# Quantitative Magnetic Resonance Imaging Biomarkers in the Evaluation of Disease Activity in Thyroid Eye Disease: A Retrospective Study

## TL;DR

This study shows that MRI measurements of eye muscles can accurately detect active thyroid eye disease, offering a non-invasive diagnostic tool.

## Contribution

The study introduces MRI-derived ratios (total EOM/exophthalmos and IR/total EOM) as novel biomarkers for assessing TED activity.

## Key findings

- Inferior rectus and medial rectus muscles were significantly larger in active TED patients.
- Total EOM/exophthalmos ratio (cut-off 0.647) and IR/total EOM ratio (cut-off 0.362) showed high sensitivity and specificity for active disease.
- MRI biomarkers reliably distinguish active from inactive TED with high diagnostic accuracy.

## Abstract

Objective

The aim of this study was to determine whether quantitative measurements of extra-ocular muscles and structures on magnetic resonance imaging (MRI) correlate with disease activity in thyroid eye disease (TED).

Method

A retrospective review of 162 patients, between June 2014 and June 2024, with TED who underwent MRI orbits was conducted. The patients were categorised into two groups: those with clinical and imaging features consistent with active TED (n = 73; 45.1%) and those with inactive disease (n = 89; 54.9%). T1-weighted imaging, T2-weighted imaging, and short tau inversion recovery (STIR) on MRI sequences were used to measure exophthalmos value, extra-ocular muscle diameters, orbital fat thickness, and optic nerve diameters. An independent t-test was carried out to determine statistical significance (p < 0.05) in structural measurements between both groups. The area under the receiver operating characteristic curve (ROC) analysis was used to deduce the diagnostic accuracy of structural measurements in predicting disease activity.

Results

Inferior rectus (IR) and medial rectus (MR) muscles were significantly larger in patients with active disease (p < 0.05 and p < 0.05, respectively). Total extraocular muscle (EOM) was also significantly larger in this group (p < 0.05). Total EOM/exophthalmos and IR/total EOM ratios showed higher values in patients with active disease (p < 0.05). ROC analysis showed that the cut-off point of total EOM/exophthalmos ratio was 0.647, with a sensitivity of 90.4% and specificity of 88.8% (AUC = 0.96; 95% CI: 0.94-0.99), whilst the IR/Total EOM ratio had a cut-off point of 0.362, with a sensitivity of 90.4% and a specificity of 87.6% (AUC = 0.94; 95% CI: 0.9-0.98).

Conclusion

MRI-derived quantitative measurements, particularly the total EOM/exophthalmos and IR/total EOM ratios, reliably distinguish active from inactive TED, on MRI, with high sensitivity and specificity. These imaging biomarkers offer a valuable, non-invasive tool for assessing disease activity and monitoring progression in TED.