вход по аккаунту


Chemical pathology of acute demyelinating lesions and its correlation with disability.

код для вставкиСкачать
Chemical Pathology of Acute Demyelinating
Lesions and Its Correlation with Disabihty
Nicola De Stefano, MD," Paul M. Matthews, MD, DPhil,"? Jack P. Antel, MD," Mark Preul, MD,"
Gordon Francis, MD," and Douglas L. Arnold, MD"
We report the chemical pathological changes on magnetic resonance spectroscopic images of 4 patients, each of whom
had a single large demyelinating plaque. The patients were followed from soon after the onset of the symptoms for
a minimum of 7 months to a maximum of 3% years. We observed increases in the relative resonance intensities of
choline-containing compounds, lactate, and myo-inositol inside the lesion acutely. Decreases in relative resonance
intensities of N-acetylaspartate and creatine were seen both in and around the magnetic resonance imaging-detected
lesions. In all patients neurological deficits improved and creatine, lactate, and myo-inositol resonance intensities
normalized during the follow-up. Choline compounds recovered more slowly and were still abnormally high in 1
patient after 7 months. Partial recovery of the N-acetylaspartate resonance was seen for all patients. Evaluation of the
relationships between indices of cerebral chemical pathology, brain lesion volumes, and functional disability showed
highly significant negative correlations between N-acetylaspartate resonance intensities and both brain lesion volumes
( r = -0.80, p < 0.0001) and clinical disability ( r = -0.73, p < 0.0001). As N-acetylaspartate is localized solely in
neurons in the adult central nervous system, our results suggest that neuronal dysfunction may be a proximate mechanism of disability even in inflammatory disorders primarily affecting myelin and oligodendroglial cells.
De Stefano N, Matthews PM, Antel JP, Preul M, Francis G, Arnold DL. Chemical pathology of acute
demyelinating lesions and its correlation with disability. Ann Neurol 1995;38:901-909
Magnetic resonance (MR) imaging (MRI) has been
used to measure disease burden and activity in multiple
sclerosis (MS). Total lesion volume measured by MRI
morphometry has proved to be a useful quantitative
outcome measure in recent clinical trials 11-31. Use
of contrast agents sensitive to changes in blood-brain
barrier permeability increases the sensitivity and specificity of the images and allows better characterization
of the pathological evolution of plaques [4, 51. However, a strong relationship between clinical disability
and brain MRI abnormalities has not been found {6].
The most probable explanation for this is the known
heterogeneous nature of the pathology associated with
plaques visualized by conventional MRI [6]. Specific
measures of pathological changes associated with MS
lesions therefore may be useful for defining the relationship between lesion load and clinical status of patients with demyelinating disease.
MR spectroscopic imaging (MRSI) allows measurement in vivo of metabolites whose concentrations reflect pathological changes in the demyelinating lesions.
To determine the relationship between chemical
pathological changes in acute demyelinating lesions and
clinical disability, we performed serial MRSI examina-
tions in 4 patients, each of whom had a single large
hemispheric demyelinating plaque. The patients were
followed longitudinally from soon after the onset of
the symptoms for a minimum of 7 months to a maximum of 3% years. During the follow-up period, clinical
disability was evaluated for correlation of functional
recovery with brain MRI lesion volume and MRSI indices of cerebral chemical pathology.
'From the Department of Neurology and Neurosurgery, Montreal
Neurological Institute and Hospital, and the tDepartment of Human Genetics, McGili University, Montreal, Quebec, Canada.
Address correspondence to Dr Arnold, Montreal Neurological Institute, 3801 University Street, Montreal, Quebec, Canada H3A 2B4.
Materials and Methods
PATIENT 1. A 26-year-old woman was hospitalized because
of progressive right hemisensory loss that began 3 days earlier. Neurological examination revealed hypoesthesia over
the right half of the body and mild pyramidal signs on the
same side. The neurological deficit progressed over the next
week to an almost complete hemiplegia and hemisensory
deficit. By day 10 the patient had reached maximal disability
and was unable to walk (Expanded Disability Status Scale
[EDSS} score = 7)[7}. A conventionalbrain MRI performed
on the day of admission showed an area of increased signal
intensity in proton density and T2 weighted images centered
on the posterior limb of the left internal capsule and extending into the subcortical white matter of the frontotemporal region. A second MRI performed 2 weeks later showed
Received May 22, 1995, and in revised form Jul 31. Accepted for
publication Aug 3, 1995.
