Cardiology · Medical Literature
Beta-Blockers After Myocardial Infarction With Preserved and Mildly Reduced Ejection Fraction: A Meta-Analysis With Trial Sequential Analysis.
Abstract
We aimed to evaluate the efficacy of β-blockers after myocardial infarction (MI) across left ventricular ejection fraction (LVEF) strata and to assess the conclusiveness of the available evidence using trial sequential analysis (TSA). PubMed, Embase and ClinicalTrials.gov were searched for randomized controlled trials (RCTs) evaluating β-blockers in post-MI patients with LVEF ≥ 40%. A time-to-event meta-analysis was performed for the primary composite (as defined by each trial) and for individual endpoints. The Mantel-Haenszel method was used to pool risk ratios (RR) for major adverse cardiovascular events (MACE; death, MI or heart failure), including LVEF-stratified analyses. TSA estimated the required information size (RIS) and generated adjusted significance and futility boundaries, assuming a 5% type I error and 90% power. Across 19,826 post-MI patients (17,941 with LVEF ≥ 50% and 1885 with LVEF 40%-49%), β-blockers did not reduce time to the primary endpoint (HR 0.92, 95% CI 0.85-1.01; p = 0.08; I2 = 35%) or mortality. In patients with preserved LVEF (≥ 50%), β-blocker therapy did not reduce the risk of death or MACE (RR 0.96, 95% CI 0.87-1.05; I2 = 42%), with TSA confirming futility. Among those with mildly reduced LVEF (40%-49%), β-blockers reduced MACE (RR 0.74, 95% CI 0.58-0.94; I2 = 0%), although TSA failed to establish conclusiveness (RIS of 5717 [14.4%]). A significant interaction by LVEF subgroup was observed for cardiac death (p = 0.03). β-blockers conferred no benefit in post-MI patients with preserved LVEF, with conclusive evidence of futility, whereas therapy was associated with reduced MACE in those with mildly reduced LVEF, pending confirmation in further adequately powered randomized trials.
Clinical relevance
In post-MI patients with preserved (LVEF ≥50%) or mildly reduced (LVEF 40–49%) left ventricular ejection fraction managed in the contemporary reperfusion era, does beta-blocker therapy reduce mortality or major adverse cardiovascular events (MACE), and is the current body of randomized evidence sufficient to draw definitive conclusions? __SNAPSHOT__ {"studyDesign":"Systematic review and meta-analysis of randomized controlled trials, with trial sequential analysis (TSA) to assess statistical conclusiveness of the pooled evidence","setting":"International; contemporary RCTs conducted in the percutaneous coronary intervention (PCI) and guideline-directed medical therapy era; databases searched: PubMed, Embase, ClinicalTrials.gov","sampleSize":"n=19,826 total (17,941 with LVEF ≥50%; 1,885 with LVEF 40–49%)","population":"Post-MI patients with LVEF ≥40% enrolled in contemporary RCTs; patients without standard indications for beta-blocker therapy (e.g., reduced EF heart failure, significant arrhythmia); predominantly PCI-treated cohorts","intervention":"Oral beta-blocker therapy (agent and dose as defined by each included trial)","comparator":"No beta-blocker therapy (open-label in constituent trials)","primaryOutcome":"Time-to-primary-composite-endpoint as defined by each individual trial (trial-level heterogeneity in endpoint definition); MACE (death, MI, or heart failure) for pooled Mantel-Haenszel risk ratio analyses, including LVEF-stratified comparisons","followUp":"Not uniformly reported; varied by constituent trial (median follow-up of included trials ranged from approximately 1 to 3.5 years based on source trial data)","fundingSource":"Not reported in abstract; constituent trials were funded by non-profit and government sources (e.g., CNIC, South-Eastern Norway Regional Health Authority, Research Council of Norway)"}
Key findings
Across 19,826 post-MI patients with LVEF ≥40%, beta-blockers did not significantly reduce time to the primary composite endpoint (HR 0.92, 95% CI 0.85–1.01; p=0.08; I²=35%) or overall mortality. In the preserved LVEF subgroup (LVEF ≥50%; n=17,941), beta-blockers did not reduce the risk of death or MACE (RR 0.96, 95% CI 0.87–1.05; I²=42%), and trial sequential analysis crossed the pre-specified futility boundary, confirming that additional trials of similar design are unlikely to demonstrate a clinically meaningful benefit in this population. In the mildly reduced LVEF subgroup (LVEF 40–49%; n=1,885), beta-blockers were associated with a statistically significant reduction in MACE (RR 0.74, 95% CI 0.58–0.94; I²=0%); however, TSA determined that only 14.4% of the required information size (RIS of 5,717 patients) had been reached, rendering this finding statistically inconclusive. A significant interaction by LVEF subgroup was observed for cardiac death (p=0.03), suggesting that the relationship between beta-blocker therapy and cardiac mortality differs according to ejection fraction category. NNT for the mildly reduced LVEF subgroup cannot be calculated from the abstract alone, as absolute event rates in each arm were not reported in the provided text. No specific adverse event rates or NNH were reported in the abstract.
