Efficacy and Safety of Carvedilol in Treatment of Heart Failure with Chronic Kidney DiseaseClinical Perspective
A Meta-Analysis of Randomized Trials
Background— The safety and efficacy of different types of β-blocker therapy in patients with non–dialysis-dependent chronic kidney disease (CKD) and systolic heart failure (HF) are not well described. We assessed whether treatment of systolic HF with carvedilol is efficacious and safe in adults with CKD.
Methods and Results— We performed a post hoc analysis of pooled individual patient data (n=4217) from 2 multinational, double-blinded, placebo-controlled, randomized trials, CAPRICORN (Carvedilol Postinfarct Survival Control in Left Ventricular Dysfunction Study) and COPERNICUS (Carvedilol Prospective Randomized, Cumulative Survival study). Primary outcome was all-cause mortality. Secondary outcomes included cardiovascular mortality, HF mortality, first HF hospitalization, the composite of cardiovascular mortality or first HF hospitalization, and sudden cardiac death. Non–dialysis-dependent CKD was defined by estimated glomerular filtration rate ≤60 mL/min/1.73 m2, using the abbreviated Modification of Diet in Renal Disease equation. CKD was present in 2566 of 4217 (60.8%) of the cohort, 50.4% of whom were randomly assigned to carvedilol therapy. Within the CKD group, treatment with carvedilol decreased the risks of all-cause mortality (hazard ratio [HR], 0.76; 95% confidence interval [CI], 0.63 to 0.93; P=0.007), cardiovascular mortality (HR, 0.76; 95% CI, 0.62 to 0.94; P=0.011), HF mortality (HR, 0.68; 95% CI, 0.52 to 0.88; P=0.003), first hospitalization for HF (HR, 0.74; 95% CI, 0.61 to 0.88; P=0.0009), and the composite of cardiovascular mortality or HF hospitalization (HR, 0.75; 95% CI, 0.65 to 0.87; P<0.001) but was without significant effect on sudden cardiac death (HR, 0.76; 95% CI, 0.56 to 1.05; P=0.098). There was no significant interaction between treatment arm and study type. Carvedilol was generally well tolerated by both groups of patients, with an increased relative incidence in transient increase in serum creatinine without need for dialysis and other electrolyte changes in the CKD patients. However, in a sensitivity analysis among HF subjects with estimated glomerular filtration rate <45 mL/min/1.73 m2 (CKD stage 3b), the efficacy of carvedilol was not significantly different from placebo.
Conclusions— This analysis suggests that the benefits of carvedilol therapy in patients with systolic left ventricular dysfunction with or without symptoms of HF are consistent even in the presence of mild to moderate CKD. Whether carvedilol therapy is similarly efficacious in HF patients with more advanced kidney disease requires further study.
