Uncertainty on the Use of Aldosterone Antagonists for Primary Therapy for Sudden Cardiac Death in the Setting of Implanted Devices

Abstract

HomeCirculationVol. 115, No. 23Uncertainty on the Use of Aldosterone Antagonists for Primary Therapy for Sudden Cardiac Death in the Setting of Implanted Devices Free AccessArticle CommentaryPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessArticle CommentaryPDF/EPUBUncertainty on the Use of Aldosterone Antagonists for Primary Therapy for Sudden Cardiac Death in the Setting of Implanted Devices Robert A. Kloner, MD, PhD and David S. Cannom, MD Robert A. KlonerRobert A. Kloner From the Heart Institute, Good Samaritan Hospital, and Division of Cardiovascular Medicine, Keck School of Medicine, University of Southern California (R.A.K.), and Cardiology Division, Good Samaritan Hospital, and Cardiology Division, University of California at Los Angeles School of Medicine (D.S.C.), Los Angeles. Search for more papers by this author and David S. CannomDavid S. Cannom From the Heart Institute, Good Samaritan Hospital, and Division of Cardiovascular Medicine, Keck School of Medicine, University of Southern California (R.A.K.), and Cardiology Division, Good Samaritan Hospital, and Cardiology Division, University of California at Los Angeles School of Medicine (D.S.C.), Los Angeles. Search for more papers by this author Originally published12 Jun 2007https://doi.org/10.1161/CIRCULATIONAHA.106.684522Circulation. 2007;115:2983–2989In the early development of therapy for acute myocardial infarction, it was thought that once the necrotic process had been completed (usually within 24 hours of coronary artery occlusion), additional therapies could not affect outcome. However, after completion of the necrotic process, the myocardial infarction may thin and stretch (involving lengthwise slippage of myocytes), a phenomenon referred to as myocardial infarct expansion. This process causes local left ventricular cavity dilatation followed by gradual global left ventricular dilatation and lengthwise (eccentric) hypertrophy of the noninfarcted tissue. Apoptosis (programmed cell death) and some attempt of the myocardium to regenerate, especially at the infarct border zone, may also contribute to this remodeling process of the ventricle. If the left ventricle remodels in such a way that it becomes very dilated, then the prognosis is poor, and heart failure is more likely to occur.1 These later processes of myocardial infarct expansion and left ventricular remodeling became the target of therapies such as angiotensin-converting enzyme (ACE) inhibition that could be initiated after 24 hours of coronary occlusion. ACE inhibition, angiotensin receptor blockade, and long-term β-blockade have become standard pharmacological approaches for postinfarction left ventricular dysfunction and heart failure.Response by Pitt and Pitt p 2989Ventricular arrhythmias can occur in both the acute and chronic phases of acute myocardial infarction and can lead to sudden cardiac death (SCD). Reentrant arrhythmias may arise at the border zone of infarcts, causing monomorphic ventricular tachycardia that may occur years after the index infarction. Recurrent myocardial ischemia resulting in an unstable substrate may contribute to polymorphic ventricular tachycardia or ventricular fibrillation. Agents such as β-blockers that are anti-ischemic may reduce sudden death by quieting this unstable substrate. In the Multicenter Automatic Defibrillator Implantation Trial (MADIT) II, implantable defibrillators were shown to reduce mortality in post–myocardial infarction patients with left ventricular dysfunction entirely due to a reduction in SCD.2 It is likely that these devices did not primarily improve the arrhythmic substrate. However, long-term cardiac resynchronization will encourage reverse remodeling that might reduce the substrate for arrhythmia.ACE Inhibitors and Angiotensin Receptor Blockers After Myocardial InfarctionIn the now classic study by Pfeffer et al,3 captopril administered long-term, starting within about the first few weeks of myocardial infarction, decreased total mortality, congestive heart failure, and recurrent myocardial infarction. Echocardiographic analysis demonstrated