See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/291947643

MRI criteria for the diagnosis of multiple sclerosis: MAGNIMS consensus

guidelines

Article  in  The Lancet Neurology · January 2016

DOI: 10.1016/S1474-4422(15)00393-2

CITATIONS READS

222 3,257

16 authors, including:

Mara A Rocca Nicola De Stefano

San Raffaele Scientific Institute Università degli Studi di Siena
571 PUBLICATIONS   17,377 CITATIONS    441 PUBLICATIONS   25,373 CITATIONS   

SEE PROFILE SEE PROFILE

Ludwig Kappos Alex Rovira

Universitätsspital Basel University Hospital Vall d'Hebron
1,015 PUBLICATIONS   47,573 CITATIONS    376 PUBLICATIONS   13,541 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Optic neuritis and MS (various project). Klaus, det kunne være hyggeligt at hilse på dig igen. Skriv på [email protected] View project

Semaphorins 3A and 7A in multiple sclerosis and EAE mouse model View project

All content following this page was uploaded by Claudio Gasperini on 31 January 2016.

The user has requested enhancement of the downloaded file.

 Review

MRI criteria for the diagnosis of multiple sclerosis: MAGNIMS

consensus guidelines
Massimo Filippi, Maria A Rocca, Olga Ciccarelli, Nicola De Stefano, Nikos Evangelou, Ludwig Kappos, Alex Rovira, Jaume Sastre-Garriga,
Mar Tintorè, Jette L Frederiksen, Claudio Gasperini, Jacqueline Palace, Daniel S Reich, Brenda Banwell, Xavier Montalban, Frederik Barkhof,
on behalf of the MAGNIMS Study Group*

In patients presenting with a clinically isolated syndrome, MRI can support and substitute clinical information in the Lancet Neurol 2016
diagnosis of multiple sclerosis by showing disease dissemination in space and time and by helping to exclude disorders Published Online
that can mimic multiple sclerosis. MRI criteria were first included in the diagnostic work-up for multiple sclerosis in January 25, 2016
http://dx.doi.org/10.1016/
2001, and since then several modifications to the criteria have been proposed in an attempt to simplify lesion-count S1474-4422(15)00393-2
models for showing disease dissemination in space, change the timing of MRI scanning to show dissemination in time,
See Online/Comment
and increase the value of spinal cord imaging. Since the last update of these criteria, new data on the use of MRI to http://dx.doi.org/10.1016/
establish dissemination in space and time have become available, and MRI technology has improved. State-of-the-art S1474-4422(16)00023-5
MRI findings in these patients were discussed in a MAGNIMS workshop, the goal of which was to provide an evidence- *MAGNIMS Steering Committee
based and expert-opinion consensus on proposed modifications to MRI criteria for the diagnosis of multiple sclerosis. members are listed at the end of
the paper

Introduction have also been published and are complementary to the Neuroimaging Research Unit,
Institute of Experimental
MRI was formally included in the diagnostic work-up of recommendations in this Review.11 Neurology, Division of
patients presenting with a clinically isolated syndrome Since 2011, new data on application of MRI to show Neuroscience, San Raffaele
suggestive of multiple sclerosis in 2001 by an dissemination in space and time have become available, Scientific Institute, Vita-Salute
international panel of experts.1 Diagnosis of multiple and these data deserve consideration for future revisions San Raffaele University, Milan,
Italy (Prof M Filippi MD,
sclerosis relies on proof of disease dissemination in of the multiple sclerosis diagnostic criteria. Additionally, M A Rocca MD); Department of
space and time and exclusion of other disorders that can many improvements in MRI technology have occurred, Neuroinflammation, Queen
mimic multiple sclerosis by their clinical and laboratory which have resulted in development of innovative Square MS Centre, UCL
Institute of Neurology,
profile. MRI can support and substitute clinical acquisition sequences, identification of novel patho-
London, UK
information for multiple sclerosis diagnosis, enabling an physiological mechanisms that might help with (Prof O Ciccarelli PhD); National
early and accurate diagnosis and, as such, early treatment. differential diagnosis, and new insights into multiple Institute for Health Research
MRI criteria for multiple sclerosis are based on the sclerosis disease activity from studies using high-field (NIHR), UCL/UCLH Biomedical
Research Centre, London, UK
presence of focal lesions in the white matter of the and ultra-high-field scanners. The MAGNIMS members
(Prof O Ciccarelli); Department
CNS, which are considered typical for this disorder in felt the need for timely review of these findings and of Medicine, Surgery and
terms of distribution, morphology, evolution, and consideration of how they should be used to modify the Neuroscience, University of
signal abnormalities on conventional MRI sequences MRI criteria for diagnosis of multiple sclerosis. A Siena, Siena, Italy
(N De Stefano MD); Division of
(eg, T2-weighted and T2-weighted fluid-attenuated summary of the main proposed revisions or clarifications
Neurosciences, Queen’s Medical
inversion recovery [FLAIR] scans, and pre-contrast to the MRI component of the 2010 McDonald criteria10 Centre Campus, University of
and post-contrast T1-weighted scans).2–4 Several modifi- for multiple sclerosis is given in panel 1. Nottingham, Nottingham, UK
cations of MRI diagnostic criteria have been proposed, (N Evangelou PhD); Department
of Neurology, University of
but emphasis has consistently been that such criteria Methods Basel, Basel, Switzerland
should be applied only in patients who present with a In March, 2015, an international workshop was held in (Prof L Kappos MD); Magnetic
typical clinically isolated syndrome suggestive of Milan, Italy, under the auspices of MAGNIMS. The Resonance Unit, Department
multiple sclerosis or symptoms consistent with a CNS workshop involved clinical and imaging experts in of Radiology (IDI), Hospital
Universitari Vall d’Hebron,
inflammatory demyelinating disease. These revisions diagnosis and management of patients with multiple Universitat Autònoma de
have simplified lesion-count models for proof of sclerosis, including neurologists and neuroradiologists. Barcelona, Barcelona, Spain
dissemination in space, changed the timing of MRI Before the meeting, two co-chairs (M Filippi and (A Rovira MD); Centre
scanning to show dissemination in time, and increased F Barkhof) identified areas in which revision, clarification, d’Esclerosi Múltiple de
Catalunya (Cemcat),
the value of spinal cord imaging.5–8 In 2007, the or both might be necessary in future diagnostic criteria Department of Neurology/
European collaborative research network that studies for multiple sclerosis. Experts for each topic were invited Neuroimmunology, Hospital
MRI in multiple sclerosis (MAGNIMS) reviewed the to provide a summary during the meeting of the main Universitari Vall d’Hebron,
findings of studies that addressed these issues and findings related to their argument, based on a review of Universitat Autònoma de
Barcelona, Barcelona, Spain
proposed new MRI criteria to be applied in multiple the literature and on their personal experience. They were (J Sastre-Garriga MD,
sclerosis.9 Those MAGNIMS criteria are included in the then asked to define whether such a measure would be M Tintorè MD,
most recent diagnostic criteria for multiple sclerosis, useful in the diagnostic process, and whether it would Prof X Montalban MD);
known as the 2010 McDonald criteria.10 Consensus move the field forwards in a promising way, in order to Department of Neurology,
Glostrup Hospital and
guidelines for clinicians to optimise planning, stimulate group discussion. For each measure, a group University of Copenhagen,
performance, and interpretation of brain and spinal agreement was reached (100% agreement was achieved in Copenhagen, Denmark
cord MRI in the multiple sclerosis diagnostic process all cases) during the workshop and summarised in a first (Prof J L Frederiksen MD);