Copyright 0 1995 by the American Neurological Association 901
an increase in the area of hyperintensity. A stereotaxic needle
biopsy of the lesion was performed to differentiate idiopathic
demyelination, neoplasm, and infection. The biopsy confirmed the primary demyelinating nature of the lesion and
the patient improved after treatment with dexamethasone.
H e r neurological symptoms and signs largely resolved (EDSS
= 2). During more than 3 years of follow-up she had one
other mild attack that was not accompanied by visible
changes on cerebral MRI. Eight brain MRI and MRSI examinations of this patient were performed over 3% years.
2. A 19-year-old girl was admitted to the hospital
because of an acute onset of numbness on the right side of
her body. Symptoms progressed over the next few days to
an almost complete hemiparesis and hemisensory deficit
(EDSS = 7.5). Brain T2-weighted MRI showed an increase
of signal intensity over a large portion of the left parietal
hemisphere that enhanced after gadolinium injection. Analysis of the cerebrospinal fluid (CSF) demonstrated increased
protein and immunoglobulin-albumin ratio. No oligoclonal
bands were seen. A stereotaxic needle biopsy of the lesion,
done to rule out neoplasm or infection, showed inflammation
and demyelination only, consistent with primary demyelinating disease. Treatment with methylprednisone was started
and her clinical condition improved progressively. The patient underwent six combined brain MRI and MRSI examinations over 15 months of follow-up. No other attacks occurred
during this time. The EDSS score was 1 at the time of the
last examination.
PATIENT 3. A 27-year-old woman was admitted to a hospital because of an acute decrease of strength and sensation on
the right side of her body. Neurological examination demonstrated a spastic hemiparesis and decreased vibration sense
in the affected limbs. Speech was dysarthric. The EDSS score
was 8. A conventional MRI showed a large lesion localized
to the left parietal white matter that enhanced after injection
of gadolinium. Analysis of CSF showed an increase of the
immunoglobulin-albumin ratio and positive oligoclonal
bands. The patient was started on therapy with dexamethasone and there was only slight immediate clinical improvement. A follow-up MRI performed 1 month after the first
showed enlargement of the original lesion into the frontoparietal region. A stereotaxic needle biopsy, done to rule out
neoplasm or infection, showed localized areas of myelin loss
and gliosis compatible with a demyelinating lesion. The patient continued the same treatment with later, slow improvements of the symptoms. Three combined brain MRI and
MRSI examinations were performed over 7 months. No subsequent attacks occurred during this period. H e r EDSS score
was 3.5 at the time of last examination.
PATIENT 4. A 26-year-old man was admitted to a hospital
because of clumsiness of the left arm and leg. The neurological examination showed mild spasticity, decreased strength
and hyperreflexia in the left upper and lower extremities,
and a Babinski sign on the left. The EDSS score was 3 at
this first examination. CSF analysis revealed a slight increase
of protein, normal immunoglobulin-albumin ratio, and no
oligoclonal bands. A conventional brain MRI showed an eggsized lesion in the subcortical white matter in the parietal
902 Annals of Neurology
Vol 38 No 6 December 1995
region of the right hemisphere with no evidence of mass
effect. The patient improved spontaneously over several
days. He had a second, milder attack with dizziness and diplopia 9 months later, without visible changes on a brain MRI.
He underwent four combined brain MRI and MRSI examinations over 18 months. The EDSS score was 1 at the time
of the last examination.
A group of 20 normal volunteers underwent proton MRSI
with the same acquisition parameters and localization of the
volume of interest (VOI). All studies were approved by the
Ethics Committee of the Montreal Neurological Institute and
Hospital (MNIH). The MRSI results and clinical history of
Patient 1 have been reported partially elsewhere E8).
Magnetic Resonance Examinations
Conventional proton MRI and MRSI examinations of the
brain were performed in a single session for each examination
using a Philips Gyroscan S15 operating at 1.5 T (Philips Medical Systems, Best, Netherlands). A sagittal survey image was
used to identify the anterior commissure (AC) and posterior
commissure (PC). Multislice images were obtained in coronal
and transverse planes, perpendicular and parallel to the
AC-PC line, respectively (TR 2 100 TE 30,78 slice thickness
6 mm). These images were used to select an intracranial VOI
for spectroscopy. For the 4 patients with demyelinating lesions a VOI positioned parallel to the AC and PC line and
measuring approximately 70 mm in the anteroposterior (AP)
direction x 20 mm in the craniocaudal direction (CC) x 90
mm from left to right (LR) was positioned to include the
lesion and the homologous region of the contralateral hemisphere. The size and location of the VOIs were kept constant
in subsequent examinations of each patient.
