Extraocular muscle Diffusion Weighted Imaging as a quantitative metric of posterior orbital involvement in thyroid associated orbitopathy

Abstract

Abstract Objectives The clinical activity score (CAS) and European severity scale (ESS) are established clinical tools to assess thyroid eye disease (TED) but are limited in terms of subjectivity and their reliability in non-Caucasian individuals, and can underestimate significant disease in the posterior orbit. Preliminary data from pilot studies have shown that diffusion-weighted imaging (DWI) using extraocular muscle (EOM) apparent diffusion coefficient (ADC) measurements may provide complementary information in TED. This study expands on previous research to assess for correlations between clinical scores and EOM-ADCs in stratifying disease activity and severity in a large patient cohort from an ethnically diverse population. Methods A retrospective review of TED clinics between 2011 and 2021 identified 96 patients with a documented CAS and ESS and an orbital MRI that included DWI. From regions of interest manually placed on EOM bellies, the highest ADC was computed for each patient and analysed for correlations and associations with CAS and ESS using Spearman Rank correlation and Mann–Whitney U tests, and any potential discriminatory cut-offs using Receiver Operator Curve analyses. A p-value < 0.05 indicated statistical significance. Results EOM-ADCs showed a positive association with CAS (p ≤ 0.001). EOM-ADCs were higher in sight-threatening compared to mild disease (p ≤ 0.01). A cut-off of 995 mm2/s achieved AUC = 0.7744, equating to 77% sensitivity and 67% specificity for discrimination between mild-moderate and sight-threatening disease. Conclusion EOM-ADCs correlate with higher scores of disease severity and activity in TED. Besides providing quantitative data to support clinical tools, EOM-ADC cut-offs may identify patients at risk of developing sight-threatening diseases. Critical relevance statement This study critically evaluates the limitations of conventional clinical assessment tools for TED and demonstrates the utility of DWI scans with ADC measurements in identifying active disease, offering valuable insights to advance clinical radiology practice. Key Points Conventional tools for TED assessment have subjective limitations. ADCs from non-echoplanar diffusion-weighted imaging correlate with clinical activity. Non-echoplanar diffusion-weighted imaging offers quantitative assessment to aid clinical practice reliability. Graphical Abstract