www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2 1

 Review

characteristic for multiple sclerosis (juxtacortical,

Panel 1: Recommended 2016 MAGNIMS modifications to the 2010 McDonald periventricular, infratentorial, and spinal cord). We
criteria10 for MRI in the diagnosis of multiple sclerosis propose an increase in the number of lesions necessary
Proposed revisions to confirm involvement of the periventricular area from
• Three or more lesions are needed to define the involvement of the periventricular one to three, and to add an additional cardinal CNS
region to establish disease dissemination in space (expert consensus) location, the optic nerve (panel 2). Together with a spinal
• The presence of a lesion in the optic nerve should be added to the criteria for cord lesion, these changes increase the number of
dissemination in space as an additional CNS area (expert consensus) dissemination in space locations from four to five.
• The combined term cortical/juxtacortical is recommended to expand the concept of
juxtacortical lesion in the criteria for dissemination in space, by including all multiple Periventricular lesions
sclerosis cortical lesion types, involvement of the white matter next to the cortex, or A single lesion was deemed not sufficiently specific to
both; when available, advanced imaging sequences should be applied to visualise determine whether involvement of the periventricular
cortical lesions (expert consensus) region is due to a demyelinating inflammatory event,
• No distinction needs to be made between symptomatic and asymptomatic MRI and the use of one periventricular lesion for assessing
lesions for dissemination in time or space (evidence based)12–15 dissemination in space has never been formally
• Imaging of the whole spinal cord is recommended to define dissemination in space validated. Incidental periventricular lesions can be
(especially in patients who do not fulfil brain MRI criteria for dissemination in space); detected in healthy individuals and patients with other
spinal cord imaging has a limited role for identification of dissemination in time neurological disorders, including up to 30% of patients
(evidence based)16–19 with migraine.32 Importantly, three or more peri-
• Identical criteria for dissemination in space should be used for primary progressive ventricular lesions was the most accurate threshold
multiple sclerosis and relapse-onset multiple sclerosis (expert consensus); CSF results identified in a study using receiver-operating curve
should be considered for clinically uncertain cases of primary progressive multiple analysis by Barkhof and colleagues (known as the
sclerosis (evidence based)20 Barkhof criteria),4 and was therefore applied in the 2001
• In children aged older than 11 years with presentation that does not resemble acute and 2005 McDonald criteria.1,8 Analysis of a large cohort
disseminated encephalomyelitis (ADEM), MRI criteria used to establish dissemination of 652 patients with a clinically isolated syndrome has
in time and space in adults should be applied (evidence based)21–26 shown that, in patients who do not meet criteria for
• Caution is recommended when applying the 2010 criteria10 solely at baseline in dissemination in space for multiple sclerosis, presence
patients younger than age 11 years, even in those with a non-ADEM presentation27 of three periventricular lesions, combined with age or
• MRI criteria can be applied equally well to patients with multiple sclerosis in Asia or presence of oligoclonal bands, is helpful to identify
Latin America as to patients from Europe and North America, once alternative those at risk of multiple sclerosis.33 In a retrospective
neurological disorders (eg, neuromyelitis optica spectrum disorder) have been study34 in patients with a clinically isolated syndrome
carefully excluded (evidence based)28–31 affecting the spinal cord, a prediction model, including
• MRI criteria used to establish dissemination in time and space in multiple sclerosis those aged 40 years or younger, with three or more
should be applied for assessment of radiologically isolated syndromes; when a clinical periventricular lesions, and with intrathecal immuno-
attack occurs in patients with radiologically isolated syndromes with evidence of globulin synthesis, identified patients who would
dissemination in time (who, by definition, have dissemination in space), a diagnosis of develop multiple sclerosis with an accuracy of 78%.34 In
multiple sclerosis can be made (expert consensus) a multicentre trial35 of 468 patients with a clinically
isolated syndrome, presence of at least three
Additional clarifications and summary statements periventricular lesions had a strong prognostic value for
• MRI criteria for disease dissemination in time can remain unchanged conversion to multiple sclerosis in 3-year period. In a
• Presence of non-enhancing black holes is not useful as a potential alternative criterion study comparing patients with multiple sclerosis with
for dissemination in time in adults; the contribution of non-enhancing black holes those with primary and secondary CNS vasculitis,36
seems to be more robust in distinguishing children with multiple sclerosis from presence of three or more periventricular lesions was
children with monophasic demyelination (especially ADEM) the only individual component of the Barkhof criteria4
• In the case of atypical imaging presentation, other acquired and inherited white that could be used to distinguish patients with multiple
matter diseases should always be considered in the differential diagnosis sclerosis from those with systemic lupus erythematosus
• At present, insufficient evidence exists to support earlier diagnosis of multiple or Sjögren’s syndrome.
sclerosis when using high-field or ultra-high-field scanners However, in paediatric patients, presence of a single
periventricular lesion (and one or more T1-hypointense
Department of Neurosciences, draft of these guidelines, which was circulated among the lesions) powerfully distinguished children with multiple
San Camillo-Forlanini Hospital, meeting participants and some additional experts for sclerosis from children with monophasic demyelination.37
Rome, Italy (C Gasperini MD);
Department of Clinical
critical discussion and revision.
Neurology, University of Optic nerve lesions
Oxford Hospitals Trust, Oxford, Dissemination in space 20–31% of patients with a clinically isolated syndrome
UK (J Palace MD); Translational According to the 2010 McDonald criteria for multiple present with acute optic neuritis.38–40 Compared with
Neuroradiology Unit, National
Institute of Neurological
sclerosis,10 dissemination in space can be established other clinical presentations, adult patients with optic
with at least one T2 lesion in at least two of four locations neuritis are more likely than those with acute

2 www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2

 Review

demyelination in other CNS locations to have a least one intracortical lesion on baseline scans was Disorders and Stroke, National
monophasic illness,38,41,42 as confirmed by results of a considered.55 Cortical lesion assessment might also help Institutes of Health, Bethesda,
MD, USA (D S Reich MD);
study of 1058 patients with a clinically isolated with differential diagnosis between multiple sclerosis Division of Child Neurology,
syndrome.39 Importantly, in this cohort and in other and disorders that mimic multiple sclerosis, since The Children’s Hospital of
studies, the likelihood of optic neuritis being a cortical lesions have not been reported in patients with Philadelphia, Department of
monophasic illness was substantially reduced in the migraine with white matter T2 lesions32 or neuromyelitis Neurology, Perelman School of
Medicine, University of
presence of CSF oligoclonal bands or clinically silent optica.56 Intracortical lesions are also rare in healthy Pennsylvania, PA, USA
brain MRI lesions (with a hazard ratio [HR] of 5·1 for controls (identified in one of 30 individuals who were (Prof B Banwell MD); and MS
patients with one to three lesions and 11·3 for patients scanned with phase-sensitive inversion recovery Centre Amsterdam, VU
with ten or more lesions). Presence of even one sequences).49 University Medical Centre,
Amsterdam, Netherlands
clinically silent T2-hyperintense brain lesion in children Even with these promising results, many unsolved (Prof F Barkhof MD)
with optic neuritis is highly associated with issues remain regarding inclusion of cortical lesion Correspondence to:
confirmation of a multiple sclerosis diagnosis,43 assessment in the diagnostic work-up of patients with a Prof Massimo Filippi,
whereas the absence of brain lesions strongly predicts a clinically isolated syndrome. First, MRI sequences used Neuroimaging Research Unit,
monophasic illness.27 in research settings for identification of these lesions Institute of Experimental
Neurology, Division of
Clinical features of optic neuritis (visual impairment, might not be available and easily implementable on most Neuroscience, San Raffaele
scotoma, red–green desaturation, and pain with ocular clinical scanners. Second, the acquisition parameters for Scientific Institute, Vita-Salute
movement), MRI evidence of optic nerve inflammation these sequences still need to be standardised across San Raffaele University, Milan,
(increased T2 signal, gadolinium enhancement, and scanning systems from different manufacturers and for Italy
[email protected]
optic nerve swelling), abnormalities on optical coherence various field strengths. Third, agreement among
tomography (evidence of retinal nerve fibre layer observers in assessment of these sequences is at best
thinning), and neurophysiological abnormalities moderate (complete agreement 19% for double inversion
(especially delayed visual evoked potentials) all support recovery), and guidelines for their assessment are
inclusion of the optic nerve as an additional CNS area changing.49,57 Fourth, different criteria and terms are
that might be affected at the onset of a clinically isolated applied by different research groups for the distinction
syndrome. Clinical documentation of optic nerve between intracortical, leukocortical, mixed white matter
atrophy or pallor, neurophysiological confirmation of and grey matter, and juxtacortical lesions.47–50,55 Finally,
optic nerve dysfunction (slowed conduction), or imaging subpial demyelination, which can be quite extensive, is
features of clinically silent optic nerve inflammation usually not scored.46
(MRI lesions or retinal nerve fibre layer thinning) Since intracortical, leukocortical, and juxtacortical
support dissemination in space and, in patients without lesions cannot be distinguished reliably and consistently
concurrent visual symptoms, dissemination in time. on conventional MRI scans using most available MRI
scanners in clinical settings, expert consensus was that
Cortical lesions these lesions should be combined in a single term
Results of pathology studies have shown extensive (cortical/juxtacortical lesions) that indicates involvement
involvement of the grey matter in multiple sclerosis.44–46 of the white matter next to the cortex, the cortex, or both,
According to their location within the grey matter, thereby expanding the term juxtacortical lesion used in
different cortical lesion locations (subpial, purely the 2010 McDonald criteria10 for dissemination in space.
intracortical, and leukocortical lesions on the grey When available, advanced imaging sequences should be
matter–white matter border) have been identified.45 applied to visualise cortical lesions.
Imaging cortical lesions is challenging, especially with
conventional clinical MRI protocols. Different MRI
techniques have been proposed and are being compared Panel 2: Recommended 2016 MAGNIMS MRI criteria to
for their sensitivity for cortical lesion detection, establish disease dissemination in space in multiple sclerosis
including double inversion recovery,47 phase-sensitive
Dissemination in space can be shown by involvement* of at
inversion recovery,48–50 and magnetisation-prepared rapid
least two of five areas of the CNS as follows:
acquisition with gradient echo51 sequences (figure 1).
Despite use of these techniques, results of correlative • Three or more periventricular lesions
MRI pathology studies have shown that many cortical • One or more infratentorial lesion
lesions remain invisible on MRI, at least with 1·5 T and • One or more spinal cord lesion
3·0 T MRI scanners.52,53 • One or more optic nerve lesion
With double inversion recovery sequences, cortical • One or more cortical or juxtacortical lesion†
lesions have been identified in more than 30% of patients *If a patient has a brainstem or spinal cord syndrome, or optic neuritis, the
with a clinically isolated syndrome.54,55 In a cohort of symptomatic lesion (or lesions) are not excluded from the criteria and contribute to the
80 patients with a clinically isolated syndrome, with lesion count. †This combined terminology indicates the involvement of the white
matter next to the cortex, the involvement of the cortex, or both, thereby expanding
4-year follow-up, the accuracy of MRI diagnostic criteria the term juxtacortical lesion.
for multiple sclerosis increased when the presence of at

