Magnetic resonance imaging as a surrogate outcome measure of disability in multiple sclerosis: have we been overly harsh in our assessment?

Ann Neurol. 2006 Apr;59(4):597-605. doi: 10.1002/ana.20832.

Abstract

Objective: The validity of magnetic resonance imaging (MRI) as a surrogate outcome measure in multiple sclerosis (MS) clinical trials has been greeted skeptically both by the US Food and Drug Administration and by clinical researchers because the correlation between current MRI measures and clinical disability, although significant, has generally been low. Thus, the reported correlations have varied between rho = 0.09 and rho = 0.60, and have often been at the lower end of this range. Nevertheless, it still appears possible that this apparently poor correlation is due not to any deficiency either with our current MRI measures or with our disability scale, but rather to the intrinsic variability in the clinical expression of MS plaques in different anatomical locations.

Methods: This article explores this possibility through the development of a general mathematical model for the relation between MRI changes and clinical disability in patients with MS.

Results: Under the conditions of this general model, the maximum expected correlation between clinical disability and MRI will typically be quite low (eg, rho = 0.2-0.3), even when it is assumed that the MRI changes are the sole determinant of disability and, furthermore, that the scale used to measure disability is ideal.

Interpretation: These observations, together with the significant relations already reported between MRI and disability (with observed correlations in the range of 0.2-0.6), actually suggest that our available clinical and MRI measures are considerably better than is currently believed and, in fact, that the MRI may be a valid surrogate marker in the assessment of treatment efficacy in MS.

Publication types

  • Comparative Study
  • Review

MeSH terms

  • Disability Evaluation*
  • Humans
  • Likelihood Functions
  • Magnetic Resonance Imaging / methods*
  • Models, Statistical
  • Multiple Sclerosis / diagnosis*
  • Multiple Sclerosis / physiopathology*
  • Outcome Assessment, Health Care*
  • Reproducibility of Results