| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Hypertension. 2008;52:271.)
© 2008 American Heart Association, Inc.
Original Articles |
From the University of Michigan Health System (B.P.), Ann Arbor; School of Medicine (A.A.), University of Alabama at Birmingham, and Veterans Affairs Medical Center, Birmingham, Ala; School of Medicine, Case Western Reserve University (T.E.L.), Cleveland, Ohio; Department of Epidemiology and Preventive Medicine (H.K.), Monash University, Prahan, Australia; Heart Institute (InCOR) (J.N.), University of São Paulo Medical School, São Paulo, Brazil; Departamento de Engenharia Electrotécnica e de Computadores (J.S.C.), Faculdade de Engenharia, Universidade do Porto, Porto, Portugal; Emergency Cardiology Department (A.P.), National Institute of Cardiology, Kiev, Ukraine; Interni Klinika (M.A.), Cardiovascular Center, Prague, Czech Republic; Departamento de Enfermedades Cardiovasculares (R.C.), Hospital Clínico y Facultad de Medicina, Pontifica Universidad Catolica de Chile, Santiago, Chile; Pfizer Inc (H. Solomon, H. Shi), New York, NY; and the Clinical Investigation Center (F.Z.), INSERM-CHU de Nancy Hopital Jeanne d'Arc, Dommartin-les Toul, France.
Correspondence to Bertram Pitt, University of Michigan, 1500 E Medical Center Dr, 3910 Taubman Center, Ann Arbor, MI 48109-0366. E-mail bpitt{at}med.umich.edu
| Abstract |
|---|
|
|
|---|
Key Words: eplerenone hypertension myocardial infarction heart failure morbidity mortality
| Introduction |
|---|
|
|
|---|
| Methods |
|---|
|
|
|---|
40%) left ventricular ejection fraction (LVEF) and symptomatic heart failure (HF) were randomly assigned within 3 to 14 days of their AMI to receive eplerenone 25 mg/d titrated to 50 mg/d (n=3319) or matching placebo (n=3313). Patients were receiving standard medical therapy, including an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker (87%) and a β-blocker. Patients were followed for up to 2.5 years, with a mean follow-up of 16 months. Patients in the placebo and eplerenone groups were receiving a mean dose of 43.5 mg and 42.6 mg per day, respectively. Exclusion criteria included the use of potassium-sparing diuretics, a serum creatinine concentration >2.5 mg/dL (220 µmol/L), and a serum potassium concentration >5.0 mEq/L (mmol/L). Of the EPHESUS participants, 4007 patients had a history of hypertension (Hx-HTN), and 2625 had no Hx-HTN at the time of enrollment. Of the 4007 patients with Hx-HTN, 2024 (50.5%) were in the placebo group, and 1983 (49.5%) were in the eplerenone group. Of the 2625 patients without Hx-HTN, 1289 (49.1%) were in the placebo group, and 1336 (50.9%) were in the eplerenone group. The 2 primary end points of the EPHESUS, all-cause mortality and the combined end point of CV hospitalization or CV mortality, were also the primary end points for this analysis. Major secondary end points from EPHESUS, such as CV mortality, which included mortality because of AMI, HF, stroke, and sudden cardiac death (SCD), as well as hospitalization because of AMI and HF, were also studied. The cause of death or the primary diagnosis leading to hospitalization was adjudicated by a blinded EPHESUS critical events committee.
