Abstract 4138912: Clinical outcomes of cardiac synchronization with or without an implantable cardioverter defibrillator based on pooled data from 5 clinical trials: a patient-level meta-analysis

Background: Cardiac resynchronization therapy (CRT) is a well-established therapy for patients with heart failure with reduced ejection fraction (HFrEF) and wide QRS. Whether CRT-defibrillators (CRT-D) reduce mortality more than CRT-pacemakers (CRT-P) remains controversial. Aims: To compare the clinical outcomes of CRT-D vs CRT-P using data from 5 landmark CRT trials, both overall and stratified by etiology of cardiomyopathy (ischemic vs non-ischemic), sex (male vs female), age (≥ 70 y/o vs < 70 y/o), and QRS morphology (IVCD, LBBB, RBBB). Methods: We performed a meta-analysis of patient level data from 5 prospective CRT trials (MIRACLE, REVERSE, RAFT, COMPANION and MADIT-CRT). Inclusion criteria were CRT-P vs CRT-D status (randomized comparison only in COMPANION), age ≥ 18 y/o and LVEF ≤ 35%. Exclusion criteria included secondary prevention ICD, QRS < 120ms, pacemaker upgrade, ventricular pacing indication, or missing data. Primary outcome was composite of time to heart failure hospitalization (HFH) or all-cause death. Secondary outcomes were time to HFH and death. Outcomes were analyzed using a frequentist Cox Proportional Hazards mixed effects model adjusted for 17 variables. Results: A total of 3407 patients met inclusion criteria. Relative to patients with CRT-P (n=843), those with CRT-D (n=2564) were of similar age (66 y/o, p=0.5), less often female (24% vs 34%, p<0.001), and more often had ischemic cardiomyopathy (59.4% vs 52.4%, p<0.001), Fig 1A. Primary outcome was similar across groups (HR 0.902 [0.752, 1.081], p=0.26), but all-cause mortality was lower with CRT-D vs CRT-P (HR 0.77 [0.603, 0.983], p=0.036), Fig 1B. Interaction analyses suggested lower all-cause mortality with CRT-D vs CRT-P in patients with non-ischemic cardiomyopathy (HR 0.502 [0.346, 0.726], p=0.0003) and patients age ≥70 y/o (HR 0.679 [0.502, 0.919], p=0.012), with significance preserved after Bonferroni correction (Fig 1B). Conclusion: In patients receiving CRT for HFrEF, those with CRT-D had lower all-cause mortality than patients with CRT-P, driven mainly by a lower mortality with CRT-D in older patients and those with non-ischemic cardiomyopathy. No significant interactions were noted between ICD&sex or ICD&QRS morphology.