www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2 3

 Review

progressive disease subtypes. For that reason, their

presence in patients with a clinically isolated
syndrome is indicative of an already-established multiple
sclerosis disease process. The prevalence of non-
enhancing T1-hypointense lesions and their added value
for identification of adult patients with multiple sclerosis
was analysed in a large multicentre study61 of 520 patients
with a clinically isolated syndrome. Non-enhancing black
holes were fairly common in adult patients with a
clinically isolated syndrome (36%) and were associated
with an increased likelihood of multiple sclerosis
diagnosis. However, the value of this magnetic resonance
finding for prediction of a second clinical attack in these
patients was lost when added to the other criteria.61 Of
note, T1-hypointese lesion assessment is still subjective
and highly dependent on the type of T1-weighted
sequence and field strength. Nevertheless, in paediatric
patients with acute demyelination, presence of one or
more T1-hypointense lesions was highly associated with
subsequent confirmation of multiple sclerosis.37
The criteria for dissemination in time should therefore
remain unchanged, and the presence of non-enhancing
black holes should not be considered as a potential
alternative criterion to show dissemination in time in
adult patients with multiple sclerosis, but might be
useful to identify multiple sclerosis in paediatric patients.

Symptomatic lesions
In patients with a clinically isolated syndrome,
symptomatic lesions that align with an acute clinical
deficit do not contribute to the dissemination in time or
space components of existing multiple sclerosis
diagnostic criteria.10 Specifically, in patients with
Figure 1: Cortical and juxtacortical lesion detection with MRI
Examples of lesion classification based on integrated analysis of double inversion recovery (left column) and
brainstem or spinal cord syndromes, lesions within the
magnetisation-prepared rapid acquisition with gradient echo (MPRAGE; middle and right columns) MRI symptomatic region cannot be counted for dissemination
sequences. Top row: a hyperintense lesion close to the cortex (green arrow) is visible on double inversion recovery in space. The simultaneous presence of asymptomatic
MRI, but the MPRAGE images show that the lesion is located in the white matter. Middle row: a hyperintense gadolinium-enhancing and non-enhancing lesions at
lesion close to the cortex (green arrow) is visible on double inversion recovery MRI, and the MPRAGE images show
that the location borders the cortex (juxtacortical). Bottom row: a hyperintense lesion close to the cortex (green
any time is a criterion to define dissemination in time.
arrow) is visible on double inversion recovery MRI, and the MPRAGE images show that the lesion is intracortical. In patients with a clinically isolated syndrome
Under the proposed system, the lesions in the middle and bottom rows would be classified as cortical/juxtacortical. presenting with brainstem symptoms, results of a 2004
study62 showed that the specificity of MRI criteria for
Dissemination in time dissemination in space (Barkhof’s criteria4 at that time)
According to the 2010 McDonald criteria,10 disease was lower than that reported in other clinically isolated
dissemination in time can be established by the syndromes (myelitis and optic neuritis; 61% vs 73%).
following: presence of at least one new T2 or gadolinium- A 2014 study12 assessed the likelihood of multiple
enhancing lesion on follow-up MRI, with reference to a sclerosis confirmation in 35 (3%) of 954 patients with a
baseline scan, irrespective of the timing of the baseline single symptomatic lesion in the brainstem or spinal
MRI; or the simultaneous presence of asymptomatic cord with a follow-up of almost 8 years. The HR for a
gadolinium-enhancing and non-enhancing lesions at diagnosis of multiple sclerosis was higher for patients
any time. with a symptomatic lesion (HR 7·2) than for those with a
Non-enhancing T1-hypointense lesions (black holes) single asymptomatic lesion (5·7), or with no lesions
are chronic lesions characterised by severe axonal (1·0), in the same regions. Another retrospective study13
damage.58 In relapsing-remitting multiple sclerosis, in 146 patients with a clinically isolated syndrome who
brain T1-hypointense lesion volume increases by about fulfilled the 2010 McDonald diagnostic criteria10 reported
11% per year and is associated with long-term disability that the presence of a symptomatic lesion identified
progression.59,60 T1-hypointense lesion formation is more patients with multiple sclerosis with a high sensitivity.13
common in patients with long disease durations and In a study of 30 patients with a clinically isolated

4 www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2

 Review

syndrome who were followed up for a mean of 7·3 years sequences) is preferable to increase confidence in lesion
(SD 1·98) after onset, the sensitivity of the dissemination identification, in part because about 40% of spinal cord
in space criteria was 73% for the 2010 McDonald criteria, lesions are identified in the thoracolumbar region
80% when asymptomatic lesions in the symptomatic (figure 2).17–19
region were included, and 87% when any lesion in the The value of spinal cord imaging to establish
symptomatic region was included; specificity was 73% dissemination in time in patients without accrual of
for the 2010 McDonald criteria, 73% when asymptomatic deficits associated with the spinal cord is low, since new
lesions in the symptomatic region were included, and clinically silent cord lesions are not frequent; spinal cord
73% when any lesion in the symptomatic region was imaging in these patients is therefore not recommended
included; and accuracy was 73% for the 2010 McDonald by the expert panel.
criteria, 77% when asymptomatic lesions in the
symptomatic region were included, and 80% when any Primary progressive multiple sclerosis
lesion in the symptomatic region was included.14 These In all formulations of the diagnostic criteria, diagnosis of
results suggest that inclusion of presence of lesions in primary progressive multiple sclerosis has been
the symptomatic region in criteria for dissemination in separated from that of the more common relapse-onset
space might increase the sensitivity of MRI criteria for form of the disease. Since no evidence exists that imaging
diagnosis of multiple sclerosis, without compromising features differ substantially between patients with
specificity. relapse-onset multiple sclerosis and patients with
The diagnostic effect of counting any gadolinium- primary progressive multiple sclerosis, the expert group
enhancing and non-enhancing lesions (not only agreed that use of similar criteria would simplify the
asymptomatic, but also symptomatic) in criteria for diagnostic work-up of patients with primary progressive
dissemination in time has also been analysed.15 Inclusion multiple sclerosis. In 2009, unification of MRI criteria
of symptomatic lesions in the dissemination in time for dissemination in space was proposed for primary
criteria increased the proportion of patients meeting the progressive multiple sclerosis and relapsing-remitting
MRI diagnostic criteria for multiple sclerosis to 33%, multiple sclerosis,63 which was only partly integrated in
compared with 30% for those diagnosed without the 2010 McDonald criteria.10 According to these criteria,
inclusion of such lesions, with three additional patients dissemination in space in primary progressive multiple
meeting the 2010 McDonald criteria.10 In fact, deciding sclerosis is defined by occurrence of two of the following
what is symptomatic or not is often very difficult. This three criteria: dissemination in space in the brain, based
distinction is fairly straightforward in brainstem and on presence of at least one lesion in at least one area
spinal cord presentations but not in other clinical characteristic for multiple sclerosis (periventricular,
scenarios, and on the basis of the evidence, the expert juxtacortical, or infratentorial regions); dissemination in
panel recommends that no distinction needs to be made space in the spinal cord, based on presence of at least two
between symptomatic and asymptomatic MRI lesions to lesions in the spinal cord; and positive CSF examination.
establish dissemination in both space and time. The sensitivity of the spinal cord criteria and the
usefulness of CSF examination were retrospectively
Spinal cord imaging analysed in a cohort of 95 patients with primary
As set out in the 2010 McDonald criteria,10 clinically silent progressive multiple sclerosis.20 In that study, if the
spinal cord lesions can contribute to assessment of criterion for two or more cord lesions was changed to
dissemination in space and time, and on the basis of the
available evidence, the expert panel recommends use of
spinal cord MRI to establish dissemination in space. At
symptom onset, spinal cord imaging is recommended in
patients with clinical features suggestive of spinal cord
involvement to exclude alternative cord pathology
(eg, compression, spinal cord tumour, neuromyelitis
optica, or vasculitides) and in those with non-spinal
clinically isolated syndromes that do not fulfil brain MRI
criteria for dissemination in space. In patients with a
non-spinal clinically isolated syndrome that does not
meet brain MRI criteria for dissemination in space,
whole cord imaging showed that the presence of one
spinal cord lesion can be used to identify patients at high
risk of multiple sclerosis confirmation.16 Imaging of the
Figure 2: Spinal cord MRI in a patient with multiple sclerosis
whole cord, with at least two magnetic resonance Sagittal intermediate and T2-weighted dual echo fast-spin echo MRIs of the spinal
sequences (eg, T2 and short T1 inversion recovery, T2 cord from a female patient aged 45 years with multiple sclerosis. Abnormalities
and double inversion recovery, T2 and post-contrast T1 (arrows) are present both at the cervical and thoracic level of the cord.