that captopril reduced diastolic dilatation at 2 years, suggesting a decrease in deleterious left ventricular remodeling.4 Other studies confirmed that long-term administration of an ACE inhibitor improved cardiac outcome after myocardial infarction.5–11 Benefits of long-term therapy with the angiotensin receptor blocker valsartan12 were also reported to benefit post–myocardial infarction patients with left ventricular dysfunction. Valsartan was shown to result in similar but not superior effects on survival compared with captopril in the Valsartan in Acute Myocardial Infarction (VALIANT) study.12 Furthermore, treatment with captopril plus valsartan resulted in no advantage over treatment with either agent alone. In a head-to-head comparison of the ACE inhibitor captopril with the angiotensin receptor blocking agent losartan in patients with acute myocardial infarction and evidence of heart failure or left ventricular dysfunction, captopril was associated with a nonsignificantly lower all-cause mortality and a significantly lower cardiovascular mortality compared with losartan, but losartan was better tolerated than captopril.13 Some of these studies demonstrated less ventricular arrhythmias when an ACE inhibitor was used.11β-Blockers After Myocardial InfarctionAnother class of drugs that has been shown to reduce mortality after acute myocardial infarction is the β-blockers. The landmark Beta-Blocker Heart Attack Trial (BHAT) tested the effect of long-term propranolol therapy in patients after a myocardial infarction. More than 3800 patients were randomized to either propranolol (180 to 240 mg/d maintenance dose) or placebo starting 5 to 21 days after myocardial infarction and were followed up for ≈2 years. Total mortality was 7.2% in the propranolol group and 9.8% in the placebo group (26% reduction). Sudden cardiac death occurred in 3.3% of the propranolol patients versus 4.6% of the placebo patients (28% reduction).14 A subset of 826 of these patients also had paired ambulatory ECG monitoring at baseline and after 6 weeks of therapy. An increase in ventricular arrhythmias over the 6-week period was blunted by propranolol.15 Some but not all studies in which other β-blockers were administered after myocardial infarction showed a reduction in SCD.16 Hjalmarson17 postulated that the more lipophilic β-blockers (timolol, metoprolol, propranolol) were more likely to demonstrate this benefit because they could penetrate the brain and maintain high vagal tone during stress, thus reducing ventricular fibrillation. Of course, it is also possible that the β-blockers are primarily reducing postinfarction ischemia, which would explain why they are antiarrhythmic.Although these findings represented a major advance, as pointed out by Fonarow et al,18 many of the earlier studies with β-blockers did not include patients with heart failure. The Carvedilol Post-Infarct Survival Control in LV Dysfunction (CAPRICORN) study determined the effects of carvedilol added to standard therapy (including ACE inhibitors) for patients with an acute myocardial infarct who had an ejection fraction <0.40.19 More than 1900 patients with a mean ejection fraction of 0.33 were randomized to placebo or carvedilol and followed up for 15 months. Total mortality was reduced from 15.3% in the placebo group to 11.9% with carvedilol. SCD was reduced as well (Table 1). In the CAPRICORN study, sudden death occurred in 5% of carvedilol patients versus 7% of placebo patients (P=0.098). TABLE 1. Recent Key Randomized Controlled Trials of Nonantiarrhythmic Drugs and Effect on SCDDrugStudy CharacteristicsNo.HR for SCD (95% CI)MI indicates myocardial infarction; HOPE, Heart Outcomes Prevention Evaluation; LVEF, left ventricular ejection fraction; and CHF, congestive heart failure.Reproduced, with permission of the publisher, from Richter et al.26 Copyright © 2005, The American Heart Association, Inc.