Age-Dependent Diagnostic and Treatment Response Prediction of TBII and TSI in Graves' Orbitopathy: Integration with Orbital MRI Biomarkers

Cheng Song¹Xiao Wang¹Qintao Ma¹Genfeng Yu¹Yan Zhu¹Zimeng Huang¹Tian Chen¹Kai Huang²Yuanping Hai¹Hai-Xiong Chen²Yongbo Duan²Jie Shen^1*

• ¹Department of Endocrinology and Metabolism, Shunde Hospital, Southern Medical University (The First People’s Hospital of Shunde, Foshan), Foshan, Guangdong, China
• ²Shunde Hospital of Southern Medical University, Foshan, China

Abstract Background Thyrotropin-binding inhibitory immunoglobulin (TBII) is involved in the pathogenesis of Graves’ orbitopathy (GO). Although thyroid-stimulating immunoglobulin (TSI) may offer superior diagnostic or prognostic value, its utility remains incompletely defined. Methods This retrospective study included 177 consecutive patients with GO, comprising 128 newly diagnosed cases and 49 individuals with a history of intravenous methylprednisolone (IVMP) therapy. All participants underwent standardized evaluations, including endocrine assessments, ophthalmic examinations, and orbital magnetic resonance imaging (MRI). MRI was used to quantify the maximum signal intensity ratio of the extraocular muscles (SIR max) and the volume of the extraocular muscles (EMV). TBII levels were measured using a third-generation competitive-binding immunoassay, and TSI levels were assessed using a bridge-based chemiluminescence immunoassay. Treatment response to combined IVMP and mycophenolate mofetil (MMF) was evaluated in a subgroup of 70 newly diagnosed patients. Results In newly diagnosed patients, the TSI positivity rate was significantly higher than that of TBII (P<0.001). Notably, only within this subgroup did both antibodies show a positive correlation with the clinical activity score (CAS) (TBII: r=0.354, P<0.001, TSI: r=0.323, P<0.001) and SIR max (TBII: r=0.234, P=0.008; TSI: r=0.175, P=0.048). Multivariate analysis identified age (β=0.297, P=0.002), TBII (β=0.365, P<0.001), and TSI (β=0.325, P=0.003) as independent factors associated with CAS. An age-stratified analysis demonstrated stronger correlations between antibody levels and CAS in patients older than 45 years (TBII-CAS: r=0.410; TSI-CAS: r=0.426), with correspondingly higher areas under the curve (AUC) for identifying active disease (TBII: 0.736; TSI: 0.760). In evaluating treatment response to IVMP combined with MMF, higher baseline TSI levels (OR=1.086, 95% CI: 1.014–1.163), elevated SIR max (OR=9.205, 95% CI: 1.072–79.053), and lower HDL levels (OR=0.033, 95% CI: 0.003–0.346) were independently associated with poor outcomes. In contrast, TBII did not retain independent predictive value in this treatment context. Conclusion In newly diagnosed patients with GO, both TBII and TSI levels were associated with disease activity, with their diagnostic value being more pronounced in older individuals. TSI demonstrated a higher positivity rate than TBII and served as an independent predictor of treatment response to IVMP combined with MMF.