Personal View

MRI monitoring of pathological changes in the spinal cord in

patients with multiple sclerosis
Achim Gass, Maria A Rocca, Federica Agosta, Olga Ciccarelli, Declan Chard, Paola Valsasina, Jonathan C W Brooks, Antje Bischof, Philipp Eisele,
Ludwig Kappos, Frederik Barkhof, Massimo Filippi, for the MAGNIMS Study Group*

The spinal cord is a clinically important site that is affected by pathological changes in most patients with multiple Lancet Neurol 2015; 14: 443–54
sclerosis; however, imaging of the spinal cord with conventional MRI can be difficult. Improvements in MRI provide Published Online
a major advantage for spinal cord imaging, with better signal-to-noise ratio and improved spatial resolution. Through March 4, 2015
http://dx.doi.org/10.1016/
the use of multiplanar MRI, identification of diffuse and focal changes in the whole spinal cord is now routinely
S1474-4422(14)70294-7
possible. Corroborated by related histopathological analyses, several new techniques, such as magnetisation transfer,
See Comment page 348
diffusion tension imaging, functional MRI, and proton magnetic resonance spectroscopy, can detect non-focal, spinal
*Steering Committee members
cord pathological changes in patients with multiple sclerosis. Additionally, functional MRI can reveal changes in the listed at end of paper
response pattern to sensory stimulation in patients with multiple sclerosis. Through use of these techniques, findings Department of Neurology,
of cord atrophy, intrinsic cord damage, and adaptation are shown to occur largely independently of focal spinal cord Universitätsmedizin
lesion load, which emphasises their relevance in depiction of the true burden of disease. Combinations of Mannheim UMM, University of
magnetisation transfer ratio or diffusion tension imaging indices with cord atrophy markers seem to be the most Heidelberg, Germany
(A Gass MD, P Eisele);
robust and meaningful biomarkers to monitor disease evolution in early multiple sclerosis. Neuroimaging Research Unit,
Institute of Experimental
Introduction features of multiple sclerosis in the spinal cord, assess Neurology, Division of
Inflammatory demyelination and neurodegeneration are improvements in conventional MRI techniques, and Neuroscience and Department
of Neurology, San Raffaele
present from the very early stages of multiple sclerosis. discuss the newer MRI approaches that have provided Scientific Institute, Vita-Salute
Brain MRI has helped greatly to understand the insight into pathophysiology of the disease. We also San Raffaele University, Milan,
contribution from these processes to the pathology of address the relevance of developments in imaging and Italy (M A Rocca MD, F Agosta
multiple sclerosis, and both conventional MRI and their potential implications to help understand and MD, P Valsasina PhD, M Filippi
MD); NMR Research Unit,
quantitative approaches to assess neurodegeneration monitor spinal cord involvement in multiple sclerosis. Queen Square Multiple
have become accepted measures in clinical trials. The Sclerosis Centre (D Chard PhD)
spinal cord is also a clinically important site that is Pathological changes in the spinal cord and Department of Brain
Repair and Rehabilitation
affected by pathological changes in most patients with Focal spinal cord lesions are commonly seen in patients
(O Ciccarelli PhD); University
multiple sclerosis. MRI of the spinal cord is a recom- with multiple sclerosis, and tissue abnormalities are College London, Institute of
mended and valuable part of diagnostic investigations in reported in 80–90% of patients on conventional MRI.6–13 Neurology National Institute
patients with suspected multiple sclerosis.1–3 However, Spinal cord lesions usually do not respect the anatomical for Health Research, University
College London Hospitals,
MRI of the spinal cord is more challenging than brain grey–white matter boundaries.14–16 Most lesions are
Biomedical Research Centre,
MRI because the spinal cord is a long and thin structure located in the lateral or dorsal white matter and are only London, UK; University of
that has some mobility; MRI of the spinal cord is further infrequently seen in the anterior columns and central Bristol, Clinical Research and
complicated by artifacts (ie, from breathing, heartbeat, cord area. Most focal lesions are located in the cervical Imaging Centre, Bristol, UK
(J C W Brooks PhD); Department
CSF flow, and aortic flow).4,5 These inherent features cord. However, only 33 (16%) of 202 patients showed
of Neurology, University
make it difficult to image subtle pathological changes and exclusive cervical-cord lesions in a large cohort.9 In Hospital Basel, Basel,
to ensure comparability in follow-up examinations. patients with primary progressive multiple sclerosis, Switzerland (A Bischof MD,
Technical developments in MRI have increased the diffuse areas of mildly increased T2-weighted signal L Kappos MD); Department of
Radiology and Nuclear
signal-to-noise ratio for spinal cord MRI. The use of intensity are the predominant finding in the spinal cord,
Medicine, VU University
parallel imaging techniques, including improved coil whereas patients with secondary progressive multiple Medical Center, Amsterdam,
technology and high field strength, have improved sclerosis show mostly focal lesions at high signal Netherlands (F Barkhof MD)
conventional and quantitative MRI to become more intensity.17 Unlike white matter demyelination in all Correspondence to:
robust and improve the signal-to-noise ratio, image multiple sclerosis patients, which is widespread in the Dr Achim Gass, Department of
Neurology, Universitätsmedizin
resolution, and consequently, the sensitivity and the upper cervical part of the spinal cord, grey matter
Mannheim, University of
specificity of spinal cord MRI. New MRI techniques, demyelination has been shown to extend equally over all Heidelberg, 68135 Mannheim,
such as magnetisation transfer, diffusion tensor imaging spinal levels.18 Axonal loss occurs widely, with reduced Germany
(DTI), functional MRI (fMRI), and ¹H-magnetic axonal density of about 46%, even in normal-appearing achim.gass@medma.uni-
heidelberg.de
resonance spectroscopy (MRS), show an improved white matter. This increases up to about 59% in
picture on the presence and development of lesions, and hyperintense lesions on proton-density weighted
enable imaging of previously difficult to visualise images.19 Lateral columns are most affected, and anterior
pathological changes, enabling the assessment of their columns are least affected by axonal loss.20 In a post-
significance in relation to clinical disability. mortem study14 on 37 patients with multiple sclerosis,
We review the use of spinal cord MRI in multiple most of whom had a progressive disease course, total
sclerosis. We briefly present the main pathological neuronal number at the upper thoracic level was reduced

www.thelancet.com/neurology Vol 14 April 2015 443

 Personal View
by 30·3% correlating with the multiple sclerosis disease
state. In histopathological tests, interneuron and motor
neuron numbers were reduced to various degrees,
notably at the upper spinal cord levels. These changes
seem to occur mainly in demyelinated grey matter and
less so in myelinated grey matter.21
By contrast with demyelination, axonal pathological
changes do not correlate with findings on T2-weighted
images, but seem to occur independently of de-
myelination.20 This was corroborated by histopathological
findings that axonal loss and diameter did not correlate
with myelin density.19 High signal intensity regions on T2-
weighted images showed nearly normal axonal features
(staining intensity, diameter, regularity) on histopatho-
logical examination, suggesting that inflammation does
not necessarily lead to or precede axonal loss.20
Spinal cord atrophy might be due to focal tissue
destruction in lesions or to the degeneration of tracts
secondary to pathological changes that damage axons.
Two studies22,23 reported more pronounced atrophy in the
cervical cord compared with the thoracic and lumbar cord
segments. Cord atrophy was mostly due to white matter
volume loss, and these cord lesions have little effect on
local cord areas. Another study24 showed a clear gradient
in axonal loss in the cord, with the greatest loss in the
cervical region and the least in the lumbar region.
However, atrophy seemed to result from white matter
loss because the white matter was substantially reduced
in in the upper spinal cord levels of patients with multiple
sclerosis, whereas the grey matter area was not sub-
stantially different between patients with multiple
sclerosis and healthy controls.22 By contrast, grey and
white matter have also been reported to be affected to the
same extent.25 Spinal cord atrophy was shown to associate
with T1 and, to a lesser extent, with the increase in T2-
relaxation time and with a decrease in magnetisation
transfer ratio. Furthermore, atrophy has been shown to
correlate well with disability, but is not a direct marker of
axonal numbers.19
Studies that correlate MRI findings with histo-
pathological findings suggest that conventional MRI is
mainly sensitive to detect demyelination. From ex-vivo
MRI on the spinal cord, Gilmore and colleagues14 reported
a sensitivity of 87% for the detection of white matter
lesions at 4·7 T on proton-density weighted MRI
sequences, 73% for grey matter lesions, and a high
specificity for grey matter demyelination. In a comparison
of different sequences, T2-weighted imaging was the
strongest to predict demyelination, but also other
sequences (proton density, magnetisation transfer ratio,
relaxation-time measurement) have been shown to detect
areas of demyelination.19 At 4·7 T, the proton-density
weighted high-resolution MRI correlated well in regard to
location and shape of the histopathological findings.17
Relaxation times were shown to increase with progressive
demyelination whereas the magnetisation transfer ratio
decreased.19
444
Some limitations are known to affect the extrapolation
of these post-mortem findings to in-vivo MRI of the
spinal cord. First of all, post-mortem imaging of spinal
cord lacks motion and tissue artifacts, and surface coils
provide a better signal-to-noise ratio. Furthermore, MRI
parameters, especially T1-relaxation time and magneti-
sation transfer ratio, are affected by formalin fixation.
Spatial resolution and histopathological specificity is low
in in-vivo MRI. Therefore, extrapolation of ex-vivo data to
in vivo should be done with caution.
Conventional spinal cord MRI
Imaging of demyelination
Conventional spinal cord MRI is usually undertaken in
the sagittal plane to cover a large anatomical area. Proton-
density weighted, T2-weighted, or short tau inversion
recovery (STIR) sequences are usually used, alone or in
combination, to visualise the outline of the cord, focal
lesions, and diffuse hyperintense signal change. STIR
sequences take longer than proton-density weighted MRI
but have a higher sensitivity to detect lesions.26 Frequently,
two sets of sagittal images with different contrasts (proton-
density weighted and T2-weighted, or T2-weighted and
STIR) are used to increase confidence in lesion detection,
particularly because of spinal cord MRI motion artifacts.
The inherent contrast of spinal cord lesions against
healthy cord signal is usually lower than in the brain
parenchyma, which is one reason why fluid-attenuated
inversion recovery (FLAIR) sequences are rarely used for
multiple sclerosis lesion detection in the spinal cord.27,28
One advantage of the relative insensitivity of T1-weighted
MRI for the detection of lesions29 is its robustness for
spinal cord atrophy assessments. Whole-cord biplanar
MRI (in the sagittal and axial planes) is time-consuming,
but parallel imaging techniques with multiarray coils
enable a scan of the whole cord to take place quickly.
Imaging in the axial plane can be very useful to ascertain
suspected abnormal changes and can increase the
confidence for the lesion detection in peripheral cord
locations, where partial-volume effects (limited resolution)
make lesion detection difficult on sagittal images
(figure 1).9 Diffuse hyperintense signal change, a patho-
logical feature that is particularly frequent in patients with
primary progressive multiple sclerosis, is usually seen in
the whole-cord cross-section. Differences in the reported
frequency of these diffuse changes might be due to the
lower contrast-to-noise ratio of diffuse versus focal lesions
and patient selection in different studies.9,13 Acute lesions
are often associated with cord swelling and with acute
symptoms, whereas overall30 contrast-enhancing lesions
are less frequently seen in the spinal cord than in the
brain.31,32
The sensitivity of conventional MRI to detect focal
lesions is particularly valuable in the diagnostic con-
siderations of the early phases of multiple sclerosis. In
patients with clinically isolated syndrome (CIS) and
radiologically isolated syndrome (RIS), asymptomatic
www.thelancet.com/neurology Vol 14 April 2015

