Coronary artery disease is the underlying etiology for approximately 80% of fatal ventricular arrhythmias, the remainder being accounted for by other pathological conditions involving the myocardium (1).

The protective effect against ventricular tachycardia (VT)/ventricular fibrillation (VF) attributed to use of 3-hydroxy-3-methylglutaryl-coenzyme A reductase-inhibitors (statins) has been described in numerous studies (2). However, because these studies enrolled mainly patients with coronary artery disease, scarce data exist regarding this effect in nonischemic cardiomyopathy (NICM) patients. A sub analysis of the DEFINITE (Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation) trial suggested a trend toward a reduction in the risk of arrhythmic sudden death among statin users, yet there was no difference in the rate of appropriate implantable cardioverter-defibrillator (ICD) shocks (9).

The MADIT-CRT (Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy) enrolled mildly symptomatic heart failure patients with ischemic cardiomyopathy (n = 999) and NICM (n = 821) who were randomized to cardiac resynchronization therapy with a defibrillator (CRT-D) or to defibrillator-only therapy (10). The aim of the present study was to evaluate the effect of statin therapy on the risk for life-threatening ventricular tachyarrhythmias among NICM patients enrolled in the MADIT-CRT trial.

The design and results of the MADIT-CRT trial were previously published (10). Briefly, study patients were randomly assigned in a 3:2 ratio to receive either CRT-D or ICD. Patients who were at least 21 years of age were enrolled if they had ischemic cardiomyopathy (New York Heart Association functional class I or II) or NICM (New York Heart Association functional class II only), sinus rhythm, left ventricular ejection fraction ≤30%, and a prolonged QRS duration of >130 ms. An NICM diagnosis was based on the absence of: 1) a documented history of a prior infarction (Q-wave or enzyme positive); 2) a history of a coronary revascularization procedure (coronary artery bypass graft surgery, balloon and/or stent angioplasty); and 3) documented significant (>50%) stenosis in a major epicardial coronary artery at angiography and history of angina pectoris or other coronary-related symptoms or signs. These criteria were evaluated in each follow-up examination and by the treating physician. No patients developed ischemic coronary artery disease during follow-up according to the aforementioned definitions, and hence none were excluded from the analysis for this reason. Follow-up data included detailed information of drug therapy during the trial. Statin therapy (type of drug and dosage) was prescribed at the discretion of the treating physician.

The protocol was approved by the institutional review board at each participating organization, and each patient provided written informed consent before enrollment.

Commercially available devices (Boston Scientific, Natick, Massachusetts) were used in the trial. Device testing and programming were performed as reported (10). The configuration used a VT zone set at 170 beats/min and a VF zone set at 230 beats/min. Fast VT was defined as a ventricular rate faster than 180 beats/min. The ICDs were interrogated quarterly and after the occurrence of device therapy at a core laboratory for arrhythmia adjudication and categorization. Ventricular tachycardia was defined as ventricular rate up to 250 beats/min; VF was defined as ventricular rate faster than 250 beats/min with disorganized ventricular electrograms.

The primary endpoint of the present study was the first occurrence of appropriate defibrillator therapy for fast VT/VF or death, whichever occurred first. Death was included in the primary endpoint to avoid a possible bias of censoring fatal arrhythmic events that were not interrogated. The mode of death was adjudicated by the mortality committee with the modified Hinkle-Thaler classification (11). The secondary endpoints of the study included the first occurrence of an appropriate defibrillator shock during the trial and the first occurrence of defibrillator therapy for fast VT/VF or for VF only.

