PURPOSE: High-dose intravenous glucocorticoids are the standard first-line treatment in active, moderate to severe and severe thyroid eye disease (TED). We evaluate the usefulness of clinical activity score (CAS) and thyroid-stimulating immunoglobulin (TSI) as predictors and/or post-treatment markers of corticoresistance in patients with TED and the effect of rituximab in second-line treatment. METHODS: We enrolled 236 patients with an active TED into this retrospective single-tertiary-center cohort study. All patients were initially treated with high-dose systemic glucocorticoids. Rituximab was later administered to 29 of 42 corticoresistant patients. RESULTS: The CAS of the corticoresistant patients was significantly higher both before (p = 0.0001) and after (p = <
0.0001) first-line treatment compared to the corticosensitive group. ROC analysis established the cut-point value as CAS ≥ 2.5 with a sensitivity of 96.3%, specificity of 57.5% and area under the curve of 82.8%. In 22 patients treated with rituximab, CAS gradually decreased to zero values without reactivation during extended follow-up. There was no difference in the TSI of corticosensitive and corticoresistant patients before or after first-line therapy. CONCLUSION: CAS ≥ 2, after first-line treatment, could be used as a corticoresistance marker. Corticoresistant patients should be subject to long-term follow-up for early detection of reactivation to reduce the delay to second-line treatment. Rituximab is a well-tolerated choice of second-line treatment and has a long-lasting effect on disease activity. Although TSI is a valuable biomarker of Graves' disease and TED activity, according to our results, TSI cannot be used as a marker of corticoresistance.