Abstract
BACKGROUND: Atrial fibrillation (AF) is a major complication of hypertrophic cardiomyopathy (HCM) with significant prognostic implications. Current risk prediction models lack the integration of comprehensive cardiac magnetic resonance (CMR) metrics and subtype-specific analyses. METHODS: A retrospective study of 405 HCM patients (86 with AF) was performed from 2019 to 2024. After excluding highly correlated variables (|r| > 0.8), the cohort was split into training and validation sets in a 7:3 ratio. Least Absolute Shrinkage and Selection Operator (LASSO) regression and multivariable logistic regression analyses were used to identify predictors, with model performance assessed via receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis. Subgroup analyses were conducted for obstructive (HOCM) and non-obstructive (HNCM) subtypes. RESULTS: Independent predictors of AF in the overall HCM cohort included right atrial diameter anteroposterior (RAD anteroposterior: odds ratio (OR) = 1.819, 95% confidence interval (CI) 1.130-3.007; p = 0.016), left ventricular end-systolic volume (LVESV: OR = 0.978, 95% CI 0.963-0.991; p = 0.002), septal mitral annular plane systolic excursion (MAPSE septal: OR = 0.850, 95% CI 0.736-0.976; p = 0.023), tricuspid annular plane systolic excursion (TAPSE: OR = 0.919, 95% CI 0.852-0.987; p = 0.022), and maximum left atrial volume (MaxLAV: OR = 1.016, 95% CI 1.004-1.029; p = 0.010). The model achieved an area under the curve (AUC) value of 0.850 in the training set and an AUC of 0.861 in the validation set. The HOCM subtype predictors included septal MAPSE and left atrial ejection fraction (LAEF); meanwhile, the HNCM predictors included septal MAPSE, maximal left atrial volume (MaxLAV), and right atrial ejection fraction (RAEF). CONCLUSIONS: A validated multiparametric CMR model can accurately predict AF risk in HCM patients, with subtype-specific predictors enabling personalized monitoring and early intervention.