Statistical Analysis
Because the balance achieved by randomization in the main trial may have been lost in the groups with and without Hx-HTN, propensity scores for the receipt of eplerenone were used to assemble a balanced cohort. The propensity score for the receipt of eplerenone for a patient is defined as the conditional probability of receiving eplerenone given that patients measured covariates.2–6 Propensity scores were calculated separately for each of the 4007 and 2625 patients with and without Hx-HTN, respectively, using a nonparsimonious multivariable logistic regression model, incorporating the 36 baseline covariates (Table 1). Patients receiving eplerenone and placebo were matched based on their propensity to receive eplerenone. In all, 1838 pairs of patients with Hx-HTN and 1176 pairs of patients without Hx-HTN were matched. Absolute standardized differences were estimated to assess residual balance after matching.6,7
|
We used Kaplan–Meier plots and matched Cox regression analysis to estimate the effect of eplerenone in patients with and without Hx-HTN. We used multivariable Cox regression analyses in the prematch cohort, separately adjusting for the raw propensity scores. All of the analyses were based on intent to treat. All of the statistical tests were evaluated using 2-tailed 95% CIs.
Because the sample size of matched patients with Hx-HTN (n=3676) was larger than matched patients without Hx-HTN (n=2352), we conducted a sensitivity analysis by repeating our analysis in a smaller group of patients with Hx-HTN. In addition, we conducted a sensitivity analysis to determine the potential effects of an unmeasured covariate that may potentially invalidate our main conclusions.6,8,9
| Results |
|---|
|
|
|---|
Eplerenone and All-Cause Mortality
During a median follow-up of 16 months, 596 matched patients with Hx-HTN (16.2%) and 336 matched patients without Hx-HTN (14.3%) died of all causes. Among patients with Hx-HTN, all-cause mortality occurred in 18% of the placebo group and 14% of the eplerenone group (hazard ratio [HR]: 0.71; 95% CI: 0.59 to 0.85; P<0.0001; Figure 1A and Table 2). Among patients without Hx-HTN, all-cause mortality occurred in 14.4% of the placebo group and 14.2% of the eplerenone group (HR: 0.91; 95% CI: 0.72 to 1.15; P=0.435; Figure 1B and Table 2).
|
|
Eplerenone and CV Hospitalization or CV Mortality
Among patients with Hx-HTN, the coprimary combined end points of CV hospitalization or CV mortality occurred in 32.8% patients in the placebo group and 28.2% of patients in the eplerenone group (HR: 0.82; 95% CI: 0.72 to 0.94; P=0.003; Figure 2A and Table 2). Among patients without Hx-HTN, the coprimary combined end points of CV hospitalization or CV mortality occurred in 25.3% of patients in the placebo group and 23.6% of patients in the eplerenone group (HR: 0.91; 95% CI: 0.76 to 1.10; P=0.331; Figure 2B and Table 2).
|
Eplerenone, SCD, and Other Secondary Study End Points
SCD occurred in 6.5% of patients with Hx-HTN in the placebo group and 5.0% of patients with Hx-HTN in the eplerenone group (Figure 3A and Table 2). Among patients without Hx-HTN, SCD occurred in 5.2% patients in the placebo group and 4.4% patients in the eplerenone group (Figure 3B and Table 2). Effects of eplerenone on other secondary end points in patients with and without Hx-HTN are displayed in Table 2.
|
Eplerenone and Adverse Events
In patients with Hx-HTN, the incidence of severe hyperkalemia was significantly higher in patients receiving eplerenone (5.9% versus 4.2%; odds ratio for patients receiving eplerenone: 1.43; 95% CI: 1.06 to 1.93; P=0.019; Table 3). The incidence of severe hypokalemia was lower in patients receiving eplerenone (14.7% versus 9.6%; odds ratio for patients receiving eplerenone: 0.62; 95% CI: 0.51 to 0.76; P<0.0001). Other adverse events are displayed in Table 3.
|
Eplerenone, History of Hypertension, and Baseline Blood Pressure
Eplerenone had a favorable effect on outcomes in patients with Hx-HTN regardless of their baseline systolic blood pressure (SBP). In the subgroups of patients with Hx-HTN and a baseline SBP of
120 mm Hg, all-cause mortality occurred in 191 placebo patients (rate: 1465/10 000 person-years) and 157 eplerenone patients (rate: 1176/10,000 person-years; HR: 0.80; 95% CI: 0.65 to 0.99; P=0.043). Eplerenone also reduced CV hospitalization or CV mortality (HR: 0.86; 95% CI: 0.73 to 0.997; P=0.046) in these patients. In Hx-HTN patients with a baseline SBP of >120 mm Hg, similarly, eplerenone reduced both all-cause mortality (HR: 0.68; 95% CI: 0.52 to 0.87; P=0.002) and the coprimary composite end point (HR: 0.79; 95% CI: 0.66 to 0.94; P=0.010).
