The Role of MRI in the Diagnosis of Multiple Sclerosis

Multiple sclerosis (MS) is the most commoninflammatory demyelinating disease of theCNS and the most frequent cause of nontraumaticneurological disability in young andmiddle-aged adults.¹ Women are twice as likelyto be affected as men, and onset typicallyoccurs between the ages of 20 and 40 years.

Variability and diversity characterize MS. Deficits in sensory, motor, cerebellar, brain stem, and autonomic functions are the most common clinical manifestations. Most patients (85%) experience a relapsing-remitting course characterized by the episodic onset of symptoms followed by residual deficits or by a full recovery within a few weeks, especially in the early stages of the disease.²

Approximately 15% of patients with relapsing-remitting MS (RRMS) will remain clinically stable or nearly stable for 2 or more decades (benign MS). However, within 25 years, RRMS in most untreated patients will evolve into secondary progressive MS, characterized by a chronic and steady increase of physical symptoms and disability.

Approximately 10% to 15% of patients with MS experience a primary progressive course. Primary progressive MS (PPMS) differs from RRMS in that it affects both men and women equally, occurs in older persons, exhibits lower levels of inflammatory markers and myelopathological features, and is unresponsive to immunomodulatory agents.³ Progressive relapsing MS is uncommon and is progressive from the onset with clear, acute relapses with or without recovery, and with periods between relapses characterized by continuing progression.⁴

DIAGNOSIS OF MS

MS is a clinical diagnosis, dependent on a detailed history; careful neurological examination; and supportive paraclinical investigations, including MRI, analysis of both cerebrospinal fluid (CSF) and evoked potentials, and blood tests to exclude confounding diagnoses. The classic MS diagnostic criteria are evidence of lesions in the CNS disseminated in space and time (ie, more than 1 clinical episode involving more than 1 area of the CNS [brain, spinal cord, and optic nerves]).

The use of MRI, since its introduction by Young and colleagues,⁵ has had a major impact on allowing early and more precise diagnosis of the disease. In patients with clinically definite MS, brain MRI reveals multifocal cerebral white matter lesions in more than 95% of patients and focal spinal cord lesions in 75% to 85% of them. Cerebral white matter lesions that are indistinguishable from those seen in definite MS are observed in about two thirds of patients who have a single episode of suspected demyelination or clinically isolated syndrome (CIS).⁶ Because the presence of such lesions increases the likelihood of developing clinically definite MS, it is not surprising that standard MRI features suggesting dissemination of pathologic lesions in space and time were incorporated into the diagnostic criteria for MS by the International Panel on the Diagnosis of Multiple Sclerosis in 2001.⁷

THE ROLE OF MRI

The previous diagnostic criteria for MS, authored by Poser⁸ in 1987, were established for use in clinical trials of MS and included clinically definite MS, laboratory supported (ie, evoked potentials and CSF analysis, including IgG index and oligoclonal bands) definite MS, probable MS (either clinically or laboratory supported), and possible MS. Because MRI was relatively new at the time these criteria were presented, the technology was included as a paraclinical element but was not further defined.

According to the new McDonald criteria, the diagnosis of MS requires objective evidence of lesions disseminated in space and time.

• MRI findings may contribute to the determination of dissemination in time or space.
• Other supportive investigations include evaluation of CSF and the visual evoked potential (VEP).
• Diagnostic categories include possible MS, MS, or not MS.

For dissemination in space, 3 of the following 4 features are required, based on criteria established by Barkhof and colleagues⁶ and Tintore and colleagues⁹:

• At least 1 gadolinium-enhancing lesion or 9 T2-weighted hyperintense lesions.
• At least 1 infratentorial lesion.
• At least 1 juxtacortical lesion.
• At least 3 periventricular lesions.

A spinal cord lesion can substitute for any of the above brain lesions. If immunoglobulin abnormalities are detected in the CSF, then the MRI criteria are relaxed to require that only 2 T2-weighted lesions typical of MS be present.

For dissemination in time, a new gadolinium-enhancing lesion appearing 3 or more months after the initial clinical event indicates a new CNS inflammatory event (the duration of gadolinium enhancement in MS is usually less than 6 weeks). If no gadolinium- enhancing lesions are detected but a new T2-weighted lesion is, then MRI should be repeated after another 3 months to detect the presence of a new T2-weighted lesion or a gadolinium-enhancing lesion.

