Overall Study Question
This randomized, multicentre, double blind, placebo-controlled trial (denoted the “Copernicus” study) was designed to determine the effects of the beta-blocker carvedilol on the survival of patients with severe heart failure as a result of ischemic or non-ischemic cardiomyopathy.
Patient were included in this study if they demonstrated symptoms of heart failure at rest or on minimal exertion, were clinically euvolemic, and had an ejection fraction of less than 25 percent, despite being treated with conventional therapy. Conventional therapy was defined as treatment with diuretics (in a dose which achieved the absence of rales, ascites, and minimal peripheral edema) and an angiotensin-converting-enzyme inhibitor (or an angiotensin II receptor blocker). Treatment with nitrates, hydralazine, digitalis, spironolactone, and amiodarone was allowed, but not required.
There were several exclusion criteria employed in this study. Patients were excluded if they required recent intensive care or intravenous inotropes or vasodilators within four days of screening. Patients were also excluded if they had heart failure as a result of valvular disease, correctible cardiomyopathy, or were likely to receive a cardiac transplant in the near future. Patients were also excluded if they had recently undergone coronary revascularization, experienced an acute MI, stroke, or cardiac arrhythmia. Patients who had received an alpha-adrenergic blocker, a calcium channel blocker, or a class I antiarrhythmic drug within the four weeks, or a beta-blocker within two months, were also excluded; along with patients with a systolic BP <85mmHg, a HR <68bpm, a SCr >247.5umol/L, or a serum potassium <3.5mmol/L or >5.2mmol/L.
Patients meeting study criteria were assigned to receive either oral carvedilol or matching placebo. The dose of each drug was started at 3.125mg po twice daily, and was doubled every two weeks up to a target dose of 25mg twice daily as tolerated by the patient. Doses of all concomitant drugs could be adjusted at the discretion of the investigator. The patients were evaluated every two months until the end of the trial. The primary outcome of the study was death from any cause, and the composite of death or hospitalization for any reason was used as a secondary endpoint.
Are the Results of the Study Valid?
1. Was assignment of patients randomized?
Yes. Patients who met the entry criteria were randomly assigned in a 1:1 ratio to receive either intervention. The study was executed by an independent steering committee who ensured the random allocation of treatments.
2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?
Yes. After the trial was discontinued prematurely, the investigators had a mean of 10.4 months of follow-up data for all 2,289 patients who were enrolled. No patients were lost to follow-up. In addition, the primary data analysis was performed on an intention-to-treat basis.
- 3. Were patients, their clinicians, and study personnel ‘blind’ to treatment?
- Yes. This was a double-blind trial. Patients, clinicians, and study personnel were blinded to treatment. The trial was designed, executed, and analyzed by a steering committee, an end-points committee, a biostatistics center, and a data and safety monitoring board, all who conducted their roles independently from the sponsors.
- 4. Were the groups similar at the start of the trial?
- Yes. The groups appear to be similar at baseline with respect to disease severity (ejection fraction, previous hospitalizations, blood pressure, heart rate, and serum creatinine) and concomitant treatments (ACE inhibitor, digitalis, diuretic, spironolactone, and amiodarone use). They did not report the proportion of patients receiving ASA and cholesterol lowering therapies; however, considering these patients were extremely high-risk cardiac patients, it might be assumed that most were receiving these therapies. The authors also did not report if the mean dose of ACE inhibitor therapy was equivalent between the groups, which potentially may bias the results; however, the authors addressed this point in a subsequent letter to the editor in which they report that the average doses were indeed similar between groups.
- 5. Aside from the experimental intervention, were the groups treated equally?
- Yes. Both groups were followed-up at 2-month intervals until the discontinuation of the trial, and fewer than 5% of patients in both groups received open-label treatment with a beta-blocker at some point in the trial.
What were the Results?
1. How large was the treatment effect?
Recruitment was terminated prematurely after a mean follow-up of 10.4 months. This was undertaken when it was determined by interim analysis that the patients in the carvedilol treatment group were having a significant beneficial effect from the intervention compared to the placebo group.
According to the intention to treat analysis, there was a statistically significant benefit in the carvedilol group for the primary outcome of all-cause mortality. The placebo group had 190 of the original 1,133 patients die during the 10-month follow-up, which is a crude mortality rate of 16.8%. The carvedilol group had 130 of the original 1,156 patients die during the same timeframe, which is a crude mortality rate of 11.2%. This represents an absolute reduction in all-cause mortality of 5.6% (p=0.00013), a relative risk reduction of 35%, and a corresponding number needed to treat (for approximately 4 months) of 18. According to Kaplan-Meier analysis, the cumulative risk of death at one year was 18.5% in the placebo group and 11.4% in the carvedilol group (p<0.05).
A similar beneficial effect was observed for the secondary outcome of the combined risk of death or hospitalization. According to the intention to treat analysis, 44.7% of the patients in the placebo group died or were hospitalized over the 10-month follow-up, compared to 36.8% in the placebo group (p<0.01). This represents an absolute reduction of 7.9% (relative risk reduction of 18%) and a corresponding number needed to treat (for approximately 4 months) of 13.
These benefits were seen across all subgroups defined according to sex, age, left ventricular ejection fraction, cause of heart failure, location of study center, and history with respect to hospitalization in the previous year.
The carvedilol also appeared to be better tolerated than placebo, with cumulative withdrawal rates due to adverse effects or for reasons other than death being 18.5% in the placebo group and 14.8% in the carvedilol group (p=0.02). It is unexpected, and difficult to rationalize, that the placebo group actually had more withdrawals than carvedilol in this trial. However, this difference could be explained by more patients in the placebo group discontinuing therapy due to poor disease control, as opposed to poor tolerability of the placebo. Unfortunately the authors did not provide specific reasons for drug therapy withdrawal, making it difficult to explain this result. Nevertheless, it is comforting from a clinician’s perspective that as many as 85% of patients with severe heart failure will tolerate carvedilol if it is initiated and titrated conservatively.
2. How precise was the estimate of the treatment effect?
The reported confidence intervals were fairly narrow and did not approach zero (no effect). The relative risk reduction of 35% for the primary endpoint of all-cause mortality had a 95% CI of 19 – 48%. This suggests the estimate of the treatment effect was very precise, as the minimal benefit that we could expect according to this trial is still a 19% relative risk reduction in all-cause mortality.
Will the Results Help Me in Caring for My Patients?
- 1. Can the results be applied to my patient care?
The results of this study suggest that long-term treatment with carvedilol results in a significant benefit in patients with severe chronic heart failure, regardless of what subgroup they are from. These results can be applied to my patients only to the extent that they would have fit into the rather extensive inclusion/exclusion criteria of this trial. Patients who required intensive care, had marked fluid retention, were receiving IV inotropes or vasodilators, had a recent MI or stroke, were hypotensive, or had severe renal dysfunction were not included in this study. It cannot be assumed that these patients would receive the same benefits from treatment with carvedilol, and therefore should be stabilized before being considered for this therapy. It is possible that in patients who are severely decompensated, activation of the sympathetic nervous system is critical to maintain cardiovascular stability, and administration of a sympathetic nervous system antagonist may not be advantageous.
Interestingly, only 19-20% of the patients were receiving concomitant therapy with spironolactone during this trial. Considering the results of the RALES Trial, most of these patients with severe heart failure would benefit from this intervention. Therefore, there is the possibility that these results with carvedilol may not be the same in a present day cohort of patients also receiving spironolactone. However, the RALES Trial did report that the beneficial effects of spironolactone were similar in subgroups of patients who did, and who did not receive concomitant beta-blockers. Also, the authors of the current study reported in their subsequent letter to the editor that a post-hoc analysis of patients who were, and who were not receiving spironolactone, suggested that the benefits of carvedilol were similar in both cohorts. Therefore, we can be fairly confident that the benefits seen with carvedilol in this study can be applied to our severe chronic heart failure patients today, most of whom are receiving spironolactone.
2. Were all clinically important outcomes considered?
The endpoints of mortality and hospitalization were considered, which are the outcomes of clinical importance to patients suffering from severe chronic heart failure.