The chi-square test was used for the comparison of categorical variables, and the nonparametric Wilcoxon rank sum test was used for continuous variables (assessed as mean ± SD). The cumulative probabilities for the first occurrence of the endpoints by treatment with statins were assessed according to the method of Kaplan and Meier, with comparison of cumulative events by the log-rank test and comparison at specific time points with a z-test. Kaplan-Meier analyses were carried out as follows: 1) for patients who were treated with statins at baseline, follow-up in the survival analysis was censored at the time in which statin therapy was terminated during follow-up; and 2) for patients who did not receive statin therapy at baseline, follow-up was censored at the time in which statin therapy was initiated during follow-up. Multivariate regression models were developed with Cox proportional hazards regression. Covariates included in the model were identified with a best subset procedure among variables that were unbalanced between statin users and nonusers as predictive of the primary endpoint, fast VT/VF, or death. The covariates used in the multivariate model were age, diabetes mellitus, hypertension, left ventricular end-systolic volume indexed to body surface area, and resynchronization treatment. Statin therapy was assessed in the multivariate model in a time-dependent manner (i.e., by incorporating in the Cox model data for each patient that identifies the effect of each follow-up time “on” and “off” statin therapy during the trial). The effects of time-dependent statin therapy on the endpoints in each treatment arm were assessed with interaction-term analysis. All p values were 2-sided, and a p value <0.05 was considered significant. Analyses were conducted with SAS software (version 9.2, SAS Institute, Cary, North Carolina).

The baseline clinical and echocardiographic patient characteristics stratified by statin therapy are shown in (Table 1). Among MADIT-CRT patients with NICM, 499 (61%) were statin users. Statin users were older and more frequently had diabetes and hypertension. Statin nonusers more frequently smoked and used digoxin, and their indexed left atrial volume was higher.

During follow-up, a total of 158 confirmed events of VT and VF were documented, of which 49 were further categorized as fast VT; a total of 139 appropriate shocks were delivered and 32 (3.9%) patients died.

The mean tachycardia cycle length was significantly longer among statin users (308 vs. 268 ms, p = 0.02).

The cumulative probabilities of time to the combined endpoint by statin usage at enrollment and follow-up censored upon change in medication usage is presented in (Figure 41_gr1). Statin users had a lower cumulative probability of fast VT/VF or death (11% compared with 19% at 4 years of follow-up; unadjusted p log-rank = 0.006). The corresponding results for the combined outcome of fast VT/VF or sudden cardiac death were very similar (10% compared with 17.5% for nonusers, p = 0.007). The cumulative probabilities of time to fast VT/VF and to VF only by statin usage are presented in (Figure 41_gr2)A and (Figure 41_gr2)B, respectively. Statins users had a lower cumulative probability of fast VT/VF events (6% compared with 10% in nonusers, unadjusted p log-rank = 0.003) and lower probability of VF events (6% compared with 9% among nonusers, unadjusted p log-rank = 0.045). Accordingly, patients treated with statins during the trial experienced a significantly lower cumulative probability of appropriate shocks (15% at 4 years of follow-up) as compared with statin nonusers (22%; unadjusted p log-rank = 0.01) (Figure 41_gr3B).

Multivariate Cox proportional hazards regression analysis showed that time-dependent statin therapy was independently associated with a 77% (p < 0.001) reduction in the combined endpoints of fast VT/VF or death, when adjusted for clinical and echocardiographic parameters (Table 2). Similar results were observed when the effects of statins on VF episodes were evaluated and on fast VT/VF; however, possibly due to sample size limitations, the effect of statin therapy on this endpoint did not reach statistical significance (hazard ratio: 0.57; 95% confidence interval: 0.31 to 1.05, p = 0.07). The reduction in the risk of ventricular arrhythmias was accompanied by an independent 46% (p = 0.015) reduction in the risk for the occurrence of appropriate device shocks (Table 3).

The reduction in arrhythmic risk associated with statin therapy was consistent for both the CRT-D and ICD-only arms (all p values for statin therapy × treatment arm interactions >0.10), suggesting that the effect of statin therapy on the risk of ventricular tachyarrhythmias in NICM patients was independent of the implanted device.

