DNA Methylation Predicts Adverse Outcomes of Coronary Artery Disease

Background Adverse outcomes including myocardial infarction and stroke render coronary artery disease (CAD) a leading cause of death worldwide. Biomarkers that predict such adversity enable closer medical supervision and opportunities for improved outcomes. Methods and results We present a study of genome-wide DNA methylation profiling in 933 CAD patients with up to 13 years of clinical follow-up. We discovered 115 methylation sites associated with poor prognosis and inferred that cellular senescence, inflammation, and high-density lipoprotein mediated the adversity. We built succinct prognostic models combining a few methylation sites and clinical features, which could stratify patients of different risks. Furthermore, we assessed genetic regulation of the differential methylation by interrogating QTL effects. Prognostic genes such as FKBP5 , UBE2E2 and AUTS2 appeared recurrently in various analyses and were validated in patients of myocardial infarction and stroke. Conclusions Our study provides prognostic models for clinical application and revealed methylation biomarkers and mechanisms of CAD adverse outcomes.