Write click: Editor's choice

摘要

We validated surrogacy of MRI lesions/relapses for EDSS worsening, the only clinical measure accepted as a true measure of outcome by regulatory agencies. We agree that EDSS is largely insensitive (i.e., noisy) and this makes finding surrogacy for MRI lesions/relapses more difficult. Ebers et al. keep confounding correlation with surrogacy. A surrogate is a marker to predict the effect of a treatment on the clinical endpoint of interest and surrogacy cannot be evaluated outside of a treatment trial. Therefore, the lack of correlation in uncontrolled studies, as in those cited by Ebers et al., is not informative: the absence of evidence cannot be confounded with the evidence of absence.6 We agree with Ebers et al. that it is time to rework the outcomes used in MS. This should start from clinical measures, which are largely more insensitive and noisy than MRI measures, especially if they are coming from old databases.