Evaluation of MADIT II risk stratification score among registry of heart failure patients with primary prevention ICD/CRTD device

FUNDING ACKNOWLEDGEMENTS: Type of funding sources: None. BACKGROUND: Current guidelines advocate primary prevention ICD implantation for all symptomatic heart failure (HF) patients with low LVEF. As most patients will not use their device during lifetime, a score delineating subgroups with differential ICD benefit is crucial. The MADIT Risk Stratification Score (MRSS) based on 5 parameters including: age>70, Creatinine >1.4 mg/dl, QRS width >120ms, presence of AF, and NYHA >2, was developed for this purpose. OBJECTIVE: Evaluate MRSS among Israeli nationwide registry of HF patients implanted with prophylactic ICD/CRTD. Endpoints included overall mortality, sustained ventricular arrhythmia (VA), and competing risk of arrhythmic (VA-related) versus non-arrhythmic death, a surrogate for potential ICD survival benefit. METHODS: Study based on comprehensive registry of ICD/CRTDs implanted in Israel between 2011-2018. All registry patients were categorized into MRSS-based risk-groups (low-risk- MRSS 0, intermediate-risk- MRSS 1-5, very high-risk (VHR) group defined by Creatinine >2.5mg/dL). Univariate and Kaplan Meier analysis used to evaluate the association of risk groups with study endpoints. RESULTS: 2177 HF patients, implanted with a primary prevention ICD (1255, 58%) or CRTD (922, 42%) were included. There were 189 (8.7%) patients with sustained VA and 316 (14.5%) deaths during median follow-up (F/U) period of 2.5 years. A significant correlation was found between MRSS-based risk subgroups and overall mortality (p 0.001). However, MRSS association with sustained VA was weak (p 0.2). Notably, the MRSS-based very high-risk (VHR) subgroup had an exceptionally high mortality but low VA incidence. Competing risk of arrhythmic versus non-arrhythmic death revealed a large and significant ICD survival benefit among the low and intermediate MRSS-based risk subgroups, with 10.2 and 12.3 months survival gained over 3-year F/U period among these subgroups, respectively (p 0.001). The VHR subgroup in contrast, showed a minimal non-significant ICD survival benefit. CONCLUSIONS: Use of MRSS among a contemporary real-world registry of HF patients revealed subgroups with differing ICD survival benefit, suggesting it as a universal tool to predict ICD survival benefit. MRSS-based VHR subgroup may not gain survival benefit from prophylactic ICD implant.