Review

Clinically isolated syndromes suggestive of multiple

sclerosis, part 2: non-conventional MRI, recovery processes,
and management
David Miller, Frederik Barkhof, Xavier Montalban, Alan Thompson, Massimo Filippi Lancet Neurol 2005; 4: 341–48
See Review Lancet Neurol 2005;
The onset of multiple sclerosis (MS) in 85% of young adults is with a subacute clinically isolated syndrome (CIS) of 4: 281–88

the optic nerves, brainstem, or spinal cord. Whereas multifocal brain lesions are present on MRI in many patients MS NMR Research Unit,
Institute of Neurology,
with a CIS, some patients have additional abnormalities on quantitative MRI in otherwise normal-appearing white
University College London,
and grey matter that suggest an extensive pathological process. Functional outcome for patients with symptomatic London, UK (D Miller FRCP,
CIS lesions is determined by the interplay of inflammation, demyelination, axonal damage, remyelination, and A Thompson FRCP);
cortical adaptation. Recovery of function may be accelerated by high dose corticosteroids, and although interferon Department of
Neuroradiology, VU Medical
beta delays the development of a second relapse, its long-term effect is unknown. A better understanding of Centre, Amsterdam,
pathological and pathogenetic processes in patients with a CIS will facilitate the development of disease-modifying Netherlands (F Barkhof MD);
treatments for patients with MS before they become disabled. Continued clinical and laboratory investigation of Department of
patients with a CIS should be encouraged. Neuroimmunology, Hospital
Vall d’Hebron, Barcelona, Spain
(X Montalban MD);
Introduction Conventional MRI shows T2-visible lesions but does Neuroimaging Research Unit,
In 85% of young adults who develop multiple sclerosis not show other abnormalities. However, several Scientific Institute and
(MS), onset is with an acute, clinically isolated syndrome quantitative MRI techniques can show abnormalities in University Ospedale San
Raffaele, Milan, Italy
(CIS) of the optic nerves, brainstem, or spinal cord. In otherwise normal-appearing brain tissue; these (M Filippi MD)
part 1 of this review, published in the May 2005 issue, techniques have the potential to increase knowledge of Correspondence to:
the clinical presentation, pathogenesis, diagnosis, and the pathobiology of MS. Conventional MRI lesion Prof D H Miller, NMR Research
prognosis of CISs were covered. In part 2 we review non- measures are of limited value for prognosis in the early Unit, Department of
conventional MRI findings, recovery processes, and stages of the disease, and quantitative MRI techniques Neuroinflammation, Institute of
Neurology, University College
management of patients. have potential to give a clearer prognosis. London, London WC1N 3BG, UK
Magnetisation-transfer MRI has been one of the most d.miller@ion.ucl.ac.uk
Abnormalities on non-conventional MRI widely used methods for the assessment of patients
Progressive brain atrophy is a well-known feature of with MS and can detect “occult” tissue damage in the
MS and is thought to be a marker of irreversible tissue brain and cervical cord.7 Low magnetisation-transfer
damage.1 However, the exact causes and timing of ratios have been detected in normal-appearing brain
atrophy in MS are unknown. The assessment of brain tissue of patients with CIS at presentation, and in one
and spinal-cord atrophy in patients with a CIS should study, the extent of these abnormalities was reported to
help to define how early irreversible tissue loss occurs be an independent predictor of subsequent disease
in MS. Dalton and co-workers2 prospectively followed- progression.8 However, other studies9,10 with region-
up 55 patients with CISs for 3 years. After 1 year, of-interest or whole-brain-histogram analysis of
patients with MS (according to the McDonald criteria3) magnetisation-transfer MRI data did not find that low
had substantially more ventricular enlargement than magnetisation-transfer ratios predict disease
those without such disease progression. After 3 years, progression. Another study11 showed magnetisation-
29 (53%) of the patients had progressed to MS and transfer-ratio abnormalities in a group of patients with
ventricular volume and grey-matter atrophy were a recent onset of CIS and in a group with a remote
greater in those with MS than in those who did not onset of CIS; the similar findings in both groups, and
have MS. White-matter volume did not change over in those with normal and abnormal conventional
time in either subgroup.4 Measures of lesion load and scans, prompted the researchers to propose that
atrophy measures were moderately related, suggesting magnetisation-transfer MRI abnormalities in CIS
that T2-visible lesions account only partially for the suggest susceptibility to demyelination. Recent studies
observed changes in brain volume. Similar results were of patients with CISs has found no abnormality
also found when brain atrophy was measured in the in cervical-cord magnetisation-transfer-ratios12 but
Early Treatment of MS (ETOMS) trial.5 In 43 patients abnormal measures of diffusion in normal-appearing
with a CIS, atrophy of the spinal cord was also white matter;13 however these findings were not
assessed;6 although the area of the spinal cord in predictive of subsequent lesion dissemination (as
patients with a CIS and an abnormal brain MRI at defined by the McDonald criteria) at 3 months and
presentation was slightly smaller than that of healthy 12 months.13 Overall, the published findings on
controls, there was no measurable change in patients magnetisation-transfer ratio and diffusion suggest that
or controls over 1 year. subtle white-matter damage might occur at a very early