spinal cord lesions have been described in 30–40% of
patients.33 People with RIS seem to be at a high risk of
progression to CIS.33 Some patients with longer disease
duration also show more focal lesions; in particular,
patients with primary progressive multiple sclerosis
might also show an increase in diffuse signal intensity
on T2-weighted and proton-density weighted images
throughout a large portion of the cord.34,35 In patients
with primary progressive multiple sclerosis, cord abnor-
malities are frequently more pronounced than brain
abnormalities. Clinical correlations with MRI spinal cord
lesion counts have been repeatedly low. Most studies
have focused on the analysis of focal lesions, and
attempts to include both focal lesions and diffuse signal
abnormalities have not been any more revealing.9 Even
in a study with a large group of patients, correlations
were not strong enough to make meaningful assessments
on an individual patient basis.9
Spinal cord atrophy assessments
Quantification of the spinal cord cross-sectional area at a
high cervical level has been the most reported MRI
measure of cord atrophy.36 Spinal cord atrophy has been
detected in patients at high risk of developing multiple
sclerosis (ie, those with cord lesions and additional brain
lesions) after a CIS.37 In patients with secondary
progressive multiple sclerosis, the atrophy in the cervical
cord area is about 16–28%, whereas in patients with
clinically benign disease, spinal cord atrophy was
considerably less (around 5–8%)36,38. A study on a large
and diverse cohort (143 healthy controls, 22 CIS,
101 relapsing remitting multiple sclerosis, 79 secondary
progressive multiple sclerosis, 58 benign multiple
sclerosis, 75 primary progressive multiple sclerosis) has
substantiated the presence of cervical cord atrophy in
people with progressive disease, and the relative absence
in patients with benign multiple sclerosis. Spinal cord
atrophy was not detected in patients with CIS or with
relapsing remitting multiple sclerosis.39 Klein and
colleagues40 reported increased tissue volumes (when
normalised for intracranial volumes) in low cervical and
thoracic cord in patients with relapsing remitting
multiple sclerosis, perhaps attributable to inflammation
or oedema associated tissue swelling. Other work has
suggested a rate of cord atrophy in the order of 1–5% per
year in people with multiple sclerosis,36,41–43 and cord
atrophy seems to be faster in patients with secondary
progressive multiple sclerosis compared with primary
progressive multiple sclerosis.32,44 However, as highlighted
by the varying results in CIS and relapsing remitting
multiple sclerosis cohorts, spinal cord atrophy might not
be a linear process.
A study on 440 patients with multiple sclerosis
(311 relapsing remitting, 92 secondary progressive, and
37 primary progressive) confirmed that spinal cord
atrophy is more pronounced in those with progressive
disease.45 Although there was no association between
www.thelancet.com/neurology Vol 14 April 2015
Personal View
A B
Figure 1: Biplanar conventional MRI of a 41-year-old female patient with relapsing remitting multiple sclerosis
(A) Sagittal proton-density weighted MRI of the whole spinal cord, with lesions seen as hyperintense areas.
(B) The axial images confirm the location of the pathology; lesions are marked with green arrow heads.
spinal cord lesion load and spinal cord atrophy,
conventional brain MRI lesion burdens were related to
spinal cord atrophy to various extents, which suggests
that the reduction in upper cervical cord cross-sectional
area could be a result of T1-hypointense lesions in the
brain rather than lesions of the spinal cord. Furthermore,
the cervical spinal cord cross-sectional area and brain T1-
weighted lesion load were the strongest MRI predictors of
disability on the expanded disability status scale (EDSS)
score, but atrophy of the spinal cord was more strongly
related to disability (R²=0·564) than spinal cord lesion
load.45 Few studies have assessed to what extent spinal
cord atrophy contributes to disability when brain lesion
load and atrophy has been accounted for. Two studies46,47
reported spinal cord atrophy to be associated with physical
disability in regression models, suggesting that it makes
a pronounced independent contribution. A 2 year follow-
up study in 352 patients with multiple sclerosis measured
conventional brain and spinal cord MRI parameters in
addition to the annualised percentage brain volume
change (aPBVC), and the annualised percentage upper
cervical cord cross-sectional area change (aUCCA)
obtained from the caudal region of the three-dimensional
T1-weighted brain MRI acquisition. aUCCA over
24 months was highest in patients with secondary
progressive multiple sclerosis (–2·2% per year), and was
significantly higher in patients with disease progression
(–2·3% per year) than in stable patients (–1·2% per year;
p=0·003). Whereas aPBVC did not differ between
subtypes (relapsing remitting multiple sclerosis: –0·42%
per year; secondary progressive multiple sclerosis: –0·6%
per year; primary progressive multiple sclerosis: –0·46%
per year), or between progressive and stable patients
(p=0·055).48 Baseline UCCA and aUCCA over 24 months
445
 Personal View
were reported to contribute to EDSS at month 24, whereas
baseline UCCA and number of spinal cord segments
with lesions at baseline, but not aUCCA over 24 months,
were relevant contributors to disease progression.48 This
study further underlines the feasibility of follow-up
atrophy assessments and their value in provision of
clinically important information not obtained from brain
MRI analyses.
A method for voxel-wise statistical analysis of cord
volume has been developed and validated.49 The
investigators showed the regional distribution of cord
atrophy in patients with relapsing remitting multiple
sclerosis versus secondary progressive multiple sclerosis.50
Patients with relapsing remitting multiple sclerosis had
localised clusters of atrophy in the posterior cord, whereas
patients with secondary progressive multiple sclerosis
showed a more widespread pattern of cord atrophy,
predominantly in the posterior and lateral cord columns
(figure 2). The same approach was used to map the
regional distribution of atrophy and T2-hyperintense
lesions in the cervical cord of patients with multiple
sclerosis with different clinical phenotypes.51 Notably, the
spatial distributions of cord atrophy and cord lesions were
not the same for any phenotype,51 suggesting that atrophy
is not necessarily the local result of focal demyelination.
Correlations between cord lesion loads and EDSS scores
have been inconsistent and of low strength when
compared with associations of cord atrophy and EDSS
(where reported, Spearman correlations range from 0·2
to 0·7).36,43 Measures of spinal cord atrophy have the
A A B C
C1/C2
L R
C3
L R A L
C3
C4
L R
L R
C5
C7
L R
L R
C6
P
L R
C7
L R
A P A P A P
P P
t value 0 6 t value 0 4 t value 0
Figure 2: Voxel-wise assessment of the regional distribution of spinal cord atrophy
Sagittal and axial MRI at different cervical cord levels on 45 patients with relapsing remitting multiple sclerosis,
26 patients with secondary progression multiple scelerosis and 67 age-matched healthy control subjects. Colour
coded for t values; p=0·001, data uncorrected only for display reasons. (A) Patients with all subtypes of multiple
sclerosis compared with healthy control subjects. (B) Patients with relapsing remitting multiple sclerosis compared
with healthy control subjects. (C) Patients with secondary progressive multiple sclerosis compared with those with
relapsing-remitting multiple. A=anterior. C=cervical region. L=left. R=right. P=posterior.
446
potential to provide a useful and clinically relevant marker
of disease progression and offer additional information
when acquired alongside brain lesion load and brain
atrophy measures.
New techniques for spinal cord imaging
Magnetisation transfer imaging
Magnetisation transfer imaging of the spinal cord has
provided substantial insights into focal and diffuse tissue
abnormalities associated with multiple sclerosis because
of its sensitivity to microstructural damage. Magnetisation
transfer imaging is a quantitative MRI technique based
on the interactions between protons in a free environment
and motional restricted protons that are bound to
macromolecules. With an off-resonance radiofrequency
pulse, the magnetisation of bound protons is saturated
and transferred to the mobile protons, thus reducing the
signal intensity of the observable magnetisation.52 Since
in the CNS the bound protons are mainly associated with
myelin lipids and proteins, the amount of signal decrease
is thought to suggest damage to myelin or to the axonal
membranes.19,52,53
The magnetisation transfer ratio, calculated from the
magnetisation transfer-on image and magnetisation
transfer-off image, is the most common quantitative
metric used to assess the magnitude of the magnetisation
transfer effect on a pixel-by-pixel basis (figure 3).
Magnetisation transfer maps can subsequently be
compared between controls and patients with multiple
sclerosis either by region of interest or by histogram
analysis.54 Several studies55–58 have shown that measure-
ment of the magnetisation transfer ratio in the cervical
A
C1/C2 cord of patients with multiple sclerosis is feasible and
leads to reliable collection of quantitative MRI data,
L R
which is not the case for conventional MRI. These
C3 studies have shown that changes in the magnetisation
R
transfer ratio of the spinal cord probably occur late in the
course of the disease because no abnormalities were
C4 detected in patients with CIS, relapsing remitting
L R multiple sclerosis, or paediatric multiple sclerosis.
Conversely, substantial reductions in histogram-derived
C5
magnetisation transfer ratio were recorded in the spinal
L R cord of patients with primary progressive multiple
C6
sclerosis and patients with secondary progressive
multiple sclerosis. These cord magnetisation transfer
L R
ratios were much lower than those noted in patients
C7 with relapsing remitting multiple sclerosis.59 In patients
with primary progressive multiple sclerosis, disability
L R
was statistically significant according to the EDSS scores,
albeit weakly, associated with a composite MR model,
8
including measures of axonal loss and intrinsic damage
of cervical cord tissue, indicated by a reduced
magnetisation transfer ratio and signs of cord atrophy.59
The magnetisation transfer ratio was significantly
altered in patients with locomotor disability compared
with those without (p=0.001).60 Moreover, in patients
with relapsing remitting multiple sclerosis, a moderate
www.thelancet.com/neurology Vol 14 April 2015

