Amiodarone or an ICD for Congestive Heart Failure (SCD-HeFT)

Authors, Journal, Affiliations, Type, DOI

• Bardy GH, Lee KL, Mark DB, Poole JE, Packer DL, et al. for the SCD-HeFT Investigators
• New England Journal of Medicine, 2005;352:225–237
• Seattle Institute for Cardiac Research, Duke University, Mayo Clinic, and 148 other sites
• Three-arm, double-blind (placebo/amiodarone) randomised controlled trial
• DOI: https://doi.org/10.1056/NEJMoa043399
• Funding: NHLBI (principal); Medtronic (devices) and Wyeth-Ayerst (amiodarone) provided study materials; no role in design or analysis

Overview

SCD-HeFT was a landmark three-arm RCT randomising 2,521 patients with NYHA II/III HF and LVEF ≤35% (ischaemic 52%, non-ischaemic 48%) to placebo, amiodarone, or a conservatively programmed shock-only single-lead ICD. Enrolled September 1997 – July 2001; median follow-up 45.5 months. ICD therapy reduced all-cause mortality by 23% (HR 0.77; P=0.007; absolute 7.2% at 5 years) without any aetiological interaction — establishing equal benefit in ischaemic and non-ischaemic HF. Amiodarone showed no overall mortality benefit (HR 1.06; P=0.53) with a significant NYHA class interaction: neutral in NYHA II (HR 0.85), but potentially harmful in NYHA III (HR 1.44; P for interaction=0.004). SCD-HeFT directly established the LVEF ≤35% guideline threshold (MADIT-II used ≤30%) and remains the foundational RCT for primary prevention ICD in both ischaemic and non-ischaemic HF.

Keywords

ICD, primary prevention, sudden cardiac death, heart failure, amiodarone, LVEF ≤35%, non-ischaemic cardiomyopathy, ischaemic cardiomyopathy, NYHA class interaction, SCD-HeFT

Key Takeaways

Study Design and Population

• Three-arm double-blind RCT: placebo (n=847) vs amiodarone (n=845) vs shock-only single-lead ICD (n=829)
• Eligibility: age ≥18; NYHA II (70%) or III (30%); chronic stable HF; LVEF ≤35%; ischaemic or non-ischaemic aetiology
• Ischaemic defined as LV systolic dysfunction + ≥75% narrowing of a major coronary artery OR prior MI
• Enrolled September 16, 1997 – July 18, 2001; followed until October 31, 2003; all patients followed ≥2 years
• Median LVEF 25%; median follow-up 45.5 months (longest 72.6 months)
• Background therapy: ACEi required (if clinically reasonable); beta-blocker required; aldosterone/aspirin/statins when appropriate. 69% on beta-blocker at baseline. Pre-SGLT2i, pre-ARNi era.
• ICD: conservatively programmed shock-only mode; detection threshold ≥187 bpm; no ATP; no biventricular/dual-chamber pacing allowed; single Medtronic model 7223
• Amiodarone: weight-based loading (800 mg/day × 1 week → 400 mg/day × 3 weeks); maintenance 200–400 mg/day by weight; placebo identical tablets

Primary Endpoint — All-Cause Mortality

• Deaths: placebo 244 (29%); amiodarone 240 (28%); ICD 182 (22%)
• ICD vs placebo: HR 0.77 (97.5% CI 0.62–0.96; P=0.007) — 23% RRR; absolute −7.2% at 5 years
• Amiodarone vs placebo: HR 1.06 (97.5% CI 0.86–1.30; P=0.53) — no benefit

Subgroup Analyses

Aetiology (Ischaemic vs Non-Ischaemic)

• No interaction of ICD therapy with aetiology (P=0.68) — first large RCT confirming ICD primary prevention benefit extends equally to non-ischaemic HF
• No interaction of amiodarone with aetiology (P=0.93) — amiodarone neutral across both aetiologies

