Author + information
- Received February 14, 2006
- Revision received May 5, 2006
- Accepted May 15, 2006
- Published online September 19, 2006.
- Jeffrey J. Goldberger, MD, FACC⁎,⁎ (, )
- Haris Subacius, MA⁎,
- Andi Schaechter, RN⁎,
- Adam Howard, BA⁎,
- Ronald Berger, MD, PhD, FACC†,
- Alaa Shalaby, MD, FACC‡,
- Joseph Levine, MD§,
- Alan H. Kadish, MD, FACC⁎,
- DEFINITE Investigators
- ↵⁎Reprint requests and correspondence:
Dr. Jeffrey Goldberger, Northwestern University Feinberg School of Medicine, 251 East Huron, Feinberg Pavilion 8-542, Chicago, Illinois 60611.
Objectives We sought to evaluate whether statins were associated with a survival benefit and significant attenuation in life-threatening arrhythmias in patients with nonischemic dilated cardiomyopathy.
Background Statins are associated with a reduction in appropriate implantable cardioverter-defibrillator (ICD) therapy in patients with coronary artery disease and improved clinical status in nonischemic dilated cardiomyopathy.
Methods The effect of statin use on time to death or resuscitated cardiac arrest and time to arrhythmic sudden death was evaluated in 458 patients enrolled in the DEFINITE (DEFIbrillators in Non-Ischemic cardiomyopathy Treatment Evaluation) study. The effect of statin use on time to first appropriate shock was analyzed only in the 229 patients who were randomized to ICD therapy.
Results The unadjusted hazard ratio (HR) for death among patients on versus those not on statin therapy was 0.22 (95% confidence interval [CI] 0.09 to 0.55; p = 0.001). When controlled for statin effects, ICD therapy was associated with improved survival (HR 0.61; 95% CI 0.38 to 0.99; p = 0.04). There was one arrhythmic sudden death in the 110 patients receiving statin therapy (0.9%) versus 18 of 348 patients not receiving statins (5.2%; p = 0.04). The unadjusted HR for arrhythmic sudden death among patients on versus those not on statin therapy was 0.16 (95% CI 0.022 to 1.21; p = 0.08). The HR for appropriate shocks among patients on versus those not on statin therapy was 0.78 (95% CI 0.34 to 1.82) after adjustment for baseline differences in the two groups.
Conclusions Statin use in the DEFINITE study was associated with a 78% reduction in mortality. This reduction was caused, in part, by a reduction in arrhythmic sudden death. These findings should be confirmed in a prospective, randomized clinical trial.
Hydroxymethylglutaryl coenzyme A reductase inhibitors (statins) are known to have effects independent of their cholesterol-lowering effects, referred to as pleiotropic effects. These effects have been implicated in protection against atrial fibrillation (1–3) and in clinical improvement in patients with heart failure (4–6). Statins also have been associated with a reduction in mortality in large-scale lipid trials and a reduction in appropriate implantable cardioverter-defibrillator (ICD) therapy in patients with coronary artery disease (7–9). Whether they have a specific effect on life-threatening ventricular arrhythmias or whether their effects on reducing these arrhythmias are predominantly mediated by an anti-ischemic effect is unknown. Of note, significant improvements in clinical status have been noted in patients with nonischemic dilated cardiomyopathy treated with statins (4–6,10). If the pleiotropic effects of these drugs have specific beneficial effects on either cardiac function or life-threatening ventricular arrhythmias, they also may be associated with a survival benefit in patients with nonischemic dilated cardiomyopathy. The aim of this report was to evaluate whether statins were associated with a survival benefit and significant attenuation in life-threatening arrhythmias in patients with nonischemic dilated cardiomyopathy enrolled in the DEFINITE (DEFIbrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation) study.