Proton spectra were acquired using a 90-180- 180-degree
sequence for volume selection @J (TR 2000lTE 272). Magnetic field homogeneity was optimized to a line width of
about 5 Hz over the VOI using the proton signal from water.
Water suppression was achieved by selective inversion of the
water resonance prior to volume selection using an adiabatic
inversion pulse and adjustment of the waiting time so that
the spectrum was acquired when the water signal passed
through zero {lo]. MRSIs were generated by twodimensional phase encoding (250 x 250-mm field of view
[FOVJ, 32 x 32 phase encoding steps, 1 signal average per
step). After a water-suppressed acquisition was completed,
another MRSl was acquired without water suppression (TR
850, TE 272, 250 X 250 FOV, 16 x 16 phase encoding
Postprocessing of the raw data was done on a SUN,
SPARC system using SUNspec 1 software (Philips Medical
System, Best, Netherlands). The non-water-suppressed
MRSIs were interpolated to 32 x 32. A mild gaussian kspace filter and an inverse two-dimensional Fourier transformation were then applied to both the water-suppressed and
unsuppressed MRSIs. Artifacts present in the time domain
water-suppressed signal due to magnetic field inhomogeneities were corrected for by dividing the water-suppressed
MRSI signal by the non-water-suppressed signal [ 1 If. The
resulting time domain signal was left shifted and subtracted
from itself to improve water suppression E12). To enhance
the resolution of the spectral peaks, a lorentzian-to-gaussian
transformation was applied prior to Fourier transformation
in the spectral domain. The result was 1,024 voxels (32 x
32), each containing a spectrum ready for quantification and
subsequently generation of the MRSIs. The nominal voxel
size in plane was approximately 8 x 8 mm and approximately
12 x 12 mm after k-space filtering. Gray-scale spectroscopic
images were created by integration between frequency
shifts were calculated relative to the NAA resonance at 2.0
Metabolite signals are expressed as ratios of the metabolite
resonance intensities in voxels ipsilateral to the lesion and
homologous voxels in the contralateral hemisphere (metaboliteyc).This ratio was not significantly different from 1 ir
normal control subjects (Table). Large decreases of creatine
(Cr) signal intensities during the acute phase of demyelination were observed in 3 of the 4 patients (see Table), making
it inappropriate to normalize metabolite signal intensities relative to Cr in the same voxel. Therefore, signal intensities
from voxels on the side of the lesion were also determined
relative to C r coming from homologous voxels of contralateral hemisphere (metabolite/Cr, j (see Table).
Brain MRI lesion volumes were measured in all patients
by projecting homologous axial T2-weighted images onto a
white surface for each different examination. Volumes were
measured in cubic millimeters by assuming spherical symmetry with dimensions determined from the greatest crosssectional diameter of the lesion. Analysis was facilitated by
careful positioning of the head in the same orientation for
each examination.
Data Analysis
Resonance intensities were determined automatically from
peak heights relative to a spline-corrected baseline. Peak
heights were chosen rather than peak areas after comparison
of results from single-voxel MR spectroscopic examinations
in 20 healthy control subjects revealed that the results were
the same, but peak heights were less variable than peak areas
(data not shown). This is consistent with the fact that the
major contribution to line width is from field inhomogeneity.
Chemical shifts were calculated relative to the N-acetylaspartate (NAA) resonance at 2.0 ppm.
Values of metabolite resonance intensities were determined from an average spectrum obtained from voxels located within a given lesion ( n = 6-7). Spectra were considered within the lesion if they were entirely filled with
abnormal signal on the first MRI. Spectra were considered
outside the lesion if there was no abnormal hyperintense
MRI signal within the voxel or in adjacent voxels. Chemical
Relative resonance intensity ratios were considered abnormal
if they were more than 2 standard deviations (SDs) outside
Relatitie Brain Metabolite Signal Intendies Measured by MRSl in the Centers of Single Demyelinating Lesions and Brain MRl
Lesion Volumes from the 4 Patients"
0. 7b
1.1 k 0.2
1.1 '' 0.2
4.3 +- 0.6
1. I b
"Values are shown from the initial examination during the acute phase of the disease and from a final follow-up study after recovery of the
major symptoms.
MRSI values are expressed as ratio of the metabolite resonance peak intensities in voxels within the lesion relative to homologous voxels in
the contralateral hemisphere, except for mI and La which are not observable in normal brain parenchyma with these technical conditions (top
ofthe table). Metabolite resonance intensities in voxels on the side of the lesion are expressed also as their ratio to Cr from homologous voxels
in the contralateral hemisphere (bottom ofthe table). Brain MRI lesion volume values are expressed in cubic millimeters (see methods).
bValues greater than 2 standard deviations from the control mean.
