Using a diagnostic threshold of at least 1 paramagnetic rim lesion, MS was identified with a sensitivity of 80% and specificity of 88%, which increased to 95% after excluding patients with RIS and CIS.
Patients who are referred for a new evaluation for clinical or radiological suspicion of multiple sclerosis (MS) may benefit from the diagnostic capabilities of paramagnetic rim lesions (PRL), new findings indicate. In the first ever multicenter study of its kind, PRL demonstrated a high specificity for diagnosing MS using the 2017 McDonald criteria.1
Lead author Brian Renner, MD, researcher, Mouth Cedar Sinai, presented these findings as part of the “Emerging Concepts in MS” session at the Americas Committee for Treatment and Research in Multiple Sclerosis (ACTRIMS) Forum 2022, held February 24-26, in West Palm Beach, Florida. The audience was captivated with Renner’s presentation as it could have potential implications on the diagnostic process of MS in the future.
In the analysis, 95 patients who had clinical or radiological suspicion of MS underwent phase imaging that included high-resolution echo-planar-imaging scans, as well as 3D T2-FLAIR images, which were acquired at 3T. Three independent raters were blinded during the initial evaluation, followed up by group consensus with an expert rater. PRLs were defined by a complete or partial hypointense rim surrounding an isointense core on phase colocalizing with a hyperintense T2-FLAIR lesion.
Of the entire cohort, 46% (n = 44) met McDonald 2017 criteria and 39% (n = 37) were given an alternative diagnosis to MS. The remaining 14 patients were considered at risk of developing MS, with diagnoses of clinically isolated syndrome (CIS) or radiologically isolating syndrome (RIS). In total, 41 of the 95 patients had at least 1 PRL recorded, whereas in the remaining 53, no PRL was observed. For those who were diagnosed with MS, a total of 122 PRLs were recorded, with 35 of the 44 (79.5%) individuals showing at least 1 PRL. In comparison, for those who did not meet an MS diagnosis, only 6 of the 51 (11.7%) individuals had 1 or more PRLs.
There were 6 patients that did not have any MS diagnosis that did have PRLs. Two patients were alternatively diagnosed with post-infectious encephalomyelitis or an unspecified demyelinating disorder and had 14 total PRL recorded. "It’s important to recognize them," Renner said. "Of course, these patient aredifficult to diagnose because they did not have blue bands and we had no evidence of clinical events during medical history." The remaining 4 included 3 patients with RIS and 1 patient with CIS, with a total of 13 PRLs observed.
The presence of 1 or more PRL resulted in high sensitivity (80%; 95% CI, 68-91), high specificity (88%; 95% CI, 79-98), high accuracy (84%; 95% CI, 75-91), and excellent area under the curve (81%) for diagnosing MS. When removing the RIS and CIS patients from the cohort, the specificity of PRL increased to 95%.
"We looked at 80% of our cohort for MS-only showing PRL, whereas in previous studies we’ve seen generally around 50%,” Renner added. "This is despite similar techniques, including grading, processing, and evaluation. We do believe that is likely related to the nature of the cohort."
There was one outlier who did not receive an MS diagnosis but had PRL counts over 10, the highest among any patient. In the Q&A session of the presentation, a member of the audience asked about the presence of central vein sign (CVS) for that patient, for which Renner responded that "they have 23% of CVS, but they didn’t meet any threshold criteria for CVS. They did have some, but this was given to a special case."
Renner and his colleagues plan on conducting future evaluations of PRL, including how these PRL levels may change over time. They will investigate the demographic and clinical characteristics of the cohort and how they compare to the presence and number of PRLs. Additionally, the research will analyze the combination of biomarkers such as PRL and central vein sign together in relation to misdiagnosis.
For more coverage of ACTRIMS 2022, click here.