MRI in detection of acute optic neuritis: a systematic review and meta-analysis

Abstract

Background MRI is frequently performed as part of the diagnostic workup in suspected acute optic neuritis (AON). Conflicting results exist regarding the overall diagnostic performance of MRI and different protocols in AON diagnosis. Purpose This meta-analysis aims to evaluate a diagnostic performance of MRI for the detection of AON. Data Sources Search was performed on PubMed/MEDLINE, EMBASE and Cochrane databases until December 2024. Study Selection Studies were assessed independently by two authors following PRISMA guidelines and were screened to exclude duplicates and review articles. Published original articles that met the following criteria were considered eligible for meta-analysis: (a) Patients with acute loss vision; (b) evaluated diagnostic performance of MRI for AON detection compared to a valid reference standard and (c) provided data for the meta-analytic calculations. We excluded studies that couldn’t allow defining how many optic nerves affected. Remaining articles were reviewed in a full-text review. Data Analysis The bivariate random-effects model was used to find the pooled sensitivity and specificity of MRI for AON and 95 % CIs. The hierarchical summary receiver operating characteristic model was used to draw the summary receiver operating characteristic curve and calculate the area under the curve (AUC). Data analysis was conducted using RStudio version 2023.06.0+421. Data Synthesis A total of 11 studies involving 5 sequences, 522 patients, and 856 optic nerves, met the inclusion criteria and were included in the meta-analysis. MRI had a pooled sensitivity and specificity of 84.0 % (95 % CI: 78.0; 88.6) and 86.4 % (95 % CI: 80.7;90.6), respectively. Per sequence, contrast-enhanced T1WI (CE-T1WI) sequence demonstrated the highest sensitivity at 81.2 % (95 % CI: 70.1; 88.9), while the T2-FLAIR sequence demonstrated the highest specificity at 83.1 % (95 % CI: 66.6; 92.4), among all sequences. The AUC of MRI was 90.6 % (95 % CI: 82.7, 91.5). Limitations There were a low available number of studies. The diagnostic accuracy in distinguishing hyper T2-FLAIR from acute optic neuritis or its principal differential diagnoses has not been clearly established. The included studies presented high heterogenicity from differences in MRI protocols, patient demographics, and study designs. Conclusions MRI has high sensitivity and specificity for the diagnosis of AON. Incorporating sensitive-liquid sequences and contrast into MRI may increase the detection of AON. Brain or orbital dedicated T2-FLAIR sequences can help to confirm clinical suspected AON.