Orbital MRI for thyroid eye disease activity staging: a systematic review and meta-analysis

OBJECTIVE: To evaluate the diagnostic efficacy of magnetic resonance imaging (MRI) for staging the activity of thyroid eye disease (TED). METHODS: A comprehensive search was conducted in the Medline, PubMed, Cochrane Library, Web of Science Core Collection, and EMBASE databases for relevant studies. The risk of bias and applicability of the included studies were assessed using the Prediction Model Risk of Bias Assessment Tool (PROBAST). Statistical analyses were performed using STATA software. Subgroup and sensitivity analyses were conducted to explore sources of heterogeneity, while publication bias was assessed using funnel plots and Egger’s test. RESULTS: This meta-analysis included 16 studies involving a total of 2,317 eyes published between 2020 and 2025, with 15 studies conducted in mainland China. All studies were found to have a high risk of bias, primarily due to retrospective design and lack of validation. The pooled area under the curve (AUC) value for MRI in differentiating active and inactive TED was 0.84 [95% CI (0.81–0.86), I2 = 66.5%, p < 0.001], indicating fair diagnostic performance. T2-weighted MRI sequences showed superior diagnostic efficacy compared to T1-weighted sequences [AUC = 0.84, 95% CI (0.80–0.87), I2 = 63.4%, p = 0.001; AUC = 0.81, 95% CI (0.77–0.85), I2 = 0%, p = 0.642; respectively]. Subgroup analysis based on orbital tissue type revealed that the extraocular muscles provided the highest diagnostic efficacy compared to the lacrimal gland and orbital fat [AUC = 0.83, 95% CI (0.80–0.87), I2 = 71.5%, p < 0.001]. CONCLUSION: Orbital MRI demonstrates a valuable diagnostic efficacy for staging TED activity to assist the CAS, with extraocular muscles serving as the most promising indicator of disease activity. This non-invasive imaging modality offers a tool for assessing pathological changes in TED. Future studies should focus on the diagnostic efficacy models for MRI in TED staging in diverse populations for expanding its global generalizability.