MRI = magnetic resonance imaging; MRSI = magnetic resonance spectroscopic imaging; NAA = N-acetylaspartate; Cr = total creatine;
Cho = choline-containing compounds; mI = inositoi; La = lactate; LV = lesion volume; C = controls; (r) = right hemisphere; (I) = left
hemisphere; i = ipsilateral; c = contralateral.
De Stefan0 et al: 'H MRSI of Demyelinating Lesions 903
the normal mean. Values of brain MRI lesion volume and
relative resonance intensities of cerebral metaboliteg, were
correlated with clinical disability using the Pearson correla-
Nomzal Control Subjects
Voxels for quantification of normal control values (n
= 6-9) were selected from the white matter of the
frontoparietal periventricular region to avoid partial
volume effects due to CSF in the ventricles.
Brain MRSI of 20 healthy subjects showed no differences between homologous voxels from the two hemispheres for resonance intensities of NAA (2.0 ppm),
tri- and tetramethylammonium compounds (primarily
choline-containing compounds [Cho]) (3.2 ppm),
or total Cr (3.0 ppm): NAArlghdleft
= 0.97 ?z 0.06,
= 0.982 0.18, CrrlRhdleft
= 0.95 f 0.15
(see Table). Metabolite resonance intensities expressed
relative to Cr were also similar in the two hemispheres
(see Table). Signals from myo-inositol (mI) (3.55 ppm)
and lactate (La) (1.4-1.2 ppm) were not detectable in
spectra from individual MRSI voxels of normal subjects.
Averaged metabolite resonance intensity values of the
4 patients for voxels entirely within the acute plaque
are shown in the Table. Values are reported for the
MRSI examinations performed during the acute phase
of the disease and the final examination at the
follow-up. Values for initial and final brain MRI lesion
volume in each patient are also reported in the Table.
All lesions showed large increases in
Cho and La resonance intensities in the initial examinations (Figs 1, 2A). In the 3 patients (Patients 1-3) with
more severe clinical symptoms the MRSI showed large
decreases of NAA (55-66s) (see Figs 1, 2A). A
smaller decrease was seen in the lesion of the patient
with milder symptoms (Patient 4, 25%). A decrease of
relative Cr signal (25-60%) was also found in the lesion of the 3 patients with severe symptoms. Reproducible increases of the signal at 3.55 ppm attributable
to mI were seen in Patients 2 and 3 (see Figs 1, 2A).
A significant, but smaller decrease of NAA was seen
at the border of the plaque in Patients 1, 2, and 3
(NAA, 0.7, 0.6 and 0.4, respectively) than in the lesion center. In Patients 1 and 3, who had progressive
enlargement of the lesion between the first two examinations, a decrease of NAA was found on the first
examination in voxels that at thar time appeared normal on conventional MRI and only in later examinations showed abnormal hyperintense signal on MRI
(NAA,,: Patient 1 = 0.7, Patient 3 = 0.8; NAMCr,:
Patient 1 = 3.3, Patient 3 = 3.0) (Figs 3A, 3B). DeINITIAL MRSI.
creases in NAA were not seen in the normal-appearing
white matter on the edge of plaques in Patients 2 and
4, who did not show enlargement of their lesion after
the initial examination.
After the acute phase of the disease,
intensities of Cr, La, and mI returned to normal over
successive examinations for all patients (see Fig 2B).
Cho resonance intensity returned to control values
more slowly than did the intensity of the other metabolites; it was still abnormally high in Patient 3 even after
7 months. NAA showed a recovery (at least partial)
(see Table, Fig 2B) in all patients with a time course
that has been described in detail elsewhere [131. This
recovery was highly correlated with the decrease of
brain lesion volumes occurring in all patients during
the follow-up period ( Y = 0.80, p < 0.0001) (Fig 4A).
There WaS
not a consistent relationship between initial lesion volume and functional impairment at the time of presentation; for example, lesion volumes of Patients 1, 2 , and
4 were similar, yet the EDSS score was between 7 and
8 for the first 2 patients and only 3 for patient 4. A
stronger trend for lower NAA with higher EDSS score
was seen (see Table and methods section). In all patients there was improvement of the neurological deficit after the acute illness (duration of maximal disability, 15-30 days) that was reflected in a progressive
decrease of the EDSS scores. Metabolite resonance intensities and brain lesion volumes measured during
each MRI/MRSI examination were correlated with the
clinical disability score at the time of each examination.