Data regarding the effect of statins on arrhythmic risk in NICM patients are scarce and limited mostly to 1 sub study directly addressing this issue (9).

Our study suggests that statin therapy was associated with a significant reduction in the risk of ventricular tachyarrhythmias and appropriate defibrillator shocks in patients with NICM enrolled in the MADIT-CRT trial.

Antiarrhythmic pleiotropic properties unrelated to the reduction of ischemic events have been attributed to statins. These include the modulation of transmembrane ion channel properties (12), improved heart rate variability due to beneficial effect on the autonomic tone (13), and a reduction in corrected QT duration and circadian variability corrected QT dispersion (14). It is feasible that the reduction in ventricular arrhythmias and shock treatment observed in our study were due to some or all of these effects.

In both the CORONA (Controlled Rosuvastatin Multinational Trial in Heart Failure) (17) and GISSI HF (Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto Miocardico–Heart Failure) (18) trials, statin therapy was not found to be associated with reduction in sudden cardiac death. However, several important differences exist between the 3 studies, including different inclusion criteria (only patients older than 60 and with an ejection fraction lower than 40% due to ischemic etiology were recruited to the CORONA trial, whereas in the GISSI HF trial only 35% of the patients had NICM), study patient age (the patients in both the CORONA and GISSI HF trials were older than the patients with NICM recruited to the MADIT-CRT), and outcome evaluation (sudden cardiac death was a secondary endpoint, and arrhythmias were not directly evaluated in either trial). Furthermore, in both the CORONA and GISSI HF trials, rosuvastatin at a dose of 10 mg was used—a dosage that many believe to be insufficient to maximize the pleiotropic effects of statins.

Our observation that statin treatment is associated with slower ventricular arrhythmias is novel. Faster ventricular tachyarrhythmias have been previously attributed to inflammatory states (19). One of the most well-established pleiotropic effects of statin is a reduction in inflammation, and therefore the slower arrhythmia rates could be partially attributed to this effect. An additional possible explanation is the modulation of the autonomic tone attributed to statin treatment. Further studies are needed to confirm this finding and provide additional underlying mechanisms.

We have recently shown that CRT-D therapy is associated with a significant reduction in the risk of VT/VF compared with ICD only among patients who showed an echocardiographic response to therapy (20). The present study shows that the effect of statin therapy on the risk of ventricular tachyarrhythmias was consistent for patients randomized to either therapy. Thus, treatment with statins might potentiate an already pre-existing risk-reducing effect of CRT-D on arrhythmic risk among patients with NICM.

The MADIT-CRT patients were selected for characteristics suggesting the possibility of a response to resynchronization therapy, and the results might not be generalizable to a broader NICM patient population.

Statins were prescribed nonrandomly by the treating physician. In addition, this was an observational analysis of the MADIT-CRT database, and thus differences between the baseline characteristics of the patients who were and were not treated with statins could have affected the occurrence of ventricular arrhythmias. Nevertheless, the favorable effects of statins persisted after multivariate adjustment for clinical factors.

Data regarding individual clinical outcomes were not provided, because the study sample was not sufficiently powered to evaluate the separate occurrence of VT or VF. Although the magnitude of risk reduction associated with statin therapy was consistent for all endpoints studied, the effect of statin therapy on the risk of the endpoint of fast VT/VF only did not reach statistical significance. These findings might be due to sample size limitations or to the fact that some lethal events (which were not included in this endpoint) were due to arrhythmic events.

Statins have also been suggested to exert direct antiarrhythmic effects, in addition to reduction in the risk of recurrent ischemic events. In this study, we extend these prior findings and show that treatment was associated with a significant reduction in the risk of ventricular arrhythmias for patients with NICM. In addition, our data suggest that statin treatment is associated with a slower cycle length at the time of the arrhythmia and with a lower incidence of VT/VF, regardless the implanted device-type. Additional studies, preferably performed in a randomized fashion, are needed to confirm these findings.