Paramagnetic rim lesions predict the development of clinical MS in radiologically isolated syndrome: results from a prospective cohort study
Jiwon Oh1, Timothy Lim2, Suradech Suthiphosuwan3, Adrian Espiritu4, Melanie Guenette1, Aditya Bharatha3, Pascal Sati5, Martina Absinta6, Daniel Reich7
1St Michael's Hospital, 2Medical Imaging, St Michael's Hospital, 3St Michael’s Hospital, Unity Health, 4BARLO MS Centre, St. Michael's Hospital, University of Toronto, 5Cedars Sinai, 6National Institutes of Health, 7National Institutes of Health, Neuroimmunology Branch, NINDS
Objective:

To determine the association of various MRI measures and subsequent development of clinical MS in people with radiologically isolated syndrome (pwRIS).

Background:

We previously found that most pwRIS have high proportions of white matter lesions positive for the central vein sign (CVS+L) and at least one paramagnetic rim lesion (PRL), representing perivenular and chronic active demyelination, respectively. Whether PRL and CVS+L relate to the risk of developing clinical MS in pwRIS prospectively is unknown.

Design/Methods:

PwRIS were recruited prospectively and underwent 3T-MRI, including 3D-FLAIR and T2*segmented EPI of the brain, and sagittal T1-PSIR of the cervical spinal cord(SC). MRIs were evaluated for presence of PRL,CVS+L, and SC lesions (SCL).

Results:

36 pwRIS (median age 43 years, 70% women, median time of clinical follow-up 6.3 years) were included in the study. Clinical MS developed in 9 (25%) subjects with a median time to first clinical event of 5.2 years. 67% developed RRMS and 33% PPMS. At baseline, pwRIS who developed clinical MS vs those who did not had a higher median number of PRL (11 vs 1, p=0.01) and CVS+L(34 vs 10, p=0.04). Longitudinally, number of new brain lesions, PRL, CVS+L (all p<0.01) and SCL (p=0.02) were higher in those who subsequently developed clinical MS. Multivariable logistic regression using backward stepwise selection identified baseline PRL count as the most predictive variable of MS development among baseline and follow-up imaging measures (p=0.01).

Conclusions:

In this prospective cohort of pwRIS, 25% developed MS within 5 years after initial diagnosis, with 33% diagnosed as PPMS. Baseline PRL count was the most predictive MRI measure for subsequent development of clinical MS in pwRIS, suggesting that PRLs may have prognostic utility. Continued follow-up of this cohort and validation in larger datasets will be important to confirm the clinical predictive value of PRL and CVS+L in pwRIS.

10.1212/WNL.0000000000203626