3. Are the likely treatment benefits worth the potential harms and costs?
Administration of carvedilol in patients with severe chronic heart failure resulted in a relative risk reduction in mortality of 35%, which translated into a number needed to treat of 18 for approximately 10 months to prevent one death, and a number needed to treat of 13 for the same duration to prevent one death or hospitalization. These outcomes are without a doubt very clinically important. The dramatic physical and financial benefits of this reduction in death and hospitalization are realized with very little cost or increased risk. The titration of carvedilol was actually better tolerated in this trial than placebo was, and could likely be done in an outpatient setting, making it safe and inexpensive. However, there were still 15% of the patients in the carvedilol group who could not tolerate the medication, which suggests that even using a very slow dose titration, many patients will not tolerate this therapy.
Metoprolol and carvedilol have been demonstrated in earlier trials to be effective in reducing the risk of hospitalization and death in patients with chronic heart failure. Unfortunately, these earlier studies primarily involved patients who had only mild-moderate heart failure (i.e. New York Heart Association (NYHA) Class II or III). Consequently, little has been known about the effects of these agents in patients with severe heart failure. This distinction is very significant, and concern has been expressed that beta-blockers may actually worsen the symptoms and prognosis of heart failure, especially in patients with very severe disease. This concern appears to be unfounded in patients with mild-moderate disease, but it remains a valid consideration in patients with severe symptoms.
The results of this study are likely to have a profound effect on the treatment of patients with severe heart failure. This trial has demonstrated that long-term treatment of patients with severe heart failure (NYHA Class III-IV symptoms and ejection fraction<25%) using carvedilol has a significant benefit in prolonging survival and preventing hospitalization. This is the first published study to reveal a beneficial effect using a beta-blocker in severe heart failure, and helps disprove the theory that it may be unsafe to use these drugs in this patient cohort. This study provides a missing piece of the therapeutic puzzle as it relates to beta-blocker therapy in the treatment of chronic heart failure. There is now evidence that, regardless of the severity of the symptoms, patients with heart failure will benefit from beta-blocker therapy. However, future research is still required to determine if the benefits in patients with severe heart failure can be extrapolated to other beta-blockers, and to a cohort of patients who are all concomitantly receiving spironolactone.
When titrated very slowly, carvedilol was found to be better tolerated than placebo in this study; however, a substantial proportion of patients (i.e. 15%) were still unable to tolerate the medication. Therefore, despite the fact that this trial demonstrates that the administration of carvedilol is safe and effective in these patients with severe chronic heart failure, this drug cannot be tolerated by all patients.
The significant question that remains is whether or not these results can be extrapolated to the use of beta-blockers other than carvedilol. As mentioned earlier, metoprolol, carvedilol and bisoprolol have all been shown to be safe and effective in patients with milder disease. A recent randomized controlled trial involving an assessment of the long-term effects of bucindolol in patients with severe heart failure showed no significant difference between bucindolol and placebo after ten months of follow-up. This suggests that perhaps the previous beneficial effects seen with metoprolol or bisoprolol in patients with mild-moderate heart failure cannot be assumed to translate into similar benefits in severe heart failure, as they have with carvedilol.
Carvedilol has been shown to have some pharmacological effects (i.e. alpha-adrenergic blockade, antioxidant effects, antiendothelin effects) in addition to beta-adrenergic blockade. This property does not appear to exist for most other beta-blockers. Since it is not known what the precise mechanism is that leads to the beneficial effects of beta-blockers in heart failure, it is plausible that these unique aspects of carvedilol may translate into an enhanced benefit in patients with severe heart failure. Therefore, based on the lack of evidence with other beta-blockers, and the existing evidence with bucindolol, it should not be assumed that any beta-blocker other than carvedilol is safe and effective in treating patients with severe heart failure. This creates an interesting therapeutic and ethical dilemma when using a beta-blocker other than carvedilol for a patient with mild-moderate heart failure, when we know they will eventually progress to a more severe state, at which time we cannot guarantee the effectiveness of their beta-blocker therapy. These questions and dilemmas will likely be clarified as results of future trials are reported which study the effects of other beta-blockers in patients with severe heart failure.