www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2 5

 Review

one or more cord lesions (whether symptomatic or not), Taiwanese,29 Argentinean (including a sub-analysis
an increased number of patients would meet the spinal applied only to non-European descendants—ie, mestizos,
cord criteria for diagnosis, with increasing sensitivity and natives, and zambos),30 and Russian31 patients with
simplification of the criteria. However, specificity of clinically isolated syndromes, after careful exclusion of
these simplified criteria still needs to be tested. alternative neurological disorders, such as neuromyelitis
In view of the evidence, the expert consensus was that optica and neuromyelitis optica spectrum disorder in
identical dissemination in space criteria should be used Korean patients.28 All these studies provided evidence
for primary progressive multiple sclerosis and relapse- that the 2010 McDonald criteria apply well irrespective of
onset multiple sclerosis, with use of CSF testing for world region, and, therefore, the consensus view was that
confirmation in uncertain cases. MRI criteria for dissemination in time and space apply
equally well to patients from Asia and Latin America as
Paediatric populations to patients from Europe and North America.
The 2010 consensus was that the proposed MRI criteria10
could be used for most paediatric patients with multiple Radiologically isolated syndromes
sclerosis. An alert specified that use of the 2010 McDonald Availability of MRI assessment for indications unrelated
criteria for multiple sclerosis at baseline was not to multiple sclerosis has led to increased recognition of
applicable for children with encephalopathy and individuals with incidental brain lesions consistent with
multifocal neurological deficits meeting criteria for acute multiple sclerosis. Criteria have been proposed to
disseminated encephalomyelitis.10 Such children have identify imaging features that are suggestive of a
many lesions, some of which might enhance; however, clinically asymptomatic demyelinating disorder,
when defined with international consensus criteria for including fulfilment of at least three of four Barkhof
acute disseminated encephalomyelitis, 95% have a criteria4 for dissemination in space.64,65 The 2010
monophasic illness.27 Diagnosis of multiple sclerosis in McDonald criteria10 concluded that “a firm diagnosis of
paediatric patients manifesting initially with a first attack [multiple sclerosis] based on incidental findings on MRI
that resembles acute disseminated encephalomyelitis alone, even with additional supportive findings on
relies on clinical or MRI evidence of further non-acute evoked potentials or typical CSF findings in the absence
disseminated encephalomyelitis attacks or accrual of of [multiple sclerosis]-relevant clinical symptoms, is
clinically silent MRI lesions. problematic.” We propose that the MRI criteria used to
Although results of a study of 52 patients25 suggested establish dissemination in time and space in multiple
that inclusion of spinal cord imaging at first attack does sclerosis should be applied for assessment of
not increase accuracy of the 2010 McDonald criteria,10 a radiologically isolated syndromes, and that when a
retrospective investigation22 of 85 patients showed that clinical attack occurs in patients with radiologically
addition of spinal cord MRI was helpful in identification isolated syndromes with evidence of dissemination in
of dissemination in time and space in 10% of cases. time (who by definition have dissemination in space), a
Several studies21–26 have confirmed that the 2010 diagnosis of multiple sclerosis can be made. Thus, we
McDonald criteria perform better than or similar to agreed that people should not be diagnosed with
previously proposed paediatric multiple sclerosis criteria multiple sclerosis on the basis of MRI findings alone,
for diagnosis of children with non-acute disseminated and at least one clinical event consistent with acute
encephalomyelitis presentations and paediatric patients demyelination remains a cornerstone for multiple
older than 11 years, and the consensus group therefore sclerosis diagnosis.
recommend caution when using these criteria in children Use of advanced MRI techniques to characterise CNS
younger than 11 years. Clinical and MRI serial assessment involvement in patients with radiologically isolated
to confirm new lesions over time might be especially syndromes has shown extensive axonal damage
important in this age group.27 (measured with magnetic resonance spectroscopy)66 and
In line with previous criteria10 and the above evidence, a perhaps surprisingly high percentage (40%) of patients
we agreed that MRI criteria for dissemination in time with cortical lesions (which were more frequent in
and space identical to those applied in adults should be patients with CSF oligoclonal bands, cervical cord
used in children aged 11 years or older who have non- lesions, and dissemination in time on brain MRI).67
acute-disseminated-encephalomyelitis-like presentation. About two-thirds of patients with radiologically isolated
syndromes develop new lesions on longitudinal MRI
Non-white populations scans and a third develop neurological symptoms within
The 2010 McDonald criteria have been developed and 5 years, especially those with gadolinium-enhancing or
tested mostly in adult white European and North spinal cord lesions.68 In people with clinically silent brain
American populations, and their formulation states that lesions consistent with multiple sclerosis, presence of
validation is needed in Asian and Latin American oligoclonal bands, younger age (≤37 years), male sex, and
populations.10 Between 2011 and 2015, performance of abnormal visual evoked potentials were predictors of
MRI diagnostic criteria has been tested in Korean,28 development of a first clinical attack. Focusing on MRI,

6 www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2

 Review

the presence of gadolinium-enhancing lesions69 and following: normal findings or only non-specific white
asymptomatic spinal cord lesions (cervical or thoracic) matter lesions; or a T2-hyperintense or T1-weighted
are predictors of clinical change.68,70 gadolinium-enhancing lesion extending for more than
At present, more specific characterisation of people half the optic nerve length or involving the optic chiasm.
with radiologically isolated syndromes is needed, and For a diagnosis of acute myelitis, presence of an associated
prospective long-term studies should be done to estimate intramedullary MRI lesion extending for more than three
the risk of development of multiple sclerosis in these contiguous segments (longitudinally extensive transverse
patients. As a result, a firm recommendation for myelitis) or three contiguous segments of focal spinal
radiologically isolated syndromes is not possible. cord atrophy in patients with a history compatible with
However, even at this stage, individuals who have several acute myelitis needs to be shown. Diagnosis of area
risk factors clearly need to be distinguished from those postrema syndrome depends on the presence of associated
without these factors, since they are likely to have a dorsal medulla or area postrema lesions. Finally,
prodromal disorder, and specific criteria are needed for associated peri-ependymal brainstem lesions are needed
prompt diagnosis when the first symptom of CNS for a diagnosis of an acute brainstem syndrome.
involvement occurs.
T2*-weighted Phase
Differential diagnosis
Exclusion of alternative diagnoses that can mimic multiple
sclerosis, including atypical demyelination and
neuromyelitis optica, is imperative when applying the
2010 McDonald criteria.10 From an imaging perspective,
many inherited and acquired disorders could manifest
with evidence of dissemination in time or space, or both,
and these disorders should be included in the differential
diagnosis of multiple-sclerosis-like lesions. A timely
recognition of imaging red flags in the work-up of patients
suspected of having multiple sclerosis should alert
clinicians to reconsider the differential diagnosis more
extensively and do some additional analyses.71 Several
reviews discuss imaging features of the main acquired
and inherited white matter disorders that can enter the
differential diagnosis of multiple sclerosis,71–73 and, in our
view, when atypical imaging presentation occurs, these
disorders should be considered.
In the 2010 McDonald criteria,10 a specific focus was the T2*-weighted
differential diagnosis between multiple sclerosis and
neuromyelitis optica spectrum disorders. Up to 70% of
patients with neuromyelitis optica spectrum disorders at
onset have brain MRI lesions. Brain, optic nerve, and
spinal cord MRI findings of patients with neuromyelitis
optica spectrum disorders have been reviewed,74 and
revised diagnostic criteria for neuromyelitis optica
spectrum disorders have been proposed.75 The
International Panel for Neuromyelitis Optica Diagnosis Phase
proposed use of the unifying term neuromyelitis optica
spectrum disorders, which was stratified further by
AQP4–IgG testing. According to this revision, for patients
with a positive AQP4–IgG test, at least one core clinical
characteristic is needed for a diagnosis of neuromyelitis
optica spectrum disorder; these characteristics include
clinical syndromes or MRI findings related to optic nerve,
spinal cord, area postrema, other brainstem, diencephalic, 1 2 3 Years
or cerebral presentations. For AQP4–IgG-negative
patients or patients with unknown AQP4–IgG status, Figure 3: Ultra-high-field brain MRI in a patient with relapsing-remitting multiple sclerosis
T2*-weighted gradient-echo and phase axial MRIs, obtained with a 7·0 T scanner, from a 36-year-old woman with
stringent clinical criteria, with additional neuroimaging
relapsing-remitting multiple sclerosis. A periventricular non-enhancing lesion with a paramagnetic rim (green
findings, are needed. Specifically, for a diagnosis of acute arrows) is shown. A prominent central vein is visible on all images. The lesion maintains the same morphological
optic neuritis, brain MRI is needed to show either of the features at medium-term (3 years) follow-up imaging (magnified boxes).

