neurons in neocortex in ALS and controls had no difference (21.7 × 109 and 22.3 × 109, respectively) as well as in motor cortex (1.33 × 109 and 1.29 × 109, respectively). Neuronal metabolic dysfunction, instead of neuronal loss, may be the real cause of NAA reduction. The changes included motor neuron loss, degeneration, tissue edema, and macrophage infiltration, and pathological changes of CST included demyelination, edema, and gliosis.

Some studies showed quantitative relationship between MRS changes and clinical function score of ALS. NAA value had a correlation with Norris score (r = 0.30; P < 0.00) but failed to correlate with ALSFRS (12). Abe et al (13) also reported that NAA/Cr of motor cortex correlated with Norris score (r = 0.50; P < 0.01). In our study, no statistical correlation was found between motor cortex NAA/Cr and modified Norris score (r = 0.55; P = 0.06), or between NAA/Cr and ALSFRS (r = 0.53; P = 0.07). The difference between our results and those of others may result from different samples.

In addition to NAA, Cho, and Cr, Glu changes have drawn more attention recently. The excitotoxicity of Glu was thought to be the most important pathogenesis of ALS. Glu elevation in plasma and CSF of ALS patients was reported (1, 2). In pathological conditions, postsynaptic receptors cannot re-uptake Glu promptly; therefore, excessive Glu keeps activating excitatory amino acid receptors, thus causes overloading of Ca2+ in neurons. Excessive intracellular Ca2+ is lethal: it impedes oxidative phosphorylation in mitochondria and causes ATP deficit. Hyperactivity of the ATP enzyme in muscle fiber, cytoplasmic reticulum, and mitochondria consumes more ATP and causes damage to cellular structures and function. Ca2+ can also activate catabolic enzymes, stimulating protein kinase, phospholipase, nucleate endonuclease, nitrogen monoxide enzymes, which directly damage the cellular structure or produce many free radicals that destroy the cytomembrane, RNA, and protein through oxidization. All these lead to neuron death (14).

In this study, we used specially designed TE-averaged PRESS, which could separate Glu peaks and Glx (Glu+Gln) peaks from other metabolite peaks, and the two peaks showed good signal to noise ratio. TE-averaged PRESS sequence depends on the J coupling function of H atom. When collecting the spectra according to different TE, there are border peaks around the main peaks. When adding and averaging all the spectra together, the border peaks become zero, then the main peaks are more prominent. Glu shows solo peak at 2.35 ppm, and overlaps with Gln at 3.75 ppm forming a Glx peak. Prominent peaks of Glu and Glx were acquired in this study. The peak values of Glx/Cr and Glu/Cr were 0.40 ± 0.04 and 0.34 ± 0.05 in motor cortex separately, and the values were 0.33 ± 0.06 and 0.29 ± 0.06 in the posterior limb of the internal capsule, which were higher than those of controls, and the differences showed statistical significance (P < 0.01).

Pioro et al (15) detected Cho, Cr, Gln, Glu, NAA, and NAAG in medulla oblongata of ALS patients using the short TE 1H-MRSI technique, and 17% lower Nax (NAA+NAAG) and 55% higher Glx compared with controls were revealed. They also found that Glx had negative correlation with ALSFRS, especially with medulla function score (r = −0.68; P = 0.03). Although it remained unclear whether the elevation of Glx is the cause of ALS or the consequence of it, they speculated that Glx could be an important marker in monitoring disease progression, in clinical diagnosis, and in understanding the pathogenesis of ALS.

In this study, we detected Glx values of the motor cortex and posterior limb of the internal capsule in ALS patients and studied the correlations of Glx with ALSFRS and modified Norris score. Glx/Cr and Glu/Cr values in the motor cortex and posterior limb of the internal capsule had negative correlation trends with both ALSFRS and Norris score, but only Glx/Cr values of the motor cortex with Norris score had statistical significance (F = 16.60; P = 0.001).

The limitation of the study was that the ALS patient sample was rather small, and possible lateralization in ALS patients was not investigated. Furthermore, analysis on subgroups (definite and probable) of ALS was not performed. With more cases in the future, the unilateral differences and subgroup differences might be found, and the correlation between the metabolic changes and disease severity could be further elaborated.

In conclusion, our current study has confirmed that Glu/Cr, Glx/Cr, and NAA/Cr are ideal indexes for the clinical evaluation of ALS. The level of NAA/Cr could help in differential diagnosis, and Glx/Cr has a correlation with ALS function grading. With more studies of multicenter research, the clinical application value of MRS in ALS will become clearer, and the guidance value of MRS in ALS will be recognized. Compared with MRI and BOLD, 1H-MRS can gain quantitative results in the analysis of ALS, which makes 1H-MRS a priority choice in clinical practice.