Eplerenone and Changes in Baseline Blood Pressure
Increases in SBP from baseline (Table 1) were similar in both groups of patients with or without Hx-HTN and were slightly greater in the placebo group. The SBP increased by 6 and 3 mm Hg in the placebo and eplerenone patients, respectively, in both groups with and without Hx-HTN.
Results of Sensitivity Analyses
Because the group with Hx-HTN (n=3676) was larger than the group without Hx-HTN (n=2352), we assembled a cohort of patients with Hx-HTN that was similar in size to that of those without Hx-HTN (please see the data supplement, available online at http://hyper.ahajournals.org). In this smaller group of patients with Hx-HTN (n=2334), we observed that eplerenone use was associated with reduction in all-cause mortality (HR: 0.76; 95% CI: 0.61 to 0.95; P=0.014).
Within the Hx-HTN–matched cohort, sign-score tests for matched survival data with censoring provide evidence that the use of eplerenone, compared with placebo, was associated with decreased all-cause mortality (z=3.73; 2-tailed, P=0.0002) and decreased rates of CV hospitalization or CV mortality (z=2.92; P=0.0035). Sensitivity analyses suggest that an unmeasured binary covariate would need to increase the odds of eplerenone use by >17.6% to explain the all-cause mortality association and by >6.6% to explain the association with the coprimary end points.
| Discussion |
|---|
|
|
|---|
There are several plausible explanations for the differential benefit of eplerenone in patients with Hx-HTN. Effects of a treatment on outcomes are known to vary between groups of patients based on differences in pathophysiology, natural history, severity of disease, comorbidity, and absolute risks between groups.10 EPHESUS patients with Hx-HTN were older and had more severe disease and comorbidity burden than those without Hx-HTN. They had 14% greater risk for all-cause mortality and 28% increased risk for CV mortality or CV hospitalizations than those without Hx-HTN. Pathophysiologically, compared with patients without Hx-HTN, those with Hx-HTN are at an increased risk of incident AMI, subsequent ventricular remodeling, and CV mortality.11–14 Hypertension predisposes patients to an increase in reactive oxygen species, vascular remodeling, and myocardial collagen formation.15–18 After AMI, patients with Hx-HTN have been found to have a greater degree of ventricular remodeling than those without Hx-HTN.13,19,20 Ventricular remodeling results in myocardial stretch, which is an important stimulus for activation of various neurohormones, including angiotensin II and aldosterone.21,22 HF is associated with an upregulation of mineralocorticoid receptors and aldosterone with an increase in myocardial calcium channel expression.23–26 Recently, AMI has been shown to cause electric remodeling before mechanical remodeling and LV hypertrophy.27 The increase in intracellular calcium associated with electric remodeling has been suggested to increase the risk of ventricular arrhythmias and SCD.28,29 Alterations in intracellular calcium and potassium may be greater in patients with AMI and Hx-HTN, many of whom have LV hypertrophy, and an increase in mineralocorticoid receptors and calcium channel expression.25,26 Therefore, the effectiveness of eplerenone in patients with Hx-HTN, most of whom were treated with an angiotensin-converting enzyme inhibitor or an angiotensin receptor blocker and a β-adrenergic receptor blocker, in reducing total mortality and SCD along with a trend toward a reduction in death because of progressive HF can be explained by the effects of aldosterone blockade in preventing electric remodeling, as well as by improving sympathetic/parasympathetic balance related to a decrease in reactive oxygen species production and an improvement in NO availability.30–33 These effects are in addition to the effects of eplerenone on ventricular mechanical remodeling, LV hypertrophy, and collagen formation, all of which are likely of greater magnitude in those with Hx-HTN.34–38
Results of our subgroup analysis suggest that the effects of eplerenone were observed regardless of baseline SBP, suggesting a long-term effect of HTN on target organs rather than baseline blood pressure levels as the underlying mechanistic explanation for the differential effect of eplerenone in the group with Hx-HTN. Another plausible explanation of a differential benefit of eplerenone in patients with Hx-HTN is the larger sample of these patients; however, sensitivity analysis suggests that eplerenone was beneficial in a smaller subset of patients with Hx-HTN. In addition, statistical analysis showed no significant difference in the treatment effect of eplerenone on SBP between patients with and without Hx-HTN, suggesting that our findings were not driven by higher absolute reductions in SBP among patients with Hx-HTN.
