bjects had tolerated a 2‑week course on immediate‑release carvedilol 6.25 mg twice daily.
In each trial, there was a primary end point, either progression of heart failure (1 U.S. trial) or exercise tolerance (2 U.S. trials meeting enrollment goals and the Australia‑New Zealand trial). There were many secondary end points specified in these trials, including NYHA classification, patient and physician global assessments, and cardiovascular hospitalization. Other analyses not prospectively planned included the sum of deaths and total cardiovascular hospitalizations. In situations where the primary end points of a trial do not show a significant benefit of treatment, assignment of significance values to the other results is complex, and such values need to be interpreted cautiously.
The results of the U.S. and Australia‑New Zealand trials were as follows:
Slowing Progression of Heart Failure: One U.S. multicenter trial (366 subjects) had as its primary end point the sum of cardiovascular mortality, cardiovascular hospitalization, and sustained increase in heart failure medications. Heart failure progression was reduced, during an average follow‑up of 7 months, by 48% (P = 0.008).
In the Australia‑New Zealand trial, death and total hospitalizations were reduced by about 25% over 18 to 24 months. In the 3 largest U.S. trials, death and total hospitalizations were reduced by 19%, 39%, and 49%, nominally statistically significant in the last 2 trials. The Australia‑New Zealand results were statistically borderline.
Functional Measures: None of the multicenter trials had NYHA classification as a primary end point, but all such trials had it as a secondary end point. There was at least a trend toward improvement in NYHA class in all trials. Exercise tolerance was the primary end point in 3 trials; in none was a statistically significant effect found.
Subjective Measures: Health-related quality of life, as measured with a standard questionnaire (a primary end point in 1 trial), was unaffected by carvedilol. However, patients’ and investigators’ global assessments showed significant improvement in most trials.
Mortality: Death was not a pre-specified end point in any trial, but was analyzed in all trials. Overall, in these 4 U.S. trials, mortality was reduced, nominally significantly so in 2 trials.
The COMET Trial
In this double-blind trial, 3,029 subjects with NYHA class II-IV heart failure (left ventricular ejection fraction less than or equal to 35%) were randomized to receive either carvedilol (target dose: 25 mg twice daily) or immediate-release metoprolol tartrate (target dose: 50 mg twice daily). The mean age of the subjects was approximately 62 years, 80% were males, and the mean left ventricular ejection fraction at baseline was 26%. Approximately 96% of the subjects had NYHA class II or III heart failure. Concomitant treatment included diuretics (99%), ACE inhibitors (91%), digitalis (59%), aldosterone antagonists (11%), and “statin” lipid-lowering agents (21%). The mean duration of follow-up was 4.8 years. The mean dose of carvedilol was 42 mg per day.
The trial had 2 primary end points: all-cause mortality and the composite of death plus hospitalization for any reason. The results of COMET are presented in below. All-cause mortality carried most of the statistical weight and was the primary determinant of the trial size. All-cause mortality was 34% in the subje |
