Hypothetically, 3-hydroxy-3-methylglutaryl co-enzyme A (HMG-CoA) reductase inhibitors (statins) protect against arrhythmias in addition to their well-established secondary prevention benefits against atherosclerotic coronary artery disease. The use of statins has been suggested to protect against atrial fibrillation (AF) in some clinical studies, but remains rather inadequately explored. Observational studies have provided evidence supporting a protective role of statins against AF.1,2 The evaluation of their possible benefit on the end-point of incidence or recurrence of AF has been improved after the publication of recent data based on secondary analysis of randomised trials with statins.
Some mechanisms by which statins may prevent AF have been described. Inflammation is involved in the development, recurrence and persistence of AF.3 These conditions are associated with enhanced myocardial tissue inflammation and atrial remodelling, which may serve as a substrate for the development of AF.2 Some observational data suggest that therapy with statins in patients with coronary artery disease and AF is associated with an increase in collagen degradation and an attenuation of inflammation independently of cholesterol lowering.4 In addition, elevated C-reactive protein (CRP) levels have been shown to be independently associated with an increased risk of the development or recurrence of AF.5 The capacity of statins to reduce inflammation and CRP levels is well established,6,7 and this may also explain a potentially beneficial effect of statins against AF, although this has not been specifically demonstrated.
Risk factors for atherogenesis such as age, obesity and hypertension have been associated with an increased risk of AF,1 suggesting an association between AF and atherosclerotic vascular disease. Statins are known to improve lipid abnormalities. Whether statins have a protective role against AF development through anti-atherogenic properties is not established.
There is evidence suggesting an association between AF and enhanced renin–angiotensin system activity. Angiotensin II has a growth-enhancing effect on cardiac myocytes, vascular smooth-muscle cells and fibroblasts, thus resulting in remodelling and fibrosis of the atria that could also provide a potential arrhythmogenic substrate for the development of AF,8 and the inhibition of the renin angiotensin system may decrease the incidence of AF.9
There is also evidence suggesting an interaction between dyslipidaemia and the renin angiotensin system activity.10 Statins decrease both cholesterol levels and oxidative stress,11 and may downregulate the renin angiotensin system. This mechanism may explain a possible antiarrhythmic effect of statins against AF.10 Finally, it has been suggested that a modulation of the autonomic nervous system by statins may have a protective role against AF in post-operative patients with enhanced sympathetic activity.12
Clinical and Experimental Evidence
Several clinical studies have assessed the role of statins in protecting against AF. Observational studies provided initial evidence supporting a protective role of statins against AF. Young-Xu et al. were the first to demonstrate that statin therapy was associated with a 52% reduced incidence of AF in 449 patients with chronic stable coronary artery disease who were followed for an average of five years.13 This benefit was independent of cholesterol lowering, whereas there was a dose–response relationship between the length of statin therapy and a reduction in AF. Additionally, statin-induced reduction in the incidence of AF remained significant after adjustment for many confounding factors. Two subsequent observational studies in different populations gave similar results.14,15 Their main limitation was that these were not randomised studies.
Some well-designed, randomised studies have since been performed evaluating the protective role of statins against AF. A meta-analysis of all randomised controlled trials that compared statins with placebo or a control treatment was recently published by our group.16 We looked for randomised controlled trials that met the following criteria: direct comparison between statin and control treatment or placebo regardless of the background therapy in either group; incidence or recurrence of AF as a specified event, although not necessarily a primary end-point; and follow-up of at least three weeks. Six randomised studies of 3,557 patients in sinus rhythm were included in the analysis.17–22 Three studies investigated the use of statins in patients with a history of paroxysmal AF (n=1) or persistent AF undergoing electrical cardioversion (n=2) and three in primary prevention of AF in patients undergoing cardiac surgery (post-operative AF, n=2) or after acute coronary syndrome (new-onset AF, n=1).
Table 1 summarises the characteristics of the six trials. The six eligible trials included 1,542 patients randomised to statins and 1,559 patients randomised to placebo or control regimens. Atorvastatin was used in five of the six studies, but intervention doses were variable. Comparisons were made with placebo (n=4) or a control regimen (n=2).
Follow-up durations ranged from three to 26 weeks. Coronary artery disease was present in 3,306/3,557 patients (93%). When the Myocardial Ischemia Reduction with Aggressive Cholesterol Lowering (MIRACL) study was removed, coronary artery disease was present in 219/470 of the patients (47%). Incidence or recurrence of AF occurred in 386 patients: 165/1,775 in patients treated with statin versus 221/1,782 in controls. Overall, the use of statins was associated with a significantly decreased risk of the recurrence of AF compared with control (odds ratio [OR] 0.39, 95% confidence interval [CI] 0.18–0.85; p=0.02) (see Table 2). The benefit of statin therapy seemed to be more marked in secondary prevention of AF (OR 0.33; p=0.06) than for new-onset or post-operative AF (OR 0.60; p=0.23). When atorvastatin was considered alone, the benefit was higher (OR 0.30; p=0.01 on the end-point of both incidence or recurrence of AF). Results were similar when ORs were calculated after exclusion of the MIRACL study or when studies with the lowest-quality score were removed. Thus, our analysis suggests that the use of statins was significantly associated with a decreased risk of incidence or recurrence of AF in patients in sinus rhythm with a history of previous AF undergoing cardiac surgery or after acute coronary syndrome. This beneficial effect appeared more marked in the prevention of AF recurrences than in the primary prevention of AF, although this is not certain as none of the subset analyses was statistically significant. The lower number of patients with new-onset AF or post-operative AF may explain the lack of significance for this subgroup.