A highly significant relationship was found between
both EDSS scores and NAA values ( v = -0.73, p <
0.0001) and EDSS (Fig 4B) and brain MRI lesion volumes ( Y = 0.67, p = 0.001). No significant correlations were found between EDSS scores and the other
metabolites measured.
MRI has provided important new insights into disease
activity, but attempts to correlate clinical disability with
MRI lesion load have been disappointing [6f. One possible reason for this is the pathological heterogeneity
of the demyelinating lesion defined by MRI. Recent
MR spectroscopic studies emphasized that there is a
large variation in relative NAA concentrations between demyelinating lesions, reflecting significant differences in the extent of local axonal damage or dysfunction IS, 14-17], and suggested that neuronal
damage or dysfunction may play a more direct role
in determining functional disability in demyelinating
disease than has been generally recognized [l8]. Brain
proton MRSI offers a unique method for in vivo characterization of axonal damage or dysfunction in individ-
904 Annals of Neurology Vol 38 No 6 December 1995
Fig 1 . Brain MRI and proton MRSI images o f Patient 2 during the acute phase of disease ( I week after the onset of the
symptoms). Focal abnormalities related to the demyelinating lesion are shown in the conventional MRI (upper left image),
and in the metabolic images o f myo-inositol (ml) (upper center
image), choline-containing compounds (Cho) (upper right image), creatine (Cr) (lower left image), N-acetylaspartate
(NAA)(lower center image), and lactate (La)(lower right
image). Note the focal increases in ml, Cho, and La and decreases in NAA and Cr that colocalize with the MRI lesion.
The abnormalities extend outside the lesion as seen on MRI for
some metabolites, such as NAA.
ual lesions by means of measurement of local concentrations of the neuronal-specific metabolite NAA.
Here we used this technique to monitor chemical
pathological changes in single demyelinating lesions of
patients who presented soon after their first clinical
manifestations and had follow-up studies through a period of significant functional improvement. This gave
us the opportunity to evaluate the relationship between
chemical pathological changes in the lesions and clinical
The most significant finding of our study was a
strong correlation between clinical disability and reversible decreases in NAA in the lesions. The associa-
tion between a decrease in EDSS score and recovery
of NAA resonance intensities in patients in whom
functional disabilities can be attributed to a single lesion supports the hypothesis that repair of axonal injury or dysfunction may be a proximate mechanism of
functional recovery.
It is likely that several events may contribute to functional impairment caused by demyelinating lesions.
Nonetheless, axonal injury or dysfunction that is reflected in changes of the NAA resonance intensity may
be a more direct measure of the effective burden of
disease than is lesion volume. Previous studies found
either no relationship or only a weak correlation between clinical disability and brain lesion volume [b,
17, 17, 201. Our recent results emphasize this with a
comparison of two groups of MS patients (stratified for
clinical course either as relapsing remitting or secondary progressive) showing higher total brain lesion volume in the group with Lower EDSS scores (P. M. Matthews and colleagues, unpublished observations,
1995). In the same population of patients a highly significant correlation was found between EDSS score and
NAA (L. Fu and colleagues, unpublished observations,
Both brain MRI lesion volume and local NAA were
De Stefan0 et al: 'H MRSI of Demyelinating Lesions
Fig. 2. Spectra from Patient 2 during the acute phase of disease
(1 week afer the onset of the symptoms) (A) and 15 months
later (B). (A) There are increases in relative resonance intensities of myo-inositol (m1) (3.55 ppm), choline-containing compounds (Cho) (3.2 ppm), and lactate (La) (1.4-1.2 ppm) and
decreases in creatine (Cr) (3.0 ppm) and N-acetylaspartate
( N A A ) (2.0 ppm) signal intensities in spectra within the demyelinating lesion (spectra a and b) when compared with spectra
coming from homologous voxels in the contralateral hemisphere
(spectra c and d). (B)The evolution of the changes on a follow-up examination perfomzed I S months later. Spectra coming
from simikzr voxels to those in (A)show nomzalization of mI,
Cho, Cr, and La relative resonance intensities and a substantial
recovery of N A A signal intensity (spectra e and f). Spectra coming from homologous voxels in the contralateral nomzal hemisphere are shown in spectra g and h.