http://neurology.thelancet.com Vol 4 June 2005 341

 Review

stage in patients with a CIS but do not predict short- co-workers24 found that magnetisation-transfer ratio
term lesion development. was lower in the optic nerves of patients with MS
In patients with established MS, proton-magnetic- without recovery than in those with clinical recovery,
resonance spectroscopy studies have found low and that magnetisation-transfer ratio was similarly low
N-acetylaspartate14 and high myoinositol concen- in patients with Leber’s hereditary optic neuropathy,
trations.15 By contrast, small and preliminary studies on suggesting that axonal loss is likely to be an important
patients with CIS16,17 found no abnormalities in brain contributor to low magnetisation-transfer ratio in MS.
metabolites. More recently, one of us with other Further support for the idea that axonal loss is one of
researchers18 used a new unlocalised proton-magnetic- the pathological substrates of optic-nerve damage
resonance spectroscopy technique to show a decrease in comes from findings of nerve atrophy after an episode
the concentration of N-acetylaspartate in the whole brain of optic neuritis.26–28
in a group of 31 patients with a CIS. Fernando and co-
workers19 studied 96 patients with CIS within 6 months Recovery processes
of the clinical episode and found high myoinositol and Structural repair mechanisms and insights from MRI in
creatine in white matter, suggesting that glial optic neuritis
proliferation is an early event in MS (figure 1). The optic nerve, when studied both experimentally and
As found in patients with established MS,20 in patients clinically, serves as a window revealing general
with a CIS non-conventional MRI measures might be mechanisms of symptom onset and recovery equally
better than lesion load in reflecting clinical status. Notably, applicable to isolated and recurrent demyelination
patients with a CIS and cognitive impairment had a lower elsewhere in the CNS. Altered structure and function in
ratio of N-acetylaspartate to creatine and a lower brain the complex oligodendrocyte–axon unit should be part of
parenchymal fraction than patients with a CIS and any explanation of symptom onset and recovery.29 In
normal cognition, whereas there was no difference in T2- addition, given the capacity for CNS plasticity and
lesion and T1-lesion volumes between the two groups.21 functional adaptation, clinical recovery might not result
Patients with CIS in this study also had white-matter from structural repair within the primary lesion alone.
volume loss but not grey-matter volume loss. Acute cytokine release induces transient conduction
In about 20% of patients with MS, optic neuritis is block, probably caused by damage from nitric oxide.
the initial symptom of the disease. Although the With intact myelination and preserved axons, the
assessment of optic-nerve pathology has been recovery mechanism involves removal of inflammatory
hampered by several methodological factors (the small mediators and reversal of the functional deficit. More
size of the nerve, motion artifacts, and the effect of prolonged exposure impairs both structure and
surrounding CSF, lipids, and bony structures) it is function, leading to persistent demyelination and axonal
possible to obtain high quality T1-weighted and T2- damage. Recovery can occur through three mechanisms:
weighted images,22 as well as reliable quantitative remyelination, development of continuous conduction
data.23,24 In patients with MS, the magnetisation- through sodium channels that develop along the
transfer ratio is substantially lower in affected optic demyelinated segment, and cortical plasticity.
nerves than in unaffected ones. Furthermore, low Substantial axonal loss could occur and result in
magnetisation-transfer ratio is associated with delay in permanent loss of structure and function. Axon loss
the latency of visual evoked potentials.23,25 Inglese and might be acute, caused by the effects of inflammation, or

Choline
Choline
Total N-acetyl-aspartate
Myoinositol Total N-acetyl-aspartate
Creatine Myoinositol Creatine

Glutamate Glutamate

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 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
Chemical shift (parts per million) Chemical shift (parts per million)

Figure 1: T2-weighted MRI of a patient with a CIS

The square shows short echo-time proton spectra from a voxel of white-matter (left). Spectrum from the white-matter of a patient with a CIS (middle) and a healthy person (right). The peak in
myoinositol is larger in the patient than in the control. Reproduced with permission from Oxford University Press.19