Chronic heart failure (HF) is associated with an increased risk of morbidity and mortality.1–3 Chronic kidney disease (CKD)4 defined by estimated glomerular filtration rate (eGFR) ≤60 mL/min/1.73 m2 is a powerful risk factor for cardiovascular adverse outcomes in patients with HF.5–9 Both CKD and chronic HF are common, with prevalence rates in the US population estimated to be 20 million10–12 and 6 million,1,2,13 respectively. Not unexpectedly, the prevalence of concomitant CKD in patients with HF is also high.9,14–16
Clinical Perspective on p 26
This coexistence of CKD and HF could result in significant adverse interactive effects.17 Indeed, the presence of CKD in HF patients is associated with an increased risk of hospitalization and death from pump failure and all-cause mortality14–16,18 independent of the degree of impairment in left ventricular ejection fraction (LVEF)19 or the severity of heart failure based on New York Heart Association (NYHA) class.20 Conversely, HF and its treatment can also increase the rate of progression in CKD.5,14,17
The excess mortality in HF patients in the presence of CKD is likely to be multifactorial17 and may include low utilization rates of currently available therapeutic options either due to lack of proven efficacy,21,22 lack of randomized, controlled studies,11,23 or concerns over the increased rates of adverse events, as has been demonstrated with the use of other agents such as angiotensin-converting enzyme inhibitors (ACE-I),24 spironolactone,25 and nesiritide.26 Consequently, patients with the combination of HF and CKD are less likely to be treated with established HF therapies.21,27
β-Blockers are a cornerstone of therapy for HF, but their use in patients with HF and different degrees of CKD have not been thoroughly evaluated in randomized studies.21 Carvedilol is a nonselective vasodilating β-adrenergic blocker with α1 adrenergic-blocking activities28,29 with established efficacy in decreasing mortality and morbidity in patients with mild to severe chronic HF with systolic left ventricular dysfunction (LVD).30,31 Carvedilol decreases systolic and diastolic blood pressure without a decrease in renal blood flow or GFR while reducing renal vascular resistance.32 These hemodynamic effects of carvedilol could be particularly beneficial in patients with CKD.33 Although a small, randomized study demonstrated that carvedilol improved survival in chronic dialysis patients with severe HF34 and other studies have failed to demonstrate a benefit of other β-blockers in the dialysis population,35 there remains a paucity of data from randomized studies evaluating the efficacy and safety of carvedilol therapy in the presence of CKD. We therefore conducted a post hoc meta-analysis of the CKD subgroup among 4217 patients enrolled in 2 placebo-controlled randomized studies of carvedilol therapy in patients with different types of systolic LVD with or without symptomatic HF.30,31
We analyzed the pooled patient-level data from 2 large, multicenter, multinational, double-blind, randomized, placebo-controlled studies of carvedilol in systolic LVD with or without symptoms of HF.30,31 These studies were similar with respect to the study design, dose titration of carvedilol, primary and secondary outcomes, and duration of the follow-up. Briefly, CAPRICORN (Carvedilol Post-Infarct Survival Control in Left Ventricular Dysfunction Study) randomly assigned 1959 patients within 21 days after acute myocardial infarction with LVEF ≤0.40 with or without symptomatic HF to either carvedilol (n=975) or placebo (n=984).30 In the COPERNICUS (Carvedilol Prospective Randomized, Cumulative Survival) study, a total of 2289 patients with LVEF ≤0.25 and severe chronic HF of ischemic or nonischemic etiology while on standard therapy were randomly assigned to carvedilol (n=1156) and to placebo (n=1143).31 For CAPRICORN, carvedilol was initiated at 6.25 mg twice daily within 21 days of the acute myocardial infarction and progressed to 12.5 mg twice per day within 3 to 10 days, with further escalation to the target dose of 25 mg twice per day (maximum dose) within another 5 to 10 days as tolerated. For COPERNICUS, carvedilol was initiated at 3.125 mg twice per day and titrated at intervals of no less than 2 weeks to 12.5 mg twice daily with further escalation to the target dose of 25 mg twice daily (maximum dose) as tolerated. The study participants were already receiving conventional therapy for HF, including different doses of ACEI-I, angiotensin-receptor blockers, diuretics, spironolactone, and digitalis, as deemed necessary by the treating physician.
Definition of CKD
We used the 4 variable Modification of Diet in Renal Disease formula to estimate the eGFR expressed in mL/min/1.73 m2,4 based on the serum creatinine values at the time of enrollment. This formula has been previously used in several HF studies.5,7,18,27,36,37 CKD and non-CKD groups were defined on the basis of eGFR values of ≤60 mL/min/1.73 m2, respectively. As a sensitivity analysis, we assessed the efficacy of carvedilol in those with eGFR <45 mL/min/1.73 m2(CKD stage 3b) and eGFR ≥45 to 60 mL/min/1.73 m2(CKD stage 3a).
