Endometriosis is a heterogeneous, estrogen-dependent inflammatory disorder that affects up to 15% of reproductive age women. Progestin-based therapies are the most commonly prescribed initial treatment; however, approximately one-third of patients exhibit progestin resistance, leading to inadequate symptom relief and discontinuation. Given the role of epigenetic dysregulation in endometriosis and its impact on hormonal responsiveness, we aimed to identify if circulating leukocyte DNA methylation signatures were associated with progestin treatment response and could serve as a non-invasive predictive biomarker. We conducted a prospective cohort study of 31 women with surgically confirmed endometriosis, categorized as progestin responders (n = 10) or non-responders (n = 21) based on clinical outcomes. Buffy coat-derived leukocytes were processed for whole-genome methylation analysis using enzymatic methyl-seq. Differentially methylated CpG sites were identified using logistic regression, and candidate genes were subject to Receiver Operating Curve analysis. A stepwise logistic regression model was developed to identify the minimal methylation gene set predictive of treatment response. Internal validation included permutation and bootstrap testing. Responders and non-responders did not differ significantly in baseline demographics or clinical variables. There were 1,439 genes that were significantly differentially methylated between responders and non-responders. A three-gene methylation signature, MMP20, NRXN1, and RNA5-8SN5, distinguished responders from non-responders with high accuracy (ROC AUC = 0·952). Internal validation confirmed model robustness (bootstrap AUC = 0·907, 95% CI: 0·80–0·957; permutation p < 0·001). Circulating leukocyte methylation profiles can serve as noninvasive biomarkers of progestin responsiveness in endometriosis. Our findings suggest that blood-based epigenetic profiling may inform personalized treatment decisions, avoiding prolonged treatment with ineffective therapy.
Building similarity graph...
Analyzing shared references across papers
Loading...
Cevik et al. (Wed,) studied this question.
www.synapsesocial.com/papers/69be37726e48c4981c67715f — DOI: https://doi.org/10.1186/s40364-026-00907-1
E. Cansu Cevik
Ramanaiah Mamillapalli
Giacomo Sferruzza
Biomarker Research
Yale University
Building similarity graph...
Analyzing shared references across papers
Loading...