342 http://neurology.thelancet.com Vol 4 June 2005

 Review

chronic, caused by persistent inflammation or lack of blood-oxygenation level dependent signal in response to
trophic support from myelin. binocular stimulation. Monocular stimulation of the
In patients with optic neuritis, recovery is related to affected eye resulted in cortical activated volumes and
MRI findings for both the location and extent of BOLD signal increases that were consistent with the
abnormality. Long lesions located in the optic canal have results of visual testing. These findings suggest that
been associated with poor visual recovery. In 101 of 107 transient adaptive changes happen in the visual cortex
patients with acute optic neuritis, assessment with during recovery.
gadolinium enhanced fat-suppression MRI found that Comparison of 20 patients who had acute unilateral
the affected optic nerve was enhanced but none of the optic neuritis with 46 healthy people over 1 year by use of
unaffected nerves were enhanced.30 MRI with fMRI found differences over time in the patients’ visual
gadolinium therefore seems to be sensitive in the cortex and extraoccipital regions—including the lateral
detection of acute optic neuritis. Although lesions in the temporal cortex, insula, corpus striatum, orbitofrontal
canal or over longer segments of the optic nerve were cortex, and inferior parietal region—when either the
predictive of poor vision at onset, they were not affected or unaffected eyes were used.34 These differences
predictive of recovery. Pain was common when optic were most prominent soon after clinical onset and
neuritis involved the orbital segment of the optic nerve.31 subsided after a few months. Markers of optic-nerve
Changes in magnetisation-transfer ratio over 1 year structure (cross-sectional area and gadolinium
were assessed with magnetisation-transfer MRI in 29 enhancement), clinical function, and the fMRI response
patients with acute optic neuritis and in healthy people.25 for both the affected and the unaffected eyes were
Whereas magnetisation-transfer ratios of the unaffected associated, particularly at baseline. Visual function at
optic nerve and of the optic nerves of healthy controls baseline was inversely related to optic-nerve damage and
were stable during follow-up, the diseased optic nerve directly related to the extent of response on fMRI. At
magnetisation-transfer ratio decreased over time and baseline and 1 month later, low fMRI responses (for both
was lowest at about 240 days. The magnetisation- affected and unaffected eyes) were associated with
transfer ratio of the diseased optic nerve seems to substantial optic-nerve damage, whereas at 3 months,
increase after reaching the lowest point, although not baseline optic-nerve damage was associated with the
significantly. Time-averaged magnetisation-transfer amount of fMRI activity outside the visual cortex. Two
ratios and mean visual-evoked-potential latency were studies33,34 suggest there are prominent dynamic
related, suggesting that magnetisation-transfer ratio is a spatiotemporal functional cortical changes after acute
measure of the structural integrity of the optic nerve optic neuritis, which might have an adaptive function
and, possibly, the extent of myelination. A decrease in during early recovery.
optic-nerve magnetisation-transfer ratio early in the fMRI has also been used to assess functional cortical
disease process is consistent with demyelination and changes associated with motor tasks in patients with a
Wallerian degeneration, whereas occurrence of the CIS.35,36 When patients do a simple motor task with the
lowest magnetisation-transfer ratio late in the disease dominant hand, cortical activations in the “classic”
process might be caused by slow clearance of myelin motor areas differ,35 suggesting that cortical
debris; a subsequent rise in magnetisation-transfer ratio reorganisation might occur early in MS. The extent of
could be due to remyelination. these cortical activations is strongly associated with the
Tissue volume of optic nerves changes for up to 1 year concentration of N-acetylaspartate in brain tissue. In
after the onset of optic neuritis.28 At first, nerves swell; patients with a CIS, there is pronounced use of the
the cross-sectional area of the nerve increases by a contralateral primary sensorimotor cortex when the
median of 20%. Consistent with acute inflammation, same motor task was done with the non-dominant
swelling resolves over several months. Atrophy then hand.36 The findings are similar when these patients do
starts, with a mean decrease of 12% in nerve area after the task with the dominant foot and there is an anterior
1 year. The mean loss of nerve tissue after a single attack shift of the centre of activation.36 During a complex
of optic neuritis is small, which suggests that only a few motor task involving the dominant hand and foot,
optic nerve axons are lost. Thus, patients’ generally good patients with a CIS have an increased recruitment of a
visual outcome is not surprising. widespread network (including the frontal lobe, the
insula, and the thalamus), thought to commonly
Cortical adaptation and insights from functional MRI function in motor, sensory, and multimodal integration
One of the factors that might contribute to visual processing.36 The effect of isolated spinal-cord disease on
recovery after acute optic neuritis is cortical adaptation.32 cortical adaptation has been assessed through study of a
Study of 12 patients with acute monosymptomatic optic group of patients with a previous episode of acute
neuritis found a low functional MRI (fMRI) response, myelitis.37 Increased activation of several cortical regions
reflecting low input during acute visual loss.33 However, of the “classic” motor network was found, and the
5 weeks after onset of optic neuritis, there was a activity of some of these areas was related to the severity
significant increase in the volume of activation and in of spinal-cord damage.

http://neurology.thelancet.com Vol 4 June 2005 343

 Review

In a 1 year follow-up study of patients with a CIS, those functionally disabling or when the patient is not
who developed clinically definite MS had a different improving spontaneously. The CIS that has been most
fMRI response on motor tasks at first presentation than carefully studied in randomised-controlled trials is optic
those who did not (figure 2).38 By comparison with those neuritis. The Optic Neuritis Treatment Trial (ONTT)
who did not develop MS, those who did had substantial showed that treatment with a 3 day course of
bilateral activations of the superior frontal sulcus, the intravenous methylprednisolone and then 11 days of
superior frontal gyrus, the infraparietal sulcus, and the oral prednisone was associated with a more rapid
putamen; in addition, they had high activation of the recovery of vision compared with placebo.39 After 1 year,
ipsilateral middle frontal gyrus, superior temporal gyrus, vision was not different between the group given an
cuneus, and contralateral fusiform gyrus. Those who did active treatment and that given a placebo.40 A 2 week
not develop MS had more substantial activations than course of oral prednisone alone did not accelerate
those who did develop MS in the contralateral primary recovery compared with placebo, but was associated with
somatomotor cortex and supplementary motor area, the an increased probability of having recurrent episodes of
ipsilateral paracentral lobule, and the cerebellar optic neuritis.
hemisphere. These results suggest that activation of the The Quality Standards Subcommittee of the American
regions normally involved in a task seems to be a Academy of Neurology concluded: “Higher dose oral or
favourable prognostic factor, whereas a widespread parenteral methylprednisolone or ACTH [adrenocortico-
recruitment of additional areas seems to be associated trophic hormone] may hasten the speed and degree of
with short-term disease activity. recovery of visual function in persons with acute
Overall, fMRI studies of patients with a CIS suggest monosymptomatic optic neuritis. There is, however, no
that cortical reorganisation occurs and has the potential evidence of long-term benefit for visual function. The
to improve long-term patient outcomes, even at this decision to use these medications to speed recovery but
early stage of the disease process. not to improve ultimate visual outcome should therefore
be based on other non-evidence-based factors such as
Management quality of life, risk to the patient, visual function in the
Acceleration of recovery from the CIS unaffected eye, or other factors that the clinician deems
Many patients with a mild CIS will recover appropriate”.41
spontaneously without the need for specific treatment. Plasma exchange has been assessed in a small trial
Corticosteroids are used when the symptoms are that included a “sham-exchange” control in patients with
a severe neurological deficit due to an episode of CNS
A B inflammation that did not improve after conventional
treatment with intravenous methylprednisolone.42 Eight
(42%) of 19 patients in the treatment arm improved
compared with one (6%) of 17 patients in the control
arm. Improvement was most likely if plasma exchange
was given within 2 weeks of presentation with a severe
neurological deficit. This finding was supported by the
results of subsequent use of plasma exchange in a non-
trial setting; about half of the patients substantially
improved in the days and weeks after treatment.43
Plasma exchange is likely to be beneficial only when the
patient acquires a severe neurological deficit from a
C D single acute inflammatory CNS lesion, not if many
25 lesions cause the deficit to accumulate in small steps.
Response to plasma exchange might reflect an antibody-
20 mediated pathogenesis of the underlying disease, and
this is supported by IgG and complement deposition in
15
the lesions of responsive patients.44
A recent double-blind, placebo-controlled trial of
10
intravenous immunoglobulin in 68 patients with acute
optic neuritis showed no effect of treatment on vision
5
during 6 months of follow-up.45
0
Treatments to delay the development of MS
Figure 2: Pattern of cortical activations during a simple motor task
Although the ONTT did not aim to assess whether
The dominant, functionally unaffected right hand is used in patients with CISs suggestive of MS who developed
definite MS over a short period of follow-up (top) and those who did not (bottom). Activation of the sensorimotor treatment delayed conversion to MS, patients treated
network is most extensive and widespread in those who developed definite MS. with intravenous methylprednisolone and oral