www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2 7

 Review

patients with clinically isolated syndromes has not yet

been assessed. Several studies have identified some
Axial Sagittal interesting lesion characteristics, which can aid
differential diagnosis between multiple sclerosis and
other neurological disorders. The improved definition of
the relation between demyelinating lesions and the
intraparenchymal venous system, obtained by use of
T2*-weighted magnitude and phase imaging, confirms
results of pathological studies showing that many
multiple sclerosis plaques form around the
microvasculature.81–85,88,89 Perivenular lesion location can
help to distinguish white matter lesions in patients with
Sagittal Coronal
multiple sclerosis from incidental (ischaemic) white
matter lesions.85,89 This finding has been reinforced by
investigation of blood–brain barrier abnormalities in
multiple sclerosis at 7·0 T and 3·0 T, which showed that
most enhancing lesions are perivenular and that the
smallest lesions have a centrifugal pattern of
enhancement, suggesting that they grow outwards from a
central vein.90,91 The presence of a central small vein and a
rim of hypointensity on T2*-weighted magnitude or
Coronal Sagittal FLAIR* (combined T2*-weighted and FLAIR) imaging,85
obtained with a 7·0 T scanner, could be a distinctive
feature of multiple sclerosis white matter lesions, which
might assist in differentiation from lesions of patients
with neuromyelitis optica spectrum disorders92 or Susac
syndrome.93
A few studies have tracked the longitudinal changes of
the above-mentioned abnormalities (figure 3). Results of
Figure 4: Identification of the central vessels with MRI a longitudinal study80 of 29 patients with possible but
Non-contrast FLAIR* images (axial, sagittal, and coronal views), obtained with a 3·0 T scanner, from a 33-year-old
unclear diagnosis have shown that the presence of a
woman with multiple sclerosis. A conspicuous central vessel is clearly visible in most hyperintense lesions.
Visualisation of the central vessel in at least two perpendicular views (arrows in magnified boxes) is needed to central vein in most lesions can be used to accurately
define perivenular lesions. FLAIR*=combined T2*-weighted MRI and fluid-attenuated inversion recovery MRI. identify patients with multiple sclerosis.80 Another study94
showed that phase-ring lesions remained unchanged
Use of high-field and ultra-high-field scanners during a 2·5-year period in five patients with relapsing-
High-field scanners (3·0 T) remitting multiple sclerosis, whereas such a ring can be
Compared with 1·5 T scanners, use of high-field-strength transient in acute lesions.90,91
scanners (3·0 T) enables detection of a significantly In summary, use of high-field or ultra-high-field
higher number of lesions in patients with clinically scanners is not likely to result in an earlier diagnosis.
isolated syndromes,76,77 with improved recognition of However, lesion features distinctive of multiple sclerosis
lesions involving the cortical,78 infratentorial, and could emerge from use of these scanners and might
periventricular regions.76 Comparison of MRI criteria eventually enhance differentiation of multiple sclerosis
performance at 1·5 T versus 3·0 T in 40 patients with from other diseases.
clinically isolated syndromes showed that one additional
patient was diagnosed with dissemination in space at Future perspectives
high field, without improvement for dissemination in The expert panel noted that some promising measures
time.79 deserve further investigation before being moved (or not)
to diagnostic criteria in the future.
Ultra-high-field scanners (7·0 T) For identification of the central vein, sequences capable
Ultra-high-field MRI enables detection of a significantly of showing these features on 3·0 T and 1·5 T scanners
higher number of lesions,80 and improved definition of need to be standardised, and standardised definitions
lesions located in the white and grey matter with respect need to be created for identification of central vessels. So
to their morphology and association with the far, central vessels have been identified if they could be
vasculature,81–85 than was previously shown with 1·5 T86 or visualised in at least two perpendicular planes, appeared
3·0 T87 scanners. Whether assessment of lesion number linear in at least one plane, and were completely
and distribution with ultra-high-field MRI scanners helps surrounded by hyperintense signal in at least one plane
to make an earlier diagnosis of multiple sclerosis in (figure 4).89 Whether central veins are confirmatory for

8 www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2

 Review

multiple sclerosis lesions needs further study with

appropriate disease comparisons. Search strategy and selection criteria
For identification of the hypointense lesional rim on References for this Review were identified through searches
T2*-weighted magnitude or phase images, longitudinal of PubMed with the search terms “clinically isolated
studies need to be done at 3·0 T and 7·0 T; magnetic syndrome”, “multiple sclerosis”, “McDonald criteria”,
resonance sequences across 3·0 T scanners from “diagnosis”, “differential diagnosis”, “cortical lesions”, “white
different manufacturers need to be standardised and matter”, “lesions”, “cortical lesions”, “brain”, “spinal cord”,
clinically implemented; value in predicting conversion to “MRI”, “optic nerve”, “disease dissemination in space”,
multiple sclerosis and disability progression in patients “disease dissemination in time”, “radiologically isolated
with clinically isolated symptoms needs to be analysed; syndromes”, “pediatric MS”, “T1-hypointense lesions”,
and different multiple sclerosis clinical phenotypes and “symptomatic lesions”, “primary progressive multiple
other neurological disorders that can mimic multiple sclerosis”, “non-caucasian populations”, “neuromyelitis
sclerosis need to be studied. optica”, “neuromyelitis optica spectrum disorders”, “high
For identification of cortical pathology, high-field- field”, and “ultra-high field” from Jan 1, 1979, to Nov 15,
strength imaging is expected to enable identification of 2015. Articles were also identified through searches of our
cortical lesions more reliably than conventional MRI but own files. Only papers published in English were reviewed.
is not likely to be available in clinical practice in the near The final reference list was generated on the basis of
term. More advanced techniques for cortical lesion originality and relevance to the broad scope of this Review.
identification at 3·0 T might prove valuable in multiple
sclerosis diagnosis. The definition of standardised, up-to-
date guidelines for cortical lesion classification is also multiple-sclerosis-directed therapies at present. Since
pending. high-field-strength imaging and more recently developed
sequences enable improved pathology-level interrogation
Conclusions of the CNS, the fundamental question of what defines a
Assessment of MRI scans should be done in the disease such as multiple sclerosis will need to be
appropriate clinical context. The premise of these answered.
guidelines and criteria is that we assume a basic Contributors
knowledge of what constitutes a lesion. The largest linear MF and FB developed the idea for the meeting, organised it, chaired it,
measurement for lesion definition should be 3 mm or and framed the structure of this Review. AR, FB, JS-G, LK, MAR, MT,
NDS, NE, and OC participated in the meeting, summarised different
more in at least one plane. Therefore, lesion identification aspects for the discussions, and took part in the discussions. JLF, CG,
should be done by trained, expert personnel. Image JP, and DSR participated in the meeting and discussions. BB and XM
quality should be of a high standard. A conservative were involved after the meeting for critical discussion and revision.
approach to identification of lesions should be used. The complete Review was commented on, revised, and approved by all
authors.
In the diagnostic work-up of patients with suspected
multiple sclerosis, use of post-contrast sequences MAGNIMS Steering Committee
F Barkhof (MS Centre Amsterdam, VU University Medical Centre,
provides important information for differential Amsterdam, Netherlands), O Ciccarelli (Queen Square MS Centre, UCL
diagnosis. However, the US Food and Drug Institute of Neurology, London, UK), N De Stefano (University of Siena,
Administration has issued a safety communication for Siena, Italy), C Enzinger (Department of Neurology, Medical University For the FDA safety
the long-term effects of repeated administration of of Graz, Graz, Austria), M Filippi and M A Rocca (San Raffaele Scientific communication report see
Institute, Vita-Salute San Raffaele University, Milan, Italy), http://www.fda.gov/downloads/
gadolinium-based contrast agents, after the description J L Frederiksen (Glostrup Hospital and University of Copenhagen, Drugs/DrugSafety/UCM455390.
of deposition of gadolinium in the brains of some Copenhagen, Denmark), C Gasperini (San Camillo-Forlanini Hospital, pdf
patients who underwent several contrast-enhanced MRI Rome, Italy), L Kappos (University of Basel, Basel, Switzerland), J Palace
scans. The effect of these safety concerns is unknown at (University of Oxford Hospitals Trust, Oxford, UK), A Rovira and
J Sastre-Garriga (Hospital Universitari Vall d’Hebron, Universitat
present. Autònoma de Barcelona, Barcelona, Spain), T Yousry (Queen Square MS
MRI remains a valuable tool for identification of Centre, UCL Institute of Neurology, London, UK), H Vrenken (MS
children and adults with multiple sclerosis, both at the Centre Amsterdam, VU Medical Centre, Amsterdam, Netherlands).
time of an incident attack and when applied serially to Declaration of interests
confirm the chronic nature of this disease. Advanced Travel expenses for the workshop were paid by Novartis through a
provider. The speakers did not receive direct reimbursement. MF serves
imaging techniques provide information about regional
on scientific advisory boards for Teva Pharmaceutical Industries, has
CNS involvement with greater sensitivity than received compensation for consulting services and speaking activities
conventional MRI and might add to diagnostic specificity. from Biogen Idec, Excemed, Novartis, and Teva Pharmaceutical
Whether MRI features consistent with multiple sclerosis Industries, and receives research support from Biogen Idec, Teva
Pharmaceutical Industries, and Novartis. MAR has received speakers
in the absence of clinical involvement can confirm
honoraria from Biogen Idec, Novartis, Genzyme, Sanofi-Aventis, and
multiple sclerosis diagnosis remains an area of debate Excemed. OC serves as a consultant for GE, Biogen Idec, and Novartis,
that needs further study and deliberation, especially in and all the payments are made to the institution (Queen Square MS
view of evidence that some such individuals have focal Centre, UCL Institute of Neurology, London, UK). NDS has received
honoraria from Schering, Biogen Idec, Teva Pharmaceutical Industries,
and global loss of tissue integrity but are not eligible for

