Beta-Blockers after Myocardial Infarction with Normal Ejection Fraction

Authors, Journal, Affiliations, Type, DOI

• Kristensen AMD, Rossello X, Atar D, Yndigegn T, Kimura T, Latini R, Lindahl B, Halvorsen S, Olsen MH, Fuster V, et al.; Beta-Blocker Trialists' Collaboration Study Group
• N Engl J Med 2026;394:540–550
• Multicentre international collaboration (Denmark, Spain, Norway, Japan, Sweden, UK, New Zealand, Italy)
• Individual-patient-level meta-analysis of 5 open-label randomised controlled trials (preplanned; preregistered PROSPERO CRD420251119176)
• DOI: 10.1056/NEJMoa2512686

Overview

This preplanned individual-patient-level meta-analysis pooled data from five contemporary randomised trials — REBOOT, REDUCE-AMI, BETAMI, DANBLOCK, and CAPITAL-RCT — comprising 17,801 patients with a preserved left ventricular ejection fraction (LVEF ≥50%) after recent MI who had no other indication for beta-blockers. Over a median follow-up of 3.6 years, beta-blocker therapy did not reduce the composite of all-cause death, MI, or heart failure (HR 0.97; 95% CI 0.87–1.07; P=0.54), with consistent null findings across individual endpoints and all prespecified subgroups. This is the highest-grade evidence to date on this question and directly challenges the ACC/AHA Class I recommendation to use beta-blockers after MI regardless of LVEF.

Keywords

Beta-blockers, myocardial infarction, preserved ejection fraction, LVEF, MACE, death, heart failure, meta-analysis, individual patient data, randomised trials, REBOOT, REDUCE-AMI, BETAMI, DANBLOCK, CAPITAL-RCT

Key Takeaways

Background

• Beta-blocker therapy has been standard of care post-MI since the early 1980s based on seminal trials (propranolol BHAT, Göteborg metoprolol, ISIS-1, Norwegian timolol).
• Modern advances (routine PCI, high-intensity statins, contemporary DAPT) have significantly improved outcomes, raising uncertainty about continued need for beta-blockers in patients without HF or reduced EF.
• ESC guidelines give a Class IIA recommendation for beta-blockers post-MI without reduced LVEF; ACC/AHA gives a Class I recommendation regardless of LVEF — a clinically significant divergence.
• Five contemporary RCTs individually yielded conflicting results; none were adequately powered for definitive individual-outcome assessment, and all lacked sufficient data for robust subgroup analyses.

Methods

• Preplanned IPD meta-analysis; systematic Medline search (August 12, 2025) confirmed no additional eligible trials existed; PRISMA-compliant reporting.
• Included trials: REBOOT (n=7,459), REDUCE-AMI (n=4,967), BETAMI (n=2,441), DANBLOCK (n=2,277), CAPITAL-RCT (n=657) — all investigator-initiated open-label superiority RCTs.
• Eligibility: MI within 14 days; LVEF ≥50%; no other indications for beta-blockers (HF, AF, uncontrolled HTN etc.); no contraindications. Type and dose of beta-blocker determined by treating physician (all trials except CAPITAL-RCT, which used carvedilol exclusively).
• Primary endpoint: composite of all-cause death, MI, or heart failure (time-to-first-event).
• One-stage fixed-effects Cox proportional-hazards model, stratified by trial; intention-to-treat.
• Three sensitivity analyses: (1) multivariable adjustment (age, sex, MI type, LVEF, trial); (2) random-effects two-stage model; (3) adjudicated-events only (excluding REDUCE-AMI, which did not use blinded adjudication).

Patient Characteristics

• Total n=17,801: 8,831 (49.6%) beta-blocker group; 8,970 (50.4%) no–beta-blocker group.
• Median age 62 years (IQR 55–71); 20.7% women; 45.7% STEMI; 94.4% underwent PCI.
• 8.1% prior MI; 2.0% atrial fibrillation; baseline characteristics well balanced between groups.
• Beta-blocker doses were generally low across all trials, reflecting contemporary clinical practice.
• Approximately half of screened patients (in BETAMI and DANBLOCK) were ineligible due to established BB indications (AF, HTN, HF), limiting generalisability to the specific no-indication subgroup.

