Background: In the COMET study, carvedilol improved survival compared to metoprolol tartrate in 3029 patients with NYHA II-IV heart failure and EF <35%, followed for an average of 58 months.
Aims: To evaluate whether the effect on overall mortality was specific for a particular mode of death. This may help to identify the mechanism of the observed difference.
Methods: Of the 1112 total deaths, 972 were adjudicated as cardiovascular, including 480 sudden, 365 circulatory failure (CF) and 51 stroke deaths. For each mode of death, the effect of pre-specified baseline variables was assessed, including sex, age, NYHA class, aetiology, heart rate, systolic blood pressure, EF, atrial fibrillation, previous myocardial infarction or hypertension, renal function, concomitant medication, and study treatment allocation.
Results: In multivariate Cox regression analyses, compared to metoprolol, carvedilol reduced cardiovascular (RR 0.80, CI 0.7-0.91, p=0.0009), sudden (RR 0.77, CI 0.64-0.93, p=0.0073) and stroke deaths (RR 0.37, CI 0.19-0.71, p=0.0027) with a non-significant trend for CF death (RR 0.83, CI 0.66-1.04, p=0.07). Treatment benefit with carvedilol did not differ between modes of death, except for a greater reduction in stroke death with carvedilol (competing risk analysis, p=0.0071 vs CF death). There were no interactions between treatment allocation and baseline characteristics.
Conclusion: Mortality reduction with carvedilol compared to metoprolol appears relatively non-specific and could be consistent with a superior effect of carvedilol on cardiac function, arrhythmias or, in view of the greater reduction in stroke deaths, on vascular events.