Application of these criteria in several natural history and treatment trial cohorts indicated that they were more robust in facilitating an earlier diagnosis of MS than could be achieved through other diagnostic methods.^10-12 The criteria also proved robust in predicting the likelihood of conversion to clinically definite MS in patients with a CIS in whom MRI ultimately showed evidence of dissemination in space and time.^10-12 Specificity was high, in particular when dissemination in time was present; dissemination in space per se was less specific.

The requirement that a gadolinium-enhancing lesion be present 3 months after the initial event to fulfill dissemination in time had poor sensitivity. This was ameliorated by requiring that either a new gadolinium-enhancing or T2-weighted lesion be present instead.¹³

In the light of subsequent studies, and in view of criticism from the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology (AAN), the 2001 criteria were revised by a reconvened International Panel in 2005 (Tables 1, 2, and 3).¹⁴ The AAN subcommittee had recommended that the presence of 1 to 3 lesions was sufficient evidence for an MS diagnosis.¹⁵

The criteria published in 2001 and in 2005 draw on the criteria established by Barkhof and colleagues⁶ and Tintore and colleagues.⁹ The criteria differ in what extent a spinal cord lesion can assist in fulfillment of dissemination in space. In the 2001 McDonald criteria, only 1 cord lesion could substitute for a brain lesion, whereas in the 2005 criteria, any number of cord lesions can substitute for brain lesions and a cord lesion is also assigned the same status as an infratentorial lesion.

Although the MRI features of MS in children are similar to those of adult-onset MS, children tend to have fewer lesions. Lesions also have a lower propensity to enhance with gadolinium in children. A Canadian study of 20 children in whom MS was diagnosed demonstrated that although all children had 1 or more T2-weighted lesions, only 53% met McDonald criteria for lesion dissemination in space at the time of their initial demyelinating event.¹⁶ This finding suggests that MRI criteria specific to pediatric-onset MS needs to be developed.

MRI AND PPMS

Patients with PPMS are clinically distinguished by the absence of the key hallmarks of MS: relapses and remissions. This, together with a predilection for the spinal cord, resulting in a slowly progressive paraparesis, tends to characterize this subtype of MS.

PPMS has a relatively later age at onset and a lesser female preponderance than other MS subtypes. On MRI, patients present with fewer lesions, which have a lower formation rate and lower level of gadolinium enhancement than lesions seen in other MS subtypes.

In addition, affected patients show more spinal cord involvement, possibly with a greater incidence of diffuse change in the cord.³ The clinical and MRI characteristics of PPMS make diagnosis difficult, especially with the fulfillment of the criteria of dissemination in time.

The original McDonald criteria for PPMS were based on the work of Thompson and colleagues,¹⁷ which was established for rigorous research purposes rather than for the clinical setting. Briefly, 3 levels of certainty were defined:

1.Definite MS. Clinical progression for at least 1 year and positive CSF and MRI findings or equivocal MRI findings and a delayed VEP.

2.Probable MS. Either clinical progression for at least 1 year and positive CSF and equivocal MRI findings or a delayed VEP or else clin- ical progression for at least 1 year and positive or equivocal MRI findings and a delayed VEP.

3.Possible MS. Patients with clinical progression for at least 1 year and equivocal MRI evidence or delayed VEP.

In the revised McDonald criteria, the presence of CSF oligoclonal bands is no longer required.¹⁴ In their place, the presence of at least 2 spinal cord lesions and either 9 brain lesions or 4 to 8 brain lesions plus abnormal VEPs are required (Tables 2 and 3).

Although a positive CSF finding increases the comfort level for a diagnosis of PPMS, such CSF findings are not specific and may be commonly detected in patients with progressive myelopathies attributable to other causes, particularly those associated with infection (eg, human T-cell lymphotrophic virus type 1). It should be noted that although the dissemination in space criteria can be met by brain or spinal cord MRI alone, in most cases, both should be obtained. Even in the presence of abnormal findings on brain MRI scans, spinal cord imaging is helpful in ruling out alternative diagnoses, such as spinal stenosis or low-grade tumors in patients with a progressive spinal cord syndrome.

