MRI activity in pre-symptomatic multiple sclerosis: the role of paramagnetic rim lesions beyond the 2024 McDonald diagnostic criteria.

Abstract

The 2024 McDonald criteria allow multiple sclerosis (MS) diagnosis in pre-symptomatic individuals but their association with MRI activity remains uncertain. We assessed whether paramagnetic rim lesions (PRL) predict future MRI activity beyond the 2024 McDonald criteria. We retrospectively included individuals with incidental MRI findings suggestive of MS, ≥6 months of radiological follow-up, 3D T2*-weighted echo-planar-imaging MRI. Clinico-radiological data included PRL, central vein sign (CVS) Select-6, cerebrospinal fluid positivity, optic nerve, spinal cord involvement and disease-modifying treatment. Time to first MRI activity and annualised lesion rate (ALR) were analysed using multivariable-penalised Cox and mixed-effects Poisson models. Among 68 patients (mean age 38 years; 79.4% female; mean follow-up 55.7 months), 21 showed MRI activity. PRL were associated with shorter time to first MRI activity (HR 5.66, 95% CI 1.33 to 28.39, p=0.019). PRL (RR 3.45, 95% CI 1.40 to 8.47, p=0.007) and CVS positivity (RR 4.09, 95% CI 1.17 to 14.27, p=0.027) predicted higher ALR. Fulfilment of the 2024 McDonald criteria associated with increased risk of future MRI activity (HR=11.95, 95% CI 2.77 to 51.51, p=0.001; Cox regression). Adding PRL presence to the criteria changed classification performance (χ²=18, p<0.001; McNemar's test), improving specificity and accuracy while reducing sensitivity. PRL are associated with subsequent radiological activity in pre-symptomatic MS, featuring complementary predictive value beyond biomarkers currently included in the 2024 McDonald criteria.