Primary Results

• Primary composite (death/MI/HF): 717/8,831 (8.1%) vs 748/8,970 (8.3%); 2.37 vs 2.45 events per 100 patient-years.
• HR 0.97 (95% CI 0.87–1.07); P=0.54 — no significant difference; median follow-up 3.6 years.
• Between-trial variance 0.005 (I² = 20%) — low heterogeneity across trials.
• All three sensitivity analyses consistent with primary result.
• Landmark analysis at 12 months: HR 0.88 (95% CI 0.74–1.05) — a non-significant trend toward early benefit that was not sustained.

Individual Components (Secondary Endpoints)

• All-cause death: HR 1.04 (95% CI 0.89–1.21); 335 vs 326 events — NS.
• MI: HR 0.89 (95% CI 0.77–1.03); 360 vs 407 events — NS; numerically fewer MIs with BB.
• Heart failure: HR 0.87 (95% CI 0.64–1.19); 75 vs 87 events — NS.
• Cardiac death: HR 1.26 (95% CI 0.94–1.70) — numerically higher in BB group, NS.
• Unplanned coronary revascularisation: HR 1.03 (95% CI 0.88–1.20) — NS.

Safety

• Ischemic stroke: 115 vs 94 patients (BB vs no-BB); between-group difference in restricted mean event-free survival 2.6 days (95% CI −0.73 to 4.4) — NS, but numerically more strokes in BB group.
• Advanced AV block: HR 1.03 (95% CI 0.73–1.44) — NS; equal rates.

Subgroup and Stratified Analyses

• Consistent null effect across all prespecified subgroups: age, sex, MI type (STEMI vs NSTEMI), LVEF, individual trial.
• A sex–beta-blocker interaction found in REBOOT alone was not confirmed in this larger IPD-MA.
• No apparent differences across beta-blocker type (selective vs nonselective) or dose, though these were not randomised.

Contextual Comparison: LVEF 40–49% Subgroup

• The companion Rossello 2025 Lancet IPD-MA (same 4 trials except REDUCE-AMI; patients with LVEF 40–49%) showed a 25% relative risk reduction (HR 0.75; 95% CI 0.58–0.97) for the same composite — the mildly reduced LVEF group appears to benefit from beta-blockers.
• This LVEF-based divergence is biologically plausible: preserved-EF patients had lower event rates (2.41 vs 3.76 per 100 person-years), smaller infarctions, less myocardial scarring, and lower susceptibility to ventricular arrhythmias and sudden cardiac death.

Limitations of the Document

• All trials were open-label; 11–18% crossover at 6 and 12 months may bias results toward equipoise.
• REDUCE-AMI used non-blinded, non-adjudicated MI and HF events — sensitivity analysis excluding it showed consistent results.
• Heart-failure definition differed across trials; partially harmonised but not fully standardised.
• Population predominantly European and Japanese; only 20.7% women — limits generalisability.
• Beta-blocker type and dose were not randomised — stratified analyses for these are observational.
• Results apply only to patients without other indications for beta-blockers (AF, angina, uncontrolled HTN, HF) — these patients were excluded, and the absence of benefit in this study cannot be applied to them.

Key Concepts Mentioned

• concepts/Beta-Blocker-Post-MI — LVEF-stratified framework; IPD-MA confirms no benefit in LVEF ≥50%

Key Entities Mentioned

• entities/Acute-Coronary-Syndrome — index event; long-term beta-blocker management post-MI
• entities/Chronic-Coronary-Disease — stable post-MI chronic phase management

Wiki Pages Updated

• wiki/sources/bb-mi-nejm-2026.md — created
• wiki/concepts/Beta-Blocker-Post-MI.md — IPD-MA data added; LVEF ≥50% section definitively updated; ACC/AHA Class I contradiction formalised
• wiki/entities/Acute-Coronary-Syndrome.md — beta-blocker section updated with IPD-MA reference
• wiki/sourceindex.md — new entry added
• wiki/wikiindex.md — Beta-Blocker-Post-MI description updated