www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2 9

 Review

Novartis, Genzyme, and Merck Serono SA for consulting services, (FP7 European committee) and the Dutch Foundation for MS Research—
speaking, and travel support. He serves on advisory boards for Biogen centre grant 2010–14.
Idec, Merck Serono SA, and Novartis. NE has received honoraria from
References
Biogen Idec, Novartis, and Genzyme for consulting services, speaking, 1 McDonald WI, Compston A, Edan G, et al. Recommended
and travel support. He serves on advisory boards for Biogen Idec, diagnostic criteria for multiple sclerosis: guidelines from the
Merck Serono, and Novartis. LK’s institution (University Hospital Basel) International Panel on the diagnosis of multiple sclerosis.
has received in the past 3 years and used exclusively for research support Ann Neurol 2001; 50: 121–27.
steering committee, advisory board, and consultancy fees from Actelion, 2 Paty DW, Oger JJ, Kastrukoff LF, et al. MRI in the diagnosis of MS:
Addex, Bayer HealthCare, Biogen Idec, Biotica, Genzyme, Eli-Lilly & a prospective study with comparison of clinical evaluation, evoked
Company, Merck, Mitsubishi, Novartis, Ono Pharma, Pfizer, Receptos, potentials, oligoclonal banding, and CT. Neurology 1988; 38: 180–85.
Sanofi-Aventis, Santhera, Siemens, Teva Pharmaceutical Industries, UCB, 3 Fazekas F, Offenbacher H, Fuchs S, et al. Criteria for an increased
and Xenoport, speaker fees from Bayer HealthCare, Biogen Idec, Merck specificity of MRI interpretation in elderly subjects with suspected
Serono, Novartis, Sanofi-Aventis, and Teva Pharmaceutical Industries, multiple sclerosis. Neurology 1988; 38: 1822–25.
support of educational activities from Bayer HealthCare, Biogen Idec, CSL 4 Barkhof F, Filippi M, Miller DH, et al. Comparison of MRI criteria
Behring, Genzyme, Merck Serono, Novartis, Sanofi-Aventis, and Teva at first presentation to predict conversion to clinically definite
Pharmaceutical Industries, royalties from Neurostatus Systems GmbH, multiple sclerosis. Brain 1997; 120: 2059–69.
and grants from Bayer HealthCare, Biogen Idec, Merck Serono, Novartis, 5 Swanton JK, Fernando K, Dalton CM, et al. Modification of MRI
Roche, and Roche Research Foundations. AR serves on scientific advisory criteria for multiple sclerosis in patients with clinically isolated
boards for Biogen Idec, Novartis, Genzyme, and OLEA Medical, has syndromes. J Neurol Neurosurg Psychiatry 2006; 77: 830–33.
received speaker honoraria from Bayer HealthCare Pharmaceuticals, 6 Swanton JK, Rovira A, Tintore M, et al. MRI criteria for multiple
Genzyme, Bracco, Merck Serono, Teva Pharmaceutical Industries, OLEA sclerosis in patients presenting with clinically isolated syndromes:
Medical, Stendhal, Novartis, and Biogen Idec, and has research a multicentre retrospective study. Lancet Neurol 2007; 6: 677–86.
agreements with Siemens AG. JS-G has received compensation for 7 Rovira A, Swanton J, Tintore M, et al. A single, early magnetic
serving on scientific advisory boards or on speaker’s bureaus from Biogen resonance imaging study in the diagnosis of multiple sclerosis.
Arch Neurol 2009; 66: 587–92.
Idec, Merck Serono, Novartis, Teva Pharmaceutical Industries, and
Sanofi-Aventis. MT has received compensation for consulting services and 8 Polman CH, Reingold SC, Edan G, et al. Diagnostic criteria for
multiple sclerosis: 2005 revisions to the “McDonald Criteria”.
speaker’s fees from Bayer Schering, Merck Serono, Biogen Idec, Teva
Ann Neurol 2005; 58: 840–46.
Pharmaceutical Industries, Sanofi-Aventis, and Novartis. JLF has served
9 Montalban X, Tintore M, Swanton J, et al. MRI criteria for MS in
on scientific advisory boards for, and received funding of travel for
patients with clinically isolated syndromes. Neurology 2010; 74: 427–34.
participation in scientific advisory boards and honoraria from, Biogen
10 Polman CH, Reingold SC, Banwell B, et al. Diagnostic criteria for
Idec, Genzyme, Merck Serono, Novartis, Sanofi-Aventis, Takeda, and Teva
multiple sclerosis: 2010 Revisions to the McDonald criteria.
Pharmaceutical Industries. CG has received compensation for consulting Ann Neurol 2011; 69: 292–302.
from Bayer HealthCare Pharmaceuticals and Biogen Idec and as a
11 Rovira A, Wattjes MP, Tintore M, et al. Evidence-based guidelines:
speaker for lectures from Biogen Idec, Bayer HealthCare Pharmaceuticals, MAGNIMS consensus guidelines on the use of MRI in multiple
Genzyme, Merck Serono, Novartis, and Teva Pharmaceutical Industries. sclerosis-clinical implementation in the diagnostic process.
JP reports personal fees from Biogen Idec, Teva Pharmaceutical Nat Rev Neurol 2015; 11: 471–82.
Industries, Merck Serono, Bayer Schering, Novartis, Chugai Pharma, 12 Tintoré M, Rovira A, Río J, et al. Factors that determine disease
Ono Pharmaceuticals Co, and CI consulting, and grants from Merck course: the symptomatic lesion matters. 1000 CIS subgroup
Serono, Bayer Schering, and Novartis. DSR reports grants from Vertex analysis. Mult Scler 2014; 20 (SI): 400.
Pharmaceuticals and patents PCT/US2012/067997 and 13 Caucheteux N, Maarouf A, Genevray M, et al. Criteria improving
PCT/US2013/033334 pending for development of lesion segmentation multiple sclerosis diagnosis at the first MRI. J Neurol 2015;
algorithms. BB serves as a centralised MRI reviewer for Novartis, and an 262: 979–87.
unpaid advisor regarding paediatric multiple sclerosis clinical trial design 14 Brownlee WJ, Swanton JK, Miszkiel KA, Miller DH, Ciccarelli O.
for Novartis, Biogen Idec, and Teva Neuroscience. XM has received Should we include lesions in the symptomatic site in dissemination
speaker’s honoraria and travel expenses for scientific meetings, has been in space in patients with clinically isolated syndromes? Mult Scler
a steering committee member of clinical trials or participated in advisory 2015; 23: 65.
boards of clinical trials in the past 5 years with Actelion, Almirall, Bayer, 15 Kang H, Metz LM, Traboulsee AL, et al. Application and a proposed
Biogen Idec, Genzyme, Merck Serono, Novartis, Octapharma, Receptos, modification of the 2010 McDonald criteria for the diagnosis of
Roche, Sanofi-Genzyme, Teva Pharmaceutical Industries, and Trophos. multiple sclerosis in a Canadian cohort of patients with clinically
FB serves as a scientific consultant to Bayer Schering Pharma, isolated syndromes. Mult Scler 2014; 20: 458–63.
Sanofi-Aventis, Biogen Idec, Teva Pharmaceutical Industries, Merck 16 Sombekke MH, Wattjes MP, Balk LJ, et al. Spinal cord lesions in
Serono, Novartis, Roche, Synthon BV, Janssen, Genzyme, and Toshiba patients with clinically isolated syndrome: a powerful tool in
diagnosis and prognosis. Neurology 2013; 80: 69–75.
Medical systems, and has served on speakers’ bureaus for Serono
Symposia Foundation and MedScape. 17 Bot JC, Barkhof F, Polman CH, et al. Spinal cord abnormalities in
recently diagnosed MS patients: added value of spinal MRI
Acknowledgments examination. Neurology 2004; 62: 226–33.
The workshop that formed the basis of this Review was supported by an 18 Weier K, Mazraeh J, Naegelin Y, et al. Biplanar MRI for the
unrestricted educational grant from Novartis. The funding source had no assessment of the spinal cord in multiple sclerosis. Mult Scler 2012;
role in the preparation of this article. We are very grateful to 18: 1560–69.
Douglas Arnold (Brain Imaging Center, Montreal Neurological Institute, 19 Nair G, Absinta M, Reich DS. Optimized T1-MPRAGE sequence for
McGill University, Montreal, QC, Canada) for his fruitful discussion better visualization of spinal cord multiple sclerosis lesions at 3T.
during the meeting and his subsequent thoughtful comments on the AJNR Am J Neuroradiol 2013; 34: 2215–22.
manuscript. We also thank Martina Absinta (Neuroimaging Research 20 Kelly SB, Kinsella K, Duggan M, Tubridy N, McGuigan C,
Unit, Institute of Experimental Neurology, Division of Neuroscience, Hutchinson M. A proposed modification to the McDonald 2010
San Raffaele Scientific Institute, Vita-Salute San Raffaele University, criteria for the diagnosis of primary progressive multiple sclerosis.
Mult Scler 2013; 19: 1095–100.
Milan, Italy) and Pascal Sati (Translational Neuroradiology Unit, National
Institute of Neurological Disorders and Stroke, National Institutes of 21 Bigi S, Marrie RA, Verhey L, Yeh EA, Banwell B. 2010 McDonald
criteria in a pediatric cohort: is positivity at onset associated with a
Health, Bethesda, MD, USA) for providing figures 3 and 4. LK has
more aggressive multiple sclerosis course? Mult Scler 2013; 19: 1359–62.
received research grants from the Swiss MS Society, the Swiss National
22 Hummel HM, Bruck W, Dreha-Kulaczewski S, Gartner J, Wuerfel J.
Research Foundation, and the European Union. JP has received research
Pediatric onset multiple sclerosis: McDonald criteria 2010 and the
grants from the Multiple Sclerosis Society UK, and the Guthy-Jackson
contribution of spinal cord MRI. Mult Scler 2013; 19: 1330–35.
Foundation. FB receives research support from Neugrid4you

