1. The Role of MRI in MS Diagnosis and Management

2. In most cases, the first clinical neurologic manifestation of MS is a clinically isolated syndrome (CIS). CIS is a solitary, inflammatory, demyelinating syndrome of acute onset in the CNS in an individual who has no prior history of demyelination and in whom alternative diagnoses have been excluded through appropriate clinical and laboratory investigations.
Certain patient characteristics (such as age) and certain neurologic signs and symptoms are more frequently associated with the development of MS. The characteristic syndromes associated with a first attack of MS include optic neuritis, brain stem dysfunction, and transverse myelitis.[1] The optic neuritis tends to be unilateral, retrobulbar, and painful in nature. In general, the patient can expect good recovery from these symptoms. Brain stem dysfunction is most often an ocular motor syndrome such as internuclear ophthalmoplegia (INO) and nystagmus. Hemisensory abnormalities as well as hemiparesis and trigeminal neuralgia may be seen. In cases of transverse myelitis, the syndrome is more often partial sensory than partial motor, with or without bowel and bladder involvement. The patient can also present with Lhermitte’s sign or band-like abdominal or chest pressure. MRI findings suggestive of MS and elimination of other possible causes would be indicative of CIS.
Here, axial T2W fluid attenuated inversion recovery (FLAIR) sequence eliminates the hyperintense signal from CSF, allowing a clear visualization of hyperintense lesions. This image shows MS lesions in the periventricular and deep white matter.

3. To fulfill the 2005 McDonald criteria for dissemination in time, any new T2 lesion detected following the baseline study (defined as a study performed at least 30 days following the initial clinical event) satisfies the dissemination in time and space criteria. A new gadolinium (Gd)-enhancing lesion detected 3 months after the initial clinical event at a different site from the clinical event also fulfills those criteria.[2]
Compared with the baseline scan in a patient with a CIS (left), a new Gd-enhancing lesion was identified in a follow-up scan done 3 months later. Based on the new MRI findings, the patient meets 2005 McDonald criteria for clinically definite multiple sclerosis (CDMS).The latest revision of the McDonald diagnostic criteria enables clinicians to make the diagnosis of MS earlier.[3,4] Among the changes:
Concomitant enhancing and non-enhancing lesions on the baseline MRI suffice to demonstrate dissemination in time.[4] This lends increasing sensitivity but also increases risk of misdiagnosis.
Clinicians should carefully evaluate clinical and radiological red flags in order to avoid diagnostic mistakes, which may lead to erroneous therapeutic choices.[4]

4. The absence of characteristic findings or the persistence of uncharacteristic symptoms -- such as the red flags listed here -- necessitates the reconsideration of a diagnosis of MS. For example, a patient presenting with optic neuritis without MRI findings would be indicative of a diagnosis other than MS. When in doubt, observe and repeat MRI in 3 to 6 months.

6. Case Study: Patient A
A 25-year-old woman with a 2-day history of left hemiparesis presents to the clinic for evaluation. Examination confirms left hemiparesis and reveals left upper and lower extremity hyperreflexia and left Babinski sign. The rest of her neurologic examination is normal. CSF is positive for 2 oligoclonal bands with no other abnormalities. The rest of her workup is normal. Her brain MRI shows 3 periventricular lesions ranging in size from 6 to10 mm, 1 with enhancement accounting for dissemination in time and space according to the 2010 McDonald criteria. Although she does not fulfill the criteria for dissemination in time according to the 2005 McDonald criteria, she does according to the 2010 revision. Her history, the findings on physical and neurological examination, and ancillary testing all indicate that she is having a CIS (first attack of MS), making her a candidate for immunomodulatory therapy. The alternative would be to wait for the development of a new lesion on follow-up MRI to confirm the diagnosis of MS by the 2005 McDonald dissemination in time criteria.
Axial T2W FLAIR image and T1W post-Gd image show lesions typical of MS. Three hyperintense lesions are seen on FLAIR with a corresponding enhancing lesion on the post-contrast image (arrow) and one T1 hypointense lesion (black hole). These findings fulfill the 2010 MRI McDonald criteria for dissemination in time and space.

