MicroRNA signature predicts post operative atrial fibrillation after coronary artery bypass grafting
Srinivasulu Yerukala Sathipati, Tonia Carter, Deepa Soodi, Nwaedozie Somto, Sanjay K. Shukla, John Petronovich, Glurich Ingrid, John H. Braxton, Param Sharma
Abstract
Early detection of atrial fibrillation (AFib) is crucial for altering its natural progression and complication profile. Traditional demographic and lifestyle factors often fail as predictors of AFib. This study investigated pre-operative, circulating microRNAs (miRNAs) as potential biomarkers for post-operative AFib (POAF) in patients undergoing coronary artery bypass grafting (CABG). We used an array polymerase chain reaction method to detect pre-operative, circulating miRNAs in seven patients who subsequently developed POAF after CABG (cases) and eight patients who did not develop POAF after CABG (controls). The top 10 miRNAs from 84 candidates were selected and assessed for their performance in predicting POAF using machine learning models, including Random Forest, K-Nearest Neighbors (KNN), XGBoost, and Support Vector Machine (SVM). The Random Forest and XGBoost models showed superior predictive performance, with test area under the curve (AUC) values of 0.76 and 0.83, respectively. Differential expression analysis revealed four upregulated miRNAs-hsa-miR-96-5p, hsa-miR-184, hsa-miR-17-3p, and hsa-miR-200-3p-that overlapped with the POAF-miRNA signature. The POAF-miRNA signature was significantly associated with various cardiovascular diseases, including acute myocardial infarction, hypertrophic cardiomyopathy, and heart failure. Biological pathway analysis indicated these miRNAs target key signaling pathways involved in cardiovascular pathology, such as the MAPK, PI3K-Akt, and TGF-beta signaling pathways. The identified miRNAs demonstrate significant potential as predictive biomarkers for AFib post-CABG, implicating critical cardiovascular pathways and highlighting their role in POAF development and progression. These findings suggest that miRNA signatures could enhance predictive accuracy for POAF, offering a novel, noninvasive approach to early detection and personalized management of this condition.