Predictive value of cardiac magnetic resonance imaging for fatal arrhythmias in structural and nonstructural heart diseases

The risk stratification for fatal arrhythmias remains inadequate. Cardiac magnetic resonance (CMR) imaging provides a detailed evaluation of arrhythmogenic substrates. This study investigated the predictive capacity of multiparametric CMR for fatal ventricular arrhythmias (VAs) in a heterogeneous disease cohort. The study included 396 consecutive patients with structural heart disease (SHD, n = 248) and non-apparent SHD (n = 148) who underwent CMR scans between 2018 and 2022. The primary endpoint was fatal composite arrhythmias. Thirty-three patients (8.3 %) experienced fatal arrhythmias (25 with SHD, 8 with non-apparent SHD) over a median follow-up of 24 months. The independent risk factors for patients with SHD included syncope (hazard ratio HR = 5.347; P < 0.001), VA history (HR = 3.705; P = 0.004), right ventricular ejection fraction (RVEF) ≤ 45 % (HR = 2.587; P = 0.039), and the presence of late gadolinium enhancement (LGE) (HR = 4.767; P = 0.040). In the non-apparent SHD group, fatal arrhythmias were independently correlated with VA history (HR = 10.23; P = 0.005), RVEF ≤ 45 % (HR = 8.307; P = 0.015), and CMR myocardial abnormalities (HR = 5.203; P = 0.033). Patients at high risk of fatal arrhythmia in the SHD and non-apparent SHD groups exhibited 3-year event-free survival rates of 69.4 % and 83.5 %, respectively. CMR provides effective prognostic information for patients with and without apparent SHD. The presence of LGE, CMR myocardial abnormalities, and right ventricular dysfunction are strong risk markers for fatal arrhythmias.