It has increasingly been recognized that in both the brain and spinal cord, a diffusely abnormal signal intensity can be present in the white matter. In a recently published study, Lycklama à Nijeholt and colleagues¹⁸ described diffuse abnormalities in addition to focal lesions in 43% of patients with progressive MS. However, in 14% of patients, only diffuse abnormalities in the spinal cord were found, correlating with a PPMS diagnosis.

Although the revised MRI criteria recognize that diffuse cord changes occur in MS, cord changes are not considered to be sufficiently reliable to be incorporated into the diagnostic criteria. At this time, therefore, there are no specific alternative recommendations for imaging patients with a diagnosis of PPMS.

MRI AND NMO

Neuromyelitis optica (NMO) is an inflammatory demyelinating disease with severe synchronous or sequential clinical optic nerve and spinal cord involvement that typically spares the brain and that, unlike MS, tends to leave severe residual injury early with each attack. A serum autoantibody marker of NMO (NMO IgG) that distinguishes NMO from MS recently has been identified.¹⁹

Recent revised diagnostic criteria for NMO²⁰ incorporate both MRI and laboratory data in a multiparametric model that yields a 90% specificity for the differential diagnosis of MS. In contrast to MS, most patients with NMO will have a normal brain MRI scan or only few nonspecific T2 hyperintensities early in the disease course. In a recent study of 60 patients who had NMO, 36 (60%) had MRI evidence of brain abnormalities. In addition, brain abnormalities developed in 15 of 30 patients who had a normal initial brain MRI scan.²¹

Spinal cord involvement in NMO, as seen on an MRI scan, typically differs from that seen in MS. The vertical extent of T2-weighted visible lesions (usually 3 or more segments) is typically increased. On axial sections, lesions commonly involve much of the cross-section. Unlike in MS, extensive "syrinx-like" longitudinal swelling and T1 hypointensity are typical in the acute and chronic stages, respectively.

MRI PROTOCOL FOR THE CNS

A standardized clinical protocol for MRI in MS or suspected MS has been recommended by an international group of neurologists and radiologists convened by the Consortium of MS Centers (CMSC).²² The recommended field strength is 1 tesla or greater. MRI sequences recommended for lesion identification (Table 4) include:

• Axial proton-density and T2-weighted (PD/T2W) spin echo and fluid attenuated inversion recovery (FLAIR).
• Sagittal FLAIR sequence covering the corpus callosum.
• Precontrast axial T1-weighted sequence.
• Postcontrast axial T1-weighted sequence. For each sequence, contiguous 3-mm-thick slices with no gap are recommended.

PD/T2W images. MS plaques, which appear as areas of focal hyperintensity on PD/T2W images (Figure 1A), can be found throughout the brain, with a predilection for periventricular white matter, the corpus callosum, U-fibers, the brain stem, and optic nerves. At the initial stage, the lesions are typically thin and linear (Dawson fingers) and are probably associated with inflammatory changes around the long axis of the medullary vein. The focal demyelinating lesions located along the lateral borders of the corpus callosum are best depicted by sagittal FLAIR imaging. These lesions together with abnormalities in the corpus callosum, U-fibers, and optic nerves may allow for the differentiation of MS from cerebrovascular disease.

Figure 1. Axial T2-weighted (A) and axial post-contrast T1-weighted (B) imagesfrom a patient with multiple sclerosis. The T2-weighted image showsseveral typical hyperintense lesions in the periventricular white matter. Onthe post-contrast T1-weighted image, 1 of these T2-weighted visible lesionsenhances after the injection of gadolinium-chelate, indicating an active inflammatorylesion with leakage of the blood-brain barrier.

Although MS predominantly affects white matter, lesions occur in gray matter and are better detected on FLAIR imaging than conventional MRI. Their small size, less severe inflammation, and partial volume artifacts in CSF and adjacent white matter make gray matter lesions difficult to detect on conventional MRI.

