Abstract
Graves' disease (GD) and Thyroid Eye Disease (TED) are autoimmune disorders characterized by significant heterogeneity in treatment response. Up to 50% of GD patients relapse after antithyroid drug (ATD) withdrawal, and a substantial portion of TED patients (20-50%) are resistant to first-line glucocorticoid (GC) therapy. This review evaluates the current evidence on epigenetic modifications as predictive biomarkers to guide personalized treatment. We synthesized recent findings (up to 2025) from PubMed, focusing on DNA methylation and microRNAs (miRNAs). For GD, ATD relapse risk is linked to a persistent "epigenetic memory" in T cells, notably the hypomethylation of Th17-associated genes. Circulating miRNA signatures, including miR-346, miR-23b-5p, and miR-92a-3p, also show promise in predicting remission. For TED, GC sensitivity is strongly correlated with specific circulating miRNAs. High pre-treatment levels of miR-146a predict a positive response (100% positive predictive value), while low levels of miR-224-5p predict non-responsiveness. While DNA methylation is confirmed in TED pathogenesis, its predictive role is unstudied. Major research gaps persist, particularly the near-total absence of data on histone modifications as predictive markers and the lack of epigenetic predictors for new biologics treatments, which currently rely on genetic or pharmacokinetic markers. Epigenetic biomarkers represent a promising frontier for stratifying patients and optimizing therapeutic strategies in Graves' autoimmunity.