Predicting dysthyroid optic neuropathy in moderate-to-severe thyroid eye disease: a clinically applicable nomogram.

Objective: Dysthyroid optic neuropathy (DON) is a severe complication of thyroid eye disease (TED) with limited early detection methods. This study aimed to investigate the clinical characteristics of patients with TED who developed DON and to establish a predictive model for early identification of high-risk cases. Methods: Herein, 257 TED patients were prospectively included, of whom 68 (26.5%) developed DON. All patients were divided into derivation and validation cohorts, and Least Absolute Shrinkage and Selection Operator (LASSO) regression and logistic regression analyses were applied to identify clinical factors and construct a prediction model. Results: In the derivation cohort (185 TED patients), 49 (26.5%) developed DON. DON patients showed significantly higher prevalence of pretibial myxedema (PTM) (22.4 vs 5.9%, P = 0.001), diabetes mellitus (18.4 vs 7.4%, P = 0.029), older age (58.04 ± 11.30 years vs 47.99 ± 10.65 years, P < 0.001), higher CAS (5 vs 4, P < 0.001), elevated triglyceride (TG) levels (1.44 mmol/L vs 1.15 mmol/L, P = 0.042), and lower visual functioning (VF) (43.75 vs 62.50, P < 0.001). LASSO regression analysis identified age, PTM, TG, VF, and CAS as independent predictors of DON. The developed nomogram presented AUCs of 0.853 (95% CI: 0.792-0.914) and 0.856 (95% CI: 0.762-0.950) in the derivation and validation cohorts, respectively. Conclusions: Altogether, the findings of this study identify advanced age, elevated CAS, increased TG, lower VF, and PTM as significant predictors of DON in patients with TED. The proposed nomogram offers a practical clinical tool for risk stratification, providing clinicians with an approach for individualized risk assessment and timely therapeutic intervention.