The DEFINITE trial is described in detail elsewhere (11). In brief, the DEFINITE trial was a randomized, prospective investigator-initiated study. Inclusion criteria were: age between 21 and 80 years, nonischemic cardiomyopathy with a left ventricular ejection fraction (LVEF) ≤35%, history of symptomatic heart failure, and the presence of one of the following within the past six months; nonsustained ventricular tachycardia on telemetry monitoring or Holter monitoring and/or an average of 10 premature ventricular beats per hour on a 24-h Holter monitor. Patients were excluded from enrollment if they had New York Heart Association (NYHA) functional class IV heart failure, were candidates for an ICD, or had a permanent pacemaker. Each patient was randomized to standard oral medical therapy for heart failure or standard oral medical therapy plus an ICD. The primary end point of the study was death from any cause. A prespecified secondary end point was sudden death from an arrhythmia. The cause of death was determined by an events committee whose members were unaware of the patients’ treatment assignments according to classification proposed by Epstein et al. (12). All shocks were reviewed by a separate committee. Shocks were classified by the committee as appropriate if delivered for ventricular tachycardia or fibrillation.
In this trial, more than 85% of patients received beta-blockers, and more than 90% received angiotensin-converting enzyme (ACE) inhibitors/angiotensin receptor blockers. The DEFINITE study protocol did not specify treatment for hypercholesterolemia, but these data were collected at enrollment, at the time of an arrhythmic event, and at study termination. Cholesterol levels were not routinely collected from all study participants.
The baseline characteristics of the two groups were compared with the use of two-sample ttests for continuous variables and chi-square tests for categorical variables. The log-rank test was used to compare Kaplan-Meier survival curves in the 2 groups and the Cox proportional hazards survival model was used to adjust for covariates as well as estimate the hazard ratio (HR) and corresponding 95% confidence intervals (CIs). All covariates were entered in the first step of the regression model using a forward stepwise method. Statin therapy status was entered in a separate step after the significant covariates had been determined. The impact statin therapy had in addition to the combined effect of all significant covariates was evaluated using a likelihood ratio test.
Three separate end point analyses were performed to evaluate the effect of statin therapy in this trial. In the first analysis, time to death or resuscitated cardiac arrest was analyzed. Sixty-eight deaths and 2 resuscitated cardiac arrests took place during the trial for a total of 69 events in this analysis (1 patient who had a resuscitated cardiac arrest subsequently died—the patient was coded as having reached the end point at the time of the first event). The second analysis used an end point of arrhythmic sudden death, which included 17 deaths that had been adjudicated by the events committee to be arrhythmic sudden deaths and two resuscitated cardiac arrests. These two analyses were performed in all 458 patients in the trial. The final analysis evaluated time to first appropriate shock in the 229 patients randomized to receive an ICD. One additional arrhythmic death is included as an end point in a patient whose ICD did not discharge despite the appearance of a lethal ventricular arrhythmia. Four patients who crossed over from the ICD group to the standard therapy group were censored from analysis at the time of crossover. The average number of appropriate shocks experienced by 34 patients also was compared using the ttest. All p values were 2-tailed; p < 0.05 was considered significant.
Four-hundred fifty-eight patients with nonischemic dilated cardiomyopathy were enrolled. There were 110 patients who were receiving statin therapy at the time of their first event (death or resuscitated cardiac arrest) or at the end of the trial for those patients who did not reach the end point during the study. The baseline patient characteristics based on statin therapy at the time of an event or study end are summarized in Table 1.There were no significant differences in baseline characteristics between the 2 groups.
There were 5 deaths in the 110 patients receiving statin therapy (4.6%) versus 64 deaths in the 348 patients not receiving statin therapy (18.4%; p < 0.001). Figure 1depicts the Kaplan-Meier survival curve for freedom from death. The unadjusted HR for death among patients on statin therapy versus those not on statin therapy was 0.22 (95% CI 0.09 to 0.55; p = 0.001).