Primary and Secondary Outcomes
The primary outcome in this analysis was all-cause mortality. Secondary outcomes were cardiovascular mortality, HF mortality, first hospitalization for HF, composite of cardiovascular mortality or hospitalization for HF, and sudden cardiac death. The outcomes in the present analysis were identical to those specified in the original protocols.30,31 Clinical outcomes were adjudicated by the end point committees for the respective studies.
The data from the individual patients who participated in the CAPRICORN and COPERNICUS trials were pooled for the present analysis. Baseline characteristics were summarized descriptively using summary statistics according to the variable distribution (ie, mean and standard deviation for continuous variables; frequencies for categorical variables).
Analysis of the time to each end point was performed using Cox proportional hazards regression. For all the end points considered, the null hypothesis assumed that carvedilol therapy was similar to placebo in HF patients with concomitant CKD (eGFR ≤60 mL/min/1.73 m2). The same hypothesis was tested for the non-CKD (>60 mL/min/1.73 m2) subgroup.
The following covariates were considered in the CKD subgroup analyses of these end points: baseline age (≥65 versus > 65 years), sex, race (Caucasian, African American, others), LVEF (<20% versus ≥20%), and NYHA classification. Other covariates included history (yes or no) of diabetes mellitus, hypertension, coronary artery bypass graft, percutaneous coronary angioplasty, and stroke; and use (yes or no) of ACE-I or angiotensin-receptor blockers, spironolactone, digitalis, and diuretics.
Initially for each end point, covariates were evaluated one by one, with treatment arm and study type in the model. Treatment by study and treatment by covariate interactions were explored sequentially in this model. Main effects were tested at the 0.05 level and interactions at a 0.10 level. If the covariate or interaction effect was not significant, it was dropped and not included in the main analysis with treatment and study effect in the CKD subgroups (parsimonious model).
Additionally, to verify the robustness of the treatment effect, for each end point, CKD subgroup analyses models were examined that included: age (≤65 versus >65 years), sex, race (Caucasian, African American, others), diabetes mellitus (yes or no), LVEF (<20% versus ≥20%), treatment arm, and study type. These covariates were selected because of prior evidence suggesting their association with outcomes in congestive HF as well as in CKD. When compared with a parsimonious model including only treatment arm and study type, the treatment effect observed in these subgroup analyses models was nearly identical. As such, primary inferences in the CKD subgroup analyses for these end points were based on the parsimonious models.
Finally, a sensitivity analysis was performed in the subgroups with eGFR <45 mL/min/1.73 m2 (CKD stage 3b) and eGFR ≥45 to 60 mL/min/1.73 m2 (CKD stage 3a). These sensitivity analyses models included treatment arm and study type.
All analyses were performed with SAS, version 8.2 (SAS institute, Inc, Cary, NC).
Baseline Characteristics of the Cohort
A total of 4217 subjects were included in this analysis. The mean (±SD) eGFR was 57.2 (±18.7) mL/min/1.73 m2. The eGFR values at baseline as well as at the last follow-up were similarly distributed in the carvedilol and placebo groups (Figure). CKD was present in 2566 of 4217 (60.8%) of all patients enrolled in these 2 trials. Among individuals with CKD, 1293 (50.4%) were randomly assigned to carvedilol therapy and 1273 (49.6%) to placebo. There were 1651(39.2%) non-CKD patients at baseline. Of these, 822 (49.8%) received carvedilol.
Baseline characteristics of the patients are summarized in Table 1 and in Supplemental Table 1. In both CKD strata, comorbidities as well as the use of medications such as ACE-I or angiotensin-receptor blockers and anticoagulation therapy were similarly distributed in patients randomly assigned to carvedilol or placebo. Conversely, patients with CKD appeared older and to have more severe HF as assessed by NYHA classification than non-CKD individuals. In addition, a higher number of CKD patients were treated with a diuretic (84% versus 54%), digitalis (48% versus 28%), or spironolactone (14% versus 7%).