association was identified between baseline spinal cord
magnetisation transfer ratio and disability changes and
relapse rate over 18 months of follow-up.61
Some technical developments have been introduced to
improve both magnetisation transfer acquisition and
analysis. To overcome the restrictions due to the small size
of the spinal cord and its sensitivity to motion, the use of
three-dimensional gradient-echo sequences, which obtain
high-resolution images with a high signal-to-noise ratio
and a strong magnetisation transfer saturation effect,62 has
been proposed. This approach was applied in a preliminary
study,62 in which magnetisation transfer on and
magnetisation off images were acquired within a single
sequence together to obtain inherently coregistered
magnetisation transfer saturated and unsaturated images.
Despite the small patient cohort included in the study, this
method was sufficiently sensitive to reveal differences in
the average magnetisation transfer ratio and peak location
of seven patients with multiple sclerosis compared with
healthy control participants. Magnetisation transfer
analysis improved the regional assessment of cord
damage, by measuring the cord grey matter in isolation
from white matter. With this approach, a substantial
reduction of average grey matter magnetisation transfer
ratio was recorded in 18 patients with relapsing remitting
multiple sclerosis without T2-visible cord lesions,63 which
was also associated with patients’ disability. A further novel
approach that has been developed to overcome cord
motion-related problems, which might cause an imperfect
subtraction between magnetisation-saturated and
magnetisation-unsaturated images leading to an unreliable
magnetisation transfer ratio, is a new metric to measure
the magnetisation transfer effect in the cord tissue by
using the CSF as an internal magnetisation transfer signal-
intensity reference for the spine. This technique needs the
acquisition of only one magnetisation-saturated image,
and measures the strength of the signal at each voxel by
dividing voxel-signal intensity by CSF-signal intensity.
This new metric (MTCSF64) was successfully used to
measure tissue damage in the spinal cord of 42 patients
with multiple sclerosis.65 With a region-of-interest-based
approach, this study showed a substantial MTCSF change
in the dorsal and in the lateral columns of these patients,
which were both associated with sensorimotor impairment
and disability scores.
In a further study, magnetisation transfer ratio abnor-
malities were reported in an area corresponding to the
expected location of the pia mater and subpial region in
the outer-cervical spinal cord.66 These outer spinal cord
abnormalities were also seen early in the course of
multiple sclerosis before cord atrophy was pronounced,
and a greater reduction in magnetisation transfer ratio
values was noted in progressive multiple sclerosis. A
composite measure, which took into account the cervical
cord magnetisation transfer ratio and cross-sectional
area, improved the strength of the correlation between
cervical cord MRI findings and the EDSS score.67
www.thelancet.com/neurology Vol 14 April 2015
Personal View
A B
MT off
MT on
MTR
MD FA Diffusion directions
Figure 3: Magnetisation transfer ratio and diffusion tensor MRI maps of the cervical cord
(A) Magnetisation transfer MRI: axial gradient-echo images of the cervical cord from a patient with relapsing
remitting multiple sclerosis. (B) Diffusion tensor MRI obtained from a patient with relapsing remitting multiple
sclerosis showing sagittal MD, left; FA, middle; and diffusion directions (right panel, blue shows diffusion along the
caudo-rostral and rostro-caudal direction). Diffusion maps were obtained from a patient with relapsing remitting
multiple sclerosis. MT=magnetisation transfer. MTR=magnetisation transfer ratio. MD=mean diffusivity.
FA=fractional anisotropy.
Diffusion tensor imaging
Diffusion MRI has been widely used to study multiple
sclerosis because it can detect and quantify CNS disease-
related pathological changes.68 The characteristics of
diffusion, which can be defined as the random
translational motion of molecules in a fluid system, are
changed by several CNS tissue components, including
cell membranes and organelles. The MRI-measured
diffusion coefficient of healthy CNS tissues—the
apparent diffusion coefficient (ADC)—is lower than the
ADC of free water.68 Pathological processes that lead to a
decrease of barriers that restrict diffusion can lead to an
increase in ADC values.68 Because the magnitude of
diffusion is also dependent on the direction in which it is
measured, a more complete characterisation can be
obtained from a diffusion tensor (DT),69 a matrix that
accounts for the correlation between molecular displace-
ments along orthogonal directions. From the DT,
parameters that are quantitative and invariant to the
choice of reference frame can be derived, including the
mean diffusivity, which measures the average molecular
motion independent of any tissue directionality, and the
fractional anisotropy (FA), which shows the diffusivity
along one spatial direction (figure 3).69 Post-mortem
studies in multiple sclerosis have shown that the main
pathological correlates of diffusivity changes are
demyelination and axonal loss,53,70 with the correlation
with pathological features being stronger for FA than for
diffusivity indices.53,70 One study showed that the DT
indices in longitudinal diffusivity and radial diffusivity
show a pathological increase, and FA shows a pathological
decrease in T2-hyperintense lesions. These effects were
mainly caused by demyelination, which was corroborated
by an immunofluorescent assay in which axons were still
relatively preserved.71
447
 Personal View
Using a region-of-interest approach72 in three patients
with multiple sclerosis, increased mean diffusivity and a
tendency towards decreased FA in cord lesions of
patients, in comparison with the cord of healthy controls,
was reported. By applying sensitivity encoding echo-
planar imaging to DT MRI of the cervical cord,73 the
extent of cervical cord damage was estimated in patients
with multiple sclerosis with different clinical phenotypes.
With a histogram-based approach, 44 patients with
relapsing remitting multiple sclerosis and secondary
progressive multiple sclerosis had increased mean
diffusivity and decreased FA of the cervical cord
compared with healthy controls. However, no substantial
differences in cord DT MRI measures were seen between
the two patient cohorts.74 In the same study, using a
multivariate linear regression model, cervical cord
average FA and brain average mean diffusivity were
independent determinants of clinical disability.
Patients with primary progressive multiple sclerosis
also had abnormal mean diffusivity and FA of the cervical
cord.40 When compared with secondary progressive
multiple sclerosis,75 patients with benign multiple
sclerosis had milder cervical cord damage. Moreover, in
patients with benign multiple sclerosis, both cord cross-
sectional area and FA were independently associated
with the extent of clinical disability,75 thus suggesting that
DT MRI could contribute to a comprehensive assessment
of spinal cord damage.
Several reports76–78 used a region-of-interest-based
approach to quantify damage within the different cervical
spinal cord compartments (ie, white matter of the different
cord columns and central grey matter). These studies
showed substantial DT MRI changes in diffusion in both
the normal-appearing white matter and grey matter in the
spinal cord of patients with multiple sclerosis compared
with healthy controls. These changes were more severe in
cord macroscopic lesions seen on the MRI than in the
normal-appearing white matter.
Cervical spinal cord DT tractography has been combined
with proton MR spectroscopy to estimate cord damage
occurring in 14 patients with spinal relapses.79 Compared
with controls, these patients had reductions in
N-acetylaspartate concentrations, which is a specific
marker in vivo for neuronal health and integrity, con-
nectivity, and FA of the lateral corticospinal tracts and
posterior tracts. These measurements were shown to
correlate with acute disability. The 6 month longitudinal
assessment of these patients showed that low (more
healthy) radial diffusivity of the corticospinal tract at
baseline was associated with improved clinical outcome.80
Another study compared different quantitative cervical
spinal cord metrics with system specific measurements
and general clinical measures in 129 patients with
multiple sclerosis.81 The authors calculated spinal cord
cross-sectional area, FA, mean diffusivity, perpendicular
diffusivity (λ⊥), parallel diffusivity (λ||), and magnetisation
transfer ratio on axial slices at C3–C4. All MRI indices
448
were substantially different, indicating pathological
tissue changes, in patients with multiple sclerosis versus
healthy controls, and between the progressive multiple
sclerosis subtypes versus relapsing multiple sclerosis
subtypes, with the exception of parallel diffusivity. In
multivariable regression models adjusted for age, sex,
brain parenchymal fraction, and spinal cord cross-
sectional area, the MRI indices independently explained
variability in hip flexion strength (p<0·01
for mean diffusivity, perpendicular diffusivity, and
parallel diffusivity, p=0·07 for FA), vibration sensation
threshold (p=0·04 for FA, p=0·05 for magnetisation
transfer ratio, and p=0·06 for perpendicular diffusivity),
and EDSS scores (p=0·003 for FA, p=0·03 for mean
diffusivity, p=0·005 for perpendicular diffusivity, and
p=0·02 for magnetisation transfer ratio). The same
method was used on 124 patients with multiple sclerosis82
who were dichotomised into a low-lesion or high-lesion
group. Of the patients with low lesion counts, the
microstructural abnormalities (such as small amounts of
demyelination, gliosis and neuronal loss) assessed by
magnetisation transfer ratio, cross-sectional area, and
diffusion imaging, were more pronounced than in
patients with a high EDSS (≥6) compared with those with
a lower EDSS (<6). In this context, the information on
microstructural abnormalities were helpful to understand
the clinical differences that was not accounted for with
the conventional MRI approach.82
At present, only one 2-year follow-up study has assessed
longitudinal changes by use of cord DT MRI in a large
sample of patients with multiple sclerosis.83 The authors
showed that cervical cord cross-sectional area and FA
decreased, and cervical cord mean diffusivity increased
during follow-up. Cord FA decrease, but not cord cross-
sectional area and mean diffusivity, was substantially
higher in patients with primary progressive multiple
sclerosis than in those with either relapsing remitting or
secondary progressive disease. In this study, multiple
sclerosis cord pathological changes were also proved
independent of concomitant brain changes and were
associated with medium-term disability accrual.
fMRI of the spinal cord
Despite being technically challenging, a spinal cord spin-
echo based fMRI with signal enhancement by
extravascular proton contrast has been shown to provide
reliable data in human beings.84–86 Cord signal changes
can be detected during different motor and sensory
paradigms85 (figure 4). Spinal cord fMRI in patients with
relapse-onset multiple sclerosis showed an altered
pattern of functional activations in the cervical cord
during both tactile and proprioceptive stimuli of the right
upper limb.84,85 Functional changes in the cervical cord
associated with tactile stimulation of the right hand were
assessed in a cohort 49 of patients with relapsing
remitting or secondary progressive multiple sclerosis.86
Greater cervical cord activation was present in patients
www.thelancet.com/neurology Vol 14 April 2015