Practice implications
This meta-analysis with TSA provides the strongest available aggregate evidence that beta-blocker therapy after MI does not reduce mortality or MACE in patients with preserved LVEF (≥50%), with TSA confirming futility — meaning that even if more similar trials were conducted, a meaningful treatment effect in this population is unlikely to emerge from the current study design and assumptions. This is highly concordant with other recently published aggregate and individual patient-level meta-analyses (e.g., the NEJM 2025 IPD meta-analysis in preserved LVEF). Conversely, for patients with mildly reduced LVEF (40–49%), a 26% relative risk reduction in MACE was observed; this signal is statistically significant but not TSA-confirmed (only ~14% of the RIS achieved), indicating that the mildly reduced LVEF finding is promising but not yet conclusive. The significant LVEF–treatment interaction for cardiac death (p=0.03) reinforces that these two populations should not be grouped together clinically. Patients well-represented include those managed with contemporary PCI and GDMT in the modern reperfusion era. Patients not adequately represented include: those without PCI access, patients with significant comorbidities (e.g., COPD, diabetes, severe renal impairment), women (historically underrepresented in MI trials), elderly patients, and those with other standard indications for beta-blockers (arrhythmia, hypertension, angina), who were excluded. Because only an abstract was available for full review, additional methodological detail — including exact trial inclusion criteria, risk-of-bias assessment, specific beta-blocker agents and doses, and full sensitivity analyses — could not be independently verified. Routine initiation of long-term beta-blocker therapy solely for secondary prevention in post-MI patients with preserved LVEF is not supported by this evidence. The mildly reduced LVEF finding justifies continued treatment in current practice and further dedicated RCTs, but does not yet constitute definitive proof of benefit. __LIMITATIONS__ Only abstract available for full methodological review; complete risk-of-bias assessment, sensitivity analyses, and subgroup pre-specification details cannot be independently verified.,Constituent trials were open-label, introducing potential performance and detection bias; no placebo-controlled blinded trials of modern beta-blockers in this population exist.,Marked heterogeneity in primary endpoint definitions across trials (each trial used its own composite) complicates interpretation of the time-to-event pooled HR.,The mildly reduced LVEF subgroup (n=1,885) is substantially underpowered relative to the TSA-estimated RIS of 5,717 patients; the MACE reduction in this group must be considered hypothesis-generating rather than confirmatory.,I²=42% in the preserved LVEF MACE analysis signals moderate statistical heterogeneity, potentially reflecting differences in trial design, patient selection, beta-blocker agent/dose, and duration of therapy.,Subgroup analysis by LVEF was post-hoc in several constituent trials and should be interpreted with caution for the interaction p-value (p=0.03), which does not correct for multiple comparisons.,Exclusive inclusion of patients with LVEF ≥40% without standard beta-blocker indications limits generalizability to the broader post-MI population.,Absolute event rates and NNT/NNH not reported in the abstract, precluding direct assessment of clinical magnitude of effect in the mildly reduced LVEF group.,Long-term follow-up beyond 3–4 years is not available in most constituent trials; late divergence of outcomes cannot be excluded.,Patient-level covariates (age, sex, diabetes, renal function, specific beta-blocker agent) are not individually adjustable in aggregate-level meta-analysis, limiting precision of subgroup inferences. __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ In post-MI patients with preserved LVEF (≥50%), this meta-analysis with TSA provides conclusive evidence of futility for beta-blocker therapy in reducing MACE or mortality in the contemporary PCI era — routine long-term beta-blocker therapy for this indication is not supported. In those with mildly reduced LVEF (40–49%), a promising but TSA-inconclusive 26% relative MACE reduction was observed, and adequately powered dedicated RCTs in this subgroup are urgently needed before practice can be definitively changed.