Overall Clinical Outcomes
The median and mean (±SD) duration of actual follow-up were 13.5 months and 13.6 (±7.9) months, respectively. Discontinuation of carvedilol was similarly frequent in CKD and non-CKD subgroups (24% versus 22%). The patients with CKD had increased rates of primary as well as secondary outcomes on univariate analysis (Table 2) and were further confirmed on multivariate analysis (data not shown).
Outcomes of Carvedilol Therapy
Random assignment to carvedilol decreased the risks for the primary as well as all secondary outcomes in both groups of patients (Table 3 and Table 4). Among the CKD group, treatment with carvedilol was associated with decreased risks of all-cause mortality (hazard ratio [HR], 0.76; 95% confidence interval [CI], 0.63 to 0.93; P=0.007); cardiovascular mortality (HR, 0.77; 95% CI, 0.62 to 0.94; P=0.011); HF mortality (HR, 0.68; 95% CI, 0.52 to 0.88; P=0.003); first hospitalization for HF (HR, 0.74; 95% CI, 0.62 to 0.88; P=0.0009); and the composite of cardiovascular mortality or hospitalization for HF (HR, 0.75; 95% CI, 0.65 to 0.87; P<0.001). However, treatment with carvedilol did not have a statistically significant impact on sudden cardiac death in HF patients with CKD (HR, 0.76; 95% CI, 0.56 to 1.05; P=0.0980). The magnitude of the relative risk reduction in the incidence of the primary and secondary outcomes was similar in the CKD and non-CKD groups. For each clinical outcome, there were no statistically significant interactions observed between treatment (carvedilol, placebo) and study type (CAPRICORN, COPERNICUS).
Furthermore, the robustness of the treatment effect for each end point in the CKD subgroup remained consistent when other important covariates were added to the Cox models (Supplemental Table 2). The interaction between treatment and CKD group was nonsignificant for all outcomes.
A sensitivity analysis (Table 4) demonstrated that efficacy of carvedilol was not significantly different from placebo for the primary or secondary outcomes in the eGFR (<45 mL/min/1.73 m2, CKD stage 3b) subgroup. The overlapping confidence intervals for these outcomes in this subgroup of heart failure patients may be related to relatively small sample size (CKD 3b, carvedilol versus placebo: n=544 versus 572).
Investigator-reported adverse events were adjudicated by the end point committee and the rates of cardiac, neurological, vascular, and other miscellaneous adverse events with the use of carvedilol compared with placebo in patients with CKD as well as non-CKD are presented in Table 5. The cumulative incidence of transient changes in serum creatinine in the CKD (carvedilol versus placebo) patients was 4.6% versus 1.8% (P<0.001). However, none of these patients required dialysis therapy. The use of carvedilol was associated with an increase in the relative incidence of headache, orthostatic hypotension, hyperkalemia, and hyperglycemia in HF patients with CKD. The overall discontinuation rate of carvedilol was similar in the CKD and non-CKD subgroups (24% versus 22%).
CKD is an independent risk factor for cardiovascular morbidity and mortality in patients with HF.5–9 With the evolving epidemic of CKD and our aging population, the prevalence of patients with concomitant heart failure and CKD is increasing.13 Hence, it is important to understand whether current standard therapies have a similar degree of efficacy and safety profiles in HF patients with different degrees of CKD.11,23
This post hoc analysis of patient-level data from 2 large, randomized, placebo-controlled studies with HF demonstrated that more than 60% of HF patients had mild to moderate CKD at the time of enrollment. Overall, treatment with carvedilol resulted in similar degrees of reduction in the relative risks for all-cause, cardiovascular, and HF mortality in the group of HF patients with CKD. As one would expect, this reduction translates to a greater absolute benefit with carvedilol in CKD patients.
