Predicting the severity of relapsing-remitting MS: the contribution of cross-sectional and short-term follow-up MRI data

Abstract

Background and objective: Predicting the long-term clinical course of multiple sclerosis (MS) is difficult on clinical grounds. Recent studies have suggested magnetic resonance imaging (MRI) metrics to be helpful. We wanted to confirm this.

Methods: Contactable individuals (N=84) from an initial 99 patients with relapsing-remitting MS (RRMS) who had undergone careful baseline and 2-year follow-up examinations including MRI were reassessed after a mean of 10.8±2.7 years. We investigated using multivariate linear regression analyses if clinical and MRI data obtained at the prior time-points and the rates of change in morphologic variables over a mean observational period of 2.5 years could have served to predict a patient's MS severity score (MSSS) 11 years later. Conversion to secondary progressive MS (SPMS) was a further outcome variable.

Results: In univariate analyses, the 'black hole ratio' (BHR) at baseline (p=0.017, beta=0.148) and at first follow-up (p=0.007, beta= -0.154) was the only MRI parameter showing a significant correlation with the MSSS. In a multiple regression model, the independent predictive value of imaging variables became statistically non-significant and the latest MSSS was predicted primarily by the baseline EDSS (r (2)=0.28; p<0.001). The BHR at baseline explained 9.4% of variance of conversion to SPMS (p=0.033). Over the observational period the MSSS remained stable in patients remaining RRMS, but increased in converters to SPMS from 4.0 to 6.4.

Conclusions: We failed to confirm a clear independent contribution of cross-sectional and short-term follow-up MRI data for the prediction of the long-term clinical course of MS. The MSSS is not a stable indicator of disease severity but may increase in converters to SPMS.