β-BlockerCAPRICORN: acute MI (3–21 days), LVEF ≤0.4019590.24 (0.11–0.49)ACE inhibitorMeta-analysis of 15 randomized controlled trials (patients with CHF)15 1040.80 (0.70–0.92)ACE inhibitorHOPE population (patients without CHF or LV dysfunction)92970.79 (0.64–0.98)SpironolactoneRALES: CHF, LVEF ≤0.3516630.70 (0.54–0.95)EplerenoneEPHESUS: acute MI (3–14 days), LVEF ≤0.4066320.79 (0.64–0.97)Aldosterone Antagonists After Myocardial InfarctionWhat is known about the use of aldosterone blockers in patients with myocardial infarction and postmyocardial dysfunction? The crucial study is the Eplerenone Post–Acute Myocardial Infarction Heart Failure Efficacy and Survival (EPHESUS) study by Pitt et al.20 The EPHESUS trial was a double-blind, placebo-controlled study of the selective aldosterone blocker eplerenone, examining morbidity and mortality in post–myocardial infarction patients with heart failure and left ventricular dysfunction (left ventricular ejection fraction of ≤40%). Therapy was started 3 to 14 days after acute myocardial infarction, and patients received placebo (n=3319) or eplerenone (25 mg/d; n=3313) for 4 weeks, after which the dose was increased to 50 mg/d. If the potassium concentration was >5.5 mmol/L, the dose of study drug was reduced or treatment stopped temporarily. Patients were already on standard optimal medical therapy including ACE inhibitors, angiotensin receptor blockers, β-blockers, diuretics, and reperfusion therapy. Follow-up was for 16 months. Fewer patients died in the eplerenone group (478; 14.4%) than in the placebo group (554 patients; 16.7%; relative risk [RR]=0.85; 95% confidence interval [CI], 0.75 to 0.96; P=0.008). Death due to cardiovascular causes occurred in 407 patients in the eplerenone group versus 483 patients in the placebo group (RR=0.83; 95% CI, 0.72 to 0.94; P=0.005). Sudden death from cardiac causes occurred in 162 of the eplerenone group versus 201 in the placebo group (RR=0.79; 95% CI, 0.64 to 0.97; P=0.03). Death from cardiovascular causes or hospitalization for cardiovascular events, death from any cause or any hospitalization, and hospitalization for heart failure were also reduced by eplerenone. At 1 year, potassium levels increased by 0.2 mmol/L in placebo-treated patients versus 0.3 mmol/L in the eplerenone group (P<0.001). Serious hypokalemia (potassium <3.5 mmol/L) occurred in 8.4% of eplerenone patients versus 13.1% in the placebo group (P<0.001). Hyperkalemia, with a serum potassium level ≥6.0 mmol/L, occurred in 5.5% of eplerenone-treated versus 3.9% of placebo-treated patients (P=0.002). Twelve patients in the eplerenone group versus 3 in the placebo group were hospitalized for the condition; 1 patient in the placebo group died of the condition.A number of proposed mechanisms for the reduction in mortality in the eplerenone group were suggested, including effects of eplerenone on plasma volume and electrolyte excretion, reductions in coronary vascular inflammation, improvements in endothelial function, attenuation of platelet aggregation, improvements in ventricular remodeling with a decrease in activation of matrix metalloproteinases, and a decrease in interstitial fibrosis. Besides these direct effects on the vasculature and myocardium, the authors pointed out that aldosterone blockade decreased sympathetic drive in experimental animal studies, improved norepinephrine uptake in heart failure victims, and improved heart rate variability.One of the simplest explanations for the benefit of eplerenone in reducing sudden death is its prevention of hypokalemia, a known trigger of ventricular arrhythmias, especially in patients also taking digitalis preparations. In a letter to the editor, Coca and Buller21 raised the issue that the 21% decrease in the rate of sudden death from cardiac causes associated with eplerenone in the EPHESUS trial may have been attributable to the reduction of hypokalemia. However, Pitt responded that “a preliminary analysis of data from EPHESUS reveals a significant reduction in the risk of sudden death from cardiac causes, which is independent of the effects of eplerenone in preventing hypokalemia.” It is still conceivable that some of the benefit of eplerenone in reducing sudden death was related to preventing hypokalemia. Subsequent analyses revealed that eplerenone reduced the risk of sudden death by 33%22 in patients with baseline left ventricular ejection fraction of ≤30% and that eplerenone reduced the early incidence of sudden death by 37% within 30 days of randomization in this trial.23Although the results