To our knowledge, this is the first report in which individual patient data on a large cohort has been used to analyze the efficacy and safety of carvedilol in HF patients with concomitant CKD. The presence of CKD can worsen the systemic effects of HF via adverse effects on cardiac, hemodynamic, and neurohormonal adaptive (or maladaptive) responses, including increased sympathetic activity, and these could portend a poor prognosis.6,38,39 Accordingly, different degrees of CKD are independently associated with an increased risk for all-cause mortality as well as HF progression.5–9 In addition, the synergism between HF and CKD could also potentiate other risk factors frequently associated with mild to moderate degrees of CKD such as anemia36,40 and chronic inflammation.41
In individuals without CKD, prolonged use of carvedilol has been shown to improve symptoms of HF, increase LVEF, mitigate neurohormonal activation and peripheral vasoconstriction, and decrease sympathetic overactivity as well as salt and water retention.6,38 These events may be particularly relevant in the setting of CKD, where carvedilol therapy could potentially improve renal hemodynamics and kidney function. Similarly, although the benefits of antioxidant properties of carvedilol are not fully understood,42 it is possible that these properties may have a special role in patients with a combination of HF and substantial CKD.
In this study population, the addition of carvedilol to conventional HF therapy was generally well tolerated. The rate of discontinuation of carvedilol based on the investigator-reported adverse events was similar in both groups of patients. Changes in renal function without need for dialysis therapy were more frequent in the HF patients with concomitant CKD on treatment with carvedilol despite careful titration of carvedilol dose. Similar changes in kidney function have been reported with other types of HF therapy.24 However, it is essential that HF patients with CKD when treated with carvedilol should be carefully monitored and dose optimization–tailored, based on individual patient responses in serum creatinine and/or eGFR values during the follow-up period. In addition, CKD patients are at increased risk for development of orthostatic hypotension, hyperkalemia, and hyperglycemia after carvedilol therapy; thus, careful monitoring for these events is clearly warranted.
These findings provide strong evidence to suggest that addition of carvedilol to the conventional HF therapy in CKD patients is beneficial and safe. Ghali et al43 and Erdmann et al44 demonstrated similar benefits with metoprolol and bisoprolol, respectively, in patients with HF and CKD. However, in a sensitivity analysis of HF patients with advanced CKD (stage 3b), the efficacy of carvedilol was not different from placebo. Whether this reflects a lack of sufficient power due to the relatively small sample size in the subgroup with advanced CKD or a different biological response to carvedilol in the setting of more advanced CKD cannot be determined in this post hoc analysis and merits additional studies.
This study has several limitations: First, almost all serum creatinine–based estimating equations for GFR are inherently imprecise, and this is specifically true when serum creatinine is measured in different laboratories or in those with lower than expected muscle mass. This is an important concern in patients with HF who may have lower muscle mass than those in the general population. Conversely, HF itself or therapy with ACE-I and diuretics may result in transient changes in GFR without concomitant changes in the underlying kidney function.16 These issues could potentially have resulted in overestimation or underestimation of eGFR with misclassification of CKD stages.45 Second, our analysis is a post hoc analysis of the original studies. We had the ability to analyze individual-level data from 2 randomized, double-blind, placebo-controlled studies, with similar homogeneity in the study design, dose titration of carvedilol, and rigorous uniform definitions as well as rigorous ascertainment of prespecified outcomes. Nevertheless, our results should still be considered hypothesis forming and must be extrapolated cautiously to those with advanced CKD or severe HF. Third, this meta-analysis involved a homogenous population of systolic HF with a relatively low number of African Americans. Our population may not accurately represent the HF patient population in the community, and our results should be generalized cautiously to those not well represented in our analysis.46
A number of strengths of our analysis are also worth noting. We analyzed a large number of patients with CKD and thus provide the most robust evidence to date to suggest that carvedilol therapy has beneficial effects in patients with mild to moderate CKD and HF. An additional randomized study with adequate sample size is needed to determine whether the benefits of carvedilol therapy extend to individuals with advanced CKD (eGFR <45 mL/min/1.73 m2). Additionally, we analyzed results of treatment with carvedilol in both ischemic and nonischemic HF patients; hence, these results are generalizable to CKD patients with LVD of either etiology.