344 http://neurology.thelancet.com Vol 4 June 2005

 Review

prednisone had a lower risk of developing clinically the individual patient. Three issues are generally relevant:
definite MS over the next 2 years than patients given a the existing chance of conversion to clinically definite MS
placebo (9% vs 17%).46 However, after the third year of or development of disability, the usefulness of the
follow-up the treatment and placebo groups’ risks of diagnostic test, and the effectiveness of treatment.
conversion to clinically definite MS did not differ.47 Evidence is accumulating that treatment may delay the
Two randomised, double-blind, placebo-controlled onset of clinically definite MS, although data about the
trials of interferon beta-1a in patients with a CIS and an effect on disability are lacking. One study suggests early
abnormal MRI brain scan found an association between treatment might decrease cerebral atrophy,5 which is
treatment with interferon beta-1a and a delay in the notable because cerebral atrophy could signify disability.54
development of clinically definite MS.48,49 Most clinicians who see patients with a CIS or suspected
Subgroup analysis in the ETOMS trial found that early MS would now agree that it is important to diagnose
conversion to clinically definite MS was more common early, and having done so, to weigh-up the potential
as the number of Barkhof MRI criteria met increased, benefits, risks, and uncertainties of disease modifying
and treatment was more likely to delay conversion to MS treatment, while ensuring the patient is fully informed
in those who fulfilled many of the MRI criteria. Thus, in and participates in the decision-making process.
patients who met one or two Barkhof criteria at
presentation, clinically definite MS developed in five When to diagnose MS
(23%) of 22 who took placebo and in four (21%) of 19 A single contrast-enhanced scan from a patient who
who took interferon beta-1a; in those who met three or presents with a CIS could be sufficient to diagnose MS,
four criteria, these figures were 64 (49%) of 132 for some have argued, because the presence of both
placebo and 48 (35%) of 135 for interferon beta-1a.50 In enhancing and non-enhancing lesions strongly suggests
the Controlled High Risk Subjects Avonex MS simultaneous dissemination in space and time.55
Prevention Study (CHAMPS), development of clinically However, MRI criteria from the scan at presentation
definite MS was more common in patients with have been only modestly reliable in the prediction of
gadolinium enhancing lesions at presentation;51 the clinically definite MS,56,57 and monophasic acute
treatment effect was also larger in this patient subgroup. disseminated encephalomyelitis—an important,
Follow-up in the ETOMS trial and CHAMPS was not although less common, clinical and radiological
long enough to assess whether treatment delayed the differential diagnosis in this setting—can not be
development of irreversible disability. Nevertheless, in excluded. An alternative strategy is to require follow-up
the ETOMS study, a low, weekly dose of interferon beta- MRI scans to show dissemination in time by the
1a reduced the brain tissue lost in 2 years by about 30%.5 development of new, subclinical lesions at least
Findings from longer follow-up (over 5 years) of some of 3 months after the CIS presents, as recommended by the
the patients in the CHAMPS study suggest that the McDonald criteria,3 especially if a new T2 lesion is
disease course was improved in those who received included at any time after 3 months of follow-up.58 This
immediate treatment with interferon beta-1a.52 However, approach is not likely to delay diagnosis for long—with
about 30% of patients were lost to follow-up. these criteria about 50% of patients with a CIS are
More frequent treatment with interferon beta-1b diagnosed as having MS within 1 year. Use of MRI to
(8 mU on alternate days) is now being trialled in patients assess dissemination in space and time—which we
with a CIS; the primary endpoint of this study, like that favour (panel 1)—will also reduce diagnostic and
of earlier CIS trials, is clinically definite MS. In addition, treatment errors from undue hastiness, and this
patients are being recruited for an extension study approach is supported by prospective group studies that
designed to compare immediate and delayed interferon- show a high specificity for clinically definite MS.56,57
beta treatment and also to investigate the development A more conservative approach to diagnosis would be to
of disability. require the traditional criteria for clinical dissemination in
A trial of glatiramer acetate is also in progress in space and time. Delaying the diagnosis might further
patients with CISs. Intravenous immunoglobulin has reduce the risk of false-positive diagnosis—eg, in patients
been studied in a placebo-controlled trial in 91 patients with recurrent cerebrovascular insults mistaken for
presenting with a first episode of suspected demyelination. However, there are disadvantages to a
demyelination; after 1 year of active treatment patients conservative approach: patients who clearly have MS are
were less likely to develop clinically definite MS; denied the diagnosis; and delayed diagnosis increases the
however, the short follow-up and small size of this study risk of cerebral atrophy and disability, particularly in high-
limit the interpretation.53 risk patients (eg, with a CIS that is severe clinically and has
many cerebral lesions on MRI) in whom the treatment
Diagnosis and treatment of individual patients benefit may be greater in the short term. 50
The questions coming from group studies of diagnosis, An early MRI-assisted diagnosis might confront
natural history, and treatment trials are how, and when, patients with the implications of a serious, chronic
these studies’ findings should affect the management of disease, and although these patients might not have been