Up to 90% of clinically definite MS cases show T2-weighted lesions on brain MRI scans. Preexisting T2-weighted lesions can reactivate with re-enhancement, enlargement, or both. Eventually, after many reactivations, single lesions will fuse with those that are adjacent, forming confluent lesions. A net accumulation of new and enlarging lesions increase the total T2 volume by 5% to 10% per year, with a large variability among patients. Despite their exquisite sensitivity, however, T2-weighted visible lesions lack pathological specificity.

T1-weighted images. Approximately a third of T2-weighted visible lesions will appear hypointense (ie, black holes) on corresponding T1-weighted images. Although acute MS lesions also may appear hypointense on T1 images, this is a result of transient edema. They do not have the same pathology or significance as true chronic T1 black holes. A T1 hypointensity may last months after an acute event and may evolve into an isointense lesion, in which the edema is resolved and repair is suggested, or else may persist as a chronic, permanent hypointensity or true black hole.

Chronic black holes are focal areas of relatively severe tissue injury, including axonal injury, matrix destruction, and myelin loss. These lesions cannot be determined with complete certainty on a single MRI scan because, by definition, they should be persistent for at least 6 months. In routine clinical practice, however, T1 black holes are assumed to be any lesions that are hypointense but non-enhancing on post-gadolinium-enhanced T1-weighted scans. Such hypointensities are unlikely to be acute on a contrast-enhanced scan, the exception being if the patient has received high-dose corticosteroids (which suppress gadolinium enhancement by resolving the leaky blood-brain barrier) within hours to weeks of MRI.

Because of greater pathological specificity for severe axonal loss, black holes have a stronger correlation with clinical disability. In addition, T1-weighted images allow for the assessment of atrophy. Brain and spinal cord atrophy in MS are more severe in patients with secondary progressive disease. Although atrophy is measured using automated computer software, volume loss also is evaluated qualitatively and may be described by using an ordinal scale (mild-moderate-severe) based on global assessment of ventricle size and sulcal width.²³

Gadolinium-enhanced T1-weighted images. Gadolinium-enhanced T1-weighted imaging is extremely useful for identifying new lesion activity (Figure 1B). Enhancing lesions are a surrogate marker for focal disruption of the blood-brain barrier associated with macroscopic inflammation, an early stage in focal MS lesions. New enhancing lesions usually last 4 weeks (range 1 to 16 weeks).²⁴ In addition, confounding diagnoses such as leptomeningeal disease, meningioma, other mass lesions, and vascular malformation may be less well visualized or even missed without contrast-enhanced MRI scans.

Finally, the identification of enhancing lesions is an important component of the revised McDonald criteria, providing evidence for disease dissemination in space and time. Conventional doses of gadolinium-chelate (0.1 mmol/kg) are recommended with a minimum delay of 5 minutes following injection. The enhancement pattern can change with the evolution of inflammation. It is more often solid and homogeneous with acute lesions and may appear ring-like in larger and older lesions. Although enhanced MRI is considered optional for the follow-up of MS,²³ it should be strongly encouraged.

MRI protocol for spinal cord. The CMSC MRI Working Group recommends using the following sequences for MRI of the cervical and thoracic spinal cord:

• Pre- and post-contrast sagittal T1-weighted scan.
• Sagittal fast spin echo PD/T2W scan.
• Axial PD/T2W scan as well as post-contrast T1W scan of suspicious lesions.

Slice thickness should be 3 mm or less without a gap between slices. If spinal MRI immediately follows contrast brain MRI (the gadolinium-chelate dose of which should be 0.1 mmol/kg), no additional gadolinium is required and the pre-contrast sagittal T1-weighted images are not included (Table 4).

For patients with symptoms involving the spinal cord, both brain and spinal cord MRI scans are recommended to exclude MS mimics such as vascular malformations and neoplasms, especially if the symptoms have not resolved.

Spinal MRI may be very useful when findings on the brain MRI scan are normal. It is recommended for patients with suspected MS when the brain MRI results are equivocal. The importance of spinal MRI has been emphasized in the revised McDonald criteria where spinal cord lesions can be included with brain lesions to fulfill the criteria of dissemination in space and time. In contrast to brain lesions, spinal cord T2-weighted visible lesions do not develop with normal aging or chronic hypertension and diabetes. Spinal cord lesions can be found in 50% to 90% of patients with clinically definite MS.

