Prevalence and Incidence of Paramagnetic Rim Lesions in Multiple Sclerosis: A Systematic Review and Meta-Analysis

Abstract

Background Paramagnetic rim lesions (PRLs) hold implications for diagnosis and prognosis of multiple sclerosis (MS). However, the factors influencing their identification have yet to be fully clarified. Purpose To conduct a systematic review and meta-analysis assessing clinical and imaging-related factors affecting PRL identification, incidence, prevalence, evolution over time, and topography in patients with MS. Materials and Methods An English literature search was performed using PubMed and Embase databases to identify eligible studies published by February 8, 2024, with no restrictions on the earliest publication date. Risk of bias was assessed with a customized 12-item checklist. Meta-analyses of proportions were used to derive PRL-pooled prevalence, incidence, and CIs, at both the lesion and patient levels. Heterogeneity was assessed using the Cochran Q test and I² statistic. Sources of heterogeneity were explored via subgroup analyses and meta-regression analyses. Results Fifty-eight studies investigating 4532 patients with MS (mean age, 39 years ± 1.1 [SD]) were included. PRL prevalence was 0.12 (95% CI: 0.09, 0.16; 32 studies) at the lesion level and 0.52 (95% CI: 0.47, 0.58; 37 studies) at the patient level. Incidence was 0.23 (95% CI: 0, 0.75; three studies) at the lesion level and 0.11 (95% CI: 0, 0.47; two studies) at the patient level. No difference was observed for the lesion-level prevalence according to the PRL definition used (P = .66) or brain compartment (P = .34), and for the patient-level prevalence according to age (P = .29), disease duration (P = .72), or MS phenotype (P = .29). A subgroup analysis revealed differences in lesion-level prevalence based on lesion stage, field strength, MRI sequence, and resolution (P < .001 for all). The heterogeneity in prevalence across studies was high on average (I² = 88.8% and 99.2% at patient-wise and lesion-wise level, respectively). In 21% (12 of 58) of the studies, a clear definition of PRLs was not provided. Conclusion Although PRLs could be identified in approximately half of patients with MS undergoing imaging, the incidence of new PRLs was relatively low, and PRL prevalence was heterogeneous across studies. © RSNA, 2025 Supplemental material is available for this article.