show that eplerenone reduced sudden death in the post–myocardial infarction patients with left ventricular dysfunction, many questions in this field remain unanswered. Was this benefit primarily due to a reduction in hypokalemia? Would eplerenone provide this benefit to patients with heart failure but not in the post–myocardial infarction setting? Would eplerenone benefit patients with heart failure who had automatic implantable defibrillators and/or patients with biventricular pacing for cardiac resynchronization therapy?Aldosterone Antagonists in Patients With Chronic Heart FailureThe RALES (Randomized Aldactone Evaluation Study)24 was a double-blind, randomized study of 1663 patients with severe heart failure and a left ventricular ejection fraction ≤35% who were already on an ACE inhibitor, a loop diuretic, and in many cases digoxin. Patients were randomized to 25 mg of the aldosterone antagonist spironolactone (n=822) versus placebo (n=841). The study was stopped at 24 months with 386 deaths (46%) in the patients receiving placebo versus 284 deaths (35%) in the spironolactone group (RR of death=0.70; 95% CI, 0.60 to 0.82; P<0.001). Spironolactone was associated with a lower risk of death from progressive heart failure as well as a lower rate of sudden death. Sudden death due to cardiac cause occurred in 82 of 822 spironolactone-treated patients versus 110 of 841 placebo-treated patients (RR=0.71; 95% CI, 0.54 to 0.95; P=0.02). Spironolactone also reduced all cardiac causes for hospitalization as well as hospitalization for worsening heart failure. The median potassium concentration increased by 0.30 mmol/L in the spironolactone group but did not increase in the placebo group. Serious hyperkalemia was observed in 10 placebo patients (1%) and 14 spironolactone patients (2%; P=NS).Again, although the exact mechanism by which the aldosterone antagonist reduced SCD in RALES is unknown, prevention of hypokalemia cannot be ruled out entirely, despite the rather small increase in potassium levels in the treated group. In addition, it is unknown whether spironolactone could also reduce sudden death in post–myocardial infarction patients with left ventricular dysfunction with or without clinical congestive heart failure or in patients already receiving automatic implantable cardioverter-defibrillators (ICDs) and/or biventricular pacing for resynchronization therapy.A small study by Ramires et al25 randomized 35 patients with class III congestive heart failure due to dilated or ischemic cardiomyopathy and mean ejection fraction of 33% to spironolactone in addition to standard medical therapy for 16 weeks. Spironolactone was initiated at 50 mg/d until week 12 and then was decreased to 25 mg/d until the end of 16 weeks. After 16 weeks, ambulatory ECG monitoring revealed a lower frequency of ventricular premature beats and episodes of nonsustained ventricular tachycardia in the spironolactone group compared with the control group. Spironolactone was also associated with an improvement in ventricular arrhythmias during treadmill exercise. The authors observed that before administration of spironolactone and after adjustment for baseline drug therapy, there was a reduction in serum sodium, potassium, and magnesium that was corrected after 16 weeks of spironolactone therapy. The authors postulated that “a possible explanation for the reduced frequency of ventricular arrhythmia could be related to electrolyte regulation promoted by spironolactone in combination with ACE inhibitors.” They described the concern that hypokalemia could have contributed to increased arrhythmias in the setting of digoxin, which was then corrected by the aldosterone antagonist.Clinical trials are currently lacking of patients with chronic heart failure (not related to the postinfarct setting) who received eplerenone. Table 1 summarizes some recent key randomized trials of nonantiarrhythmic drugs and their effect on SCD.26 Several agents used for the treatment of heart failure (as well as hypertension) have demonstrated this benefit: β-blockers, ACE inhibitors, and aldosterone antagonists. Again, a host of mechanisms have been postulated, including improvements in ventricular remodeling and endothelial function, a