Using Cox proportional hazards survival models, other significant univariate predictors of death were NYHA functional class III (HR 2.14; 95% CI 1.30 to 3.53; p = 0.003), diabetes (HR 1.79; 95% CI 1.07 to 3.00; p = 0.03), hypertension (HR 2.34; 95% CI 1.27 to 4.29; p = 0.006), beta-blocker therapy (HR 0.49; 95% CI 0.29 to 0.85; p = 0.01), duration of congestive heart failure (CHF) longer than 1 year (HR 1.62; 95% CI 1.00 to 2.61, p = 0.05), QRS duration (HR 1.01; 95% CI 1.00 to 1.02; p = 0.04), and treatment assignment (HR 0.61; 95% CI 0.38 to 0.99; p = 0.04). After accounting for these covariates, statin therapy remained a significant factor (adjusted HR 0.23; 95% CI 0.09 to 0.58; p = 0.002). Diabetes and CHF duration were no longer significant in this multivariate model. When considered along with the effects of statins, ICD therapy was associated with improved survival (HR 0.61; 95% CI 0.38 to 0.99; p = 0.04), virtually the same HR reported for the trial (11). Age, gender, smoking history, ACE inhibitor/angiotensin receptor blocker therapy, and aspirin were significant predictors of outcome. Although ACE inhibitor/angiotensin receptor blocker therapy was not a significant covariate, it should be noted that 95% of patients were on this therapy; the small sample size of patients not on this therapy virtually precludes the detection of an effect.
Arrhythmic sudden death
There was one arrhythmic sudden death in the 110 patients receiving statin therapy (0.9%) versus 18 arrhythmic sudden deaths in the 348 patients not receiving statin therapy (5.2%; p = 0.04). Figure 2depicts the Kaplan-Meier survival curve for freedom from an arrhythmic sudden death. The unadjusted HR for arrhythmic sudden death among patients on statin therapy versus those not on statin therapy was 0.16 (95% CI 0.022 to 1.21; p = 0.08).
Using Cox proportional hazards survival models, we found that the only covariate that was significantly related to the outcome in addition to statin therapy was ICD treatment (HR 0.17; 95% CI 0.05 to 0.60; p = 0.005). Although the regression coefficient for statin therapy remained a nonsignificant factor when modeled along with ICD treatment (HR 0.16; 95% CI 0.02 to 1.21; p = 0.08), the likelihood ratio test showed that the addition of statin therapy status to the model that already contains ICD treatment resulted in a significant improvement of model fit (p = 0.02). Neither age, gender, NYHA functional class, LVEF, QRS duration, smoking history, diabetes, hypertension, CHF duration longer than 1 year, beta-blocker therapy, ACE inhibitor/angiotensin receptor blocker therapy, nor aspirin therapy were significant predictors of arrhythmic sudden death.
Of the 229 patients randomized to receive an ICD, 56 were receiving statin therapy at the time of their first appropriate shock or at the end of the trial. The baseline patient characteristics for this group based on statin therapy are summarized in Table 2.More patients who were not on statin at the time of an arrhythmic event were on ACE inhibitors/angiotensin receptor blockers (96.5%) as compared with patients who were on statin therapy (89.3%; p = 0.03). There were no other significant differences in baseline characteristics. There were 7 patients who received appropriate shocks among the 56 patients receiving statin therapy (12.5%) versus 27 patients with appropriate shocks among the 173 patients not receiving statin therapy (15.6%; p = 0.60 for log-rank test statistic; HR 0.80; 95% CI 0.35 to 1.84; p = 0.80). Figure 3depicts the Kaplan-Meier survival curve for freedom from appropriate shocks. The HR for appropriate shocks among patients on statin therapy versus those not on statin therapy was 0.78 (95% CI 0.34 to 1.82) after adjustment for baseline ACE inhibitor/angiotensin receptor blocker therapy status.
Figure 4shows the distribution of number of shocks each patient received stratified by statin therapy. Patients who were not on statin therapy did not receive multiple shocks more frequently than those on statin therapy (15 of 27 or 55.6% vs. 3 of 7 42.9%, p = 0.68 via Fisher’s exact test). The mean number of shocks was 2.1 ± 1.6 shocks for those not on statin therapy versus 1.7 ± 1.1 shocks for those on statin therapy (p = 0.50). There was no significant difference in the incidence of inappropriate shocks among those ICD patients who were treated with statins (21.4%) and those not treated with statins (25.4%).
The present report demonstrates that statin use in patients with nonischemic cardiomyopathy may have dramatic effects on survival. The observed 78% reduction in mortality associated with statin therapy was larger than the benefit attributable to ICD therapy in this trial, though there was a significant benefit to ICD therapy even after accounting for the benefit of statin therapy. The effects of statins on total mortality are due, in part, to a reduction in arrhythmic sudden death. These findings should be confirmed in a prospective, randomized clinical trial.