10 www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2

 Review

23 Sadaka Y, Verhey LH, Shroff MM, et al. 2010 McDonald criteria for 46 Bo L, Vedeler CA, Nyland HI, Trapp BD, Mork SJ. Subpial
diagnosing pediatric multiple sclerosis. Ann Neurol 2012; 72: 211–23. demyelination in the cerebral cortex of multiple sclerosis patients.
24 Tantsis EM, Prelog K, Brilot F, Dale RC. Risk of multiple sclerosis J Neuropathol Exp Neurol 2003; 62: 723–32.
after a first demyelinating syndrome in an Australian Paediatric 47 Geurts JJ, Pouwels PJ, Uitdehaag BM, Polman CH, Barkhof F,
cohort: clinical, radiological features and application of the Castelijns JA. Intracortical lesions in multiple sclerosis: improved
McDonald 2010 MRI criteria. Mult Scler 2013; 19: 1749–59. detection with 3D double inversion-recovery MR imaging. Radiology
25 Kornek B, Schmitl B, Vass K, et al. Evaluation of the 2010 McDonald 2005; 236: 254–60.
multiple sclerosis criteria in children with a clinically isolated 48 Nelson F, Poonawalla AH, Hou P, Huang F, Wolinsky JS,
syndrome. Mult Scler 2012; 18: 1768–74. Narayana PA. Improved identification of intracortical lesions in
26 Williams MT, Tapos DO, Juhasz C. Use of the 2010 McDonald multiple sclerosis with phase-sensitive inversion recovery in
criteria can facilitate early diagnosis of pediatric multiple sclerosis combination with fast double inversion recovery MR imaging.
in a predominantly black cohort. Pediatr Neurol 2014; 51: 826–30. AJNR Am J Neuroradiol 2007; 28: 1645–49.
27 Banwell B, Bar-Or A, Arnold DL, et al. Clinical, environmental, and 49 Sethi V, Yousry TA, Muhlert N, et al. Improved detection of cortical
genetic determinants of multiple sclerosis in children with acute MS lesions with phase-sensitive inversion recovery MRI.
demyelination: a prospective national cohort study. Lancet Neurol J Neurol Neurosurg Psychiatry 2012; 83: 877–82.
2011; 10: 436–45. 50 Favaretto A, Poggiali D, Lazzarotto A, Rolma G, Causin F, Gallo P.
28 Huh SY, Kim SH, Kim W, et al. Evaluation of McDonald MRI The parallel analysis of phase sensitive inversion recovery (PSIR)
criteria for dissemination in space in Korean patients with clinically and double inversion recovery (DIR) images significantly improves
isolated syndromes. Mult Scler 2014; 20: 492–95. the detection of cortical lesions in multiple sclerosis (MS) since
29 Hsueh CJ, Kao HW, Chen SY, et al. Comparison of the 2010 and 2005 clinical onset. PLoS One 2015; 10: e0127805.
versions of the McDonald MRI criteria for dissemination-in-time in 51 Nelson F, Poonawalla A, Hou P, Wolinsky JS, Narayana PA.
Taiwanese patients with classic multiple sclerosis. J Neurol Sci 2013; 3D MPRAGE improves classification of cortical lesions in multiple
329: 51–54. sclerosis. Mult Scler 2008; 14: 1214–19.
30 Patrucco L, Rojas JI, Miguez JS, Cristiano E. Application of the 52 Seewann A, Vrenken H, Kooi EJ, et al. Imaging the tip of the
McDonald 2010 criteria for the diagnosis of multiple sclerosis in an iceberg: visualization of cortical lesions in multiple sclerosis.
Argentinean cohort of patients with clinically isolated syndromes. Mult Scler 2011; 17: 1202–10.
Mult Scler 2013; 19: 1297–301. 53 Seewann A, Kooi EJ, Roosendaal SD, et al. Postmortem verification
31 Belova AN, Shalenkov IV, Shakurova DN, Boyko AN. Revised of MS cortical lesion detection with 3D DIR. Neurology 2012;
McDonald criteria for multiple sclerosis diagnostics in central 78: 302–08.
Russia: sensitivity and specificity. Mult Scler 2014; 20: 1896–99. 54 Calabrese M, De Stefano N, Atzori M, et al. Detection of cortical
32 Absinta M, Rocca MA, Colombo B, et al. Patients with migraine do inflammatory lesions by double inversion recovery magnetic
not have MRI-visible cortical lesions. J Neurol 2012; 259: 2695–98. resonance imaging in patients with multiple sclerosis. Arch Neurol
33 Ruet A, Arrambide G, Brochet B, et al. Early predictors of multiple 2007; 64: 1416–22.
sclerosis after a typical clinically isolated syndrome. Mult Scler 2014; 55 Filippi M, Rocca MA, Calabrese M, et al. Intracortical lesions:
20: 1721–26. relevance for new MRI diagnostic criteria for multiple sclerosis.
34 Ruet A, Deloire MS, Ouallet JC, Molinier S, Brochet B. Predictive Neurology 2010; 75: 1988–94.
factors for multiple sclerosis in patients with clinically isolated 56 Calabrese M, Oh MS, Favaretto A, et al. No MRI evidence of cortical
spinal cord syndrome. Mult Scler 2011; 17: 312–18. lesions in neuromyelitis optica. Neurology 2012; 79: 1671–76.
35 Moraal B, Pohl C, Uitdehaag BM, et al. Magnetic resonance 57 Geurts JJ, Roosendaal SD, Calabrese M, et al. Consensus
imaging predictors of conversion to multiple sclerosis in the recommendations for MS cortical lesion scoring using double
BENEFIT study. Arch Neurol 2009; 66: 1345–52. inversion recovery MRI. Neurology 2011; 76: 418–24.
36 Kim SS, Richman DP, Johnson WO, Hald JK, Agius MA. Limited 58 van Walderveen MA, Kamphorst W, Scheltens P, et al.
utility of current MRI criteria for distinguishing multiple sclerosis Histopathologic correlate of hypointense lesions on T1-weighted
from common mimickers: primary and secondary CNS vasculitis, spin-echo MRI in multiple sclerosis. Neurology 1998; 50: 1282–88.
lupus and Sjogren’s syndrome. Mult Scler 2014; 20: 57–63. 59 Giorgio A, Stromillo ML, Bartolozzi ML, et al. Relevance of
37 Verhey LH, Branson HM, Shroff MM, et al. MRI parameters for hypointense brain MRI lesions for long-term worsening of clinical
prediction of multiple sclerosis diagnosis in children with acute disability in relapsing multiple sclerosis. Mult Scler 2014; 20: 214–19.
CNS demyelination: a prospective national cohort study. 60 Enzinger C, Fuchs S, Pichler A, et al. Predicting the severity of
Lancet Neurol 2011; 10: 1065–73. relapsing-remitting MS: the contribution of cross-sectional and
38 Miller D, Barkhof F, Montalban X, Thompson A, Filippi M. short-term follow-up MRI data. Mult Scler 2011; 17: 695–701.
Clinically isolated syndromes suggestive of multiple sclerosis, 61 Mitjana R, Tintore M, Rocca MA, et al. Diagnostic value of brain
part I: natural history, pathogenesis, diagnosis, and prognosis. chronic black holes on T1-weighted MR images in clinically isolated
Lancet Neurol 2005; 4: 281–88. syndromes. Mult Scler 2014; 20: 1471–77.
39 Tintore M, Rovira A, Rio J, et al. Defining high, medium and low 62 Sastre-Garriga J, Tintore M, Rovira A, et al. Specificity of Barkhof
impact prognostic factors for developing multiple sclerosis. Brain criteria in predicting conversion to multiple sclerosis when applied
2015; 138: 1863–74. to clinically isolated brainstem syndromes. Arch Neurol 2004;
40 Kuhle J, Disanto G, Dobson R, et al. Conversion from clinically 61: 222–24.
isolated syndrome to multiple sclerosis: a large multicentre study. 63 Montalban X, Sastre-Garriga J, Filippi M, et al. Primary progressive
Mult Scler 2015; 8: 1013–24. multiple sclerosis diagnostic criteria: a reappraisal. Mult Scler 2009;
41 Sorensen TL, Frederiksen JL, Bronnum-Hansen H, Petersen HC. 15: 1459–65.
Optic neuritis as onset manifestation of multiple sclerosis: 64 Okuda DT, Mowry EM, Beheshtian A, et al. Incidental MRI
a nationwide, long-term survey. Neurology 1999; 53: 473–78. anomalies suggestive of multiple sclerosis: the radiologically
42 Tintore M, Rovira A, Rio J, et al. Is optic neuritis more benign than isolated syndrome. Neurology 2009; 72: 800–05.
other first attacks in multiple sclerosis? Ann Neurol 2005; 65 Okuda DT. Unanticipated demyelinating pathology of the CNS.
57: 210–15. Nat Rev Neurol 2009; 5: 591–97.
43 Wilejto M, Shroff M, Buncic JR, Kennedy J, Goia C, Banwell B. 66 Stromillo ML, Giorgio A, Rossi F, et al. Brain metabolic changes
The clinical features, MRI findings, and outcome of optic neuritis suggestive of axonal damage in radiologically isolated syndrome.
in children. Neurology 2006; 67: 258–62. Neurology 2013; 80: 2090–94.
44 Kidd D, Barkhof F, McConnell R, Algra PR, Allen IV, Revesz T. 67 Giorgio A, Stromillo ML, Rossi F, et al. Cortical lesions in
Cortical lesions in multiple sclerosis. Brain 1999; 122: 17–26. radiologically isolated syndrome. Neurology 2011; 77: 1896–99.
45 Peterson JW, Bo L, Mork S, Chang A, Trapp BD. Transected 68 Okuda DT, Siva A, Kantarci O, et al. Radiologically isolated
neurites, apoptotic neurons, and reduced inflammation in cortical syndrome: 5-year risk for an initial clinical event. PLoS One 2014;
multiple sclerosis lesions. Ann Neurol 2001; 50: 389–400. 9: e90509.