The most common site of presentation is the cervical cord. Typical MS lesions do not extend beyond 2 vertebral segments, tend to involve the posterior and lateral regions, and occupy less than half the area of the cord on axial images.

DIFFERENTIAL DIAGNOSIS

Although the diagnosis of MS relies on the demonstration of disease dissemination in space and time, the exclusion of other neurological disorders is also essential. The International Panel emphasized that even if the clinical evidence and paraclinical studies are strongly indicative of MS, there must be "no better explanation" than MS for a secure diagnosis to be made.⁷ However, the limited specificity of T2-weighted visible lesions may increase the likelihood of an MS diagnosis in patients affected by other disorders (Table 5).

Because the current diagnostic criteria have not explored the ability of MRI to detect features not suggestive of MS, a workshop of the European Magnetic Resonance Network in MS (MAGNIMS) was held "to define a series of MRI red flags in the setting of clinically suspected MS."²⁵ Such red flags, derived from evidence-based findings and educated guesses, are summarized in Table 5 in a way that more closely conforms to everyday clinical practice. Recognition of such features in the workup of patients in whom MS is suspected may reduce the likelihood of a false-positive diagnosis.

The MS experts participating in the MAGNIMS workshop agreed that if a patient meets International Panel criteria and there are no MRI red flags, a diagnosis of MS is certain. No additional tests are needed beyond routine examinations.²⁵

If the patient meets the International Panel criteria and there is at least 1 MRI red flag, diagnosis of MS can only be made after appropriate additional test results are negative. In the case of equivocal findings, repeated imaging and laboratory tests might be needed and a wait-and-see approach should be taken before starting treatment with immunomodulatory agents.

Hypoxic-ischemic lesions. The most common MS mimics radiologically are hypoxic-ischemic cerebral small-vessel disorders. They usually are asymptomatic but can present with migraine, transient ischemic attacks, stroke, or subcortical arteriosclerotic encephalopathy. In contrast to the periventricular distribution of MS lesions, hypoxic-ischemic lesions can be dominated by arterial anatomy. Lesions can be cortical infarcts, border zone, or watershed lesions; lacunes; or multifocal basal ganglia lesions.

In addition, infratentorial lesions are specific for MS, but also occur in small-vessel disorders. Although these lesions are typically located at the surface of the pons, at the base of the fourth ventricle, and in the intra-axial tri-geminal tract in MS, they are usually centrally located in the pons in subcortical arteriosclerotic encephalopathy.

Acute disseminated encephalomyelitis (ADEM). ADEM, another common MS mimic, is a monophasic immune-mediated demyelinating disease that usually occurs in response to an infection or vaccination. In the acute phase, multifocal and commonly symmetrically distributed white matter lesions tend to uniformly enhance on MRI. Dawson fingers, corpus callosum, and periventricular white matter lesions are relatively absent compared with such signs in MS. Rather, lesions are typically located in the basal ganglia and thalamus.

Follow-up MRI can be very useful in the differential diagnosis because lesions in ADEM tend to resolve or remain unchanged with no new lesions being formed, whereas the appearance of new lesions is common in MS, although cases of recurrent ADEM have been reported.²⁶

Immune-mediated diseases. In systemic immune-mediated diseases, such as systemic lupus erythematosus and Behçet disease, MRI abnormalities may be indistinguishable from those of MS. However, the predominance of lesions located at the cortical or subcortical junction, as well as the concomitant finding of brain infarcts, calcification, or hemorrhages, should always raise the suspicion of neuropsychiatric, systemic immune-mediated diseases; small-vessel vasculitides; or antiphospholipid antibody syndrome. In these disorders, enhancing lesions and T1 black holes are much less common than in MS. Also, spinal cord lesions are rare and can completely disappear after corticosteroid or immunosuppressive treatment.^27,28

In patients with Behçet disease and CNS involvement, brain stem and basal ganglia lesions can be extensive and may be associated with swelling and enhancement in the acute phase and shrink at follow-up, leading to regional atrophy, which is less common in MS.²⁸

Progressive multifocal leukoencephalopathy (PML). PML needs particular mention because of recent reports regarding its development in patients taking the humanized monoclonal antibody natalizumab (Tysabri) and interferon beta 1a (Avonex).^29,30 PML, which is caused by infection of oligodendrocytes by the JC virus, usually occurs in immunocompromised patients.