This analysis suggests that patients with AMI complicated by a low LVEF and symptomatic HF should be risk-stratified based on Hx-HTN. Those with Hx-HTN should be prescribed eplerenone to improve outcomes. Although we did not observe a significant effect of eplerenone on mortality in patients without Hx-HTN, we did observe a significant reduction in HF hospitalization, likely mediated by a reduction in ventricular remodeling. There was a trend toward a reduction in mortality, but analysis of this subgroup may have been underpowered to detect a significant difference in mortality and other outcomes, which is not surprising given the low baseline risk in these patients.11 Reduction in HF hospitalization would suggest a long-term effect of eplerenone on mortality in these patients.
Eplerenone was well tolerated in patients with and without Hx-HTN. Although more patients receiving eplerenone experienced hyperkalemia (>6 mEq/L), overall absolute rates were low, and no deaths were attributed to hyperkalemia in patients receiving eplerenone. Patients receiving eplerenone had a lower risk of developing hypokalemia (<3.5 mEq/L). This is important, because the overall absolute rate of hypokalemia was higher than that of hyperkalemia, which has been associated with increased mortality.39–42
Perspectives
This analysis suggests that patients with AMI, reduced LVEF, and symptomatic HF should be risk stratified based on their Hx-HTN; and those with Hx-HTN should be treated early post-AMI with eplerenone to prevent death, especially SCD. Because HF hospitalization is an important predictor of CV death and eplerenone reduced HF hospitalization in patients without Hx-HTN, eplerenone should also be initiated in these patients.
| Acknowledgments |
|---|
Sources of Funding
This study was funded by Pfizer Inc. Editorial support was provided by PAREXEL and was funded by Pfizer Inc. A.A. is supported by the National Institutes of Health through grants from the National Heart, Lung, and Blood Institute (5-R01-HL085561-02 and P50-HL077100) and a generous gift from Jean B. Morris of Birmingham, Alabama.
Disclosures
J.S.C. has served on an advisory board for Pfizer Inc; H. Solomon is an employee of Pfizer Inc; H. Shi is an employee of Pfizer Inc; and B.P. and F.Z. have received honoraria from and served on advisory boards for Pfizer Inc. The remaining authors report no conflicts.
| Footnotes |
|---|
Received January 6, 2008; first decision January 17, 2008; accepted May 19, 2008.
| References |
|---|
|
|
|---|
Related Article:
This article has been cited by other articles:
![]() |
B. Pitt The measurement of plasma aldosterone in patients post-myocardial infarction Eur. Heart J., October 2, 2008; 29(20): 2451 - 2452. [Full Text] [PDF] |
||||
![]() |
L. M. Ruilope Aldosterone, Hypertension, and Cardiovascular Disease: An Endless Story Hypertension, August 1, 2008; 52(2): 207 - 208. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Hypertension Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2008 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |