Using Cox regression models, T2 predicted death in AL amy-loidosis (hazard ratio, HR,1.48, 95% CI 1.20-1.82) and remained significant after adjusting for EF and ECV (HR 1.31, 95% CI 1.04-1.66) (Abstract 1. Figure 2). Conclusion Patients with AL amyloidosis have a worse prognosis compared to ATTR despite having less cardiac amyloid infiltration. T2 was significantly higher in untreated AL amy-loidosis consistent with oedema, and was an independent pre-dictor of prognosis. The higher ECV in ATTR was consistent with higher amyloid infiltration. These findings highlight the unique role of CMR with multiparametric mapping for char-acterising the cardiac effects of systemic amyloidosis and risk stratification in this population. Background The DANISH trial emphasised that the selection of patients with dilated cardiomyopathy (DCM) for implant-able cardioverter defibrillators (ICD) needs to be improved. Registries demonstrate that the major burden of sudden cardiac death (SCD) occurs in those with a left ventricular ejection fraction (LVEF) >35%. Those at high-risk of SCD with milder reductions in LVEF may gain greater quality-adjusted life years from successful ICD therapy compared to those with more severe reductions, due to a lower risk of death from competing non-sudden causes. Variables that identify patients with milder reductions in LVEF at high-risk of SCD are required. Methods We prospectively investigated the utility of mid-wall late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) to predict SCD and aborted SCD in consecutive patients with DCM and LVEF >40% seen in our cardi-omyopathy service or referred for CMR between 2000 and 2011. Those with potential pre-existing indications for ICD implantation were excluded. The presence of LGE was determined by a specialist blinded to clinical data. A panel blinded to CMR data adjudicated end-point occurrences. Results Of 399 patients (145 women, median age 50 years, median LVEF 50%) followed for a median of 4.6 years, 18 of 101 (17.8%) with LGE reached the pre-specified end-point, compared to 7 of 298 (2.3%) without (HR 9.2; 95% CI 3.9-21.8; p<0.0001) (Figure 1). Nine patients (8.9%) with LGE compared to 6 (2.0%) without (HR 4.9; 95% CI 1.8-13.5; p=0.002) died suddenly, whilst 10 patients (9.9%) with LGE compared to 1 (0.3%) without (HR 34.8; 95% CI 4.6-266.6; p<0.001) had aborted SCD. Following adjustment based on propensity score, LGE predicted the composite end-point (HR 8.0; 95% CI 3.3-19.5; p<0.0001), SCD (HR 4.6; 95% CI 1.6-13.1; p=0.005) and aborted SCD (HR 32.9; 95% CI 4.3-249.9; p<0.001). Estimated hazard ratios for the primary end-point for patients with a LGE extent of 0%-2.5%, 2.5%-5% and >5% compared to those without LGE were 10.6 (95%CI 3.9-29.4), 4.9 (95% CI 1.3-18.9) and 11.8 (95% CI 4.3-32.3). Conclusion Mid-wall LGE identifies patients with DCM and a LVEF >40% with an 8-fold increased risk of SCD and aborted SCD, who may benefit from ICD implantation. Acknowledgements BPH is supported by a British Heart Foundation Clinical Resea...

Appropriate ICD shocks occur more frequently than SCD in patients with nonischemic cardiomyopathy. This suggests that episodes of nonsustained ventricular tachycardia frequently terminate spontaneously in such patients.