Sudden Death in Patients With Coronary Heart Disease Without Severe Systolic Dysfunction

Abstract

Importance: The majority of sudden and/or arrhythmic deaths (SAD) in patients with coronary heart disease occur in those without severe systolic dysfunction, for whom strategies for sudden death prevention are lacking.

Objective: To provide contemporary estimates of SAD vs other competing causes of death in patients with coronary heart disease without severe systolic dysfunction to search for high-risk subgroups that might be targeted in future trials of SAD prevention.

Design, setting, and participants: This prospective observational cohort study included 135 clinical sites in the United States and Canada. A total of 5761 participants with coronary heart disease who did not qualify for primary prevention implantable cardioverter defibrillator therapy based on left ventricular ejection fraction (LVEF) of more than 35% or New York Heart Association (NYHA) heart failure class (LVEF >30%, NYHA I).

Exposures: Clinical risk factors measured at baseline including age, LVEF, and NYHA heart failure class.

Main outcomes and measures: Primary outcome of SAD, which is a composite of SAD and resuscitated ventricular fibrillation arrest.

Results: The mean (SD) age of the cohort was 64 (11) years. During a median of 3.9 years, the cumulative incidence of SAD and non-SAD was 2.1% and 7.7%, respectively. Sudden and/or arrhythmic death was the most common mode of cardiovascular death accounting for 114 of 202 cardiac deaths (56%), although noncardiac death was the primary mode of death in this population. The 4-year cumulative incidence of SAD was lowest in those with an LVEF of more than 60% (1.0%) and highest among those with LVEF of 30% to 40% (4.9%) and class III/IV heart failure (5.1%); however, the cumulative incidence of non-SAD was similarly elevated in these latter high-risk subgroups. Patients with a moderately reduced LVEF (40%-49%) were more likely to die of SAD, whereas those with class II heart failure and advancing age were more likely to die of non-SAD. The proportion of deaths due to SAD varied widely, from 14% (18 of 131 deaths) in patients with NYHA II to 49% (37 of 76 deaths) in those younger than 60 years.

Conclusions and relevance: In a contemporary population of patients with coronary heart disease without severe systolic dysfunction, SAD accounts for a significant proportion of overall mortality. Moderately reduced LVEF, age, and NYHA class distinguished SAD and non-SAD, whereas other markers were equally associated with both modes of death. Absolute and proportional risk of SAD varied significantly across clinical subgroups, and both will need to be maximized in future risk stratification efforts.

Figure 1.. Cumulative Incidence of Sudden and/or Arrhythmic Death (SAD) and Non-SAD Across Left Ventricular Ejection Fraction (LVEF) and New York Heart Association (NYHA) Functional Class

The 4-year cumulative incidence of each mode of death (ie, SAD and non-SAD), accounting for the competing risk of death, are shown in strata of LVEF (A and B) and NYHA functional class (C and D). Differences in cumulative incidence across strata are shown (Gray P value).

Figure 2.. Differential Association of Clinical Risk Factors With Sudden and/or Arrhythmic Death (SAD) vs Competing Modes of Death

The relative incidence of each mode of death (ie, SAD and non-SAD), accounting for the competing risk of other deaths, is shown for clinical subgroups of interest. P values for the differential association of each clinical subgroup with mode of death are shown.

Figure 3.. Absolute and Proportional Risk of Sudden and/or Arrhythmic Death (SAD)

AF indicates atrial fibrillation; DM, diabetes mellitus; LVEF, left ventricular ejection fracture; NYHA, New York Heart Association. Implantable cardioverter defibrillator (ICD) benefit throughout the median follow-up of the study (4 years) was estimated using 2 metrics: the number needed to treat to save 1 life (A) and percentage reduction in total mortality (B). Implantable cardioverter defibrillator benefit was modeled in patients without an ICD. Subgroups are ordered by increasing absolute incidence of SAD (A) and increasing proportional risk of SAD (B).