highly correlated during the period of functional recovery in the present study. However, the relationship
berween lesion volume and NAA is heterogeneous in
general. A trend for lower NAA with higher EDSS
score was seen at the initial examination in the 4 patients studied here. In contrast, lesion volume did not
show a consistent relationship to disability at presentation, even in this group of patients in whom all disability could be associated with a single lesion. Previous
studies of individual lesions demonstrated that some
may have only minimal changes in NAA while others
show large decreases 114, 16, 21-23]. We concluded
from these data that the hyperintense MRI lesions
must be pathogenetically heterogeneous (a finding supported by contrast, magnetization transfer, and MR
spectroscopic assessments) and may be responsible for
variable degrees of functional impairment depending
on the functional integrity of the axons traversing the
There is no compelling evidence that neurons or
their processes are primary targets of inflammation in
MS, but neuronal damage or dysfunction may occur
secondary to actions of inflammatory mediators and
changes in the local cellular environment associated
with inflammation {24, 251. Recovery from functional
impairment therefore may result not only from redistribution of axonal ion channels and remyelination
{25-281, but also from intrinsic metabolic changes in
neurons. In the mitochondrial metabolic disease
MELAS (mitochondrial encephalopathy, lactic acidosis,
and strokelike episodes), recovery of NAA in regions
of strokelike episodes parallels improvement in clinical
deficits 1131. Use of MRSI therefore may offer a
unique window for monitoring a key aspect of neuronal metabolism.
Changes in metabolites other than NAA were seen
also in the lesions. Acutely, the most severely affected
patients had a decrease in Cr in the lesion centers. Cr
is relatively homogeneously distributed in normal brain
and has been used as an internal concentration standard
in many spectroscopy studies. However, changes (either decreases or increases) of this metabolite have
been observed in patients with MS {23, 29, 307. The
decreases in Cr that we observed (25-60%) are much
greater than those that could be expected from edema
alone (6-12%) 1311, and are in agreement with recent
studies in vitro, which intrinsically controlled for any
possible changes in signal intensity related to MR relaxation time variations {321. The normalization of values
observed with progressive resolution of the lesion suggests that a decrease in Cr is most likely to be seen
in a severe acute plaque, where it probably reflects
metabolic changes, perhaps in glia 1337. The reported
increases in Cr in MS lesions {30) may be associated
with a later stage of lesion evolution and changes of
the cell population with astrogliosis.
M 6 Annals of Neurology Vol 38 N o 6 December 1995
Fig 3. Brain MRI and spectra from Patient 1 in two different
examinations performed on 3 (A)and 19 (B) days after the on-
set of symptoms. (A) Thefocal MRI abnomlity caused by the
demyelinating lesion and the signrjcant decrease in N-acetylaspartate (NAA) resonance intensity (right spectrum) with respect t o the homologous voxel in the contralateral hemisphere
(left spectrum). The abnormal NAA intensity signal comes
from a voxel positioned well beyond the demyelinating lesion as
visualized by conventional MRI on this examination and that
only i n the examination performed 2 wee& later (B) showed abn o m l hyperintense signal on MRI. Cr = creatine; Cho =
choline-containing compounds.
De Stefano et al: 'H MRSI of Demyelinating Lesions
Lesion Volume
m C
4 ’0.41
Fig 4. Cowelations o f N-acetylaspartate (NAA) with volume of
demyelinating lesions (r = - 0.80, p < 0.0001) (A)and clinical disability (Expanded Disability Status Scale [EDSS)) (r =
-0.73, p < 0.0001) (B). Valuesfrom each examination performedfor the patients during the follow-up (n = 21) are represented. NAA is expressed as the ratio of resonance intensities in
central lesion voxels relative to those in homologous voxels in the
contralateral hemisphere (NAAipsdcontra).Lesion volumes are
expressed in cubic millimeters (see methods).
Increases in mI in MS patients have been noted previously in short-echo-time spectroscopy studies performed using both single-voxel 134, 351 and MRSI
techniques C221. We used a relatively long echo time
in our study (272 msec) but could still detect resonances with a chemical shift suggesting m1, even with
the low signal-to-noise ratio of the 1-cm3 voxels of
our MRSIs suggesting considerable increases in local
concentrations. The small but reproducible mI peak
was clearly shown within the plaque during the acute
phase of the disease. This may indicate the presence
of a very large concentration of mI due to particularly
active myelin breakdown confirmed in these patients
by needle biopsy of the lesions. With such severe demyelination, it seems somewhat unlikely that a major
component of functional recovery can be attributed to
remyelination, which may only rarely be associated
with clear clinical improvements 128) and most often
appears to be a rather limited process 1361.
