Key Indicators for Guiding Tocilizumab Therapy to Prevent Orbital Decompression Surgery in Hormone-Resistant Dysthyroid Optic Neuropathy

Wenjun Shu¹Mingxu Yan¹Lu Gan²Lu Li³Hongyue Tao³Zhiyu Peng⁴Jinghan Wang²Xiaofeng Li²Xintong Lin²Haifeng Chen²Jie Guo¹Kang Xue²Hongguang Cui⁴Heng Lou²Binbin Xu²Jinwei Cheng²Hongying Ye³Yiming Li³Jiang Qian²Rui Zhang^2*Ruiqi Ma^2*Fangfang Zeng^3*

• ¹Fudan University, Shanghai, China
• ²Eye and Ent Hospital, Fudan University, Shanghai, Shanghai Municipality, China
• ³Huashan Hospital, Fudan University, Shanghai, Shanghai Municipality, China
• ⁴The Hospital of Zhejiang University, Hangzhou, Zhejiang Province, China

Introduction: Tocilizumab (TCZ) has been demonstrated to be effective in treating thyroid-associated ophthalmopathy (TAO); however, its efficacy in hormone-resistant dysthyroid optic neuropathy (DON) remains unclear. This study aims to identify baseline and post-treatment indicators that can predict the necessity for surgical intervention following TCZ therapy.Methods: Thirty-one hormone-resistant DON patients treated with TCZ were categorized into surgery (n=15, 7 males, 8 females) and non-surgery (n=16, 8 males, 8 females) groups based on their post-TCZ surgical status. Retrospective comparisons between the two groups were performed using pre- and post-treatment ophthalmic assessments, biomarkers, and orbital magnetic resonance imaging (MRI) scans. The predictive value of identified variables was evaluated through receiver operating characteristic (ROC) curve analysis and logistic regression analysis.Results: The surgery group exhibited a significantly higher baseline thyroid-stimulating hormone receptor antibody (TRAb) (P=0.001) and a positive change in the maximal signal intensity ratio of extraocular muscle to temporalis muscle (SIR(EOM/temporalis)MAX), whereas the non-surgery group demonstrated a negative change (P<0.001). A TRAb cut-off value of ≤5.07 IU/L predicted non-surgery with 93.3% sensitivity and 81.2% specificity, while a SIR(EOM/temporalis)MAX cut-off value of ≤-1.83 had 86.7% sensitivity and 87.5% specificity. The area under the curve was 0.846 for TRAb and 0.863 for SIR(EOM/temporalis)MAX. Multivariate regression confirmed SIR(EOM/temporalis)MAX change (P=0.017) as an independent predictor of surgical intervention. Linear regression revealed a significant correlation between SIR(EOM/temporalis)MAX change and TCZ dosage in the non-surgery group (P=0.005), but not in the surgery group.Discussion: The percentage change in SIR(EOM/temporalis)MAX following TCZ treatment and baseline TRAb can serve as predictors of the necessity for surgical intervention in hormone-resistant DON. These indicators assist in the personalization of TCZ therapy and in the avoidance of unnecessary surgical procedures.