http://neurology.thelancet.com Vol 4 June 2005 345

 Review

treatment be offered? Prospective follow-up studies are

Panel 1: Merits of MRI assessment of dissemination in needed to address these issues.
space and time for diagnosis of MS in patients with a CIS Whether treatment after a first attack has any greater
Facilitates open discussion of potential for MS and appropriate effect than delaying treatment until a second episode has
assessment with MRI to clarify prognosis (abnormal MRI=high occurred is unknown. The traditional criterion for
risk; normal MRI=low risk) or to make the diagnosis treatment in many countries is a second episode. MRI
lesion load at presentation62,68 and the increase in lesion
Early MRI-assisted diagnosis with existing criteria3 is highly
load over the next 5 years62 is associated with future
specific for clinically definite MS56,57
disability. Because interferon beta reduces the
Making the diagnosis of MS can reduce the patient’s anxiety accumulation of new MRI lesions, the potential for
about the unknown diagnosis and facilitate support from MS benefit may be greater if the drug is given earlier.
nurses A balance between treating all patients with a CIS and
an abnormal MRI or none of these patients seems
Opportunity to introduce an education programme at an
sensible. Treatment could be offered to those patients
early stage so that patients can think about the implications
who have had a clinical episode of moderate or substantial
of diagnosis and participate in establishing a management
severity—particularly if multifocal or associated with poor
plan that incorporates decisions on lifestyle, work, and family
recovery—and are diagnosed with MS on the basis of
Opportunity for patient and physician to discuss use of MS early MRI dissemination in space and time, noting that
disease-modifying treatment before more CNS tissue such patients have a high risk of relapse.56–58 We find this
damage occurs approach to be satisfactory (panel 2); it seems acceptable
to begin treatment after MS is diagnosed.

diagnosed until some years later with the old diagnostic Summary and future research
criteria, surveys consistently report a reduction in anxiety Important progress has been made in the study of CISs.
when the diagnosis is made.59–61 In the present era of The relation between CIS and MS has been elucidated
openness in clinical practice, it seems inappropriate to and diagnostic advances now provide a reliable
withold information on the likelihood of developing MS prediction, soon after presentation, of the risk of MS.
from patients with a CIS, even when formal criteria for The mechanisms of acute dysfunction, subsequent
diagnosis are not fulfilled. The potential diagnosis has recovery, and persistent deficit from the CIS are better
important practical implications for patients, including understood. In addition to multifocal white-matter
lifestyle decisions and access to life insurance. lesions, abnormalities are present in the normal-
appearing white and grey matter, and progressive brain
When to start disease-modifying treatments atrophy—suggesting neuroaxonal loss—develops early
Although there is strong evidence that patients with a CIS in those who develop MS. Tools for treatment
and an abnormal MRI are likely to develop clinically monitoring—both clinical and paraclinical—provide a
definite MS, and that interferon beta delays the more efficient and comprehensive assessment of
development of clinically definite disease, the long-term therapies that target inflammation, axonal protection,
effect of the treatment is unknown. Many patients with a and remyelination than was possible 10 years ago.
CIS and MRI lesions develop MS but have a benign The potential for a better understanding of pathogenetic
course, with little or no disability for the next mechanisms and effective disease-modifying treatments
10–14 years.62,63 Moreover, not all of these patients will
relapse further. In the ONTT, 44% had not developed
clinically definite MS after 10 years.64 Interferon-beta Panel 2: Reasons for and against the use of disease-
treatment of all patients with a CIS who have an abnormal modifying treatments in patients with a CIS
scan would include many patients who would have had a For Against
good prognosis without treatment. Old65,66 and more Clinically severe CIS with Clinically mild CIS with recovery
recent67 natural history studies have shown that many persistent disability
patients experience a benign, long-term course of MS,
MRI criteria for MS fulfilled Normal MRI or few lesions
especially if they are free of disability 5 years after onset of
(dissemination in space
CIS. Together these findings suggest that an initial
and time)
approach of wait-and-see for the use of disease-modifying
treatment will serve the best interests of patients with CIS Delay time to next relapse Not known if disability is
in whom the clinical episode was mild and reversible, prevented in the long term
even if accompanied by an abnormal scan. However, the Early treatment offers Early treatment will include
best follow-up of such patients is unknown. For example, better chance of long- patients who would otherwise
should these patients have repeat MRI studies, and if term benefit have had a benign long-term course
these show new lesions, should disease-modifying