NYHA Class — Critical Prespecified Interaction

• ICD × NYHA interaction P<0.001:
  • NYHA II (70%): ICD HR 0.54 (97.5% CI 0.40–0.74); absolute −11.9% at 5 years — highly significant benefit
  • NYHA III (30%): ICD HR 1.16 (97.5% CI 0.84–1.61) — no benefit
• Amiodarone × NYHA interaction P=0.004:
  • NYHA II: amiodarone HR 0.85 (97.5% CI 0.65–1.11) — neutral
  • NYHA III: amiodarone HR 1.44 (97.5% CI 1.05–1.97) — increased mortality risk
• Investigators note: NYHA III ICD attenuation was not anticipated; other trials (MADIT-II, DEFINITE) do not support withholding ICD from NYHA III

Other Subgroups (ICD vs placebo)

• Male sex HR 0.73; Age <65 HR 0.68; LVEF ≤30% HR 0.73; QRS <120ms HR 0.84; QRS ≥120ms HR 0.67
• 6-min walk <950 ft (most impaired): ICD HR 1.14 (NS); >1275 ft: HR 0.45 (0.27–0.76)
• Benefit with beta-blocker HR 0.68; without BB HR 0.92 (trend to attenuate)

ICD Complications and Device Events

• 5% acute device-related complications; 9% later complications
• 31% of ICD patients had shocks for any cause; 21% had appropriate shocks (rapid VT/VF)
• Annual appropriate shock rate: 5.1%
• Defibrillation testing: no patient required more than 30-J shock (max device output)

Amiodarone Compliance and Safety

• Non-compliance rate 27%: placebo 22%, amiodarone 32%
• Complications vs placebo: tremor +4% (P=0.02); hypothyroidism +6% (P<0.001)
• 11% of ICD patients crossed over to open-label amiodarone; 125 drug-group patients crossed over to ICD during follow-up

Limitations of the document

• Pre-GDMT era: No SGLT2i, no ARNi, no CRT-D; background therapy is ACEi + beta-blocker only; modern quadruple GDMT independently reduces arrhythmic substrate
• Shock-only single-lead ICD: Results cannot be extrapolated to dual-chamber or CRT-D devices — investigators explicitly warn against this
• NYHA III ICD attenuation: Unanticipated prespecified subgroup finding; investigators recommend against withholding ICD in NYHA III; competing evidence (DEFINITE) shows class III benefit
• NYHA III amiodarone signal: HR 1.44 (1.05–1.97); prespecified interaction was significant; biological plausibility uncertain — could reflect selection bias or disease severity
• Open-label crossovers: 11% ICD patients received amiodarone; 11% drug-group patients crossed to ICD — likely attenuates effect estimates in ITT analysis
• Aetiology definition: Ischaemic defined broadly (≥75% stenosis OR prior MI without clinical documentation); may include overlap with non-obstructive ischaemic disease

Key Concepts Mentioned

• entities/ICD — primary prevention ICD evidence base; ≤35% threshold source
• entities/HFrEF — ICD and amiodarone positioning in HFrEF management
• entities/Amiodarone — amiodarone neutrality/harm in HFrEF; NYHA III caution

Key Entities Mentioned

• entities/ICD — SCD-HeFT established ≤35% threshold; aetiology-agnostic benefit; NYHA III attenuation
• entities/HFrEF — foundational trial for ICD indication; amiodarone NYHA III harm
• entities/Amiodarone — NYHA class interaction; NYHA III HR 1.44

Wiki Pages Updated

• wiki/sources/icd-hfref-scdheft-nejm-2005.md — created
• wiki/entities/HFrEF.md — ICD ischaemic/non-ischaemic bullets enriched; amiodarone bullet updated with NYHA subgroups; two contradictions updated/added; source_count 19→20
• wiki/entities/ICD.md — SCD-HeFT added as foundational trial to CAD and NICM primary prevention; contradiction updated; source_count 4→5
• wiki/entities/Amiodarone.md — SCD-HeFT amiodarone NYHA III harm enriched; contradiction updated; source_count 6→7
• wiki/concepts/VA-Risk-Stratification-DCM.md — SCD-HeFT NICM ICD benefit added to LVEF Limitations; source_count 6→7
• wiki/sourceindex.md — new entry added
• wiki/wikiindex.md — ICD and HFrEF entries updated