Because axons project through lesion volumes, any
axonal dysfunction or loss should extend well beyond
the borders of a lesion and into normal-appearing
white matter. Decreases in NAA were seen in this
study both within and beyond the border of the lesion
from the early stage of the disease. In Patients 1 and
3 the decrease extended well outside of the plaque (as
visualized by conventional MRI) at the time of the first
examination, into voxels that became part of the plaque
as shown by a subsequent examination performed a
few days later. Decreases in NAA beyond the border
of the plaques have been measured in vitro 1321
(where potential artifacts from “bleeding” of the signal
into adjacent voxels cannot occur) and extend in our
patients well beyond limits of the point spread function
(see Fig 3). It is intriguing that abnormal NAA outside
the acute plaques was seen only in those of our patients
who showed an enlargement of the plaque on subsequent MRIs and who therefore may have had the most
severe inflammation. The decrease in NAA was not
shown in Patient 2, who had a major enhancement of
the MRI lesion after injection of gadolinium, but did
not show any enlargement of the plaque in the subsequent follow-up. Decreases in relative NAA signal also
were not seen in the normal-appearing white matter
of the contralateral hemisphere, confirming that such
abnormalities noted in more advanced disease l 3 0 )
arise with increased lesion load.
In conclusion, our findings illustrate the potential
usefulness of proton MRSI for analysis of evolution of
chemical pathological changes in demyelinating lesions.
The strong correlation between NAA and EDSS score
suggests that specific chemical markers of pathological
changes in lesions may allow quantitative relationships
between new measures of disease burden and disability
to be defined. They also lend support to the notion
that impairment of neuronal metabolism or neuronal
loss may be more directly responsible for functional
disability in demyelinating diseases than has been appreciated generally.
908 Annals of Neurology Vol 38 No 6 December 1995
The study was supported by grants from the Multiple Sclerosis Society and the Medical Research Council of Canada.
Dr Matthews thanks the Medical Research Council of Canada for
personal support. The authors are grateful to Ms Arlene Cohen for
coordinating the study and to Mr Gilles Leroux and Andre Cormier
for providing excellent technical support.
1. Paty DW, Li DKB, The UBC MS/MRI Study Group, The
IFNB Multiple Sclerosis Study Group. Interferon beta-lb is
effective in relapsing-remitting multiple sclerosis. 11. MRI analysis results of a multicenter, randomized, double-blind, placebocontrolled trial. Neurology 1993;43:662-667
2. Kastrukoff LF, Oger JJ, Hashimoto SA, et al. Systemic lymphoblastoid interferon therapy in chronic progressive multiple
sclerosis. I. Clinical and MRI evaluation. Neurology 1990;40:
3. Paty DW, Li DK, Oger JJ, et al. Magnetic resonance imaging in
the evaluation of clinical trials in multiple sclerosis. Ann Neurol
4. Harris JO, Frank JA, Patronas N, et al. Serial gadoliniumenhanced magnetic resonance imaging scans in patients with
early, relapsing-remitting multiple sclerosis: implications for
clinical trials and natural history. Ann Neurol 1991;29:548-555
5. Miller DH, Barkhof F, Nauta JJ. Gadolinium enhancement increases the sensitivity of MRI in detecting disease activity in
multiple sclerosis. Brain 1993;116: 1077-1094
6. Miller DH. Magnetic resonance in monitoring the treatment of
multiple sclerosis. Ann Neurol 1994;36(Suppl):S91-S94 (Review)
7. Kurtqke JF. Rating neurologic impairment in multiple sclerosis:
an expanded disability status scale (EDSS). Neurology 1983;33:
8. Arnold DL, Matthews PM, Francis GS, et al. Proton magnetic
resonance spectroscopic imaging for metabolic characterization
of demyelinating plaques. Ann Neurol 1992;31:235-241
9. Ordidge RJ, Mansfield P, Lohman JA, Prime SB. Volume selection using gradients and selective pulses. Ann N Y Acad Sci
10. Luyten PR, Groen JP, Vermeulen JW, den Hollander JA. Experimental approaches to image localized human 3 l P NMR
spectroscopy. Magn Reson Med 1989;ll:l-21
11. den Hollander JA, Oosterwaal B, van Vroonhoven H , Luyten
PR. Elimination of magnetic field distortions in 'H NMR spectroscopic imaging. Proc SOCMagn Reson Med 1991;1:472 (Abstract)
12. Roth K, Kimber BJ, Feeney J. Data shift accumulation and alternate delay accumulation techniques for overcoming dynamic
range problems. J Magn Reson 1980;41:302
13. De Stefano N, Matthews PM, Arnold DL. Reversible decreases
in N-acetylasparrate following acute brain injury. Magn Reson
Med (in press)
14. Matthews PM, Francis G, Antel J, Arnold DL. Proton magnetic
resonance spectroscopy for metabolic characterization of
plaques in multiple sclerosis. Neurology 1991;41:125 1-1256
(published erratum appears in Neurology 1991;4 1:1828)
15. Miller D H , Austin SJ, Connelly A, et al. Proton magnetic resonance spectroscopy of an acute and chronic lesion in multiple
sclerosis. Lancet 1991;337:58-59 (Letter)
16. Davie CA, Hawkins CP, Barker GJ, et al. Serial proton magnetic resonance spectroscopy in acute multiple sclerosis lesions.
