Rim lesions are demonstrated in early relapsing-remitting multiple sclerosis using 3 T-based susceptibility-weighted imaging in a multi-institutional setting

Abstract

Purpose: Rim lesions, characterised by a paramagnetic rim on susceptibility-based MRI, have been suggested to reflect chronic inflammatory demyelination in multiple sclerosis (MS) patients. Here, we assess, through susceptibility-weighted imaging (SWI), the prevalence, longitudinal volume evolution and clinical associations of rim lesions in subjects with early relapsing-remitting MS (RRMS).

Methods: Subjects (n = 44) with recently diagnosed RRMS underwent 3 T MRI at baseline (M0) and 1 year (M12) as part of a multi-centre study. SWI was acquired at M12 using a 3D segmented gradient-echo echo-planar imaging sequence. Rim lesions identified on SWI were manually segmented on FLAIR images at both time points for volumetric analysis.

Results: Twelve subjects (27%) had at least one rim lesion at M12. A linear mixed-effects model, with 'subject' as a random factor, revealed mixed evidence for the difference in longitudinal volume change between rim lesions and non-rim lesions (p = 0.0350 and p = 0.0556 for subjects with and without rim lesions, respectively). All 25 rim lesions identified showed T1-weighted hypointense signal. Subjects with and without rim lesions did not differ significantly with respect to age, disease duration or clinical measures of disability (p > 0.05).

Conclusion: We demonstrate that rim lesions are detectable in early-stage RRMS on 3 T MRI across multiple centres, although their relationship to lesion enlargement is equivocal in this small cohort. Identification of SWI rims was subjective. Agreed criteria for defining rim lesions and their further validation as a biomarker of chronic inflammation are required for translation of SWI into routine MS clinical practice.

Keywords: Magnetic resonance imaging; Multiple sclerosis; Rim lesions; Susceptibility-weighted imaging.

Fig. 1

Rim lesions observed at 3 T. SWI and corresponding FLAIR images showing representative examples of rim lesions (indicated by red arrow) in different brain locations including periventricular (a, d), juxtacortical (b, e) and infratentorial (c, f) regions

Fig. 2

T1 and FLAIR appearance of an SWI rim lesion. All rim lesions observed in our MS cohort, including this white matter lesion (indicated by red arrow) exhibited T1 hypointensity. The lesion shown above is one of ten rim lesions that on FLAIR also appeared hyperintense with a core isointense or in this case, hypointense, relative to extralesional white matter

Fig. 3

Longitudinal lesion volume change. Boxplots comparing volume changes over 1 year across the three different lesion groups, namely, rim lesions (left), non-rim lesions in subjects with rim lesions (middle) and non-rim lesions in subjects without rim lesions (right)