Eplerenone Use for Mild Symptomatic Heart Failure?

Abstract & Commentary

By Michael H. Crawford, MD

Sources: Zannad F, et al. Eplerenone in patients with systolic heart failure and mild symptoms. N Eng J Med. 2011;364:11- 21; Armstrong PW. Aldosterone antagonists — last man standing? N Eng J Med. 2011;364:79-80.

Mineralocorticoid receptor blockers have demonstrated improved survival in class III-IV heart-failure patients (spironolactone) and post-acute myocardial infarction patients with systolic dysfunction heart failure (eplerenone). This study was a placebo-controlled, randomized trial of eplerenone for patients with class II symptoms due to systolic heart failure. Eligibility criteria included age > 54 years, left ventricular ejection fraction (EF) < 30%, or 30%-35% with an ECG QRS duration > 130 m/sec, and adequate doses of standard heart-failure therapy (ACEI/ARB and beta blocker). Patients could be enrolled if they were six months from a hospital admission for cardiovascular reasons or had a BNP > 250 pg/mL. The eplerenone dose was titrated from 25 mg daily to 50 mg daily, if tolerated. Patients with a glomerular filtration rate (GFR) between 30-49 mL/min received 25 mg every other day to start. Dosage was adjusted to keep serum potassium < 5.0 mmol/L. The primary outcome endpoint was cardiac death or hospitalization for heart failure.

Results: From 278 centers in 29 countries, 2,737 patients were recruited. Baseline characteristics in the two randomized groups were well balanced. The trial was stopped early because of a statistically significant difference at 21 months of follow-up. The primary outcome endpoint was reached in 18% of the eplerenone patients and 26% of the placebo patients (HR 0.63, 95% CI 0.54-0.74, p < 0.001). Similar results were seen in most prespecified subgroups, including ischemic etiology, EF > 30%, and diabetics. Secondary analyses were also significantly impacted by eplerenone, such as death from any cause (HR 0.76), hospitalization (HR 0.77), and hospitalization for heart failure (HR 0.58). The number needed to treat for the primary endpoint was 19, and for death, per year of follow-up, was 51. Discontinuation of study drug for adverse events was 14% with eplerenone and 16% with placebo (p = NS). Renal failure, hypotension, and gynecomastia were not statistically more frequent with eplerenone, but increases in serum creatinine and potassium were. The authors concluded that eplerenone therapy in patients with systolic heart failure, class II symptoms, and adequate doses of standard therapy reduced the risk of death and hospitalization.


There is a solid mechanistic basis for this trial. Aldosterone antagonists have at least three beneficial properties. First, they are diuretics that promote volume homeostasis, which will keep heart-failure patients out of the hospital. Second, they increase potassium and magnesium levels, which would have antiarrhythmic properties in heart-failure patients and reduce arrhythmic deaths. Third, animal studies have shown that aldosterone antagonists reduce myocardial fibrosis, which could reduce ventricular remodeling and improve left ventricular function. Also, two previous large randomized trials in systolic heart failure (RALES, EPHESUS) showed reduced mortality with aldosterone-blocking drugs. Thus, it is not surprising that some benefit would be shown in a less symptomatic population. What is surprising is the magnitude of the effect; a 37% reduction in death and hospitalization.

Don't expect these results at home! There are several unique features of this study that probably augmented the observed benefit. This was a sicker, higher-risk population than the class II symptoms would suggest. All the patients were > 55 years old. Most had previous admissions for heart failure and prior myocardial infarction. Many had diabetes, atrial fibrillation, and hypertension. Most had an EF < 30%; the mean EF was 26%, which was the same as in the RALES trial. This fact emphasizes the well-known disconnection between symptoms and EF. Thus, in many ways, this population was on the sicker end of class II patients.

One-quarter of the patients had a QRS duration of > 130 m/sec, and one-quarter had left bundle branch block (LBBB). The mean QRS for the group was 122 m/sec, yet few of the patients had ICDs or resynchronization devices implanted. In the United States, these therapies would be expected to be deployed in higher numbers, which would likely reduce the impact of aldosterone antagonist therapy on death and hospitalizations. Another problem is that the study was stopped early for prespecified criteria. This tends to inflate the benefits of whatever is being studied, and many experts think that this should rarely be done.

How should we apply this study to our practice? We could just add an aldosterone blocker to all symptomatic patients with systolic heart failure. This would significantly increase the complexity of patient management for many, since serum potassium has to be monitored frequently, as well as blood pressure and potential adverse events. A more selective approach would likely be more cost-effective. Obviously, patients with potassium levels > 5.0 mmol/L and an estimated GFR < 30 mL/min should be excluded. If we follow the enrollment criteria for the trial, then only those > 55 years old, with EFs < 30%-35%, should be considered. Also, patients with LBBB or a QRS duration > 130 m/sec would be favored, especially if they were not candidates for resynchronization therapy. The final issue is whether we have to use eplerenone, a branded product, or will spironolactone do? Common sense, and the previous studies would suggest, this is a class effect, about which we don't really know for sure. In addition, in men, painful gynecomastia can occur in about 10%. With eplerenone, in this trial, it was < 1% for both sexes, and rarely caused withdrawal of the study drug. Thus, if the patient could afford it, especially in men, I would use eplerenone.