Correlation between orbital imaging features of thyroid-associated ophthalmopathy and pupillary light reflex measurement.

Purpose: This study aimed to assess the factors associated with quantitative pupillary light reflex analysis and orbital magnetic resonance imaging (MRI) indicators in thyroid-associated ophthalmopathy (TAO) patients with different TAO severities, and their diagnostic significance of dysthyroid optic neuropathy (DON).

Methods: A retrospective cross-sectional analysis was conducted on 57 patients with TAO, involving 109 orbits. Using the EUGOGO severity grading system, patients were categorized into three groups: Mild TAO (45 orbits), Moderate-to-Severe TAO (48 orbits), and DON (16 orbits). All participants underwent comprehensive ophthalmological assessments, pupillary light reflex analysis using the RAPDx device (Konan Medical), and MRI imaging (GE 3.0 Signa Creator, GE Medical Systems). MRI measurements included orbital bone wall area, extraocular muscle area, and proptosis. Differences in clinical characteristics, pupillary function indicators, and MRI-derived indicators were analyzed using Generalized Estimating Equations (GEE). Correlations and trends between Latency Onset of Constriction (LOC) and MRI indicators were assessed through Pearson multivariate analysis and linear regression models. The diagnostic value of LOC and the Volume of the Medial Orbital Wall (VMW) for diagnosing DON was further evaluated using Receiver Operating Characteristic (ROC) curve analysis.

Results: The results revealed that LOC was significantly prolonged in the DON group compared to both the Mild TAO and Moderate-to-Severe TAO groups (p < 0.05 for both). LOC demonstrated strong positive correlations with Inferior Orbital Nerve Signal Loss (IONSL) (r = 0.494, p < 0.001), Proptosis (r = 0.448, p < 0.001), and Medial Rectus Area (MRA) (r = 0.428, p < 0.001). Multivariate binary logistic regression analysis identified LOC and VMW as independent predictors of DON. A predictive model combining LOC and VMW showed excellent diagnostic performance, with an Area Under the Curve (AUC) of 0.886 (p < 0.001), sensitivity of 90.5%, and specificity of 82.4%.

Conclusion: These findings underscore the critical roles of pupillary light reflex analysis and MRI in diagnosing and evaluating TAO. The significant correlations of LOC with IONSL, Proptosis, and MRA, along with its strong predictive value alongside VMW, highlight their utility as reliable diagnostic markers for DON.