7. A sagittal T2W FLAIR image showing typical Dawson's fingers can be helpful for diagnostic purposes. T1W pre-contrast images may also provide important information suggestive of MS, such as evidence of black holes. Typical MS lesions are ovoid in shape, periventricular, perpendicular to the ventricles, and more than 3 mm in size. Typical locations include:
Periventricular
Corpus callosum [sagittal FLAIR/PD]
Callosal-septal interface
Subcortical [FLAIR]
Optic nerves/visual pathways
Posterior fossa, brain stem [better visualized on PD/T2]

8. Case Study: Patient B
A 42-year-old man presents with numbness from mid-chest to toes over the past 3 days. Examination shows mild leg weakness; decreased pain and temperature sensation from T4 level; brisk leg deep tendon reflex; and bilateral Babinski signs. Brain MRI shows 2 periventricular T2 lesions. Cord MRI shows 3 enhancing partial cord lesions, and CSF is positive for oligoclonal bands.
While this patient is older than most who receive a diagnosis of MS, he is not outside of the age range for MS. The presentation of a partial acute transverse myelitis with sensory and motor involvement, the findings on spinal cord and brain MRI, and the CSF abnormalities are highly suggestive of MS, and he meets 2010 McDonald MRI criteria for dissemination in time and space. This patient would be considered to have CIS and could be offered treatment. Alternatively, he can be followed with a repeat MRI in 1 to 3 months to detect new lesions and confirm the diagnosis by 2005 McDonald criteria.

9. The Consortium of MS Centers, an international group of neurologists and radiologists, issued consensus guidelines in 2001, updated in 2006, to standardize the imaging protocol for diagnosis and follow-up of MS patients.[5] Some of the recommendations are as follows:
Protocol: When available, MRI should be part of initial evaluation after a CIS or suspicious history to meet standardized protocol. Patients who already have been diagnosed with MS should have a baseline MR evaluation in addition to a complete neurologic history and examination. MS cannot be diagnosed with MRI alone; clinical signs and symptoms must also be present. If the main symptoms occur at the level of the spinal cord, both spine and brain imaging are required.
Technical/acquisition standards: When obtaining the axial Gd-enhanced T1 sequence, use a standard dose of 0.1 mmol/kg injected over 30 seconds; start the scan no sooner than 5 minutes after injection.
Magnet size: 3T imaging of MS is actively under investigation and may detect more lesions than 1.5T. However, when evaluating a case, comparative studies should preferably be made on scanners of the same field strengths.
Other
Referral indications and documentation should be written on the request form, stating either "Suspected MS," "Baseline exam for MS," or "Follow-up MS."
There should be standardized protocols to image these patients.
A new MRI lesion (instead of 2 clinical attacks) is acceptable as establishing dissemination in time after a clinically isolated event.
MRI allows acceleration of diagnosis, may be used in decisions to treat without other evidence, can measure subclinical disease, can be used to follow disease activity, and is important for clinical trials.

10. T2 FLAIR suppresses the CSF hyperintense signal seen in T2 fast spin echo (FSE) and proton density (PD) sequences and provides clear details on periventricular and deep white matter hyperintense lesions as well as juxtacortical lesions. Thin-slice sagittal FLAIR is sensitive to early MS pathology, such as that found in the corpus callosum and at the callosal-septal interface. The exact relationship between MRI findings and clinical status of patients is unknown.[5]

11. PD and T2 series are more sensitive to infratentorial lesions that may be missed by FLAIR due to CSF motion.
Follow-up imaging in MS:
Same sequences as for diagnosis except for additional thin-slice sagittal FLAIR
Not recommended without clinical symptoms
Reassessment of disease burden for treatment purposes
Suspect other diagnoses
Contrast-enhanced MRI
Diagnosis
Initial diagnostic evaluation
Disease activity
Optional baseline, follow-up
Radiology report should include:
Description of findings
Comparison
Interpretation/differential diagnosis
T2 lesion number
Location of enhancing lesions

12. If brain MR results are equivocal, spine imaging may be justified to clarify the diagnosis.[5] Spinal cord imaging may increase diagnostic sensitivity, showing patients with lesions and normal brain, and it may exclude other lesions (vascular, etc). A finding of a spine lesion with brain lesions increases the likelihood of MS.
Between 5% and 24% of patients with MS may have isolated spinal cord disease, most of which develops in the cervical spine. Most spinal cord lesions tend to be peripherally located, and may cause cord expansion if active or atrophy if the lesions are old.
These lesions can range in size, but usually they are less than 2 vertebral segments. Most such lesions enhance if the patient is symptomatic.
On this slide, sagittal T2 FSE of the cervical spine shows a demyelinating lesion at the level of C4-C5, suspicious for MS.