reduction in sympathetic tone, and improved electrolyte balance, including less hypokalemia with ACE inhibitors and aldosterone antagonists.For all the aforementioned studies, the assessment of whether the precise cause of death is arrhythmic or due to heart failure, recurrent infarction, or other causes is very difficult. The ICD randomized trials have used total mortality as the end point precisely because retrospective analysis of an individual death is so difficult. Therefore, studies with β-blockers, ACE inhibitors, angiotensin receptor blockers, and aldosterone antagonists that claim to demonstrate a reduction in sudden death need to be interpreted cautiously and with the realization that all deaths, in a sense, are sudden. It is feasible that the mechanisms of benefit of agents including eplerenone and spironolactone may be directly antiarrhythmic, indirectly antiarrhythmic (for example, preventing hypokalemia), or due to a change in the cardiac substrate (for example, an anti-ischemic effect or a reduction in ventricular remodeling). The studies with the aldosterone antagonists to date do not clarify which of these mechanisms is most likely.Aldosterone Antagonists in ICD and Cardiac Resynchronization TrialsThe primary prophylactic ICD trials were initiated in the early 1990s at the same time that new data were emerging on the importance of β-blockers and ACE inhibitors in the prevention of SCD in patients with low ejection fraction.27–31 As is noted in Table 2, use of β-blockers and ACE inhibitors increased over time but was quite low in both the MADIT I and Multicenter Unsustained Tachycardia Trial (MUSTT) when these therapies had not reached wide acceptance.32–39 The use of spironolactone antagonists was very sparing in the ICD trials except for the Comparison of Medical Therapy Pacing and Defibrillation in Heart Failure (COMPANION) trial, which was conducted by heart failure specialists. It is of interest that, as the medical therapy in these trials improved, the degree of superiority of the ICD over conventional therapy declined (from MADIT I32 to Sudden Cardiac Death in Heart Failure [SCD-HeFT]),38 suggesting that survival in both arms in these trials improves as a result of contemporary background medical therapy. In all the trials except 2 (Coronary Artery Bypass Graft [CABG] Patch33 and the Defibrillator in Acute Myocardial Infarction Trial [DINAMIT]36), the ICD demonstrated a survival advantage over best medical therapy. In the Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) trial,39 implantation of ICDs into patients with nonischemic dilated cardiomyopathy and already on ACE inhibitors and β-blockers resulted in a nonsignificant trend toward a reduction in death from any cause and a significant decrease in sudden death. TABLE 2. Therapies According to Treatment Assignment in the Primary Prophylactic ICD TrialsStudy and GroupsACE InhibitorDigoxinβ-BlockerSpironolactoneValues are percentage of patients. NA indicates not applicable.*Spironolactone/potassium-sparing diuretics.MADIT32 ICD6058260 Control553880CABG Patch33 ICD5569180 Control5465240MUSTT34 ICD7655330 Control7753510MADIT II35 ICD68577014 Control72577012DINAMIT36 ICD95NA87NA Control94NA87NACOMPANION37 CRT69NA6855 Control69NA6655SCD-HeFT38 ICD83676920* Control85706919*DEFINITE39 ICD844286NA Control874284NATo conclusively determine whether aldosterone antagonists confer additional benefit on reducing SCDs in patients with ICDs, one would need to design a study in which patients with ICDs (and preferably a group without ICDs as well) were randomized to aldosterone antagonists versus placebo in addition to the usual heart failure medicines. Unfortunately, and to the best of our knowledge, such a study has not been performed. We have been able to obtain some observational retrospective data from MADIT II and the COMPANION trial that address this issue to some extent. We briefly present our findings, realizing that limitations exist that must be kept in mind when these data are viewed. The limitations of this analysis include the following: a lack of randomization for the use of spironolactone; a relatively small number of patients who were assigned to spironolactone, which resulted in the studies not