Is the level of evidence for the use of beta-blockers in acute myocardial infarction satisfactory enough?

The cornerstone of each therapy's recommendation should be the impact on survival. In the August 2004 issue of the European Heart Journal, expert consensus opinion on beta-blockers stated that during the acute phase of myocardial infarction, oral beta-blockers are indicated in all patients without contraindications (Class I, level of evidence A).1 The consensus group opinion was supported by the data from the pre-reperfusion ISIS-12 and MIAMI3 trials and post-reperfusion TIMI-II trial4 which compared two time protocols of metoprolol administration and, therefore, was not controlled with a non-beta-blocker or another beta-blocker. It is understandable that the results from pre-reperfusion studies cannot be simply incorporated into a modern concept of AMI treatment. Retrospective analysis from PCI studies (PAMI-2, Stent PAMI, Air PAMI, PAMI noSOS, and CADILLAC) was used to support the premise of mortality reduction with pre-PCI intravenous and post-PCI peroral use of beta-blockers.5,6 However, none of these studies were designed to randomly assess the effects of beta-blockers on mortality. Oral beta-blockers are further recommended for long-term use and for survival improvement in all patients who recover from AMI and do not present contraindications (Class I, level of evidence A); this recommendation is based on the data from a meta-analysis of 31 randomized trials, Hjalmarson's trial, Cooperative Cardiovascular Project, BHAT and Norwegian trial.7,8 However, all these trials are from the pre-reperfusion era or use registry as a database. None of these trials included random assessment of beta-blockers in the post-reperfusion modern algorithm, including PCI, fibrinolysis, ACE-inhibitors, aspirin, and statines. Furthermore, the document states that the benefit of beta-blockers in low-risk patients is questionable. CAPRICORN, the only randomized post-reperfusion trial in AMI patients with ventricular dysfunction, showed all-cause mortality reduction with carvedilol when compared with placebo.9 However, CAPRICORN did not reach the pre-specified higher level of statistical significance for the original primary endpoint of all-cause mortality (the primary endpoint was changed for co-primary during the study). Therefore, the statistical power of CAPRICORN is only borderline in favour of mortality reduction with carvedilol in high-risk AMI patients (P=0.031). Intravenous beta-blocker administration for primary prevention of sudden death was marked as a Class I, level of evidence B. The document admitted that post hoc analysis of GUSTO-1 trial does not support intravenous use of beta-blockers in the reperfusion era.10 In TIMI-IIB trial, early intravenous metoprolol administration had no advantage over the peroral treatment.4 Evidence for the use of beta-blockers in the early stage of AMI thus results from studies in the pre-reperfusion era, registries, meta-analyses, and post hoc analysis, not designed for beta-blockers' random assessment in regard to mortality reduction after AMI. In the absence of powerful statistical evidence for mortality reduction in the reperfusion time, we believe the impact of beta-blockers on survival after AMI has not been proved sufficiently. For a better evaluation of the role of beta-blockers in patients with AMI, randomized large trials assessing the impact on survival are obviously missing.