with relapsing remitting or secondary progressive
multiple sclerosis compared with controls.86 Moreover,
severely disabled patients showed an enhancement of
cord activity not only compared with controls, but also
compared with patients with less clinical disability (ie,
EDSS ≤3·5).86 Tactile-associated cervical cord over-
activation was also recorded in patients with primary
progressive multiple sclerosis.87
Compared with healthy controls, both relapse-onset
and patients with primary progressive multiple sclerosis
showed altered distribution of tactile-associated cervical
cord activity.84,87 In the expected regions of neuronal
involvement for a tactile stimulation of the palm of the
right hand, patients showed a functional lateralisation of
cord activity, which was predominant in the cord side
ipsilateral to the stimulus, and a more frequent activation
of the posterior quadrants than the anterior cord
quadrants.84,87 Activity of the anterior cord seemed to be
more prevalent in patients with multiple sclerosis than in
healthy controls, on both the ipsilateral and contralateral
sides.84,87 Additionally, unlike healthy controls, patients
with multiple sclerosis did not show differential activation
between the right and the left hemicords.84,87 The increased
activation of the cervical cord in patients with relapsing
remitting or primary progressive multiple sclerosis was
associated with cord structural damage, measured with
DT MRI, whereas no association was identified with cord
atrophy and cord-T2 focal-lesions.86,87
In a comparison of 17 healthy controls and patients
with progressive multiple sclerosis (18 secondary
progressive, 16 primary progressive), both progressive
multiple sclerosis forms had an over-recruitment of the
cervical cord, which was more pronounced in secondary
progressive than primary progressive multiple sclerosis,
despite similar levels of cord structural damage. However,
patients with secondary progressive multiple sclerosis
had more pronounced focal and diffuse brain damage,
which would advocate for an effect of supraspinal brain
abnormalities on the recorded cord activation changes.88
Rocca and coworkers89 investigated the association
between fatigue and cervical cord abnormalities by
obtaining spinal cord fMRI in 20 controls, 15 patients
with multiple sclerosis without fatigue, and 20 patients
with multiple sclerosis with fatigue. In response to tactile
stimulation, the patients with multiple sclerosis showed
more widespread cord activation than controls, but this
increase was less pronounced in patients with fatigue
and was related to the severity of fatigue. The authors
suggest that the differences in cord fMRI activation are
caused by changes in cortical inhibition and changes in
spinal interneuron activation, which are likely to serve as
compensatory mechanisms for the loss of resources to
maintain normal function. In patients with fatigue, such
compensatory mechanisms might be compromised, but
local structural cord damage did not seem to be a relevant
factor to explain the different activation patterns in
patients with and without fatigue.89
www.thelancet.com/neurology Vol 14 April 2015
Personal View
A C5 C5/6 C6 C6/7 C7 C7/T1 T1
L R
B
L R
1·68 4
t value
Figure 4: Activation maps of an fMRI on the spinal cord
Activation maps (colour-coded for t values) of the cervical cord projected onto
axial proton-density-weighted spin-echo images from C5 to T1 from a healthy
control (A) and a patient with multiple sclerosis (B) during a tactile stimulation
of the palm of the right hand. fMRI=functional MRI. C=cervical region.
T=thoracic region. L=left. R=right.
¹H-magnetic resonance spectroscopy
¹H-MRS of the spinal cord is technically challenging90–92
and patients with multiple sclerosis have rarely been
analysed with brain ¹H-MRS studies.79,93–97 Until now, the
metabolite that has been most studied in the spinal cord
is N-acetylaspartate (total N-acetylaspartate [tNAA], equal
to NAA + NAA-glutamate).98 One study has detected a
linear decrease in tNAA concentrations in the upper
cervical cord with ageing in healthy participants,
suggesting that this tNAA decline shows age-related
neuroaxonal degeneration.92
Despite technical differences in the methods applied to
acquire spectra from the spinal cord, which include the
use of MR scanners with different magnetic field
strengths, and differences in the clinical characteristics
of the patient groups studied (table), reports so far have
consistently shown a reduction in the tNAA concentration
in patients with multiple sclerosis compared with healthy
controls, which varied between 21% and 39%79,94,96,97 (table,
figure 5). This reduction in tNAA concentration could be
clinically relevant, as it has been shown to correlate with
the cerebellar function score94 and performance on the
nine-hole peg test.79
When the temporal concentration of tNAA after an
acute cervical cord lesion was investigated, as the acute
lesion resolved the tNAA concentration partly recovered,91
mirroring the temporal behaviour of NAA described in
brain studies.100,101 These findings suggest that reduced
tNAA indicates not only reduced neuronal integrity but
also metabolic dysfunction.102 Additional evidence for this
metabolic component of tNAA comes from the reported
absence of correlation between spinal cord tNAA
concentration and spinal cord atrophy.79,93 Thus, these
results suggest that differences in tNAA concentrations
are seen in pathological processes that are different from
neuroaxonal degeneration, which is known to be well
captured by in-vivo measurements of spinal cord atrophy.
The partial recovery of tNAA concentrations after an
acute spinal cord relapse was associated with the extent
449
 Personal View

Scanner Spinal cord Type of MS EDSS Tissue included in the Reduced NAA in patients versus
level spectroscopic voxel controls
Kendi et al (2004)93 1·5 T C3–C7 Not reported Not reported NAWM Yes, but reduction not reported
Blamire et al (2007)94 2·0 T C3 4 RR, 7 SP 4·5 Not reported 32%
Ciccarelli et al (2007)79 1·5 T C1–C3 13 RR, 1 SP 4·0 Acute lesions 39%
Marliani et al (2010)96 3·0 T C2–C3 15 RR Not reported Chronic lesions 21% (NAA:Cho)
Ciccarelli et al (2013)99 3·0 T C1–C3 10 RR, 5 NMO Not reported Chronic lesions 29%

MS=multiple sclerosis. EDSS=expanded disability status scale. NAA=N-acetylaspartate. NAWM=normal appearing white matter. RR=relapsing remitting. SP=secondary
progressive. Cho=choline-containing compounds. NMO=neuromyelitis optica.