being powered to definitely answer the question about a benefit of aldosterone antagonists in patients already treated with ICD, cardiac resynchronization therapy [CRT], or both; the possible presence of a type II or β error; and the possibility of confounding biases. For example, patients assigned to spironolactone may have been sicker. The analyses below are retrospective and must be considered exploratory and hypothesis generating, not definitive. However, when we were assigned the topic of presenting the “con” side of the aldosterone antagonist argument by Circulation, we tried to find all available data on this concept. Therefore, we briefly present our findings below, and we are not aware of more definitive data at the time of this writing.Although there has been little use of spironolactone in the ICD trials, the MADIT II Investigators have kindly provided new data on the issue of the possible effects of spironolactone in this trial (S. McNit, MS, and J. Hall, PhD, written communication, 2006). MADIT II was a study of 1232 patients with a prior myocardial infarction (≈8 years) and a reduced left ventricular ejection fraction (≤0.30). Patients were randomized to receive an implantable defibrillator or conventional medical therapy.35 No attempt was made to risk stratify by invasive electrophysiological testing. The primary end point was death from any cause. During an average of 20 months of follow-up, the mortality rates were 14.2% in the ICD group versus 19.8% in the conventional medical therapy group, representing a 31% reduction in the risk of death in the ICD patients (hazard ratio [HR]=0.69; 95% CI, 0.51 to 0.93; P=0.016). The conclusion of these investigators was that prophylactic implantation of a defibrillator improved survival and should be considered as therapy for patients with prior myocardial infarction and poor left ventricular dysfunction.At study commencement (in which patients were randomized in a 3:2 ratio to ICD versus conventional therapy), 101 patients (13.6%) randomized to ICD treatment were receiving spironolactone, and 57 (11.6%) in the conventional arm were receiving this drug. Spironolactone use was analyzed as a time-varying risk factor in proportional hazards regression analysis of the various end points in MADIT II. Because of the clinical suspicion that spironolactone may have been used as a result of a hospitalization for heart failure, the first occurrence of heart failure was also used as a time-dependent factor.40 The HR for all-cause mortality for patients while on spironolactone compared with patients and periods not on spironolactone was 1.13 (P=0.53). Thus, no overall effect of spironolactone use on all-cause mortality was found in MADIT II. The HR for spironolactone for all-cause mortality in the conventional medical arm was 1.43 versus 0.90 in the ICD arm (P=0.23 for difference).The HR for spironolactone for sudden death in the conventional arm was 1.13 (P=0.76). The HR for spironolactone for first appropriate shock in the ICD arm was 1.51 (P=0.07). The HR for spironolactone for either first appropriate shock or death in the ICD arm was 1.20 (P=0.34). Most of the HRs for spironolactone exceeded unity, suggesting a trend toward an increased risk of the end point occurring when the patients were on the drug relative to being off the drug. However, drug usage could be a proxy for heart failure risk or severity of heart failure. In summary, on the basis of a retrospective analysis, the MADIT II trial produces no evidence that spironolactone provides a benefit.Supporting evidence is provided by the COMPANION trial along similar lines. The COMPANION trial randomized >1500 New York Heart Association class III/IV heart failure patients with a prolonged QRS and ejection fraction ≤35% to optimal medical therapy, optimal medical therapy plus CRT, or optimal medical therapy plus CRT plus an ICD. Both CRT and CRT plus an ICD reduced combined all-cause mortality and hospitalization in heart failure patients. CRT plus an ICD reduced all-cause mortality; CRT alone had a trend toward reduced mortality.In the COMPANION trial, 55% of patients were treated with spironolactone, and this was not associated with risk of death in any treatment group and