DNA Methylation Alterations in Patients with Moderate-to-Severe Thyroid Eye Disease Treated with Intravenous Glucocorticoids

Background/Objectives: Genome-wide epigenetic alterations in thyroid eye disease (TED) remain insufficiently characterized, particularly in the context of intravenous glucocorticoid (ivGC) therapy. We aimed to characterize genome-wide DNA methylation (DNAm) alterations in patients with active, moderate-to-severe TED and assess differences in DNAm patterns observed following ivGC therapy and in relation to clinical response. Methods: We enrolled 24 euthyroid patients with moderate-to-severe TED and 7 healthy controls (HC). Peripheral whole-blood DNAm profiling was performed using reduced representation bisulfite sequencing before (TED-P1) and 12 weeks after ivGC therapy (TED-P2), with both time points compared with the same HC reference group. Within-patient DNAm changes were additionally explored at the individual level. Gene set enrichment (GSE) analysis investigated the biological context of identified methylation alterations. Results: Genome-wide and site-specific DNAm differences were identified between TED patients and HC, with a marked predominance of hypomethylation. Principal component analysis demonstrated separation of TED patients from HC, with differences in clustering patterns observed after treatment. GSE identified enrichment of pathways related to immune regulation, cell-cycle control and, among other categories, visual system development. Sensitivity analyses addressing smoking status and age supported the persistence of the predominant hypomethylation pattern. Exploratory analyses suggested differences in post-treatment DNAm patterns according to cumulative ivGC exposure, whereas no consistent DNAm signature associated with clinical treatment response was identified in this cohort. Conclusions: Patients with moderate-to-severe TED exhibited genome-wide and site-specific DNAm alterations characterized predominantly by hypomethylation and functionally relevant pathway enrichment. These findings require validation in larger, independent cohorts.