Long-Term Outcomes of Resynchronization–Defibrillation for Heart Failure (RAFT Long-Term)

Authors, Journal, Affiliations, Type, DOI

• Authors: John L. Sapp, Soori Sivakumaran, Calum J. Redpath, Habib Khan, Ratika Parkash, Derek V. Exner, Jeff S. Healey, Bernard Thibault, Laurence D. Sterns, Nhat Hung N. Lam, Jaimie Manlucu, Ahmed Mokhtar, Glen Sumner, Stuart McKinlay, Shane Kimber, Blandine Mondesert, Mario Talajic, Jean Rouleau, C. Elizabeth McCarron, George Wells, and Anthony S. L. Tang; for the RAFT Long-Term Study Team
• Journal: New England Journal of Medicine, 2024;390:212–220
• Affiliations: QEII Health Sciences Centre / Dalhousie University (Halifax); University of Alberta (Edmonton); University of Ottawa Heart Institute; Western University (London, ON); University of Calgary / Libin Cardiovascular Institute; McMaster University (Hamilton); Montreal Heart Institute; Royal Jubilee Hospital (Victoria); University of Toronto; King Abdulaziz University (Jeddah, Saudi Arabia)
• Type: Long-term follow-up of a multicenter, double-blind, randomized, controlled trial (RAFT; NCT00251251)
• DOI: 10.1056/NEJMoa2304542

Overview

The RAFT Long-Term study extends follow-up of the original RAFT RCT (1,798 patients at 34 centres; mean 40±20 months) to a median of 7.7 years overall and 13.9 years among survivors, using data from the 8 highest-enrolling sites (1,050 patients). Among patients with NYHA class II/III heart failure, LVEF ≤30%, and QRS ≥120ms randomised to CRT-D versus ICD alone, CRT-D conferred a sustained reduction in time-to-death (acceleration factor 0.80; 95% CI 0.69–0.92; P=0.002). This is the longest CRT survival follow-up available, complementing MADIT-CRT (7-year extended) and CARE-HF (4.4-year extended). Overall 15-year mortality remained approximately 80% in the CRT-D group, underscoring the severity of the underlying disease burden.

Keywords

Cardiac resynchronization therapy, CRT-D, ICD, biventricular pacing, RAFT, long-term outcomes, heart failure, reduced ejection fraction, QRS duration, LBBB, mortality, accelerated failure time model

Key Takeaways

Background and Study Population

• Original RAFT enrolled 1,798 patients (34 centres); long-term follow-up included 1,050 patients from 8 highest-enrolling sites
• Enrollment January 2003; all patients followed to death or December 31, 2021
• Eligibility: NYHA class II or III HF; LVEF ≤30%; intrinsic QRS ≥120ms or paced QRS ≥200ms; 1:1 randomisation to ICD vs CRT-D
• Protocol amended February 2006 to exclude NYHA class III (after guideline update and external trial evidence)
• Patients with AF, right bundle-branch block, nonspecific IVCD, and right ventricular pacing were NOT excluded

Patient Demographics

• 83.8% men; mean age 66.5±9.2 years
• 67.9% ischemic cardiomyopathy; 32.1% non-ischemic cardiomyopathy
• 76.5% NYHA class II; 23.5% NYHA class III at baseline
• QRS morphology: 70.2% LBBB, 8.4% RBBB, 12.5% NIVCD, 9.0% paced QRS
• 15.7% had persistent atrial arrhythmia at baseline

Follow-up Duration

• Median follow-up all 1,050 patients: 7.7 years (IQR 3.9–12.8)
• Median follow-up survivors: 13.9 years (IQR 12.8–15.7)
• ICD group: median 6.9 years all; 13.9 years survivors
• CRT-D group: median 8.5 years all; 14.2 years survivors

Primary Outcome: All-Cause Mortality

• ICD group: 405/530 (76.4%) died
• CRT-D group: 370/520 (71.2%) died
• Acceleration factor 0.80 (95% CI 0.69–0.92; P=0.002) — CRT-D delays time-to-death by ~20%
• Non-proportional hazards confirmed; exponential accelerated failure time model used (more appropriate than Cox HR for long follow-up)
• Overall 15-year mortality ~80% even in CRT-D arm

Secondary Outcome: Death / Heart Transplantation / LVAD Implantation

• ICD group: 412/530 (77.7%) composite event
• CRT-D group: 392/520 (75.4%) composite event
• Acceleration factor 0.85 (95% CI 0.74–0.98)
• Composite event curves began to converge after 12 years of follow-up

Subgroup Analyses

• 11 prespecified subgroups: age, sex, NYHA class, aetiology, QRS duration, LVEF, QRS morphology, atrial rhythm
• Benefit consistent across subgroups, including NYHA II and NYHA III analysed separately
• Benefit persisted despite inclusion of AF and non-LBBB QRS morphologies (subgroups known to derive less benefit)

Statistical Method

• ITT analysis throughout
• Non-proportional hazards confirmed (log-negative-log plots, time-dependent covariates, empirical score process)
• Primary analysis: exponential accelerated failure time model; results expressed as acceleration factors
• Sensitivity analyses with alternative non-proportional-hazards models performed

CRT Mechanisms Underlying Durable Benefit

• CRT improves cardiac performance without increasing cardiac work (Ukkonen et al., Circulation 2003)
• Leads to LV reverse remodeling (CARE-HF, REVERSE trials)
• Reduces new-onset ventricular arrhythmias (RAFT substudy: reduced VAs in primary prevention; MADIT-CRT: LVEF normalisation reduces VA risk)
• These early mechanistic benefits likely underlie the sustained long-term survival improvement

Limitations of the Document

• Subset trial: only 8 of 34 original RAFT centres (1,050/1,798 patients); findings should be interpreted cautiously
• Long-term follow-up precluded analysis of crossover between ICD and CRT-D groups
• Pharmacologic therapy advances (neprilysin inhibitors/sacubitril-valsartan, SGLT2 inhibitors) were not available during most of the follow-up; CRT benefit on top of optimal contemporary GDMT is uncertain
• Low enrollment of women (16.2%) and lack of racial diversity limit generalizability
• Clinical changes during follow-up (worsening HF, pharmacologic changes, lead function changes) not systematically captured
• No data safety monitoring committee reconvened for long-term phase; individual consent waived for follow-up data

Key Concepts Mentioned

• concepts/Cardiac-Resynchronization-Therapy — primary intervention studied; long-term mortality benefit confirmed
• concepts/Conduction-System-Pacing — alternative to biventricular CRT; not yet addressed in RAFT era
• concepts/Sudden-Cardiac-Death — CRT reduces VA burden (RAFT substudy)

Key Entities Mentioned

• RAFT (Resynchronization–Defibrillation for Ambulatory Heart Failure Trial) — parent trial; NCT00251251
• MADIT-CRT — companion trial; NYHA I/II; 7-year extended follow-up showed mortality benefit in LBBB
• CARE-HF — companion trial; NYHA III/IV; 4.4-year extended follow-up showed persistence of CRT-P mortality benefit

Wiki Pages Updated

• Created wiki/sources/crt-longterm-raft-nejm-2024.md
• Created wiki/concepts/Cardiac-Resynchronization-Therapy.md
• Updated wiki/sourceindex.md
• Updated wiki/wikiindex.md