Beta-Blockers after Myocardial Infarction and Preserved Ejection Fraction (REDUCE-AMI)

Authors, Journal, Affiliations, Type, DOI

• Authors: Troels Yndigegn, Bertil Lindahl, Katarina Mars, Joakim Alfredsson, and Tomas Jernberg, for the REDUCE-AMI Investigators
• Journal: New England Journal of Medicine, 2024; 390(15):1372–1381
• Affiliations: Lund University / Skåne University Hospital; Uppsala University / Uppsala Clinical Research Center; Karolinska Institutet; centres in Sweden (38), Estonia (1), New Zealand (6) — 45 total
• Type: Registry-based, prospective, open-label, parallel-group, randomized clinical trial (superiority design)
• DOI: https://doi.org/10.1056/NEJMoa2401479

Overview

REDUCE-AMI enrolled 5,020 patients with acute MI who underwent coronary angiography and had LVEF ≥50%, randomising 1:1 to long-term oral beta-blocker (metoprolol or bisoprolol) vs no beta-blocker, with a median follow-up of 3.5 years. The primary composite endpoint of death from any cause or new MI occurred in 7.9% of the beta-blocker group and 8.3% of the no-beta-blocker group (HR 0.96; 95% CI 0.79–1.16; P=0.64). No secondary endpoint — including all-cause death, cardiovascular death, recurrent MI, AF hospitalisation, or HF hospitalisation — favoured beta-blocker treatment. The trial established that, in the contemporary reperfusion era with optimal medical therapy, beta-blockers do not reduce death or reinfarction in post-MI patients with preserved LV function.

Keywords

Beta-blockers; myocardial infarction; preserved ejection fraction; metoprolol; bisoprolol; REDUCE-AMI; SWEDEHEART; secondary prevention; percutaneous coronary intervention; randomized clinical trial

Key Takeaways

Background

• Classical trials showing ~20% mortality reduction with beta-blockers post-MI (Norwegian Multicenter, MIAMI, BHAT) predominantly enrolled patients with large MIs and LV systolic dysfunction in the pre-PCI, pre-statin, pre-RAAS era.
• A meta-analysis (Bangalore 2014) suggested no significant mortality benefit of beta-blockers in the modern reperfusion era.
• Observational data and meta-analyses on patients with preserved EF were divergent; a Cochrane review (Safi 2021) explicitly called for new RCTs in this population.
• Despite absence of contemporary evidence, ACC/AHA (2013 STEMI, 2014 NSTEMI-ACS), and ESC 2023 ACS guidelines broadly recommended beta-blockers post-MI irrespective of LVEF.

Study Design and Population

• Registry-based RCT leveraging the SWEDEHEART registry for outcome ascertainment in Sweden (95.4% of patients); Estonia and New Zealand used electronic case-report forms with hospital record follow-up.
• Eligibility: Adult post-MI patients, 1–7 days after MI onset; coronary angiography completed; LVEF ≥50% by echocardiography; obstructive CAD documented (stenosis ≥50%, FFR ≤0.80, or iFR ≤0.89).
• Key exclusions: Indication for or contraindication to beta-blockers; non-residents of trial countries.
• Baseline characteristics: Median age 65 years; 22.5% women; 35.2% STEMI; 46.2% hypertension; 14.0% diabetes; 7.1% prior MI; 0.7% prior HF; one-vessel disease 55.4%, two-vessel 27.0%, three-vessel/left main 16.6%.
• Background therapy at discharge (near-universal): Aspirin 97.4%; P2Y12 blocker 95.8%; ACEi/ARB 80.2%; statin 98.5%; PCI performed 95.5%.

Interventions

• Beta-blocker group: Metoprolol (first choice; target ≥100 mg/day; median starting 50 mg; median target 100 mg) or bisoprolol (alternative; target ≥5 mg/day; median starting 2.5 mg; median target 5 mg).
  • 62.2% received metoprolol, 37.8% bisoprolol.
  • At 6–10 weeks, 90.6% adherent; at 11–13 months, 81.9% adherent.
• No-beta-blocker group: Beta-blockers actively discouraged; 11.3% at 6–10 weeks and 14.3% at 11–13 months were taking beta-blockers (crossover).
• Doses used were lower than in historical trials, mirroring current clinical practice; no dose-response relationship with primary endpoint was observed.

