P991Predicting arrhythmic risk in dilated cardiomyopathy: a systematic review & meta-analysis of clinical parameters

Abstract Background Patients with non-ischemic dilated cardiomyopathy (NIDCM) are at increased risk of ventricular arrhythmias and sudden cardiac death (SCD). However, identifying patients at high risk for life-threatening ventricular arrhythmia (LTVA) who may benefit from an implantable cardioverter defibrillator (ICD) remains challenging. Methods We searched MEDLINE and EMBASE for prognostic studies describing predictors of LTVA (defined as sustained ventricular tachycardia (VT), haemodynamically unstable VT, ventricular fibrillation, (aborted) SCD or appropriate ICD intervention) in patients with NIDCM. We excluded articles with composite heart failure and arrhythmic endpoints but lacking (subgroup) analysis for LTVA. Study quality and risk of bias was assessed using the QUIPS-tool, and articles with high risk of bias in ≥2 areas were excluded from analysis. Univariable hazard ratios of reported predictors were pooled from the remaining studies in a meta-analysis using a random-effects model and presented with 95% confidence interval (CI). Results Out of 1996 unique citations, 51 studies were included comprising 9798 patients with 1493 arrhythmic events. 28 studies were pooled for meta-analysis (mean age 55±4.1 years, 72% male) with a mean follow-up of 3.7±1.9 years. Crude event rate was 4.3% (95% CI 4.02–4.57) per year. From our meta-analysis, hypertension (HR 1.95; CI [1.26–3.00]), history of out of hospital cardiac arrest or sustained VT (HR 4.15; CI [1.32–13.02]), T-wave alternans (HR 6.50; CI [2.46–17.14]), LVEDV per 10ml/m2 increase (HR 1.10; CI [1.10–1.10]), LVESV per 10ml/m2 increase (HR 1.10; CI [1.00–1.22]) and delayed gadolinium enhancement (HR 5.55; CI [4.02–7.67]) were significantly associated with LTVA (figure). The quality of evidence was moderate and there was significant heterogeneity (median i2 57%; IQR 76%) among studies. Additionally from data that could not be pooled, decreased LVEF, history of nsVT and decreased heart rate variability were significantly associated with LTVA. Summary of meta-analysis results Conclusion The risk of LTVA in NIDCM is 4.3% per year and is considerably higher in patients with hypertension, history of LTVA, decreased LVEF, high LVEDV, high LVESV, T-wave alternans, history of nsVT, decreased heart rate variability and delayed gadolinium enhancement. These results may help determine appropriate candidates for ICD implantation. The high heterogeneity in reported results indicate the need for future multicent