13. Revised criteria for MRI diagnosis of MS have improved sensitivity while maintaining specificity.[3] However, it is important to identify atypical foci that might suggest an alternate diagnosis. Chronic MS plaques appear on T2-weighted images as well-demarcated, ovoid lesions, such as the one seen in this slide.[6]

14. A range of conditions may present with symptoms that mimic aspects of MS.[7] Here, MRI of a migraineur reveals lesions that could be confused with those seen in a patient with MS. Other rule-outs in differential diagnosis include:
Anti-phospholipid syndrome
Behçet disease
CADASIL
CNS lymphoma
Infection
Inflammation
Ischemic white matter disease
Lyme disease
Metabolic
Neurosyphilis
Other vasculitides
Primary angiitis of the CNS
Sarcoidosis
Sjögren syndrome
Systemic lupus erythematosus
Vitamin B12 deficiencies
Before a diagnosis of MS can be made, especially in the context of atypical clinical presentations, laboratory studies are advisable to exclude disorders whose symptoms and signs might mimic those of MS.[8] Some of these include:
Genetic syndrome
Infectious diseases
Lesions of the posterior fossa and spinal cord
Metabolic disorders
Neoplastic illnesses
Psychiatric disorders
Systemic autoimmune diseases
Variants of MS: Balo concentric sclerosis, Devic disease, Schilder's diffuse sclerosis, Marburg multiple sclerosis
Vascular disorders

15. EM is a 35-year-old female who presented to the hospital with difficulty walking and gradual worsening of numbness involving the right face. Over the next several days, she also noted clumsiness involving her left side, weakness of her left arm, and occasional diplopia.
Her exam showed a diplopia on right lateral gaze, diminished sensation to pinprick in the right upper lip, normal motor survey, and dysmetria on finger-to-nose and heel-knee-shin on the left side with ataxia of gait. Her imaging shows multiple focal areas of enhancement involving the brain stem.
She was treated with 5 days of high-dose intravenous methylprednisolone with marked improvement of symptoms.
Neurosarcoidosis is uncommon and usually involves the cranial nerves[9] or meninges rather than the parenchyma. Neurosarcoidosis may present with simultaneous enhancement of multiple punctate lesions often along Virchow-Robin spaces.[10] A similar pattern can be seen with neurologic Behçet disease.[10] Bickerstaff encephalitis might also be included in the differential diagnosis.
Neuroborreliosis[11] and treponemal infections of the CNS[12] (eg, syphilis) also usually involve the meninges and only rarely the parenchyma

16. Ring-enhancing lesions can be observed in MS, though often there is incomplete ring formation.[13] Note that in this image, in addition to the incomplete ring-enhancing lesion, there are multiple lesions more characteristic of disease. Additionally, T2-weighted images show a hypointense rim, possibly resulting from iron-containing macrophages,[14] which corresponds to the ring seen on T1 Gd-enhanced imaging. This low intensity rim on T2WI could potentially be mistaken for an abscess. Large ring-enhancing lesions located beneath the cortex with edema may be suspicious for primary CNS tumor or metastasis. Enhancement in primary CNS lymphoma may persist for a longer duration than typically observed in MS.

17. This image is from a patient who has Balo’s concentric sclerosis
This image is from a patient who has Balo’s concentric sclerosis. Note the large lesion with minimal edema and concentric appearance. The site most commonly affected in Balo variant of MS is the cerebral white matter, although lesions within the basal ganglia, pons, and spinal cord also have been documented.[15] Lesions are characterized by large areas of demyelination surrounded by smaller areas of preserved myelin.[15] Histopathological features of these areas of preserved myelin vary; some authors report normally myelinated axons[15,16] while others have reported either remyelinated or partially demyelinated myelin[17] and still others suggest that the alternating areas are merely a stage in the process of demyelination.[18] Additionally, factors contributing to the concentric pattern have yet to be fully determined; some experts feel that periplaque tissue might contain increased levels of stress proteins, making it more resistant to damage.[19]
Balo's concentric rings have been described in isolation associated with a fulminant clinical decline,[20] but they have also been documented to coexist in patients with an initial diagnosis of MS.[15,21] The paucity of cortical gray involvement and usual sparing of subcortical U fibers can be observed on imaging.[20] Additionally, contrast enhancement, which is often heterogeneous, can sometimes be seen limited to the outer rings,[21] though this might be in part dependent upon the timing of MRI scans.[18] No enhancement was observed in this patient.

