MAGNIMS consensus on MRI diagnosis of multiple sclerosis

Last revised by Henry Knipe on 21 Jul 2019

The magnetic resonance imaging in multiple sclerosis (MAGNIMS), which is a European collaborative research network, published in 2016 new recommendations to upgrade the imaging diagnosis criteria for multiple sclerosis (MS). These came as a consensus, based on evidence-based and expert opinions 2, aiming to improve on the previous McDonald diagnostic criteria from 2010 1. A new version of the McDonald diagnostic criteria for multiple sclerosis has been published in 2017 and should be used. 

The new recommendations are focused on the McDonald "dissemination in space" criteria, proposing simultaneously a widening and narrowing of these criteria:

  • optic neuritis was added to the previous four defined locations characteristic for MS (periventricular, juxtacortical, infratentorial and spinal cord)
    • justification: ~25% with a clinically isolated syndrome present with acute optic neuritis
  • minimum requirement of three lesions in a periventricular location
    • justification: incidental periventricular lesions are also found in healthy individuals and in other conditions (e.g. migraine); the minimum of three lesions increased the diagnostic accuracy in different studies
  • cortical location was added together to the previous juxtacortical location
    • justification: pathology studies have shown large grey matter involvement in MS and new MRI techniques have improved the sensitivity for the detection of those cortical lesions 3-7, although, there was no consensus on the optimal MRI technique for the detection of cortical lesions, and many cortical lesions remain invisible on 1.5 T and 3 T scanners
    • comment: as with the routine scanners the differentiation between cortical and juxtacortical is not reliable, the consensus was that those lesions should be combined in a single term "cortical-juxtacortical" 

Regarding both the "dissemination in space" and "dissemination in time":

  • no distinction between symptomatic and asymptomatic MRI lesions has to be made, all lesions are now considered when counting for either dissemination in space either dissemination in time
    • justification: previously in the McDonald criteria the symptomatic lesions (e.g. lesion in the spine in a patient having a brainstem spinal cord syndrome) were not counted in the diagnostic criteria, however, the decision if a lesion is symptomatic or not is frequently difficult and imprecise

The new 2016 MAGNIMS MRI criteria establish disease dissemination in space, by detecting involvement of at least two of the five following areas of the CNS:

  • periventricular: ≥3 lesions
  • cortical-juxtacortical: ≥1 lesions
  • infratentorial: ≥1 lesions
  • spinal cord: ≥1 lesions
  • optic nerve: ≥1 lesions

Dissemination in time can be established in one of two ways:

  • a new lesion when compared to a previous scan (irrespective of timing)
    • T2 bright lesion and/or gadolinium-enhancing
  • presence of enhancing lesion and a non-enhancing T2 bright lesion on any one scan

ADVERTISEMENT: Supporters see fewer/no ads

Updating… Please wait.

 Unable to process the form. Check for errors and try again.

 Thank you for updating your details.