Beta-Blockers after Myocardial Infarction and Preserved Ejection Fraction

 Most trials that have shown a benefit of beta-blocker treatment after myocardial infarction included patients with large myocardial infarctions and were conducted in an era before modern biomarker-based diagnosis of myocardial infarction and treatment with percutaneous coronary intervention, antithrombotic agents, high-intensity statins, and renin–angiotensin–aldosterone system antagonists.

CONCLUSIONS

Among patients with acute myocardial infarction who underwent early coronary angiography and had a preserved left ventricular ejection fraction (≥50%), long-term beta-blocker treatment did not lead to a lower risk of the composite primary end point of death from any cause or new myocardial infarction than no beta-blocker use. (Funded by the Swedish Research Council and others; REDUCE-AMI ClinicalTrials.gov number, NCT03278509.)

The efficacy of beta-blockers in patients with heart failure and reduced ejection fraction is well documented. Trials have also shown that long-term beta-blocker therapy after myocardial infarction reduces mortality by approximately 20%.1-3 However, these results are from trials that mainly involved patients with large myocardial infarctions and left ventricular systolic dysfunction and were conducted primarily in the 1980s. This era predates advancements such as high-sensitivity cardiac troponins, percutaneous coronary interventions, antithrombotic agents, high-intensity statins, and renin–angiotensin–aldosterone system antagonists. A meta-analysis suggested that in the era of modern reperfusion strategies, beta-blockers did not significantly reduce mortality.4 Data on the effect of long-term beta-blocker therapy in patients with acute myocardial infarction and preserved ejection fraction are lacking from contemporary, sufficiently powered, randomized clinical trials.

Divergent conclusions have emerged from extensive observational studies and meta-analyses of such studies.5-10 A Cochrane review underscored the need for new trials in this patient population to address current clinical practices.11 Despite the lack of clear evidence of benefit in the contemporary setting, current guidelines widely recommend beta-blocker use after myocardial infarction.12-14 We conducted a trial (Randomized Evaluation of Decreased Usage of Beta-Blockers after Acute Myocardial Infarction [REDUCE-AMI]) to investigate whether long-term oral beta-blocker treatment in patients with acute myocardial infarction and preserved left ventricular ejection fraction would lead to a lower risk of a composite end point of death of any cause or new myocardial infarction than no beta-blocker use.

Discussion

In this registry-based, prospective, randomized, open-label, parallel-group trial conducted across 45 centers, most of which were in Sweden, the early initiation of oral beta-blocker treatment after an acute myocardial infarction in patients with a preserved left ventricular ejection fraction did not lead to a lower cumulative incidence of death from any cause or new myocardial infarction (composite primary end point). In addition, no appreciable between-group differences were observed in the analyses of secondary efficacy and safety end points. After 1 year, the incidence and severity of symptoms appeared to be similar in the two groups among the patients in Sweden who attended registry follow-up visits and had symptoms assessed. The absence of an effect of beta-blocker treatment on the cumulative incidence of death or myocardial infarction appeared to be consistent across all prespecified subgroups.

The baseline characteristics indicated that the patients who were included in the trial were representative of the population of patients with myocardial infarction and preserved ejection fraction in the trial countries (Table S6) and were generally at low risk for new cardiac events. The patients were well treated with early revascularization procedures and received evidence-based medications at discharge. The overall annual event rates for the primary end point (2.4% in the beta-blocker group and 2.5% in the no–beta-blocker group) were lower than we had expected before the initiation of the trial. We designed the trial as a superiority trial, powered to detect a 25% lower risk of death or myocardial infarction with beta-blocker treatment (corresponding to a 0.7-percentage-point lower risk per year, given the actual annual event rates that we observed), which we regarded as a clinically relevant effect. Although the neutral result that we found in this trial does not rule out either a small beneficial or detrimental effect, the overlapping time-to-event curves that were observed throughout the follow-up period and the consistent results in all the prespecified subgroups and for the secondary end points make a clinically relevant difference unlikely. Our findings are also consistent with the results of several large observational studies and meta-analyses of such studies.5,7,9,10 The possible signal of a harmful effect of beta-blocker treatment in the subgroup of patients who were taking a beta-blocker at admission is of unclear relevance and is probably a spurious finding.

Our trial included only patients who had a left ventricular ejection fraction of at least 50%. During the planning phase, many potential investigators were hesitant to include patients who had a mid-range left ventricular ejection fraction (40 to 49%). We also wanted to keep the trial population as homogeneous as possible, since any interaction between trial group and a subgroup makes the trial results more difficult to interpret and generalize. A later meta-analysis of clinical trials involving patients with a mid-range left ventricular ejection fraction suggested a beneficial effect of beta-blockers generally, and a large Korean registry suggested a benefit specifically after myocardial infarction.7,17

We allowed only beta-1–receptor selective blockers (metoprolol and bisoprolol) because these drugs had the best documentation for long-term treatment and had been used extensively in the countries involved in the trial. Indications for beta-blockers other than secondary prevention was an exclusion criterion. We also mandated an early invasive strategy because it reflects a contemporary treatment strategy — that is, the basis for reevaluation of beta-blockers in a new trial. Three other large, ongoing trials examining long-term treatment with beta-blockers in patients with myocardial infarction and preserved fraction have defined a preserved ejection fraction of at least 40% and also are allowing the use of nonselective beta-blockers.18-20 Two of the trials also include patients being treated without an early invasive approach.18,19

The doses of beta-blockers that were used in our trial were lower than those in previous trials. However, the doses that were used in our trial mirror the current practice of beta-blocker treatment, and no apparent association between the planned target dose of beta-blocker treatment and the primary end point was observed. Results from contemporary observational studies comparing various doses of beta-blockers have not shown any clear association with outcome.21,22 A study from the SWEDEHEART registry that compared 33,126 patients who received a prescription for at least 50% of the target beta-blocker dose at discharge with 64,449 patients who received a prescription for less than 50% of that dose did not show a between-group difference in outcome.22

Our trial has several limitations. First, it was an open-label trial, because blinding was not judged to be feasible. Data on clinical end points were obtained from the SWEDEHEART registry and the Swedish Population Registry and were not centrally adjudicated. However, this approach should have had a limited effect on the hard composite primary end point, whereas results regarding softer end points such as symptoms need to be interpreted more cautiously. During follow-up, investigators reviewed electronic health records to confirm that reported new myocardial infarctions in the SWEDEHEART registry fulfilled the criteria for a myocardial infarction according to the treating physician, and any misclassification should have been equally distributed over the two randomized trial groups.

Second, only safety end points that are associated with hospitalization were assessed. Third, a limitation of pragmatic trials of routinely used therapy is the potential for crossovers. Despite strategies to mitigate this issue, among patients with available information, 14% of those who had been assigned to the no–beta-blocker group were taking beta-blockers after 1 year of follow-up, and we do not yet have information about beta-blocker use after the first year. The adherence to the assigned beta-blocker regimen mirrored patterns that are observed in everyday clinical practice23,24; however, we cannot rule out the possibility that the use of beta blockers in the no–beta-blocker group contributed to our null finding.

In this registry-based, prospective, randomized, open-label, parallel-group trial that investigated whether oral beta-blocker therapy that was initiated early in patients with myocardial infarction who underwent early coronary angiography and had a preserved left ventricular ejection fraction would improve long-term outcome, beta-blocker treatment did not result in a lower cumulative incidence of the composite primary end point of death from any cause or new myocardial infarction.