18. Case Study: Patient C
CM is a 34-year-old female with a history of SLE. Previous exam showed cutaneous manifestations and laboratory confirmed diagnosis with high titers of antinuclear and double-stranded DNA antibodies. She had presented with transient quadraparesis that significantly improved with high-dose intravenous steroids.
Imaging of her CNS revealed a hyperintense focus involving the upper cervical spine and the foci in the brain indicated by the arrows.
Parenchymal lesions are not unique to MS and might also be seen in a variety of other conditions, including SLE, Sjögren syndrome, and Adamantaiades-Behçet disease.[22] Lesions are often subcortical with an appearance of vascular pathology (thin arrows) but in rare cases might also be adjacent to ventricles (thick arrows) or involve the corpus callosum.[23,24] Similar findings have been observed in anti-phospholipid syndrome.[23]
Antinuclear antibody positivity may be seen in MS, but persistent high titers and/or presence of systemic involvement should increase suspicion for diagnosis of connective tissue disease.[25]

19. T2-hyperintensities located in subcortical or deep white matter areas often may not be specific to any disorder or may represent small vessel ischemic disease. Such lesions can be seen in disorders such as migraine headache where an association between headache, especially with aura, and increased stroke risk has been documented.[26] Evaluating lesion characteristics, as previously described, may assist in differentiating non-specific T2 foci from those consistent with MS.

20. Gd within developmental venous abnormalities can be mistaken for parenchymal enhancement. This image shows a venous angioma. Note that the area of enhancement appears tortuous and that corresponding hyperintensity in the T2-weighted images is linear and follows the contour of the vessel.

21. MRI is currently utilized as a surrogate marker of treatment efficacy in drug trials. Meta-analysis of 23 randomized placebo-controlled trials in RRMS, involving a total of 6591 patients, identified a strong correlation between the effect of treatment on relapses and number of MRI lesions measured.[27] More than 80% of the variance in treatment effects on relapses was explained by the variance in MRI measures.[28] Gd enhancement and new or newly enlarging lesions identified with T2 have greater utility in evaluating effectiveness of treatment.[28] However, MRI metrics have weak predictive value for future disability,[29,30] and correlate poorly with T2 lesion burden in progressive disease.[31]
In image 1A, the arrow indicates a T2 lesion. Image 1B shows the lesion has dramatically enlarged from baseline scan.
Image 2A shows new T2 lesions. In image 2B, arrowheads indicate T2 lesions that are unchanged in size from baseline scan. Yellow arrows indicate a few of the new T2 lesions when compared to baseline scan. Blue arrows indicate T2 lesions that have enlarged from baseline scan.

22. MRI is being evaluated as a marker to predict clinical response to treatment. Several studies have shown increased probability of relapse associated with MRI activity at 1 year following interferon beta therapy.[32-34] Development of new lesions on MRI early in the course of IFNB therapy might predict treatment non-responders.[33]
The utility of MRI metrics to predictor treatment response for other approved therapies has yet to be evaluated. However, despite data demonstrating surrogacy of MRI metrics in clinical trials, the utility of routine MRI evaluation in the clinical setting is unclear and controversial.[35]
Follow-up MRI is reasonable for unexpected clinical worsening and in re-assessment for treatment initiation or modification.[35]
Persistent hypointensities on T1-weighted images may correlate with tissue damage and have been utilized as a secondary outcome measure in clinical trials.[28,36]
True "black holes" are T1 hypointense lesions that persist for at least 6 months.[5] However, some studies have shown that only a small proportion of T2 lesions evolve to chronic "black holes."[28,37]
Brain atrophy is noted to occur at all stages of MS and occurs at a greater rate when compared to healthy controls.[38,39] The mechanisms involved in atrophy are not entirely clear and may include both focal inflammation and Wallerian degeneration.[39,40] Gray matter loss, especially in the deep gray structures, is disproportionate to white matter atrophy.[41] Automated segmentation allows for determination of whole brain volumes and atrophy measures in clinical trials.[42,43] However, the various mechanisms involved in measures of atrophy and the lack of such automated tools in the clinical setting make this MRI metric less useful in making day-to-day patient decisions.