Although the benefit of statin therapy in patients with coronary artery disease has been well established in multiple large-scale randomized clinical trials, only few studies have focused on their effects on life threatening ventricular arrhythmias (7–9) or on patients with nonischemic cardiomyopathy (4–6,10,13). A number of nonrandomized clinical studies have demonstrated dramatic reduction in arrhythmic events in patients who have ICDs inserted. In an observational study, DeSutter et al. (7) first reported that in patients with coronary artery disease receiving ICDs for secondary prevention of ventricular arrhythmias, treatment with lipid-lowering drug therapy (59% statins, 41% fibrates) resulted in a substantial reduction in appropriate shocks (22% in the group treated with lipid lowering drugs and 57% in those not treated). In another observational study in patients with coronary artery disease receiving ICDs, Chiu et al. (8) reported that 30% of patients on statins received ICD therapy versus 50% among those who did not receive statin therapy (HR 0.60). In the AVID (Antiarrhythmics Versus Implantable Defibrillators) trial (9), there was a reported 0.40 reduction in relative hazard (95% CI 0.15 to 0.58) for recurrence of ventricular tachycardia/fibrillation.
Recently, special focus has been given to the effects of statins in patients with heart failure as these patients typically were not included in the large-scale clinical trials evaluating statin therapy and there are data suggesting an inverse relationship between cholesterol levels and mortality in patients with heart failure (14,15). In the PRAISE (Prospective Randomized Amlopidine Survival Evaluation) trial (16), there was a 62% reduction in total mortality attributable to statin therapy. Statin therapy was administered to 134 of the 1153 patients in the trial, but only 23 patients with nonischemic cardiomyopathy received statins. In a large cohort study of elderly patients with heart failure (etiology not specified), a 33% reduction in total mortality was observed in patients treated with statins. Several studies show that fewer patients develop heart failure when treated with statins (17,18). Several small randomized studies (4–6) of 15 to 108 patients with nonischemic cardiomyopathy revealed that statin therapy was associated with significant improvement in clinical parameters including quality of life, exercise capacity, NYHA functional class, and LVEF. Finally, Horwich et al. (10) described the outcomes of 551 patients referred to a specialized cardiomyopathy center for heart failure or transplant evaluation; 55% of the patients had a nonischemic cardiomyopathy. Forty-five percent of the total population was treated with statins, but only 22% of those with nonischemic cardiomyopathy. There was a significant improvement in 1-year survival without the need for urgent heart transplantation in patients with nonischemic cardiomyopathy treated with statins. The current analysis of the use of statins in patients enrolled in the DEFINITE study further suggests that statin therapy in patients with nonischemic cardiomyopathy improves survival. Although some of this benefit is attributable to reduction in arrhythmic sudden death, the magnitude of its effect suggests that there are also other mechanisms for the improved survival.
Although statin therapy has been reported to reduce appropriate ICD shocks (7–9), these reports have included only patients with coronary artery disease. It is interesting to note that, in this study, there was no clear reduction in appropriate ICD shocks related to statin therapy, despite a reduction in mortality and arrhythmic sudden death. This result could reflect limited power to detect a difference due to the small sample size (n = 229). Alternatively, in this population, appropriate ICD shocks occurred more frequently than sudden cardiac death, suggesting that many of the arrhythmias may have self-terminated (19). Nevertheless, statin therapy appears to have some effect on arrhythmic sudden death, although not having a detectable effect on these latter arrhythmias. Finally, statin effects on nonarrhythmic mortality may be more prominent than the effects on arrhythmic mortality. Further studies are necessary to determine the relative contribution of statin therapy to improvement in arrhythmic and non-arrhythmic mortality.
The mechanism for the effects of statins on arrhythmic sudden death and survival in patients with nonischemic cardiomyopathy remains unclear. An anti-ischemic effect, as noted in patients with coronary artery disease, is plausible as a substantial proportion of patients with nonischemic cardiomyopathy are found to have coronary artery disease at autopsy (20). With regard to antiarrhythmic effects, statins also have been reported to be effective in prevention of atrial fibrillation (1–3). Finally, as noted previously, the small studies that have found improvements in ejection fraction and exercise capacity related to statin therapy support the notion that statins may have beneficial effects on left ventricular remodeling. Thus, it is most likely that statin therapy exerts multiple beneficial effects in patients with nonischemic cardiomyopathy by its lipid lowering, anti-inflammatory, antioxidant, autonomic, and/or other effects (21).
