MP-453090-7 IDENTIFYING PATIENTS WITH LVEF≤55% WHO MAY BENEFIT FROM A PRIMARY PREVENTION ICD USING SEATTLE PROPORTIONAL RISK MODEL FOR MILDLY REDUCED EF (SPRM-MR) IN 14,752 PATIENTS

Abstract

Patients with EF>30-35% remain at risk of sudden cardiac death (SCD) but are not currently indicated for a primary prevention ICD. Previous research has identified greater ICD benefit in those with a higher proportion of SCD. We sought to identify patients with EF≤55% who have a high proportion of SCD who may benefit from an ICD. Patients with EF≤55% not eligible for an ICD (NYHA 1- NonIschemic EF 1-55%, Ischemic EF 31-55%, NYHA 2-4 EF 36-55%) with mode of death from clinical trials (Val-HeFT, IPreserve, HAT-AED, HEAAL, GISSI-HF) and prospective observational studies (PRE-DETERMINE, UW, and Italian HF Registry). A Cox Proportional all-cause mortality (ACM) model and logistic regression SCD proportional risk models (SPRM-MR) were derived. Patients were considered potential ICD candidates if the predicted proportion of SCD was ≥45% and the estimated sudden death rate at 5 years ≥5%. ICD benefit was estimated by applying the SPRM*ICD interaction for ACM in HFrEF patients in SCD-HeFT (https://depts.washington.edu/sprm). The cohort included 14,752 patients. During a median follow-up of 3.6±1.4 years, 2,296 (15.6%) died. SCD comprised 30.5% of known deaths. Younger age, male sex, lower EF, higher SBP, and lower creatinine were associated with a higher proportion of SCD. One in 4 patients (23%) had a predicted SCD proportion ≥45% (observed SCD proportion 53%). An estimated SCD 5-year rate of ≥5% was present in 54% of patients. The potential ICD cohort (SPRM-MR≥45% and SCD 5yr ≥5%) comprised 11% of the entire study population. This was 17% with EF ≤45% vs <1% with EF >45%. In the potential ICD cohort, the 5 year Kaplan Meier ACM was 16.8% with 52% sudden death (∼8.7% at 5 years) vs. 20.9% ACM with 28% sudden death in the remaining patients. The estimated ICD benefit was a relative risk reduction in ACM of 27% with an ∼5% absolute risk reduction in ACM at 5 years. The 5 year number needed to treat of 20. Nearly 1 in 5 patients with an LVEF ≤ 45% who do not meet contemporary guideline criteria for ICD implantation demonstrate significant absolute (≥5%/5 years) and proportional (≥45%) risks of SCD. In these patients, we estimate an ICD would provide a 27% relative risk and 5% absolute risk reduction in all-cause mortality with a number needed to treat to save one life with an ICD of 20. Future efforts should focus on the use of integrating absolute and proportional risk as inclusion criteria for evaluating ICD benefit in this substantial population.