346 http://neurology.thelancet.com Vol 4 June 2005

 Review

References
Search strategy and selection criteria 1 Miller DH, Barkhof F, Frank JA, Parker GJ, Thompson AJ.
Measurement of atrophy in multiple sclerosis: pathological basis,
References for this review were identified by searches of the methodological aspects, and clinical relevance. Brain 2002;
125: 1676–95.
authors’ files; from presentations at the MAGNIMS Workshop
2 Dalton CM, Brex PA, Jenkins R, et al. Progressive ventricular
(London, January 2004; see acknowledgments); access to enlargement in patients with clinically isolated syndromes is
PubMed citations between 1985 and December 2004; and associated with the early development of multiple sclerosis.
J Neurol Neurosurg Psychiatry 2002; 73: 141–47.
references from relevant articles. A few abstracts and reports
3 McDonald WI, Compston A, Edan G, et al. Recommended diagnostic
from the MAGNIMS meeting were also included. The final criteria for multiple sclerosis: guidelines from the International Panel
reference list was generated from papers that were original on the Diagnosis of Multiple Sclerosis. Ann Neurol 2001; 50: 121–27.
and relevant to the topics covered in the review. 4 Dalton CM, Chard DT, Davies GR, et al. Early development of
multiple sclerosis is associated with progressive grey matter atrophy
in patients presenting with clinically isolated syndromes. Brain 2004;
127: 1101–07.
are compelling reasons for continuing to assess patients 5 Filippi M, Rovaris M, Inglese M, et al. Interferon beta-1a for brain
with a CIS. Diagnostic criteria will likely evolve as new atrophy tissue loss in patients at presentation with syndromes
suggestive of multiple sclerosis: a randomised, double-blind, placebo-
data emerges from imaging and CSF research studies; controlled trial. Lancet 2004; 364: 1489–96.
other plausible biomarkers should also be assessed for 6 Brex PA, Leary SM, O’Riordan JI, et al. Measurement of spinal cord
their diagnostic use. Prediction of disability and the long- area in clinically isolated syndromes suggestive of multiple sclerosis.
J Neurol Neurosurg Psychiatry 2001; 70: 544–47.
term course of MS are limited: combinations of 7 Filippi M, Bozzali M, Horsfield MA, et al. A conventional and
conventional and non-conventional imaging and other magnetization transfer MRI study of the cervical cord in patients with
biomarker measures may be informative and long-term MS. Neurology 2000; 54: 207–13.
8 Iannucci G, Tortorella C, Rovaris M, Sormani MP, Comi G,
prospective studies will be needed to establish convincing Filippi M. Prognostic value of MR and magnetization transfer
association between laboratory and clinical findings. A imaging findings in patients with clinically isolated syndromes
better understanding of early pathogenetic mechanisms suggestive of multiple sclerosis at presentation. Am J Neuroradiol
2000; 21: 1034–38.
would come with access to pathological tissue from 9 Kaiser JS, Grossman RI, Polansky M, Udupa JK, Miki Y, Galetta SL.
patients with a CIS, but because this is rarely available, Magnetization transfer histogram analysis of monosymptomatic
greater reliance will inevitably be placed on in vivo episodes of neurologic dysfunction: preliminary findings.
Am J Neuroradiol 2000; 21: 1043–47.
surrogate measures. New measures should be sought and 10 Brex PA, Leary SM, Plant GT, Thompson AJ, Miller DH.
verified as a high priority for their cellular, pathological, Magnetization transfer imaging in patients with clinically isolated
and immunological specificity. syndromes suggestive of multiple sclerosis. Am J Neuroradiol 2001;
22: 947–51.
Acknowledgments 11 Traboulsee A, Dehmeshki J, Brex PA, et al. Normal-appearing brain
Content for this review partly comes from presentations made at an tissue MTR histograms in clinically isolated syndromes suggestive of
International Workshop on Clinically Isolated Syndromes held in London, MS. Neurology 2002; 59: 126–28.
UK, in January 2004. The workshop was coordinated by MAGNIMS, a 12 Rovaris M, Gallo A, Riva R, et al. An MT MRI study of the cervical
European network of investigators who are interested in the application of cord in clinically isolated syndromes suggestive of MS.
magnetic resonance methods to MS. Sponsorship support was received Neurology 2004; 63: 584–85.
from the MS Society of Great Britain and Northern Ireland, the Italian MS 13 Gallo A, Rovaris M, Riva R, et al. Diffusion tensor MRI detects
Society, and the National MS Society (US). Workshop participants were: normal-appearing white matter damage unrelated to short-term
O Andersen (Gothenburg), D Arnold (Montreal), F Barkhof (Amsterdam), disease activity in patients at the earliest clinical stage of multiple
T Berger (Innsbruck), L Bo (Amsterdam), P Brex (London), A Compston sclerosis. Arch Neurol (in press).
(Cambridge), C Confavreux (Lyons), C Dalton (London), M Daumer 14 Arnold DL, Wolinsky JS, Matthews PM, Falini A. The use of
(Munich), F Fazekas (Graz), K Fernando (London), M Filippi (Milan), magnetic resonance spectroscopy in the evaluation of the natural
J Frederiksen (Copenhagen), S Hickman (London), R Hohlfeld (Munich), history of multiple sclerosis. J Neurol Neurosurg Psychiatry 1998;
64 (suppl 1): S94–101.
M Inglese (New York), M Johnson (Leeds), L Kappos (Basel),
M Kupersmith (New York), C Lucchinetti (Rochester, Minnesota), 15 Chard DT, Griffin CM, McLean MA, et al. Brain metabolite changes
in cortical grey and normal-appearing white matter in clinically early
G Lycklama (Amsterdam), H McFarland (Washington DC), P Matthews
relapsing-remitting multiple sclerosis. Brain 2002; 125: 2342–52.
(Oxford), D Miller (London), X Montalban (Barcelona), G Plant (London,
16 Tourbah A, Stievenart JL, Abanou A, et al. Normal-appearing
England), C Polman (Amsterdam), S Reingold (New York), H Roed
white matter in optic neuritis and multiple sclerosis:
(Copenhagen), M Rocca (Milan), M Ron (London), A Rovira (Barcelona), a comparative proton spectroscopy study. Neuroradiology 1999;
M Rovaris (Milan), M Sandberg-Wollheim (Lund), J Simon (Denver), 41: 738–43.
F Sellebjerg (Copenhagen), A Thompson (London), M Tintoré (Barcelona), 17 Brex PA, Gomez-Anson B, Parker GJ, et al. Proton MR spectroscopy
A Toosy (London), B Weinshenker (Rochester, Minnesota), and J in clinically isolated syndromes suggestive of multiple sclerosis.
Wolinsky (Houston). All authors are members of the MAGNIMS Steering J Neurol Sci 1999; 166: 16–22.
Committee. 18 Filippi M, Bozzali M, Rovaris M, et al. Evidence for widespread axonal
Authors’ contributions damage at the earliest clinical stage of multiple sclerosis. Brain 2003;
126: 433–37.
All authors contributed equally to the writing of this review.
19 Fernando KTM, McLean MA, Chard DT, et al. Elevated white matter
Conflicts of interest myo-inositol in clinically isolated syndromes suggestive of multiple
We have no conflicts of interest. sclerosis. Brain 2004; 127: 1361–69.
20 Miller DH, Thompson AJ, Filippi M. Magnetic resonance studies of
Role of the funding source
abnormalities in the normal appearing white matter and grey matter
No funding source was involved in the preparation of this review or in in multiple sclerosis. J Neurol 2003; 250: 1407–19.
the decision to submit it for publication.
21 Arevalo M, Rovira A, Porcel J, et al. Cognitive performance related to
brain MR imaging and MR spectroscopy in clinically isolated
syndromes. Mult Scler 2004; 10: S222

