Incremental prognostic value of epicardial adipose tissue volume and thickness assessed by cardiac MRI in hypertrophic cardiomyopathy

Abstract

Aims This study aimed to investigate the prognostic and incremental value of epicardial adipose tissue (EAT) parameters, including volume index (EATVI) and regional fat thickness, assessed by cardiac magnetic resonance (CMR) in patients with hypertrophic cardiomyopathy (HCM). Methods and results In this retrospective cohort of 457 HCM patients who underwent CMR between 2018 and 2024, EATVI and regional fat thickness at key cardiac grooves were quantified along with conventional clinical and imaging risk factors. Over a median follow-up of 27.5 months (interquartile range 15.6–47.6), major adverse cardiovascular events (MACE) occurred in 18.1% of patients. Significantly higher EATVI (72.21 ± 14.21 mL/m2 vs. 56.71 ± 10.36 mL/m2, P < 0.001) and increased regional fat thickness (P < 0.05) were observed in patients with MACE. In multivariable Cox regression, EATVI remained an independent predictor of MACE (hazard ratio 1.03, 95% confidence interval 1.01–1.05, P < 0.001). Incorporating EATVI into the clinical-CMR model improved discrimination (C-statistic 0.82–0.84; P = 0.002), enhanced calibration, and provided greater net clinical benefit on decision-curve analysis. When appended to the 2014 European Society of Cardiology (ESC) HCM Risk-SCD model, EATVI further improved discrimination (C-statistic 0.80 vs. 0.71; P < 0.001) and calibration. Kaplan–Meier analyses using quartiles and the 62.5 mL/m2 cutoff showed progressively worse event-free survival with higher EATVI; within the ESC-defined low-risk subgroup, curves also separated significantly (log-rank P < 0.05). Time-dependent receiver operating characteristic analyses confirmed stable predictive performance of these parameters. Conclusion EATVI and regional fat thickness derived from CMR independently predict adverse outcomes in HCM and improve risk stratification. Comprehensive EAT assessment may serve as a promising imaging biomarker for personalized management.