MRI scans show multifocal and asymmetric lesions in the cerebral white matter that typically start in a juxtacortical location and progressively enlarge within weeks. These lesions never develop in the optic nerves, very rarely develop in the spinal cord, and rarely display mass effect.³¹ Most cases show no gadolinium enhancement.³⁰ Although no imaging feature is pathognomonic of MS or PML, in the future, MRI findings could be used to help distinguish these 2 pathologies.

HIGH-MAGNETIC FIELD MRI

With the introduction of 3 tesla (3T) MRI systems into clinical practice, several questions arise, including the comparability of 3T versus lower-than-3T imaging data. Studies comparing 1.5T with high-field MRI up to 4T revealed an increased sensitivity (up to 45%) in the detection of white matter abnormalities in patients with MS at higher magnetic fields.³² A more recent study³³ has shown that high-field MRI (ie, 3T) has a substantial influence on the classification of patients with CIS according to imaging criteria. In this prospective intra-individual comparative study, 40 patients with CIS were studied consecutively with a 1.5T and 3T MRI system. Eleven patients (~ 28%) fulfilled more MRI criteria at 3T. This will have consequences for prognosis and clinical trials.

IN CONCLUSION

The past few years have seen increasing improvements in imaging for the diagnosis and management of MS. MRI criteria have been incorporated into formal clinical diagnostic criteria for MS, and the incidence of misdiagnosis is becoming less frequent. However, because the presence of multiple lesions on MRI is not specific for MS, particular caution should be exerted, especially in the presence of MRI red flags or features "not suggestive" of MS.

References:

REFERENCES

1.

Hauser SL. Multiple sclerosis and other demyelinating diseases. In: Isselbacher KJ, Braunwald E, Wilson JD, et al, eds.

Harrison's Principles of Internal Medicine.

13th ed. New York: McGraw-Hill; 1994:2287-2295.

2.

Weinshenker BG, Bass B, Rice GP, et al. The natural history of multiple sclerosis: a geographically based study. I. Clinical course and disability.

Brain.

1989;112:133-146.

3.

Montalban X. Primary progressive multiple sclerosis.

Curr Opin Neurol.

2005;18:261-266.

4.

Lublin FD, Reingold SC. Defining the clinical course of multiple sclerosis: results of an international survey. National Multiple Sclerosis Society (USA) Advisory Committee on Clinical Trials of New Agents in Multiple Sclerosis.

Neurology.

1996;46:907-911.

5.

Young IR, Hall AS, Pallis CA, et al. Nuclear magnetic resonance imaging of the brain in multiple sclerosis.

Lancet.

1981;2:1063-1066.

6.

Barkhof F, Filippi M, Miller DH, et al. Comparison of MRI criteria at first presentation to predict conversion to clinically definite multiple sclerosis.

Brain.

1997;120:2059-2069.

7.

McDonald WI, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the Diagnosis of Multiple Sclerosis.

Ann Neurol.

2001; 50:121-127.

8.

Poser CM. Diagnostic criteria for multiple sclerosis: an addendum.

Ann Neurol.

1987;22:773.

9.

Tintore M, Rovira A, Martinez MJ, et al. Isolated demyelinating syndromes: comparison of different MR imaging criteria to predict conversion to clinically definite multiple sclerosis.

AJNR Am J Neuroradiol.

2000;21:702-706.

10.

Dalton CM, Brex PA, Miszkiel KA, et al. Application of the new McDonald criteria to patients with clinically isolated syndromes suggestive of multiple sclerosis.

Ann Neurol.

2002;52:47-53.

11.

Tintore M, Rovira A, Rio J, et al. New diagnostic criteria for multiple sclerosis: application in first demyelinating episode.

Neurology.

2003;60: 27-30.

12.

Barkhof F, Rocca M, Francis G, et al. Validation of diagnostic magnetic resonance imaging criteria for multiple sclerosis and response to interferon beta1a.

Ann Neurol.

2003;53:718-724.

13.