Table: Magnetic resonance imaging studies of the spinal cord in patients with multiple sclerosis

Cho NAA
A C
Cr

Cho
D mIns

NAA
Cr

lipids

4·0 3·8 3·6 3·4 3·2 3·0 2·8 2·6 2·4 2·2 2·0 1·8 1·6 1·4 1·2 1·0 0·8 0·6 0·4

Chemical shift (ppm)

Figure 5: Cervical cord ¹H-spectroscopy with a 3 T MR system

Spectroscopic voxel located on an axial (A) and sagittal (B) plane T2-weighted images between C2 and C3. Spectra obtained with LCModel analysis of a healthy control
(C) and a patient with relapsing remitting multiple sclerosis (D). The patient spectrum showed reduced NAA/Cr and NAA/Cho ratios, and increased mIns/Cr and Cho/Cr
ratios; lipids are also visible in the spectrum of the patient with multiple sclerosis. MR=magnetisation transfer. Cho=choline. Cr=creatine plus phosphocreatine.
NAA=N-acetylasparate. mIns= myo-inositol. Red line demonstrates the identified peaks from the MRS data, the grey line defines the baseline. Reproduced with
permission from Anna Federica Marliani (Neuroradiology Department, Bellaria Hospital, Bologna, Italy).

of clinical recovery, suggesting that improved neuronal
metabolism could be a mechanism of tissue repair that
leads to clinical improvement.95
These findings support the use of spinal cord ¹H-MRS
to obtain insights into the mechanisms of damage and
repair of the CNS. None of the other metabolites that can
be measured with ¹H-MRS, such as choline-containing
compounds, creatine plus phosphocreatine and myo-
inositol, has been noted to be consistently different in
patients with multiple sclerosis when compared with
healthy controls (figure 5).79,93,96,99 However, some
associations between acute disability and these metabolite