The present findings are particularly notable for the large effects attributable to statin therapy, an adjusted HR of 0.22 for total mortality. This occurred in the setting of 85% usage of beta-blocker therapy and 95% usage of ACE inhibitors or angiotensin receptor blockers. On the basis of the available data at the time the trial was designed, a 15% 2-year mortality was postulated with 50% of the deaths being due to arrhythmia. The observed 2-year mortality in the standard therapy group was 14.1%, but only one-third of the deaths were due to arrhythmia. It is possible that the high frequency of statin use may have contributed to the reduced fraction of deaths attributable to arrhythmia.
Although the present findings are quite dramatic, they should be considered exploratory and require confirmation in a prospective randomized clinical trial. However, selection bias as an explanation for the present findings is unlikely. First, there are multiple studies that document a relatively large effect of treatment, including randomized trial data. Second, the presence of a selection bias would imply that hypercholesterolemia triggering statin therapy selects a patient population at markedly lower risk than nonhypercholesterolemic patients, an implausible assumption.
Although it is attractive to attribute the findings to the pleiotropic effects of statins, it is possible that the results are due to their cholesterol-lowering effects, with the possibility that all lipid-lowering agents might provide the same benefit in patients with nonischemic dilated cardiomyopathy. There were an additional 14 patients on other lipid-lowering agents; including those patients in this analysis did not alter the findings. Furthermore, it is unknown whether patients with nonischemic cardiomyopathy without hypercholesterolemia might benefit from statin therapy. Bleske et al. (13) showed neither a beneficial nor a detrimental effect of statin therapy on markers of inflammation, endothelial activation, and parasympathetic tone in patients with heart failure and average low-density lipoprotein levels.
Statin therapy has become a mainstay in patients with coronary artery disease. The combined data from this trial and the few other clinical studies that have evaluated the use of statins in nonischemic dilated cardiomyopathy provide strong justification for evaluating whether routine statin use in this patient population, even those without hypercholesterolemia, provides a clinical benefit.
The DEFINITE study was funded by St. Jude Medical, Sylmar, California.
- Abbreviations and Acronyms
- angiotensin-converting enzyme
- congestive heart failure
- DEFIbrillators in Non-Ischemic cardiomyopathy Treatment Evaluation
- hazard ratio
- implantable cardioverter-defibrillator
- left ventricular ejection fraction
- New York Heart Association
- Received February 14, 2006.
- Revision received May 5, 2006.
- Accepted May 15, 2006.
- American College of Cardiology Foundation
- Shiroshita-Takeshita A.,
- Schram G.,
- Lavoie J.,
- Nattel S.
- Kumagai K.,
- Nakashima H.,
- Saku K.
- Node K.,
- Fujita M.,
- Kitakaze M.,
- Hori M.,
- Liao J.
- DeSutter J.,
- Tavernier R.,
- DeBuyzere M.,
- Jordaens L.,
- DeBacker G.
- Mitchell L.,
- Powell J.,
- Gillis A.,
- Kehl V.,
- Hallstrom A.,
- AVID Investigators
- Horwich T.,
- MacLellan W.,
- Fonarow G.
- Epstein A.,
- Carlson M.,
- Fogoros R.,
- Higgins S.,
- Venditti F.
- Bleske B.,
- Nicklas J.,
- Bard R.,
- et al.
- Rauchhaus M.,
- Clark A.,
- Doehner W.,
- et al.
- Horwich T.,
- Hamilton M.,
- Maclellan W.
- Kjekshus J.,
- Pedersen T.,
- Olsson A.,
- Faergeman O.,
- Pyorala K.
- Ellenbogen K.,
- Levine J.,
- Berger R.,
- et al.
- Uretsky B.,
- Thygesen K.,
- Armstrong P.,
- et al.