Primary and Secondary Endpoints

• Primary (composite death/new MI):
  • Beta-blocker: 199/2508 (7.9%; annual rate 2.4%)
  • No beta-blocker: 208/2512 (8.3%; annual rate 2.5%)
  • HR 0.96 (95% CI 0.79–1.16; P=0.64) — no significant difference
• Secondary endpoints (all non-significant):
  • All-cause death: 3.9% vs 4.1% (HR 0.94; 95% CI 0.71–1.24)
  • Cardiovascular death: 1.5% vs 1.3%
  • New MI: 4.5% vs 4.7% (HR 0.96; 95% CI 0.74–1.24)
  • Hospitalisation for AF: 1.1% vs 1.4%
  • Hospitalisation for HF: 0.8% vs 0.9%
• Symptom endpoints (SWEDEHEART follow-up visits): Incidence and severity of angina (CCS class) and dyspnoea (NYHA class) after 6–10 weeks and 11–13 months were similar between groups.
• Subgroup analyses: No significant heterogeneity in treatment effect for composite primary endpoint or death across all prespecified subgroups. Exception: patients already on a beta-blocker at admission showed a non-significant trend toward higher risk with randomisation to beta-blocker — interpreted as likely spurious.

Safety Endpoints

• Hospitalisation for bradycardia, 2nd/3rd-degree AVB, hypotension, syncope, or pacemaker implantation: 3.4% vs 3.2% (NS)
• Hospitalisation for asthma or COPD: 0.6% vs 0.6% (NS)
• Hospitalisation for stroke: 1.4% vs 1.8% (non-proportional hazards; RMST analysis performed — 2195 vs 2188 days over 2224-day follow-up)

Discussion

• Actual annual primary event rates (~2.4–2.5%/year) were substantially lower than pre-trial estimates (7.2%/year), reflecting modern evidence-based post-MI care; trial was re-powered in July 2021 to detect a 25% lower risk (HR 0.75) requiring 379 events.
• Neutral result is consistent with several large observational studies and meta-analyses; overlapping KM curves throughout follow-up and consistent subgroup results make a clinically relevant benefit unlikely.
• Trial restricted to LVEF ≥50% to create a homogeneous population; does not address LVEF 40–49%.
• Beta-1-selective agents only (metoprolol, bisoprolol); non-selective agents and patients without early invasive strategy excluded.
• 14% crossover in no-beta-blocker group at 1 year cannot be excluded as contributing to null result, but crossover mirrors real-world clinical practice.

Limitations of the Document

• Open-label design (blinding judged infeasible); outcome data from registry (not centrally adjudicated) — appropriate for hard endpoints, less reliable for softer symptom endpoints.
• Safety endpoints assessed only if resulting in hospitalisation; ambulatory side effects not captured.
• 14% crossover (no-BB arm receiving BBs by 1 year); information on BB use beyond the first year unavailable.
• 95.4% Swedish population — reduced generalisability; predominantly white, low-risk post-MI cohort.
• Dose targets lower than historical trials (though no dose-response observed in sensitivity analyses).
• LVEF 40–49% (mid-range) not represented.
• Women underrepresented (22.5%).
• No pharmacogenomic stratification (e.g., GRK5, ADRB1 polymorphisms).

Key Concepts Mentioned

• concepts/Beta-Blocker-Post-MI — directly tested; primary evidence source for preserved-EF stratum
• concepts/Antiarrhythmic-Drugs — beta-blockers as antiarrhythmic class; role in LVEF <40% HF
• concepts/OMI-NOMI-Paradigm — STEMI vs NSTEMI population mix (35.2% STEMI); evolving MI classification

Key Entities Mentioned

• entities/SWEDEHEART — Swedish registry used for outcome ascertainment and follow-up
• entities/Acute-Coronary-Syndrome — index event population; PCI performed in 95.5%
• entities/Metoprolol — beta-1 selective blocker; first choice in trial
• entities/Bisoprolol — beta-1 selective blocker; alternative in trial

Wiki Pages Updated

• wiki/sources/bb-mi-reduceami-nejm-2024.md — created (this page)
• wiki/concepts/Beta-Blocker-Post-MI.md — updated: REDUCE-AMI as dedicated source, source_count 4→5
• wiki/sourceindex.md — new entry added
• wiki/wikiindex.md — Beta-Blocker-Post-MI concept entry updated