http://neurology.thelancet.com Vol 4 June 2005 347

 Review

22 Moseley IF, Miller DH, Gass A. The contribution of magnetic 45 Roed HG, Langkilde A, Sellebjerg F, et al. The effect of IVIG
resonance imaging to the assessment of optic nerve and spinal cord treatment on visual outcome after acute optic neuritis (ON): a double-
involvement in multiple sclerosis. J Neurol Neurosurg Psychiatry 1998; blind, randomized, placebo-controlled trial. Neurology 2004;
64 (suppl 1): S15–20. 62 (suppl 5): A261.
23 Thorpe JW, Barker GJ, Jones SJ, et al. Magnetisation transfer ratios 46 Beck RW, Cleary PA, Trobe JD, et al. The effect of corticosteroids for
and transverse magnetisation decay curves in optic neuritis: acute optic neuritis on the subsequent development of multiple
correlation with clinical findings and electrophysiology. sclerosis. The Optic Neuritis Study Group. N Engl J Med 1993;
J Neurol Neurosurg Psychiatry 1995; 59: 487–92. 329: 1764–69.
24 Inglese M, Ghezzi A, Bianchi S, et al. Irreversible disability and 47 Beck RW. The optic neuritis treatment trial: three-year follow-up
tissue loss in multiple sclerosis: a conventional and magnetization results. Arch Ophthalmol 1995; 113: 136–37.
transfer magnetic resonance imaging study of the optic nerves. 48 Jacobs L, Beck R, Simon J, et al. Intramuscular interferon beta-1a
Arch Neurol 2002; 59: 250–55. therapy initiated during the first demyelinating event in multiple
25 Hickman SJ, Toosy AT, Jones SJ, et al. Serial magnetization transfer sclerosis. N Engl J Med 2000; 343: 898–904.
imaging in acute optic neuritis. Brain 2004; 127: 692–700 49 Comi G, Filippi M, Barkhof F, et al. Effect of early interferon
26 Hickman SJ, Brex PA, Brierley CM, et al. Detection of optic nerve treatment on conversion to definite multiple sclerosis. Lancet 2001;
atrophy following a single episode of unilateral optic neuritis by MRI 357: 1576–82.
using a fat-saturated short-echo fast FLAIR sequence. Neuroradiology 50 Barkhof F, Rocca M, Francis G, et al. Validation of diagnostic
2001; 43: 123–28. magnetic resonance imaging criteria for multiples sclerosis and
27 Hickman SJ, Brierley CM, Brex PA, et al. Continuing optic nerve response to interferon beta-1a. Ann Neurol 2003; 53: 718–24.
atrophy following optic neuritis: a serial MRI study. Mult Scler 2002; 51 CHAMPS Study Group. MRI predictors of early conversion to
8: 339–42. clinically definite MS in the CHAMPS placebo group. Neurology
28 Hickman SJ, Toosy AT, Jones SJ, et al. A serial magnetic resonance 2002; 59: 998–1005.
imaging study following optic nerve mean area in acute optic 52 Kinkel RP, Kollman C, Glassman A, et al. Interferon beta-1a (avonex)
neuritis. Brain 2004; 127: 2498–505. delays the onset of clinically definite MS over 5 years of treatment:
29 Compston A, Coles A. Multiple sclerosis. Lancet 2002; 359: 1221–31. results from CHAMPIONS study. Neurology 2004;
30 Kupersmith MJ, Alban T, Zeiffer B, Lefton D. Contrast-enhanced 62 (suppl 5): A261–62.
MRI in acute optic neuritis: relationship to visual performance. Brain 53 Achiron A, Kishner I, Sarova-Pinhas I, et al. Intravenous
2002; 125: 812–22. immunoglobulin treatment following the first demyelinating event
31 Fazzone HE, Lefton DR, Kupersmith MJ. Optic neuritis: correlation suggestive of multiple sclerosis: a randomized, double-blind, placebo-
of pain and magnetic resonance imaging. Ophthalmology 2003; controlled trial. Arch Neurol 2004; 61: 1515–20.
110: 1646–49. 54 Miller DH. Brain atrophy, -interferon, and treatment trials in
32 Werring DJ, Bullmore ET, Toosy AT, et al. Recovery from optic multiple sclerosis. Lancet 2004; 364: 1463–64.
neuritis is associated with a change in the distribution of cerebral 55 Tas MW, Barkhof F, van Walderveen MA, Polman CH,
response to visual stimulation: a functional magnetic resonance Hommes OR, Valk J. The effect of gadolinium on the sensitivity and
imaging study. J Neurol Neurosurg Psychiatry 2000; 68: 441–49. specificity of MR in the initial diagnosis of multiple sclerosis.
33 Langkilde AR, Frederiksen JL, Rostrup E, Larsson HB. Functional Am J Neuroradiol 1995; 2: 259–64.
MRI of the visual cortex and visual testing in patients with previous 56 Tintoré M, Rovira A, Rio J, et al. New diagnostic criteria for multiple