www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2 11

 Review

69 Lebrun C, Bensa C, Debouverie M, et al. Association between 82 Hammond KE, Metcalf M, Carvajal L, et al. Quantitative in vivo
clinical conversion to multiple sclerosis in radiologically isolated magnetic resonance imaging of multiple sclerosis at 7 Tesla with
syndrome and magnetic resonance imaging, cerebrospinal fluid, sensitivity to iron. Ann Neurol 2008; 64: 707–13.
and visual evoked potential: follow-up of 70 patients. Arch Neurol 83 Tallantyre EC, Brookes MJ, Dixon JE, Morgan PS, Evangelou N,
2009; 66: 841–46. Morris PG. Demonstrating the perivascular distribution of MS
70 Okuda DT, Mowry EM, Cree BA, et al. Asymptomatic spinal cord lesions in vivo with 7-Tesla MRI. Neurology 2008; 70: 2076–78.
lesions predict disease progression in radiologically isolated 84 Tallantyre EC, Morgan PS, Dixon JE, et al. A comparison of 3T and
syndrome. Neurology 2011; 76: 686–92. 7T in the detection of small parenchymal veins within MS lesions.
71 Charil A, Yousry TA, Rovaris M, et al. MRI and the diagnosis of Invest Radiol 2009; 44: 491–94.
multiple sclerosis: expanding the concept of “no better explanation”. 85 Kilsdonk ID, Lopez-Soriano A, Kuijer JP, et al. Morphological
Lancet Neurol 2006; 5: 841–52. features of MS lesions on FLAIR* at 7 T and their relation to patient
72 Miller DH, Weinshenker BG, Filippi M, et al. Differential diagnosis characteristics. J Neurol 2014; 261: 1356–64.
of suspected multiple sclerosis: a consensus approach. Mult Scler 86 Tan IL, van Schijndel RA, Pouwels PJ, et al. MR venography of
2008; 14: 1157–74. multiple sclerosis. AJNR Am J Neuroradiol 2000; 21: 1039–42.
73 Aliaga ES, Barkhof F. MRI mimics of multiple sclerosis. 87 Sati P, George IC, Shea CD, Gaitan MI, Reich DS. FLAIR*:
Handb Clin Neurol 2014; 122: 291–316. a combined MR contrast technique for visualizing white matter
74 Kim HJ, Paul F, Lana-Peixoto MA, et al. MRI characteristics of lesions and parenchymal veins. Radiology 2012; 265: 926–32.
neuromyelitis optica spectrum disorder: an international update. 88 Mainero C, Benner T, Radding A, et al. In vivo imaging of cortical
Neurology 2015; 84: 1165–73. pathology in multiple sclerosis using ultra-high field MRI.
75 Wingerchuk DM, Banwell B, Bennett JL, et al. International Neurology 2009; 73: 941–48.
consensus diagnostic criteria for neuromyelitis optica spectrum 89 Tallantyre EC, Dixon JE, Donaldson I, et al. Ultra-high-field imaging
disorders. Neurology 2015; 14: 177–89. distinguishes MS lesions from asymptomatic white matter lesions.
76 Wattjes MP, Lutterbey GG, Harzheim M, et al. Higher sensitivity in Neurology 2011; 76: 534–39.
the detection of inflammatory brain lesions in patients with 90 Gaitan MI, Sati P, Inati SJ, Reich DS. Initial investigation of the
clinically isolated syndromes suggestive of multiple sclerosis using blood-brain barrier in MS lesions at 7 tesla. Mult Scler 2013;
high field MRI: an intraindividual comparison of 1·5 T with 3·0 T. 19: 1068–73.
Eur Radiol 2006; 16: 2067–73. 91 Absinta M, Sati P, Gaitan MI, et al. Seven-tesla phase imaging of
77 Nielsen K, Rostrup E, Frederiksen JL, et al. Magnetic resonance acute multiple sclerosis lesions: a new window into the
imaging at 3.0 tesla detects more lesions in acute optic neuritis than inflammatory process. Ann Neurol 2013; 74: 669–78.
at 1·5 tesla. Invest Radiol 2006; 41: 76–82. 92 Sinnecker T, Dorr J, Pfueller CF, et al. Distinct lesion morphology
78 Simon B, Schmidt S, Lukas C, et al. Improved in vivo detection of at 7-T MRI differentiates neuromyelitis optica from multiple
cortical lesions in multiple sclerosis using double inversion sclerosis. Neurology 2012; 79: 708–14.
recovery MR imaging at 3 Tesla. Eur Radiol 2010; 20: 1675–83. 93 Wuerfel J, Sinnecker T, Ringelstein EB, et al. Lesion morphology at
79 Wattjes MP, Harzheim M, Lutterbey GG, et al. Does high field MRI 7 Tesla MRI differentiates Susac syndrome from multiple sclerosis.
allow an earlier diagnosis of multiple sclerosis? J Neurol 2008; Mult Scler 2012; 18: 1592–99.
255: 1159–63. 94 Bian W, Harter K, Hammond-Rosenbluth KE, et al. A serial in vivo
80 Mistry N, Tallantyre EC, Dixon JE, et al. Focal multiple sclerosis 7T magnetic resonance phase imaging study of white matter lesions
lesions abound in ‘normal appearing white matter’. Mult Scler 2011; in multiple sclerosis. Mult Scler 2012; 19: 69–75.
17: 1313–23.
81 Ge Y, Zohrabian VM, Grossman RI. Seven-Tesla magnetic
resonance imaging: new vision of microvascular abnormalities in
multiple sclerosis. Arch Neurol 2008; 65: 812–16.

12 www.thelancet.com/neurology Published online January 25, 2015 http://dx.doi.org/10.1016/S1474-4422(15)00393-2

View publication stats