450 www.thelancet.com/neurology Vol 14 April 2015


concentrations have been described and should be studied
further.96 Additionally, reduced myo-inositol normalised to
creatine plus phosphocreatine was recorded in
neuromyelitis optica lesions of the spinal cord compared
with multiple sclerosis lesions and healthy cord, possibly
suggesting astrocytic damage in neuromyelitis optica.99
These findings are in agreement with pathological
evidence that neuromyelitis optica is an autoimmune
astrocytopathy, in which the damage to astrocytes is
substantial and widespread.103
Conclusions and future directions
The progress of MRI instrumentation provides a major
advantage for conventional spinal cord imaging with a
better signal-to-noise ratio and improved spatial
resolution, which has already helped modern quantitative
MRI techniques further our understanding of multiple
sclerosis. It is now possible to identify diffuse and focal
changes in an examination of the whole spinal cord with
conventional MRI. MRI in both axial and sagittal planes
might be best suited to detect focal lesions in the cord as
diagnostic information or indicators of disease activity. It
would be of great value to conventional and advanced
techniques if there was further technical progress to
reduce motion artifacts for spinal cord MRI.
Histopathological studies have shown that conventional
MRI is highly sensitive to focal demyelination and less
sensitive to visualise axonal loss or more subtle degrees of
demyelination. Cord atrophy, intrinsic microstructural
cord damage, and adaptation occur largely independent
of focal spinal cord lesion load, underlining their
relevance in depicting the true burden of disease. Cross-
sectional and longitudinal studies on cervical cord atrophy
have shown that large cohorts of patients with multiple
sclerosis can be useful to appreciate the information
provided by spinal cord atrophy, potentially independent
from that of brain MRI findings. The more advanced
techniques have so far focused on the technically most
accessible area—the cervical spinal cord. Although
¹H-MRS is technically challenging and has been mainly
used in small studies, it offers high specificity particularly
on axonal integrity. Magnetisation transfer and DTI
metrics are well known quantitative techniques for brain
imaging. In the spinal cord, both techniques are sensitive
to demyelination and axonal loss that is not visible on
conventional spinal cord MRI; however, their specificity
to differentiate demyelination and axonal loss is limited.
Microscopic changes and spinal cord atrophy tend to
become apparent after longer disease durations, and
could be very small or not detectable in the early stages of
multiple sclerosis. However, changes of magnetisation
transfer and DTI metrics seem to be clinically meaningful,
as they contribute to reported clinical differences in
patients with a similar lesion burden on conventional
spinal cord MRI. In this regard, quantitative MRI
measures show the ability to provide clinically relevant
information beyond MRI lesion load. Multicentre and
www.thelancet.com/neurology Vol 14 April 2015
Personal View
Search strategy and selection criteria
References for this Personal View were identified through
searches of PubMed with the search terms (and synonyms)
“spinal cord MRI”, and “spinal cord pathology”, “T2-weighted
MRI”, “spinal cord atrophy”, “magnetisation transfer”, and the
key terms “multiple sclerosis”, “diffusion weighted MRI”,
“functional MRI” and/or “¹H-spectroscopy” from 1990 to July,
2014. We also searched the abstract databases of the
European Committee for Treatment and Research in Multiple
Sclerosis congress 2012 and 2013, American Academy of
Neurology congress 2012 and 2013, and International
Society for Magnetic Resonance in Medicine congress 2013
and 2014 for relevant studies. The abstracts of retrieved
citations were reviewed and prioritised by relevant content,
and by the quality of evidence reported. From the large
number of manuscripts, we included mainly studies with a
prospective design. Only reports in English were reviewed.
longitudinal studies in large patient cohorts would be
useful to characterise the disease dynamics and support
and extend the results from cross-sectional studies.
Present findings suggest that different pathological
processes that contribute to clinical disability could be
better shown, including quantitative measures like
magnetisation transfer ratio or DTI indices with cord
atrophy markers, in addition to conventional spinal cord
MRI. One report showed high-resolution cervical cord
MRI at 7 T with an increase of nearly 4·2-times on T2-
weighted spinal cord MRI at 7 T relative to 3 T, which
should not only be useful to provide more anatomical
detail, but will surely help quantitative MRI approaches
in the future.104
Future technical developments, which will be associated
with the use of higher-magnetic fields, might make it
possible to cover not only a larger area of the spine,105 but
also to obtain spectra from the lower cervical spine and
thoracic levels,92 and to quantify new metabolites, such as
glutamate and glutamine, in a method that could be
applied in the clinic.106 One study with specially designed
motion and instability correction methods for applications
to the spinal cord and a custom-made, neck coil array,
showed a reliable quantification of glutathione in all, and
aspartate, lactate, and GABA in some of the volunteers.107
The combination of NAA concentration with other
imaging measures sensitive to structural damage might
allow exploration of the clinical significance of the
metabolic component of tNAA.108 These factors, together
with possible increases in ¹H-MRS sensitivity and spatial
resolution, might make spinal cord ¹H-MRS an invaluable
technique not only for diagnostic and prognostic reasons,
but also for understanding in vivo the neurobiological
mechanisms of injury and repair.
Contributors
AG, MF, LK, and FB decided on the structure of the Personal View. All
authors participated in the preparation of a first draft of the manuscript,
451
 Personal View
tables and figures. AG and MF prepared a first version of the
manuscript, contributed to writing of subsequent versions, helped to
select references, and reviewed the final version of the manuscript.
MAGNIMS Study Group Steering Committee
A Rovira, N de Stefano, X Montalban, F Barkhof, C Enzinger, M Filippi,
J Frederiksen, L Kappos, O Ciccarelli, J Palace, H Vrenken, M A Rocca,
T Yousry.
Declaration of interests
AG has received honoraria for lecturing, travel expenses for attending
meetings, and financial support for research from Bayer Schering,
Biogen Idec, Merck Serono, Novartis, and TEVA Neurosciences. FA has
received research supports from the Italian Ministry of Health and
AriSLA—Fondazione Italiana di Ricerca per la Sclerosi Laterale
Amiotrofica, and speaker honoraria from Biogen Idec and Serono
Symposia International Foundation. DC has received honoraria from
Bayer, Teva and the Serono Symposia International Foundation for
faculty-led education work, Teva for advisory board work, and holds stock
in GlaxoSmithKline. MAR has received speaker honoraria from Biogen
Idec, Serono Symposia International Foundation, Genzyme, and
Novartis and receives research support from Italian Ministry of Health
and Fondazione Italiana Sclerosi Multipla. MF serves on scientific
advisory boards for Teva Pharmaceutical Industries; has received
compensation for consulting services and/or speaking activities from
Bayer Schering Pharma, Biogen Idec, Merck Serono, and Teva
Pharmaceutical Industries; and receives research support from Bayer
Schering Pharma, Biogen Idec, Merck Serono, Teva Pharmaceutical
Industries, Italian Ministry of Health, Fondazione Italiana Sclerosi
Multipla, Cure PSP, Alzheimer’s and Drug Discovery Foundation, and
the Jacques and Gloria Gossweiler Foundation (Switzerland). OC has
received research funding from the Wellcome Trust, MS Society of Great
Britain and Northern Ireland, UCL Biomedical Research Centre, and
Engineering and Physical Sciences Research Council. She has also
received speaker honoraria from Bayer. AB received travel support from
Biogen Idec. PE has received travel expenses from Bayer Health Care.
FB serves as a consultant for Bayer Shering Pharma, Sanofi-Aventis,
Biogen-Idec, UCB, Merck-Serono, Novartis, and Roche. LK has
participated in the past 36 months as member or chair of planning and
steering committees or advisory boards in corporate-sponsored clinical
trials in multiple sclerosis and other neurological diseases. The
sponsoring pharmaceutical companies for these trials include: Actelion,
Addex, Advancell, Allozyne, BaroFold, Bayer Health Care
Pharmaceuticals, Bayer Schering Pharma, Bayhill, Biogen Idec, Biotica,
CLC Behring, Elan, Genentech, GeNeuro SA, Genmab, Genmark,
Genzyme, GlaxoSmithKline, Johnson & Johnson, Lilly, Merck Serono,
Mitsubishi Pharma, Novartis, Novo Nordisk, Octapharma, Peptimmune,
Praxicon, Roche, Sanofi-Aventis, Santhera, Siemens, Teva, UCB,
Xenoport and Wyeth. He is principal investigator for the following drug
studies: BOLD, BOLD EXT., EXPAND (BAF312, Novartis), DECIDE,
DECIDE EXT. (Daclizumab, Biogen Idec), ENDORSE (BG00012, Biogen
Idec), FINGORETT, FTY-UMBRELLA, INFORMS, INFORMS EXT
LONGTERM. (Fingolimod, Novartis), OCRELIZUMAB PHASE II EXT.,
OPERA, ORATORIO (Ocrelizumab, Roche), STRATA EXT.
(Natalizumab, Biogen Idec), and TERIFLUNOMIDE EXT.
(Teriflunomide, Sanofi-Aventis). He has also lectured at medical
conferences or in public on various aspects of the diagnosis and
management of multiple sclerosis. In many cases these talks have been
sponsored by unrestricted educational grants to his institution from one
or another of the above listed companies. Honoraria and other payments
for all these activities have been exclusively used for funding of research
of his department. JCWB and PV declare no competing interests.
References
1 Polman CH, Reingold SC, Banwell B, et al. Diagnostic criteria for
multiple sclerosis: 2010 revisions to the McDonald criteria.
Ann Neurol 2011; 69: 292–302.
2 Simon JH, Li D, Traboulsee A, et al. Standardized MR imaging
protocol for multiple sclerosis: consortium of MS Centers
consensus guidelines. Am J Neuroradiol 2006; 27: 455–61.
3 Bot JC, Barkhof F. Spinal-cord MRI in multiple sclerosis:
conventional and nonconventional MR techniques.
Neuroimaging Clin N Am 2009; 19: 81–99.
452
4 McGowan JC. Technical issues for MRI examination of the spinal
cord. J Neurol Sci 2000; 172 (suppl 1): S27–31.
5 Taber KH, Herrick RC, Weathers SW, Kumar AJ, Schomer DF,
Hayman LA. Pitfalls and artifacts encountered in clinical MR
imaging of the spine. Radiographics 1998; 18: 1499–521.
6 Kidd D, Thorpe JW, Thompson AJ, et al. Spinal cord MRI using
multi-array coils and fast spin echo. II. Findings in multiple
sclerosis. Neurology 1993; 43: 2632–37.
7 Tartaglino LM, Friedman DP, Flanders AE, Lublin FD, Knobler RL,
Liem M. Multiple sclerosis in the spinal cord: MR appearance and
correlation with clinical parameters. Radiology 1995; 195: 725–32.
8 Thorpe JW, Kidd D, Moseley IF, et al. Spinal MRI in patients with
suspected multiple sclerosis and negative brain MRI. Brain 1996;
119: 709–14.
9 Weier K, Mazraeh J, Naegelin Y, et al. Biplanar MRI for the assessment
of the spinal cord in multiple sclerosis. Mult Scler 2012; 18: 1560–69.
10 Lycklama à Nijeholt GJ, Barkhof F, Castelijns JA, et al.
Comparison of two MR sequences for the detection of multiple
sclerosis lesions in the spinal cord. AJNR Am J Neuroradiol. 1996;
17: 1533–38.
11 Lycklama à Nijeholt GJ, Castelijns JA, Weerts J, et al. Sagittal MR of
multiple sclerosis in the spinal cord: fast versus conventional spin-
echo imaging. AJNR Am J Neuroradiol. 1998; 19: 355–60.
12 Bot JC, Barkhof F, Lycklama à Nijeholt G, et al. Differentiation of
multiple sclerosis from other inflammatory disorders and
cerebrovascular disease: value of spinal MR imaging. Radiology
2002; 223: 46–56.
13 Bot JC, Barkhof F, Polman CH, et al. Spinal cord abnormalities in
recently diagnosed MS patients: added value of spinal MRI
examination. Neurology 2004; 62: 226–33.
14 Gilmore CP, DeLuca GC, Bö L, et al. Spinal cord neuronal
pathology in multiple sclerosis. Brain Pathol 2009; 19: 642–49.
15 Adams RD, Kubik CS. The morbid anatomy of the demyelinative
disease. Am J Med 1952; 12: 510–46.
16 Ikuta F, Zimmerman HM. Distribution of plaques in seventy
autopsy cases of multiple sclerosis in the United States. Neurology
1976; 26: 26–28.
17 Nijeholt GJ, Bergers E, Kamphorst W, et al. Post-mortem high-
resolution MRI of the spinal cord in multiple sclerosis: a correlative
study with conventional MRI, histopathology and clinical
phenotype. Brain 2001; 124: 154–66.
18 Gilmore CP, Bö L, Owens T, Lowe J, Esiri MM, Evangelou N. Spinal
cord gray matter demyelination in multiple sclerosis-a novel pattern
of residual plaque morphology. Brain Pathol 2006; 16: 202–08.
19 Bot JC, Blezer EL, Kamphorst W, et al. The spinal cord in multiple
sclerosis: relationship of high-spatial-resolution quantitative MR
imaging findings to histopathologic results. Radiology 2004;
233: 531–40.
20 Bergers E, Bot JC, De Groot CJ, et al. Axonal damage in the spinal
cord of MS patients occurs largely independent of T2 MRI lesions.
Neurology 2002; 59: 1766–71.
21 Gilmore CP, Geurts JJ, Evangelou N, et al. Spinal cord grey matter
lesions in multiple sclerosis detected by post-mortem high field MR
imaging. Mult Scler 2009; 15: 180–88.
22 Gilmore CP, DeLuca GC, Bö L, et al. Spinal cord atrophy in
multiple sclerosis caused by white matter volume loss. Arch Neurol
2005; 62: 1859–62.
23 Evangelou N, DeLuca GC, Owens T, Esiri MM. Pathological study of
spinal cord atrophy in multiple sclerosis suggests limited role of
local lesions. Brain 2005; 128: 29–34.
24 DeLuca GC, Williams K, Evangelou N, Ebers GC, Esiri MM. The
contribution of demyelination to axonal loss in multiple sclerosis.
Brain 2006; 129: 1507–16.
25 Bjartmar C, Kidd G, Mörk S, Rudick R, Trapp BD. Neurological
disability correlates with spinal cord axonal loss and reduced
N-acetyl aspartate in chronic multiple sclerosis patients. Ann Neurol
2000; 48: 893–901.
26 Philpott C, Brotchie P. Comparison of MRI sequences for
evaluation of multiple sclerosis of the cervical spinal cord at 3 T.
Eur J Radiol 2011; 80: 780–85.
27 Hittmair K, Mallek R, Prayer D, Schindler EG, Kollegger H. Spinal
cord lesions in patients with multiple sclerosis: comparison of MR
pulse sequences. AJNR Am J Neuroradiol 1996; 17: 1555–65.
www.thelancet.com/neurology Vol 14 April 2015