Brain 1994;117:49-58
17. Arnold DL, Riess GT, Manhews PM, et al. Use of proton magnetic resonance spectroscopy for monitoring disease progression
in multiple sclerosis. Ann Neurol 1994;36:76-82
18. McDonald WI. Rachelle Fishman-Matthew Moore Lecture. The
pathological and clinical dynamics of multiple sclerosis.J Neuropathol Exp Neurol 1994;53:338-343 (Review)
19. Filippi M, Paty DW, Kappos L, et al. Correlations between
changes in disability and T2-weighted brain MRI activity in multiple sclerosis: a follow-up study. Neurology 1995;45:255260
20. Khoury SJ, Guttmann CR, Orav EJ, et al. Longitudinal MRI in
multiple sclerosis: correlation betwen disability and lesion burden. Neurology 1994;44:2 120-2 124
21. Arnold DL, Matthews PM, Francis G , Antel J. Proton magnetic
resonance spectroscopy of human brain in vivo in the evaluation
of multiple sclerosis; assessment of the load of disease. Magn
Reson Med 1990;14:154-159
22. Doyle T, Narayana P, Jackson E, WolinskyJ. Longitudinal, short
echo time proton spectroscopic imaging of MS. Proc SOCMagn
Reson Med 1993;3:1551 (Abstract)
23. Bruhn H, Frahm J, Merboldt KD, et al. Multiple sclerosis in
children: cerebral metabolic alterations monitored by localized
proton magnetic resonance spectroscopy in vivo. Ann Neurol
1992;32:140- 150
24. Raine CS, Cross AH. Axonal dystrophy as a consequence of
long-term demyelination. Lab Invest 1989;60:7 14-725
25. Sobel A. The pathology of multiple sclerosis. 1995;l:l-21
26. Black JA, Felts P, Smith KJ, et al. Distribution of sodium channels in chronically demyelinated spinal cord axons: immunoultrastructural localizationand electrophysiologicalobservations.
Brain Res 1991;544:59-70
27. Moll C, Mourre C, Lazdunski M, Ulrich J. Increase of sodium
channels in demyelinated lesions of multiple sclerosis. Brain Res
28. Ghatak NR, Leshner RT, Price AC, Felton WL 3d. Remyelination in the human central nervous system (see comments).
J Neuropathol Exp Neurol 1989;48:507-518
29. Cadoux-Hudson TA, Kermode A, Rajagopalan B, et al. Biochemical changes within a multiple sclerosis plaque in vivo.
J Neurol Neurosurg Psychiatry 1991;54:1004-1006
30. Husted CA, Goodin DS, Hugg JW, et al. Biochemical alterations in multiple sclerosis lesions and normal-appearing white
matter detected by in vivo "P and 'H spectroscopic imaging.
Ann Neurol 1994;36:157-165
3 1. Barnes D, McDonald WI, Johnson G, et al. Quantitative nuclear
magnetic resonance imaging: characterisation of experimental
cerebral oedema. J Neurol Neurosurg Psychiatry 1987;50:125133
32. Davies SE, Newcombe J, Williams SR, et al. High resolution
proton NMR spectroscopy of multiple sclerosis lesions. J Neurochem 1995;64:742-748
33. Manos P, Bryan GK. Cellular and subcellular compartmentation
of creatine kinase in brain. Dev Neurosci 1993;15:271-279
34. Davie CA, Hawkins CP, Barker GJ, et al. Detection of myelin
breakdown products by proton magnetic resonance spectroscopy. Lancet 1993;34 1:630-63 1 (Letter)
35. Koopmans RA, Li DK, Zhu G, et al. Magnetic resonance spectroscopy of multiple sclerosis:in-vivo detection of myelin breakdown products. Lancet 1993;341:631-632 (Letter)
36. Prineas JW, Barnard RO, Revesz T, et al. Multiple sclerosis.
Pathology of recurrent lesions. Brain 1993;116:68 1-693
De Stefano et al: 'H MRSI of Demyelinating Lesions 707
Без категории
Размер файла
2 176 Кб
correlation, chemical, pathologic, lesions, disability, demyelination, acute
Пожаловаться на содержимое документа