Paramagnetic Rim Lesions Highly Specific in MS Diagnosis

The presence of paramagnetic rim lesions (PRLs) on MRI was highly specific for multiple sclerosis (MS) in people suspected to have the disease, a prospective multicenter study showed.

Based on 2017 McDonald criteria for MS diagnosis, the presence of one or more PRLs showed 88% specificity (95% CI 79-98%), 80% sensitivity (95% CI 68-91%), and 84% accuracy (95% CI 75-91%), reported Brian Renner, MD, of Cedars-Sinai Medical Center in Los Angeles, at ACTRIMS Forum 2022, the annual meeting of the Americas Committee for Treatment and Research in Multiple Sclerosis. The area under the receiver operating curve (AUROC) was 81%.

“The sensitivity of PRLs for MS was higher than previously reported — 80% for MS in the current study versus about 50% in prior studies — despite similar techniques in rating, processing, and evaluation, which is likely related to the nature of the cohort,” Renner noted.

PRLs reflect a subset of chronic active white MS lesions containing iron-laden macrophages and microglia, which accumulate at the lesion edge, reflecting smoldering inflammation.

“PRLs occur early in the disease course, as demonstrated in this prospective cohort with participants referred to MS centers for a diagnosis of MS,” Renner said. PRLs also are associated with disease severity, including cognitive dysfunction, physical disability, and brain atrophy, he added.

The role of PRLs in MS diagnosis or prognosis is actively being studied. A recent retrospective study reported a high specificity (93%) but relatively low sensitivity (52%) for MS diagnosis, using a threshold of one or more PRL.

Renner and colleagues evaluated the diagnostic performance of PRLs in 95 people who were evaluated for an MS diagnosis from 2018 to 2020. Participants were from the North American Imaging in MS Cooperative (NAIMS) cohort; mean age was 45 and 74% were women.

Overall, 44 of 95 (46%) people met McDonald 2017 criteria for MS and 51 people did not. The people who did not included 14 participants at risk of developing MS with diagnoses of clinically isolating syndrome (CIS) or radiologically isolating syndrome (RIS), and 37 participants with other diagnoses.

In the entire cohort, 41 people had one or more PRLs. In total, 148 PMLs were seen (median 3 PRLs; range 0-11).

Of 44 people diagnosed with MS, 35 had one or more PRLs. Of 51 people who did not meet MS diagnosis, six had one or more PRLs, including three RIS and one CIS cases.

“Most patients meeting MS diagnosis — 80% — displayed one or more PRLs,” Renner said. “Most patients not meeting MS diagnosis — 88% — did not display any PRLs. Almost all patients with other diagnoses — 95% — did not display any PRLs.”

When people with CIS and RIS were added to those with an MS diagnosis, specificity improved to 95% (95% CI 89%-100%), but sensitivity fell to 67%, accuracy to 78%, and AUROC to 77%.

Future work will look at the effects of demographic and clinical characteristics on the presence and number of PRLs, Renner said. In addition, the researchers will examine how PRLs and the central vein sign may be used to predict MS progression.