The results of this meta-analysis support treatment with carvedilol to reduce rates of all-cause mortality as well as other HF-related events in patients with LVD with or without symptomatic HF in the presence of mild to moderate CKD. The efficacy of carvedilol therapy in HF patients with advanced CKD has not been established.
Drs Lukas and Iyengar are employed by GlaxoSmithKline. Dr Cooper is a recipient of a research grant from GlaxoSmithKline.
Guest Editor for this article was Emily B. Levitan, PhD.
The online-only Data Supplement is available at http://circheartfailure.ahajournals.org/cgi/content/full/CIRCHEARTFAILURE.109.932558/DC1.
- Received December 28, 2009.
- Accepted October 15, 2010.
- © 2011 American Heart Association, Inc.
- Hillege HL,
- Girbes AR,
- de Kam PJ,
- Boomsma F,
- De Zeeuw D,
- Charlesworth A
- Hillege HL,
- Nitsch D,
- Pfeffer MA,
- Swedberg K,
- McMurray JJ,
- Yusuf S
- Kottgen A,
- Russell SD,
- Loehr LR,
- Crainiceanu CM,
- Rosamond WD,
- Chang PP
- DuBose TD Jr.
- McAlister FA,
- Ezekowitz J,
- Tonelli M,
- Armstrong PW
- Solomon SD,
- Anavekar N,
- Skali H,
- McMurray JJ,
- Swedberg K,
- Yusuf S
- Abraham WT,
- Tsvetkova T,
- Lowes BD
- Shlipak MG,
- Smith GL,
- Rathore SS,
- Massie BM,
- Krumholz HM
- Madsen BK,
- Keller N,
- Christiansen E,
- Christensen NJ
- Bongartz LG,
- Cramer MJ,
- Doevendans PA,
- Joles JA,
- Braam B
- Shlipak MG,
- Fried LF,
- Crump C,
- Bleyer AJ,
- Manolio TA,
- Tracy RP
- Ghali JK,
- Wikstrand J,
- van Veldhuisen DJ,
- Fagerberg B,
- Goldstein S,
- Hjalmarson A
- Myers GL,
- Miller WG,
- Coresh J,
- Fleming J,
- Greenberg N,
- Greene T
Chronic heart failure is a clinical syndrome associated with increased rates of morbidity, frequent hospitalizations, and increased utilization of health care costs as well as all-cause mortality. Similarly, chronic kidney disease (CKD) increases the risk for adverse cardiovascular outcomes in the general population as well as in those with underlying heart failure. There is a paucity of evidence whether therapeutic interventions that are effective for the treatment of heart failure in the general population are also effective in those heart failure patients with concomitant CKD. Consequently, clinicians may be reluctant to use these evidence-based therapies in the presence of CKD. We performed a post hoc meta-analysis of individual patient-level data from 2 large, randomized, controlled trials (CAPRICORN and COPERNICUS) of carvedilol in patients with ischemic or nonischemic left ventricular dysfunction. The data were categorized for the presence or absence of CKD, based on the estimated glomerular filtration rate (<60 or ≥60 mL/min/1.73 m2, respectively), using the Modified Diet Renal Disease equation from the serum creatinine values obtained at the time of enrollment. Our study demonstrates that carvedilol therapy leads to similar benefits in the presence of CKD as in those heart failure patients without CKD. However, the effect of carvedilol therapy in heart failure patients with advanced CKD (estimated glomerular filtration rate <45 mL/min/1.73 m2) was not different from placebo. This hypothesis-generating finding that carvedilol may not be efficacious in very advanced stages of CKD must be confirmed by future studies. We also observed that the use of carvedilol therapy in the presence of CKD can lead to transient fluctuations in renal function and increases the risk for orthostatic hypotension and other electrolyte abnormalities. Hence, patients with heart failure with concomitant CKD should have careful dose titration as well as judicious monitoring of kidney function, blood pressure, and electrolytes when treated with carvedilol.