Related Experiment Video
Updated: Jun 11, 2025

08:51
Magnetic Resonance Imaging of Multiple Sclerosis at 7.0 Tesla
Published on: February 19, 2021
8.9K
Paramagnetic Rim Lesions are Highly Specific for Multiple Sclerosis in Real-World Data.
Christopher C Hemond1,2, Sathish K Dundamadappa3, Mugdha Deshpande1
1Departments of Neurology, University of Massachusetts Memorial Medical Center and University of Massachusetts Chan Medical School, Worcester, MA, USA.
Medrxiv : the Preprint Server for Health Sciences
|October 7, 2024
Summary
Paramagnetic rim lesions (PRL) show high specificity in diagnosing multiple sclerosis (MS), distinguishing it from other neurological conditions. However, their modest sensitivity means PRL cannot be used as the sole diagnostic marker.
Area of Science:
- Neuroimaging
- Neurology
- Biomarker Discovery
Background:
- Paramagnetic rim lesions (PRL) are an emerging biomarker in multiple sclerosis (MS).
- PRL are associated with increased disease severity and may support MS diagnosis.
Purpose of the Study:
- To determine the specificity and sensitivity of PRL for differentiating MS from its mimics.
- To utilize real-world clinical and imaging data for diagnostic discrimination.
Main Methods:
- Retrospective analysis of a longitudinal cohort with prospectively collected data.
- MRI scans using a T2*-weighted sequence (SWAN protocol) were reviewed for PRL.
- PRL were categorized using modified consensus criteria (definite, probable, possible).
Main Results:
- A total of 580 patients were evaluated (473 MS, 57 NIND, 50 OIND).
- "Definite" or "probable" PRL achieved 98% specificity and 36% sensitivity in discriminating MS from NIND and OIND.
- Interrater agreement for identifying definite/probable PRL was nearly perfect.
Conclusions:
- PRL demonstrate high specificity for MS, aiding in diagnostic evaluation.
- Modest sensitivity restricts their use as standalone diagnostic indicators.
- Including "possible" PRL significantly reduces specificity, warranting cautious interpretation to avoid misdiagnosis.