28 Lycklama G, Thompson A, Filippi M, et al. Spinal-cord MRI in
multiple sclerosis. Lancet Neurol 2003; 2: 555–62.
29 Gass A, Filippi M, Rodegher ME, Schwartz A, Comi G,
Hennerici MG. Characteristics of chronic MS lesions in the
cerebrum, brainstem, spinal cord, and optic nerve on T1-weighted
MRI. Neurology 1998; 50: 548–50.
30 Silver NC, Good CD, Sormani MP, et al. A modified protocol to
improve the detection of enhancing brain and spinal cord lesions in
multiple sclerosis. J Neurol 2001; 248: 215–24.
31 Thorpe JW, Kidd D, Moseley IF, et al. Serial gadolinium-enhanced
MRI of the brain and spinal cord in early relapsing-remitting
multiple sclerosis. Neurology 1996; 46: 373–78.
32 Kidd D, Thorpe JW, Kendall BE, et al. MRI dynamics of brain and
spinal cord in progressive multiple sclerosis.
J Neurol Neurosurg Psychiatry 1996; 60: 15–19.
33 Okuda DT, Mowry EM, Cree BA, et al. Asymptomatic spinal cord
lesions predict disease progression in radiologically isolated
syndrome. Neurology 2011; 76: 686–92.
34 Lycklama à Nijeholt GJ, Barkhof F, Scheltens P, et al. MR of the
spinal cord in multiplesclerosis: relation to clinical subtype and
disability. AJNR Am J Neuroradiol. 1997; 18: 1041–48.
35 Nijeholt GJ, van Walderveen MA, Castelijns JA, et al. Brain and
spinal cord abnormalities in multiple sclerosis. Correlation between
MRI parameters, clinical subtypes and symptoms. Brain 1998;
121: 687–97.
36 Losseff NA, Webb SL, O’Riordan JI, et al. Spinal cord atrophy and
disability in multiple sclerosis. A new reproducible and sensitive
MRI method with potential to monitor disease progression. Brain
1996; 119: 701–08.
37 Brex PA, Leary SM, O’Riordan JI, et al. Measurement of spinal cord
area in clinically isolated syndromes suggestive of multiple
sclerosis. J Neurol Neurosurg Psychiatry 2001; 70: 544–47.
38 Filippi M, Campi A, Colombo B, et al. A spinal cord MRI study of
benign and secondary progressive multiple sclerosis. J Neurol 1996;
243: 502–05.
39 Rocca MA, Horsfield MA, Sala S, et al. A multicenter assessment of
cervical cord atrophy among MS clinical phenotypes. Neurology
2011; 76: 2096–102.
40 Klein JP, Arora A, Neema M, et al. A 3T MR imaging investigation
of the topography of whole spinal cord atrophy in multiple sclerosis.
AJNR Am J Neuroradiol 2011; 32: 1138–42.
41 Furby J, Hayton T, Altmann D, et al. A longitudinal study of MRI-
detected atrophy in secondary progressive multiple sclerosis.
J Neurol 2010; 257: 1508–16.
42 Rashid W, Davies GR, Chard DT, et al. Upper cervical cord area in
early relapsing-remitting multiple sclerosis: cross-sectional study of
factors influencing cord size. J Magn Reson Imaging 2006; 23: 473–76.
43 Rashid W, Davies GR, Chard DT, et al. Increasing cord atrophy in
early relapsing-remitting multiple sclerosis: a 3 year study.
J Neurol Neurosurg Psychiatry 2006; 77: 51–55.
44 Stevenson VL, Leary SM, Losseff NA, et al. Spinal cord atrophy and
disability in MS: a longitudinal study. Neurology 1998; 51: 234–38.
45 Lukas C, Sombekke MH, Bellenberg B, et al. Relevance of spinal
cord abnormalities to clinical disability in multiple sclerosis: MR
imaging findings in a large cohort of patients . Radiology 2013;
269: 542–52.
46 Bonati U, Fisniku LK, Altmann DR, et al. Cervical cord and brain
grey matter atrophy independently associate with long-term MS
disability. J Neurol Neurosurg Psychiatry 2011; 82: 471–72.
47 Cohen AB, Neema M, Arora A, et al. The relationships among MRI-
defined spinal cord involvement, brain involvement, and disability
in multiple sclerosis. J Neuroimaging 2012; 22: 122–28.
48 Lukas C, Knol DL, Sombekke MH,et al. Cervical spinal cord volume
loss is related to clinical disability progression in multiple sclerosis.
J Neurol Neurosurg Psychiatry 2014; published online Jun 27.
DOI: 10.1136/jnnp-2014-308021.
49 Valsasina P, Horsfield MA, Rocca MA, Absinta M, Comi G,
Filippi M. Spatial normalization and regional assessment of cord
atrophy: voxel-based analysis of cervical cord 3D T1-weighted
images. AJNR Am J Neuroradiol 2012; 33: 2195–200.
50 Valsasina P, Rocca MA, Horsfield MA, et al. Regional cervical cord
atrophy and disability in multiple sclerosis: a voxel-based analysis.
Radiology 2013; 266: 853–61.
www.thelancet.com/neurology Vol 14 April 2015
Personal View
51 Rocca MA, Valsasina P, Damjanovic D, et al. Voxel-wise mapping of
cervical cord damage in multiple sclerosis patients with different
clinical phenotypes. J Neurol Neurosurg Psychiatry 2013; 84: 35–41.
52 Ropele S, Fazekas F. Magnetization transfer MR imaging in
multiple sclerosis. Neuroimaging Clin N Am 2009; 19: 27–36.
53 Mottershead JP, Schmierer K, Clemence M, et al. High field MRI
correlates of myelin content and axonal density in multiple
sclerosis—a post-mortem study of the spinal cord. J Neurol 2003;
250: 1293–301.
54 Filippi M, Rocca MA. Magnetization transfer magnetic resonance
imaging of the brain, spinal cord, and optic nerve. Neurotherapeutics
2007; 4: 401–13.
55 Rovaris M, Gallo A, Riva R, et al. An MT MRI study of the cervical
cord in clinically isolated syndromes suggestive of MS. Neurology
2004; 63: 584–85.
56 Filippi M, Bozzali M, Horsfield MA, et al. A conventional and
magnetization transfer MRI study of the cervical cord in patients
with MS. Neurology 2000; 54: 207–13.
57 Rovaris M, Judica E, Ceccarelli A, et al. Absence of diffuse cervical
cord tissue damage in early, non-disabling relapsing-remitting MS:
a preliminary study. Mult Scler 2008; 14: 853–56.
58 Mezzapesa DM, Rocca MA, Falini A, et al. A preliminary diffusion
tensor and magnetization transfer magnetic resonance imaging
study of early-onset multiple sclerosis. Arch Neurol 2004; 61: 366–68.
59 Rovaris M, Bozzali M, Santuccio G, et al. In vivo assessment of the
brain and cervical cord pathology of patients with primary
progressive multiple sclerosis. Brain 2001; 124: 2540–49.
60 Bozzali M, Rocca MA, Iannucci G, Pereira C, Comi G, Filippi M.
Magnetization-transfer histogram analysis of the cervical cord in
patients with multiple sclerosis. AJNR Am J Neuroradiol 1999;
20: 1803–08.
61 Charil A, Caputo D, Cavarretta R, Sormani MP, Ferrante P,
Filippi M. Cervical cord magnetization transfer ratio and clinical
changes over 18 months in patients with relapsing-remitting
multiple sclerosis: a preliminary study. Mult Scler 2006;
12: 662–65.
62 Hickman SJ, Hadjiprocopis A, Coulon O, Miller DH, Barker GJ.
Cervical spinal cord MTR histogram analysis in multiple sclerosis
using a 3D acquisition and a B-spline active surface segmentation
technique. Magn Reson Imaging 2004; 22: 891–95.
63 Agosta F, Pagani E, Caputo D, Filippi M. Associations between
cervical cord gray matter damage and disability in patients with
multiple sclerosis. Arch Neurol 2007; 64: 1302–05.
64 Smith SA, Golay X, Fatemi A, et al. Magnetization transfer
weighted imaging in the upper cervical spinal cord using
cerebrospinal fluid as intersubject normalization reference
(MTCSF imaging). Magn Reson Med 2005; 54: 201–06.
65 Zackowski KM, Smith SA, Reich DS, et al. Sensorimotor
dysfunction in multiple sclerosis and column-specific
magnetization transfer-imaging abnormalities in the spinal cord.
Brain 2009; 132: 1200–09.
66 Kearney H, Yiannakas MC, Samson RS, Wheeler-Kingshott CA,
Ciccarelli O, Miller DH. Investigation of magnetization transfer
ratio-derived pial and subpial abnormalities in the multiple
sclerosis spinal cord. Brain 2014; 137: 2456–68.
67 Lycklama à Nijeholt GJ, Castelijns JA, Lazeron RH, et al.
Magnetization transfer ratio of the spinal cord in multiple sclerosis:
relationship to atrophy and neurologic disability. J Neuroimaging
2000; 10: 67–72.
68 Rovaris M, Agosta F, Pagani E, Filippi M. Diffusion tensor MR
imaging. Neuroimaging Clin N Am 2009; 19: 37–43.
69 Pierpaoli C, Basser PJ. Toward a quantitative assessment of
diffusion anisotropy. Magn Reson Med 1996; 36: 893–906.
70 Schmierer K, Wheeler-Kingshott CA, Boulby PA, et al. Diffusion
tensor imaging of post mortem multiple sclerosis brain.
Neuroimage 2007; 35: 467–77.
71 Zollinger LV, Kim TH, Hill K, Jeong EK, Rose JW. Using diffusion
tensor imaging and immunofluorescent assay to evaluate the
pathology of multiple sclerosis. J Magn Reson Imaging 2011;
33: 557–64.
72 Clark CA, Werring DJ, Miller DH. Diffusion imaging of the spinal
cord in vivo: estimation of the principal diffusivities and application
to multiple sclerosis. Magn Reson Med 2000; 43: 133–38.
453
 Personal View
73 Cercignani M, Horsfield MA, Agosta F, Filippi M. Sensitivity-
encoded diffusion tensor MR imaging of the cervical cord.
AJNR Am J Neuroradiol 2003; 24: 1254–56.
74 Valsasina P, Rocca MA, Agosta F, et al. Mean diffusivity and
fractional anisotropy histogram analysis of the cervical cord in MS
patients. Neuroimage 2005; 26: 822–28.
75 Benedetti B, Rocca MA, Rovaris M, et al. A diffusion tensor MRI
study of cervical cord damage in benign and secondary progressive
multiple sclerosis patients. J Neurol Neurosurg Psychiatry 2010;
81: 26–30.
76 Ohgiya Y, Oka M, Hiwatashi A, et al. Diffusion tensor MR imaging
of the cervical spinal cord in patients with multiple sclerosis.
Eur Radiol 2007; 17: 2499–504.
77 Ohgiya Y, Oka M, Hiwatashi A, et al. Diffusion tensor MR imaging
of the cervical spinal cord in patients with multiple sclerosis.
Eur Radiol 2007; 17: 2499–504.
78 Cruz LC Jr, Domingues RC, Gasparetto EL. Diffusion tensor
imaging of the cervical spinal cord of patients with relapsing-
remising multiple sclerosis: a study of 41 cases. Arq Neuropsiquiatr
2009; 67: 391–95
79 Ciccarelli O, Wheeler-Kingshott CA, McLean MA, et al. Spinal cord
spectroscopy and diffusion-based tractography to assess acute
disability in multiple sclerosis. Brain 2007; 130: 2220–31.
80 Freund P, Wheeler-Kingshott C, Jackson J, Miller D, Thompson A,
Ciccarelli O. Recovery after spinal cord relapse in multiple sclerosis
is predicted by radial diffusivity. Mult Scler 2010; 16: 1193–202.
81 Oh J, Zackowski K, Chen M, et al. Multiparametric MRI correlates
of sensorimotor function in the spinal cord in multiple sclerosis.
Mult Scler 2013; 19: 427–35.
82 Oh J, Saidha S, Chen M, et al. Spinal cord quantitative MRI
discriminates between disability levels in multiple sclerosis.
Neurology 2013; 80: 540–47.
83 Agosta F, Absinta M, Sormani MP, et al. In vivo assessment of
cervical cord damage in MS patients: a longitudinal diffusion tensor
MRI study. Brain 2007; 130: 2211–19.
84 Agosta F, Valsasina P, Caputo D, Stroman PW, Filippi M. Tactile-
associated recruitment of the cervical cord is altered in patients with
multiple sclerosis. Neuroimage 2008; 39: 1542–48.
85 Agosta F, Valsasina P, Rocca MA, et al. Evidence for enhanced
functional activity of cervical cord in relapsing multiple sclerosis.
Magn Reson Med 2008; 59: 1035–42.
86 Valsasina P, Agosta F, Absinta M, Sala S, Caputo D, Filippi M.
Cervical cord functional MRI changes in relapse-onset MS patients.
J Neurol Neurosurg Psychiatry 2010; 81: 405–08.
87 Agosta F, Valsasina P, Absinta M, Sala S, Caputo D, Filippi M.
Primary progressive multiple sclerosis: tactile-associated functional
MR activity in the cervical spinal cord. Radiology 2009; 253: 209–15.
88 Valsasina P, Rocca MA, Absinta M, et al. Cervical cord fMRI
abnormalities differ between the progressive forms of multiple
sclerosis. Hum Brain Mapp 2012; 33: 2072–80.
89 Rocca MA, Absinta M, Valsasina P, et al. Abnormal cervical cord
function contributes to fatigue in multiple sclerosis. Mult Scler 2012;
18: 1552–59.
90 Hock A, Henning A, Boesiger P, Kollias SS. 1H-MR spectroscopy in
the human spinal cord. AJNR Am J Neuroradiol 2013; 34: 1682–89.
91 Marliani AF, Clementi V, Albini-Riccioli L, Agati R, Leonardi M.
Quantitative proton magnetic resonance spectroscopy of the human
cervical spinal cord at 3 Tesla. Magn Reson Med 2007; 57: 160–63.
454
92 Abdel-Aziz K, Solanky BS, Yiannakas MC, et al. Age-related changes
in metabolite concentrations in the normal spinal cord. PLoS ONE
2014; published online Oct 13. DOI:10.1371/journal.pone.0105774.
93 Kendi AT, Tan FU, Kendi M, Yilmaz S, Huvaj S, Tellioğlu S. MR
spectroscopy of cervical spinal cord in patients with multiple
sclerosis. Neuroradiology 2004; 46: 764–69.
94 Blamire AM, Cader S, Lee M, Palace J, Matthews PM. Axonal
damage in the spinal cord of multiple sclerosis patients detected by
magnetic resonance spectroscopy. Magn Reson Med 2007;
58: 880–85.
95 Ciccarelli O, Altmann DR, McLean MA, et al. Spinal cord repair in
MS: does mitochondrial metabolism play a role? Neurology 2010;
74: 721–27.
96 Marliani AF, Clementi V, Albini Riccioli L, et al. Quantitative
cervical spinal cord 3T proton MR spectroscopy in multiple
sclerosis. AJNR Am J Neuroradiol 2010; 31: 180–84.
97 Henning A, Schär M, Kollias SS, Boesiger P, Dydak U. Quantitative
magnetic resonance spectroscopy in the entire human cervical
spinal cord and beyond at 3T. Magn Reson Med 2008; 59: 1250–58.
98 Moffett JR, Ross B, Arun P, Madhavarao CN, Namboodiri AM.
N-Acetylaspartate in the CNS: from neurodiagnostics to
neurobiology. Prog Neurobiol 2007; 81: 89–131.
99 Ciccarelli O, Thomas DL, De Vita E, et al. Low myo-inositol
indicating astrocytic damage in a case series of neuromyelitis
optica. Ann Neurol 2013; 74: 301–05.
100 De Stefano N, Matthews PM, Arnold DL. Reversible decreases in
N-acetylaspartate after acute brain injury. Magn Reson Med 1995;
34: 721–27.
101 Mader I, Roser W, Kappos L, et al. Serial proton MR spectroscopy of
contrast-enhancing multiple sclerosis plaques: absolute metabolic
values over 2 years during a clinical pharmacological study.
AJNR Am J Neuroradiol 2000; 21: 1220–7.
102 Bates TE, Strangward M, Keelan J, Davey GP, Munro PM, Clark JB.
Inhibition of N-acetylaspartate production: implications for
1H-MRS studies in vivo. Neuroreport 1996; 7: 1397–400.
103 Lucchinetti CF, Guo Y, Popescu BF, Fujihara K, Itoyama Y, Misu T.
The pathology of an autoimmune astrocytopathy: lessons learned
from neuromyelitis optica. Brain Pathol 2014; 24: 83–97.
104 Zhao W, Cohen-Adad J, Polimeni JR, et al. Nineteen-channel receive
array and four-channel transmit array coil for cervical spinal cord
imaging at 7 T. Magn Reson Med 2014; 72: 291–300.
105 Edden RA, Bonekamp D, Smith MA, Dubey P, Barker PB. Proton
MR spectroscopic imaging of the medulla and cervical spinal cord.
J Magn Reson Imaging 2007; 26: 1101–05.
106 Solanky BS, Abdel-Aziz K, Yiannakas MC, Berry AM, Ciccarelli O,
Wheeler-Kingshott CA. In vivo magnetic resonance spectroscopy
detection of combined glutamate-glutamine in healthy upper
cervical cord at 3 T. NMR Biomed 2013; 26: 357–66.
107 Hock A, Wilm B, Zandomeneghi G, et al. Detection of GABA,
aspartate and glutathione in the human spinal cord.
Joint annual meeting ISMRM, Milan, Italy; May 10–16 Poster
number 1712.
108 Ciccarelli O, Toosy AT, De Stefano N, Wheeler-Kingshott CA,
Miller DH, Thompson AJ. Assessing neuronal metabolism in vivo
by modeling imaging measures. J Neurosci 2010; 30: 15030–33.
www.thelancet.com/neurology Vol 14 April 2015
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.