optic neuritis. Eur J Neurol 2002; 9: 277–86. sclerosis: application in first demyelinating episode. Neurology 2003;
34 Toosy AT, Hickman SJ, Plant GT, et al. Adaptive cortical reorgani- 60: 27–30.
zation in acute optic neuritis: longitudinal functional magnetic 57 Dalton CM, Brex PA, Miszkiel KA, et al. Application of the new
resonance imaging study. Ann Neurol 2004; 56 (suppl 8): S58. McDonald criteria to patients with clinically isolated syndromes
35 Rocca MA, Mezzapesa DM, Falini A, et al. Evidence for axonal suggestive of multiple sclerosis. Ann Neurol 2002; 52: 47–53.
pathology and adaptive cortical reorganization in patients at 58 Dalton CM, Brex PA, Miszkiel KA, et al. New T2 lesions to enable an
presentation with clinically isolated syndromes suggestive of multiple earlier diagnosis of multiple sclerosis in clinically isolated
sclerosis. Neuroimage 2003; 18: 847–55. syndromes. Ann Neurol 2003; 53: 673–76.
36 Filippi M, Rocca MA, Mezzapesa DM, et al. Simple and complex 59 Mushlin AI, Mooney C, Grow V, et al. The value of diagnostic
movement-associated functional MRI changes in patients at information to patients with suspected multiple sclerosis.
presentation with clinically isolated syndromes suggestive of multiple Arch Neurol 1994; 51: 67–72.
sclerosis. Hum Brain Mapp 2004; 21: 108–17. 60 Strasser-Fuchs S, Fazekas F, Flooh E, Kapeller P, Poltrum B,
37 Rocca MA, Mezzapesa DM, Ghezzi A, et al. Cord damage elicits Offenbacher H. Die Einstellung von Patienten mit multipler Sklerose
brain functional reorganization after a single episode of myelitis. zur Krankheitsaufklärung. Nervenarzt 1997; 68: 963–66.
Neurology 2003; 61: 1078–85. 61 Heesen C, Kolbeck J, Gold SM, et al. Delivering the diagnosis of
38 Rocca MA, Mezzapesa DM, Ghezzi A, et al. A widespread pattern of MS—results of a survey among patients and neurologists.
cortical activations in patients at presentation with CIS is associated Acta Neurol Scand 2003; 107: 363–68.
with evolution to definite MS. Am J Neuroradiol (in press). 62 Brex PA, Ciccarelli O, O’Riordan JI, et al. A longitudinal study of
39 Beck RY, Cleary PA, Anderson MM, et al. A randomised controlled abnormalities on MRI and disability from multiple sclerosis.
trial of corticosteroids in the treatment of acute optic neuritis. N Engl J Med 2002; 346: 158–64.
N Engl J Med 1992; 326: 581–88. 63 Optic Neuritis Study Group. Neurologic impairment 10 years after
40 Beck RW, Cleary PA. Optic neuritis treatment trial: one-year follow- optic neuritis. Arch Neurol 2004; 61: 1386–89.
up results. Arch Ophthalmol 1993; 111: 773–75. 64 Optic Neuritis Study Group. High- and low-risk profiles for the
41 Kaufman DI, Trobe JD, Eggenberger ER, Whitaker JN. Practice development of multiple sclerosis within 10 years after optic neuritis.
parameter: the role of corticosteroids in the management of acute Arch Ophthalmol 2003; 121: 944–49.
monosymptomatic optic neuritis. Report of the Quality Standards 65 Kurtzke JF, Beebe GW, Nagler B, Kurland LT, Auth TL. Studies on
Subcommittee of the American Academy of Neurology. Neurology the natural history of multiple sclerosis, 8: early prognostic features
2000; 54: 2039–44. of the later course of the illness. J Chronic Dis 1977; 30: 819–30.
42 Weinshenker BG, O’Brien PC, Petterson TM, et al. A randomised 66 Miller DH, Hornabrook RW, Purdie G. The Natural History of
trial of plasma exchange in acute central nervous system multiple sclerosis: a regional study with some longitudinal data.
inflammatory demyelinating disease. Ann Neurol 1999; 46: 878–86. J Neurol Neurosurg Psychiatry 1992; 55: 341–46.
43 Keegan M, Pineda AA, McClelland RL, Darby CH, Rodriguez M, 67 Pittock SJ, McClelland RL, Mayr WT, et al. Clinical implications of
Weinshenker BG. Plasma exchange for severe attacks of CNS benign multiple sclerosis: a 20-year population-based follow-up
demyelination: predictors of response. Neurology 2002; 58: 143–46. study. Ann Neurol 2004; 56: 303–06.
44 Keegan M, Konig F, Bitsch A, et al. Multiple sclerosis pathological 68 Minneboo A, Barkhof F, Polman CH, et al. Infratentorial lesions
subtype predicts response to therapeutic plasma exchange. Neurology predict long term disability in patients with initial findings suggestive
2004; 62 (suppl 5): A259–60. of multiple sclerosis. Arch Neurol 2004; 61: 217–21.

348 http://neurology.thelancet.com Vol 4 June 2005