Dalton CM, Brex PA, Miszkiel KA, et al. New T2 lesions enable an earlier diagnosis of multiple sclerosis in clinically isolated syndromes.

Ann Neurol.

2003;53:673-676.

14.

Polman CH, Reingold SC, Edan G, et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the "McDonald Criteria."

Ann Neurol.

2005;58:840-846.

15.

Frohman EM, Goodin DS, Calabresi PA, et al. The utility of MRI in suspected MS: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology.

Neurology.

2003;61:602-611.

16.

Banwell B, Shroff M, Ness JM, et al. MRI features of pediatric multiple sclerosis.

Neurology.

2007;68:S46-S53.

17.

Thompson AJ, Montalban X, Barkhof F, et al. Diagnostic criteria for primary progressive multiple sclerosis: a position paper.

Ann Neurol.

2000; 47:831-835.

18.

Lycklama à Nijeholt GJ, Barkhof F, Scheltens P, et al. MR of the spinal cord in multiple sclerosis: relation to clinical subtype and disability.

AJNR Am J Neuroradiol.

1997;18:1041-1048.

19.

Lennon VA, Wingerchuk DM, Kryzer TJ, et al. A serum autoantibody marker of neuromyelitis optica: distinction from multiple sclerosis.

Lancet.

2004;364:2106-2112.

20.

Wingerchuk DM, Lennon VA, Pittock SJ, et al. Revised diagnostic criteria for neuromyelitis optica.

Neurology.

2006;66:1485-1489.

21.

Pittock SJ, Lennon VA, Krecke K, et al. Brain abnormalities in neuromyelitis optica.

Arch Neurol.

2006;63:390-396.

22.

Simon JH, Li D, Traboulsee A, et al. Standardized MR imaging protocol for multiple sclerosis: Consortium of MS Centers consensus guidelines.

AJNR Am J Neuroradiol.

2006;27:455-461.

23.

Traboulsee AL, Li DK. The role of MRI in the diagnosis of multiple sclerosis.

Adv Neurol.

2006; 98:125-146.

24.

Cotton F, Weiner HL, Jolesz FA, Guttmann CR. MRI contrast uptake in new lesions in relapsing-remitting MS followed at weekly intervals.

Neurology.

2003;60:640-646.

25.

Charil A, Yousry TA, Rovaris M, et al. MRI and the diagnosis of multiple sclerosis: expanding the concept of "no better explanation."

Lancet Neurol.

2006;5:841-852.

26.

Wingerchuk DM. The clinical course of acute disseminated encephalomyelitis.

Neurol Res.

2006; 28:341-347.

27.

Bot JC, Barkhof F, Lycklama à Nijeholt G, et al. Differentiation of multiple sclerosis from other inflammatory disorders and cerebrovascular disease: value of spinal MR imaging.

Radiology.

2002; 223:46-56.

28.

Lee SH, Yoon PH, Park SJ, Kim DI. MRI findings in neuro-behçet's disease.

Clin Radiol.

2001; 56:485-494.

29.

Kleinschmidt-DeMasters BK, Tyler KL. Progressive multifocal leukoencephalopathy complicating treatment with natalizumab and interferon beta-1a for multiple sclerosis.

N Engl J Med.

2005; 353:369-374.

30.

Langer-Gould A, Atlas SW, Green AJ, et al. Progressive multifocal leukoencephalopathy in a patient treated with natalizumab.

N Engl J Med.

2005;353:375-381.

31.

Thurnher MM, Post MJ, Rieger A, et al. Initial and follow-up MR imaging findings in AIDS- related progressive multifocal leukoencepha- lopathy treated with highly active antiretroviral therapy.

AJNR Am J Neuroradiol.

2001;22:977-984.

32.

Keiper MD, Grossman RI, Hirsch JA, et al. MR identification of white matter abnormalities in multiple sclerosis: a comparison between 1.5 T and 4 T.

AJNR Am J Neuroradiol.

1998;19:1489-1493.

33.

Wattjes MP, Harzheim M, Kuhl CK, et al. Does high-field MR imaging have an influence on the classification of patients with clinically isolated syndromes according to current diagnostic mr imaging criteria for multiple sclerosis?

AJNR Am J Neuroradiol.

2006;27:1794-1798.