Long-term treatment with appropriate doses of carvedilol or metoprolol is currently recommended for patients with heart failure with reduced ejection fraction (HFrEF) to decrease the risk of death, hospitalizations, and patients’ symptoms. It remains unclear if the β blockers used in patients with HFrEF are equal or carvedilol is superior to metoprolol types. We performed a meta-analysis of the comparative effects of carvedilol versus metoprolol tartrate and succinate on all-cause mortality and/or hospitalization. We conducted an Embase and MEDLINE search for prospective controlled trials and cohort studies of patients with HFrEF who were received to treatment with carvedilol versus metoprolol. We identified 4 prospective controlled and 6 cohort studies with 30,943 patients who received carvedilol and 69,925 patients on metoprolol types (tartrate and succinate) with an average follow-up duration of 36.4 months. All-cause mortality was reduced in prospective studies with carvedilol versus metoprolol tartrate. Neither all-cause mortality nor hospitalizations were significantly different between carvedilol and metoprolol succinate in the cohort studies. In conclusion, in patients with HFrEF, carvedilol and metoprolol succinate have similar effects in reducing all-cause mortality.
Long-term treatment with appropriate doses of specific β blockers is currently recommended for patients with heart failure with reduced ejection fraction (HFrEF) to reduce the risk of death, hospitalizations, and patients’ symptoms independently of race, co-morbidities, and the presence of coronary artery disease. However, these are not class effects and have only been reported with 2 selective β1-receptor blockers (sustained release metoprolol, bisoprolol, and carvedilol ). Another selective β1 blocker, the short-acting metoprolol tartrate, has been shown to be less effective in clinical trials. Several cohort studies and registries have reached to different conclusions regarding the comparative effects of different β blockers in patients with HFrEF. The pleiotropic actions of carvedilol (including vasodilating, antioxidant, metabolic, and antiarrhythmic actions) were hypothesized to confer additional benefit on cardiovascular outcomes compared to metoprolol. The present meta-analysis was designed to systematically evaluate prospective controlled trials and observational cohorts and assess the effects of carvedilol versus metoprolol types (succinate and tartrate) on all-cause mortality and rehospitalization.
Methods
We systematically searched the electronic databases, MEDLINE, PubMed, Embase, and the Cochrane Library for Central Register of Clinical Trials, using the MESH terms, “b-blockers,” “heart failure with reduced ejection fraction,” “heart failures,” the names of individual β blockers (carvedilol, metoprolol succinate, metoprolol tartrate), “randomized trial,” “registry,” “cohort study.” We limited our search to studies in human subjects and English language in peer-reviewed journals published until November 2014. Additionally, a manual search of all relevant references from the screened articles and reviews of β blockers and heart failure (HF) was performed for additional clinical studies.
We included only prospective trials and cohort studies published as original articles in peer-reviewed scientific journals in English. We excluded those trials that did not include a detailed description of the cohort characteristics, concomitant therapy, and end-of-treatment outcomes. We did restrict eligibility according to left ventricular ejection fraction <40%.
The primary outcome measure was all-cause mortality. We also examined the difference in rehospitalization and the composite end point of all-cause mortality and rehospitalization if the later was studied. Only 4 studies included rehospitalization as part of their secondary end points. Two studies used a composite outcome of death and rehospitalization.
The data were independently extracted by 2 authors (MP and AB) using standardized protocol. We extracted study characteristics (type of design with duration of intervention and methods), baseline demographics, ejection fraction and functional status at baseline and at the end of the study, and number of clinical outcomes from each trial. Cochrane risk-of-bias tool has been used to assess the individual risk of bias of each prospective randomized study. The Newcastle–Ottawa tool was used for the quality assessment of cohort studies. An intention-to-treat traditional meta-analysis was performed in line with recommendations from the Cochrane Collaboration and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Statement. All analyses were performed by the Review Manager (RevMan) 5.2 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2011). Statistically significant heterogeneity was defined as a chi-square p value <0.05 or an I 2 statistic 50%.
Risk estimates (all represented as hazard ratios [HRs]) and associated 95% confidence interval (CI) were abstracted from models that were at least adjusted for age, gender, race, duration of treatment, co-morbidities, concomitant heart failure treatment, and other additional cardiovascular risk factors. For the cohort studies, we extracted the HR, associated 95% CI, participants, follow-up, outcomes, and events. Not all prospective controlled studies provided HRs or data to indirectly calculate HR. For these studies, in the absence of heterogeneity, pooled estimates of odd risks (ORs) with their 95% CIs were calculated using the Mantel–Haenszel method. Reported values are 2 tailed, and hypothesis-testing results were considered statistically significant at p <0.05. The small study effect, including publication bias, was tested using funnel plot, the Begg log-rank test, and the Egger test.
Results
We identified 4 prospective controlled studies of carvedilol treatment compared to metoprolol tartrate or both succinate and tartrate (tartrate prescribed in 12% of patients in this study). We also selected 6 observational cohort studies, 3 of which compared carvedilol to metoprolol succinate, 1 study used both types, and 2 studies compared carvedilol to the tartrate formulation. The baseline characteristics of the included studies are summarized on Table 1 . They were all reported from 2000 to 2014. These studies enrolled 30,943 patients on carvedilol and 69,925 patients who received metoprolol with an average follow-up duration of 36.4 months. The mean doses of metoprolol and carvedilol were not provided in some of the cohort studies. Therefore, we conducted sensitivity analysis excluding studies that used predominantly low doses defined as carvedilol ≤12.5 mg/day and metoprolol ≤100 mg/day. In the first group of studies (3 prospective and the post hoc analysis of Multicenter Automatic Defibrillator Implantation With Cardiac Resynchronization Therapy [MADIT-CRT]), the average dose of carvedilol was 39.7 mg/day and metoprolol types 93.8 mg/day.
| Study | Design | Sample Size | Mean age (years) | Average dose | Inclusion criteria | Primary end points | Secondary end points | Metoprolol type | Follow-up (months) | Risk of bias | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Carvedilol | Metoprolol | Carvedilol (mg/day) | Metoprolol (mg/day) | |||||||||
| Metra et al (2000) | RCT, double-blinded | 75 | 75 | 56.5 | 49 | 124 | CHF≥ 6 months, NYHA class II, III, or IV, LVEF ≤ 35%, peak VO2 ≤ 25 mL/kg/min, on diuretics, ACEi. | LVEF | Mortality, exercise tolerance, QOL, NYHA class, | T | 23 | High |
| Piccirillo et al (2002) | RCT, single-blind | 42 | 40 | 60 | 50 | 100 | CHF≥ 6 months, NYHA class II, III, or IV, LVEF ≤ 35%, peak VO2 ≤ 25 mL/kg/min, on diuretics, ACEi. | QT variability index | T | 12 | High | |
| COMET (2003) | RCT, double-blinded | 1511 | 1518 | 62 | 41.8 | 85 | CHF≥ 6 months, NYHA class II, III, or IV, LVEF ≤ 35%, peak VO2 ≤ 25 mL/kg/min, on diuretics, ACEi. | All-cause and cardiovascular mortality, re-hospitalization, stroke | Medication side-effects | T | 58 | Low |
| MADIT-CRT (2013) | Post-hoc analysis of RCT | 1077 | 438 | 64 | 18 | 66 | CHF, NYHA class I or II, LVEF ≤ 30%, QRS duration ≥130 ms and GDMT. | All cause mortality or nonfatal CHF events. | VT/VF | Mainly S (88%) | 41 | High |
| Pasternak et al (2014) | NROS | 6026 | 5638 | 69.3 | 50 (max dose reached by 52% of patients) | 200 (max dose reached by 12% of patients) | LVEF ≤ 40% and GDMT | All-cause mortality. | Cardiovascular mortality | S | 29 | Low |
| Boiling et al (2014) | NROS | 12 363 | 36314 | 73 | ∗ 10.2, 26.4 and 55 | ∗ 64.4, 144.3, 200 | Age ≥35 years, with a primary discharge diagnosis of heart failure. | All-cause mortality and hospitalization. | – | S&T | 49 | Low |
| Delea et al (2004) | NROS | 887 | 887 | 71.4 | 14 | 44 | CHF on carvedilol or metoprolol tartrate. | All-cause mortality, hospitalization, and death or hospitalization. | – | T | 11 | Low |
| Lazarus et al (2011) | NROS | 2140 | 14372 | 75 | 50 | 200 | Hospitalized adult with a primary discharge diagnosis of heart failure, discharged on b-blockers | All-cause mortality | Mortality or HF readmission | S | 21.6 | Low |
| Rector et al (2008) | NROS | 17429 | 8683 | 74.5 | 12.5 | 50 | Heart failure patients receiving carvedilol or metoprolol tartrate. | All-cause mortality and all-cause hospitalization. | S | 60 | Low | |
| Shore et al (2012) | NROS | 1756 | 1960 | 61 | Doses >12.5 and ≤12.5 used | Doses >100 and ≤100 used | LVEF <40 % on b-blockers and GDMT for ≥4 weeks | All-cause mortality. | HF Readmissions, LVEF | S | 60 | Low |
∗ Patients in the study were divided into 3 groups according to the doses of medications (low, intermediate and high).
On the basis of quality assessment, 3 studies were deemed to be at high risk of bias and the remaining studies to be at low risk ( Table 1 ). A significant reduction in all-cause mortality with carvedilol compared to metoprolol (mainly tartrate) was noted in the prospective controlled studies (odds ratio [OR] 0.80, 95% confidence interval [CI] 0.70 to 0.91; p <0.0001; Figure 1 ) without significant heterogeneity between trials ( I 2 0%, p = 0.95). The results were mainly driven by the COMET study, which was the largest in that group. After exclusion of MADIT-CRT, which is a post hoc analysis studying both metoprolol types, there was a significant difference in all-cause mortality between carvedilol and metoprolol tartrate (OR 0.79, 95% CI 0.68 to 0.91; p = 0.001, without heterogeneity between studies). In the subgroup of cohort studies, no difference between carvedilol and metoprolol types was found (HR 1.02, 95% CI 0.90 to 1.16; p = 0.75; Figure 1 ) with significant heterogeneity between trials ( I 2 92%, p <0.001). After excluding studies that used metoprolol tartrate, all-cause mortality was not different between carvedilol and metoprolol succinate (HR 1.12, 95% CI 0.91 to 1.39; p = 0.29) with significant heterogeneity between trials ( I 2 95%, p <0.001). After exclusion of cohort studies that used low-dose metoprolol succinate and carvedilol, no difference in all-cause mortality was noted.
