Specific Imaging Findings

Multiple sclerosis (MS) lesions are T2 hyperintense and primarily located in the white matter. MS plaques are of varying shapes and sizes with the classical ovoid lesions radiating perpendicular from the ventricular wall (“Dawson's fingers”). This classic appearance needs to be present on axial images. Corpus callosum lesions at the calloso-septal interface are highly suggestive of MS, as are the ones within the brainstem and middle cerebellar peduncles. Juxtacortical white matter involvement is typical, while optic radiations and optic nerves are commonly affected. Multiple discreet lesions may coalesce and become smudgy. Active MS plaques may show reduced diffusion and, more reliably, enhancement with contrast. “Black holes”, corresponding to chronic MS lesions, are very dark on T1WI, frequently with a thin peripheral bright rim, better seen with magnetization transfer. Larger demyelinating plaques show low attenuation on CT. The revised McDonald criteria require presence of at least two T2 hyperintense lesions in at least two of the four locations – juxtacortical, periventricular, infratentorial, and spinal cord, in the appropriate clinical setting. If brain MRI is not conclusive, spinal MRI may be helpful, as around 25% of patients present with isolated spinal cord lesions.

Pertinent Clinical Information

Patients are more commonly young females presenting with various symptoms and signs, such as tingling, numbness, weakness, fatigue, coordination and balance problems. Optic neuritis, a frequent presenting sign, is present in up to 50% of cases. CSF examination typically shows increased protein and oligoclonal bands. Symptoms can spontaneously resolve and come back or present in another part of the body. In Asian patients, the optic–spinal type is more frequent, older age group is affected, fewer cases are positive for oligoclonal bands, and total CSF protein is higher. MR imaging has an important role, excluding alternative diagnoses, and characterizing dissemination in space and time; however, it is currently not reliable for predicting the clinical disease evolution.