did not protect against appropriate shocks in the patients with CRT plus an ICD. However, β-blockers and ACE inhibitors did afford such protection (L. Saxon, MD, written communication, 2006, and Saxon et al41).SummaryIn summary, recent studies have shown the usefulness of eplerenone for post–myocardial infarction patients with heart failure and spironolactone for patients with chronic congestive heart failure. In these 2 studies (which lacked use of ICDs), the aldosterone antagonists reduced SCD. There are a number of potential explanations for the mechanism of this benefit, including protection against hypokalemia. In recent retrospective analyses of MADIT II and COMPANION trials of patients with left ventricular dysfunction/heart failure in which ICDs were used, no evidence was provided that spironolactone afforded a survival benefit or reduced the need for appropriate ICD shocks. Aldosterone antagonists may still benefit heart failure patients who have ICDs independently of reduction of arrhythmias, for example, by reducing heart failure symptoms and/or hospitalizations. Thus, if patients with heart failure on spironolactone receive an ICD, we do not suggest that the spironolactone be stopped. Should an aldosterone antagonist be added for patients with severe heart failure who already have a defibrillator? In this case, spironolactone may reduce heart failure symptoms, but whether spironolactone will further reduce total mortality or sudden death is uncertain. Prospective, adequately powered, randomized, blinded trials are needed to examine the interaction or possibly the lack of interaction between ICD, CRT, and both with the aldosterone antagonists. Specifically, a study is needed in which patients who are already on standard medical therapy plus ICD, CRT, or both are randomized to an aldosterone antagonist versus placebo to determine whether this class of drugs further reduces mortality, SCD, and hospitalizations for heart failure. Unfortunately, we think it is unlikely that any agency or industry would fund such a study, and therefore it is possible that a definite answer to whether aldosterone antagonists confer benefits in addition to those of implantable devices alone may never be known.DisclosuresDr Kloner is a consultant and speaker for Pfizer. Dr Cannom is a consultant and speaker for Medtronic and Boston Scientific.FootnotesCorrespondence to Robert A. Kloner, MD, PhD, Heart Institute, Good Samaritan Hospital, 1225 Wilshire Blvd, Los Angeles, CA 90017. E-mail [email protected] References 1 Kloner RA, Leor J, Birnbaum Y. Remodeling, hibernation, and stunning. In: Poole-Wilson PA, Colucci WS, Massie BM, Chatteerjee K, Coats AJS, eds. Heart Failure. New York, NY: Churchill Livingstone; 1997: 109–125.Google Scholar2 Goldenberg H, Moss AJ, Hall WJ, McNitt S, Zareba W, Andrews ML, Cannom DS; for the Multicenter Automatic Defibrillator Implantation Trial (MADIT) II Investigators. Causes and consequences of heart failure after prophylactic implantation of a defibrillator in the MADIT II. Circulation. 2006; 113: 2810–2817.LinkGoogle Scholar3 Pfeffer MA, Braunwald E, Moye LA, Basta L, Brown EJ Jr, Cuddy TE, Davis BR, Geltman EM, Goldman S, Flaker GC, Klein M, Lamas GA, Packer M, Rouleau J, Rouleau JL, Rutherford J, Wertheimer JH, Hawkins CM; on behalf of the SAVE Investigators. Effect of captopril on mortality and morbidity in patients with left ventricular dysfunction after myocardial infarction: results of the survival and ventricular enlargement trial. N Engl J Med. 1992; 327: 669–677.CrossrefMedlineGoogle Scholar4 St John Sutton M, Pfeffer MA, Moye L, Plappert T, Rouleau JL, Lamas G, Rouleau J, Parker JO, Arnold MO, Sussex B, Braunwald E. Cardiovascular death and left ventricular remodeling two years after myocardial infarction: baseline predictors and impact of long-term use of captopril: information from the Survival and Ventricular Enlargement (SAVE) trial. Circulation. 1997; 96: 3294–3299.CrossrefMedlineGoogle Scholar5 Gruppo Italiano per lo Studio della Sopravvivenza nell’infarto Miocardico. GISSI-3: effects of lisinopril and transdermal glyceryl trinitrate singly and together on 6-week mortality and ventricular function after acute myocardial infarction. Lancet. 19