Scar characterisation to predict arrhythmia and death in ischaemic cardiomyopathy: insights from the REVIVED trial

Abstract Background Ischaemic cardiomyopathy continues to have unacceptably high mortality despite contemporary medical and device therapy (1). Ventricular arrhythmias are a leading cause of both death and morbidity in this population. Current guidelines recommend implantable cardioverter defibrillator (ICD) implantation in patients with a left ventricular ejection fraction (LVEF) below 35% but many who meet criteria never receive a therapy and there is a persistent sudden cardiac death risk in those who do not. Scar characteristics are not currently included in risk stratification algorithms, however both volume and entropy, a measure of myocardial heterogeneity, have been linked to arrhythmic risk. Purpose To assess whether left ventricular scar burden and morphology predicts fatal and non-fatal arrhythmia, independent of LVEF in severe ischaemic cardiomyopathy. Methods REVIVED-BCIS2 was a prospective, randomized controlled trial, which recruited 700 patients from 40 centres across the United Kingdom (1). Trial eligibility required the presence of ≥4 dysfunctional but viable myocardial segments; the majority of patients had viability assessed by cardiac magnetic resonance imaging (CMR). Semi-quantitative CMR scar burden was determined through consensus by two CMR experts in a blinded core laboratory. Scar morphology was further characterised by left ventricular entropy measured using novel dedicated software (TexRAD Ltd., Feedback plc, UK). The primary outcome was the composite of all-cause death and aborted sudden death (in turn, defined as appropriate ICD therapy or resuscitated cardiac arrest). Cox proportional hazard models were used to determine the effect of scar burden and entropy on clinical outcomes, adjusted for baseline LVEF, extent of coronary disease, type of implantable device and randomized treatment. Results Four hundred and ninety one patients underwent viability assessment by CMR and 478 were included in this analysis (Figure 1). Mean age was 69 ± 9; median follow up duration was 3.3 [2.3-4.9] years. By the end of trial follow up, 257/478 had a device in situ (CRT-D 99, ICD 147, CRT-P 11). The primary outcome occurred in 182 patients (38.1%). Each 10% increase in scar volume was associated with a 24% increase in the incidence of the primary outcome (adjusted HR per 10% 1.24, 95% CI: 1.11 to 1.39, p = 0.0002)(Figure 2). LVEF at baseline was not predictive (adjusted HR per 5% increase 0.94, 95% CI 0.86 – 1.02, p = 0.13). The relationship between LV entropy and outcomes will be presented at ESC. Conclusions In patients with severe ischaemic cardiomyopathy, scar volume was predictive of death and potentially fatal arrhythmias. Left ventricular ejection fraction was not a significant predictor in this cohort. Scar characteristics should form an integral part of risk stratification.Consort diagramPrimary outcome by scar burden