Review of the COMMIT Trial
Early intravenous then oral metoprolol in 45,852 patients with acute myocardial infarction: a randomized placebo-controlled trial
Background Beta blockers were routinely used for the early management of patients with AMI; however, their efficacy and safety remained uncertain in this clinical scenario.
Recall that in the BHAT trial, propranolol reduced death 2 years following an AMI with an NNT of approximately 33 but the cohort was highly selected, the drug was started, on average, 14 days following admission and those over the age of 70 were excluded. ISIS-1 found that atenolol reduced death with an NNT of approximately 100 when started immediately in lower risk AMI patients but treatment effect heterogeneity was observed in patient subgroups with higher risk features.
The ClOpidogrel and Metoprolol Infarction Trial (COMMIT) sought to test the hypothesis that early beta-blockade with metoprolol would reduce cardiac events and death in a broad population of patients with AMI.
Patients Patients with suspected myocardial infarction, who presented with ST elevation, left-bundle branch block, or ST depression within 24 hours of symptom onset were eligible. Patient eligibility was ultimately determined by the responsible physician but protocol guidance suggested the following relative contraindications: persistently low blood pressure (SBP <100 mmHg) or low heart rate (<50 bpm), heart block, or cardiogenic shock. Unlike previous trials, evidence of moderate heart failure (Killip II or III) was not an exclusion criteria.
Baseline characteristics The average age of participants was 61 and the majority were male (72%). The mean SBP and heart rate at entry were 128 mmHg and 82 beats per minute, respectively. Less than 10% had a previous MI. ST elevation was present in 87% of patients, ST depression in 7%, and bundle branch block in 6%. Thirty percent of patients had heart rate ≥90 bpm upon admission and 7% of those had a heart rate ≥110 bpm. Thirty three percent of patients had SBP <120 mmHg at study entry. Approximately 25% of patients were classified as Killip class II (20%) corresponding to the presence of rales, crackles, an S3, elevated JVP or pulmonary venous congestion on chest x-ray or Killip class III (5%) corresponding to frank pulmonary edema.
Procedures This trial had a 2x2 factorial design but we will focus only on the metoprolol comparison. Following randomization, patients were immediately given metoprolol or placebo via the following sequence of interventions:
An immediate intravenous injection of 5 mg of metoprolol or matching placebo, given over 2- 3 minutes.
About 2-3 minutes later, if the heart rate was >50 bpm and the SBP was >90 mmHg, another 5 mg of metoprolol or matching placebo was injected.
About 2-3 minutes later, another 5 mg of metoprolol or matching placebo was injected if the heart rate and SBP met the above parameters.
15 minutes after these intravenous doses, a 50 mg metoprolol or placebo tablet was to be given and repeated every 6 hours during days 0-1.
From day 2 onward, 200 mg of sustained-release metoprolol or placebo was to be given for up to 4 weeks or until hospital discharge (whichever came first) unless some definite contraindication was thought to have arisen.
At hospital discharge or at day 28 (whichever came first), a single-sided follow-up form was to be completed. No further follow-up was sought. Post-discharge use of aspirin, beta-blocker, and other established therapies was encouraged but not monitored.
Endpoints There were 2 prespecified co-primary outcomes. One was a composite of death, reinfarction or cardiac arrest and the other was all-cause death during the scheduled treatment period (i.e., until hospital discharge or day 28).
The investigators estimated the event rate in the placebo group for the primary composite endpoint would be 14% prior to the start of the study; however, that estimate was reduced to 10% as the study proceeded. Thus, they estimated a sample size of 45,000 patients would be needed to detect a 10% reduction in the primary composite endpoint with 95% statistical power and a p value of 0.05.
Results 45,852 participants were randomized from 1250 hospitals. 90% of beta blocker-allocated participants received all 3 doses of IV medicine and 86% completed treatment with an oral medication. The average treatment duration was 15 days.
There were 22,929 patients in the metoprolol group and 22,923 in the placebo group. Compared to placebo, metoprolol did not significantly reduce the primary composite endpoint (OR 0.96; 9.4% vs 9.9%; 95% CI 0.90-1.01; p=0.10) or all-cause death (OR 0.99; 7.7% vs 7.8%; 95% CI 0.92-1.05; p=0.69). Metoprolol use reduced reinfarction (2.0% vs 2.5%) and ventricular fibrillation (2.5% vs 3.0%) but increased cardiogenic shock (5.0% vs 3.9%). Metoprolol use also increased heart failure without shock (14.1% vs 12.7%), persistent hypotension without shock (6.0% vs 2.9%), and bradycardia (5.4% vs 2.2%).
Treatment effect heterogeneity was evident based on exploratory subgroups. Patients whose MI’s were complicated by hemodynamic instability or heart failure and who were higher risk for experiencing adverse events did worse with metoprolol. For example, for patients with SBP <120 mmHg, death rates were 9.6% in the metoprolol group compared to 8.8% in the placebo group but were 6.8% and 7.6% in those whose SBP was 120-139 mmHg.
A similar but more striking trend was evident based on heart rate and Killip class. For patients with heart rate ≥110 bpm, death rates were 20.3% in the metoprolol group and 18.3% in the placebo group but were 6.0% and 6.6% in those whose heart rate was 70-89 bpm. For patients classified as Killip class III, death rates were 19.7% in the metoprolol group and 16.5% in the placebo group but were 5.8% and 6.1% in those classified as Killip class I.
Conclusions IV metoprolol followed by oral administration over approximately 15 days did not reduce cardiovascular events or death in patients presenting with AMI. Unlike the ISIS-1 trial, patients presenting with heart failure and hemodynamic instability were NOT excluded from COMMIT and this explains the divergent results and this fact is critically important for clinical translation. In the discussion section of the manuscript the authors present data from a meta-analysis of effects of IV followed by oral beta-blocker therapy. It shows fairly consistent results across trials for death, reinfarction and VF when the low-risk patients from COMMIT are combined with prior studies including ISIS-1 and many other smaller trials.
The COMMIT trial is an important trial in the management of acute coronary syndrome as it addressed a critical gap in the evidence. In low-risk, hemodynamically stable AMI patients, beta blocker use confers a clinically small but statistically significant reduction in death, reinfarction and VF; however, in higher-risk, unstable patients with heart failure, beta blocker use increases death and cardiogenic shock and should be used with extreme caution or avoided altogether.