Radiologically isolated syndrome (RIS) describes individuals with incidental findings highly suggestive of demyelinating lesions on magnetic resonance imaging (MRI) — i.e., periventricular, ovoid, juxtacortical, gadolinium-enhancing, or posterior fossa lesions (Neurology 2009; 72:800). Whether to treat such asymptomatic or "presymptomatic" individuals remains controversial.

To examine the predictive value of spinal cord lesions in patients with RIS, researchers identified 71 patients who had spinal cord imaging, out of 102 who had received a diagnosis of RIS after brain MRI for headaches, trauma, syncope, volunteering as a "healthy" control for research, or other reasons. Cervical cord was imaged in all 71 patients and thoracic cord in 17 of the 71. Spinal lesions had to be well circumscribed, distinct, and observed on more than one MRI sequence or plane.

Cervical cord abnormalities were present in 35% (25/71) of patients, with enhancement observed in 6 of those 25. The thoracic cord was abnormal in 6 of 17; 5 of those 6 had concomitant cervical lesions. Of the 25 patients with cervical lesions, 84% developed clinical progression to a clinically isolated syndrome (CIS; n=19) or primary progressive multiple sclerosis (n=2) within a median of 1.6 years (interquartile range, 0.8–3.8 years). Clinical progression occurred in only 3 of 46 patients without a cervical cord abnormality. The presence of a clear spinal cord lesion was associated with progression to a clinical event with 88% sensitivity, 92% specificity, and a positive predictive value of 84%.

Comment: These findings support obtaining a cervical spinal cord MRI scan for asymptomatic individuals whose brain MRI suggests demyelinating lesions. Thoracic cord imaging may not be necessary for all cases, as 83% of those with a thoracic lesion also had a cervical lesion. One study limitation is the small number of cases compared with modern CIS cohorts. Also, not all individuals with RIS underwent spinal cord imaging; although no differences were reported between those with and without cord imaging, a selection bias may still be possible. Confirmation of these findings in a separate, larger cohort would be helpful. For now, we can advise patients of the high sensitivity in predicting a clinical demyelinating event when a spinal cord lesion is found with an asymptomatic brain MRI scan compatible with demyelination. Although treatment of RIS will remain controversial, this study may help with a treatment decision in select cases.

— Robert T. Naismith, MD

Published in Journal Watch Neurology February 8, 2011

http://neurology.jwatch.org/cgi/content/full/2011/208/1?q=featured_jn