Duration of follow-up in the six studies was variable, and may seem relatively short. However, different types of AF have varying times to development or onset. In each study patients were appropriately monitored based on the type of AF they had. Recurrences of paroxysmal AF or AF after cardioversion frequently occur within the first month, and all of the patients with recurrent AF included in our analysis had a follow-up period longer than one month (six weeks to six months). Post-operative AF patients were followed for at least three days and for up to 30 days.23
The MIRACL study was the only trial not to show a clear reduction in AF with atorvastatin use, particularly in the subgroup of patients in whom new-onset AF was analysed with a relatively short follow-up of 16 weeks. This shorter duration of follow-up (considering new-onset AF) may explain why a beneficial effect against AF was not observed with statins in the MIRACL study. This relatively inadequate follow-up duration may also explain the lack of benefit of statin use on the primary prevention of AF (post-operative AF or new-onset AF) since these results were essentially driven by the MIRACL study.
Effect of Statins on Atrial Fibrillation – Possibly Not Dose-dependent or Related to Low-density Lipoprotein Decrease
Our meta-analysis ignored varying doses of statins and durations of therapy, as did most of the meta-analyses. We were not able to assess the degree of low-density lipoprotein (LDL) lowering versus incidence or recurrence of AF as it was done with other events with statin therapy.24 We cannot determine from our analysis whether the benefit was seen because some type or dose of statins were used or because low LDL levels were achieved. It was not established whether a patient achieved a certain goal of LDL (<100 or <70mg/dl) using moderate-dose statin or whether outcomes would be better if a higher-dose statin was used. As populations were different, we think that, in our meta-analysis, it was rather inappropriate to compare the OR in each trial and draw precise conclusions on dose effect. The benefit against AF did not seem to be clearly related to statin dose, particularly for atorvastatin use as the OR was not lower in the Atorvastatin for Reduction of Myocardial Dysrhythmia After Cardiac Surgery (ARMYDA-3) and the MIRACL studies, in which high doses of atorvastatin were used (40 and 80mg per day, respectively).
In order to elucidate whether absolute reduction in LDL cholesterol might be related to benefit against arrhythmia, we performed a secondary analysis on these data. Equivalence doses of the statins in the six randomised trials were calculated according to the meta-analysis published by Law.25 There was no significant correlation between OR and statin dose used (rho=0.73; p=0.10), equivalent atorvastatin dose used (rho=-0.03; p=0.95) or decrease in LDL during follow-up (rho=0.00; p=0.99). The lack of relationship between this effect and an increased dose of statin or LDL decrease rather suggests a benefit beyond the lipid-lowering effect.
An important part of the answer to this question has probably been found in the work by McLean et al.26 They analysed two large, randomised trials, the Pravastatin or Atorvastatin Evaluation and Infection Therapy – Thrombolysis in Myocardial Infarction 22 (PROVE IT-TIMI 22) study and phase Z of the A to Z Trial, which compared lower- and higher-intensity statin therapy among 8,659 patients to evaluate whether higher-intensity statin therapy lowered the risk of AF onset during two years of follow-up after acute coronary syndrome. The hypothesis was that higher-intensity statin therapy would decrease the risk of AF compared with lower-intensity statin therapy. From each trial patients experiencing the onset of AF during follow-up were identified from the adverse event reports. Neither study showed a decreased AF risk with higher-dose statin. In PROVE IT-TIMI 22, 2.9 versus 3.3% in the high- versus standard-dose statin therapy, respectively, experienced the onset of AF over two years (OR 0.86; p=0.41). In the A to Z Trial, rates were 1.6 versus 0.99%, respectively (OR 1.58; p=0.096). Thus, randomised comparison found that higher-dose statin therapy did not reduce the short-term incidence of AF among patients after acute coronary syndromes compared with standard-dose statin treatment. Similarly, we were not able to establish a clear relationship between the decrease in CRP with statin use and the protective benefit obtained against AF.16 In contrast, in both trials analysed by McLean, CRP levels (plasma or serum) tended to be higher among patients experiencing the onset of AF.26
Further Considerations on Statins and Atrial Fibrillation
A beneficial effect on AF was not found in the only randomised study performed with pravastatin, in contrast to those in which atorvastatin was used. Whether the benefit obtained against AF, at least to some extent, in almost all of the populations studied with atorvastatin may be obtained with other statins is unknown. A trial comparing 10mg of rosuvastatin with placebo in heart failure of ischaemic aetiology, the Controlled Rosuvastatin Multinational Study in Heart Failure (CORONA) study, has recently been published.27 AF is a common event in these patients. This trial should provide further information if results on the end-point of incidence of AF become available. Results about statins and AF from new randomised studies will be difficult to obtain in the particular population of patients with coronary artery disease as it seems unethical to build a study including a control arm without any statin in these patients.
Patients with coronary heart disease are currently treated with statins in most cases. In contrast, but interestingly, it remains to be seen whether statins provide benefits to patients with AF without any type of established atherosclerotic disease or with a low risk of atherogenesis. There is no doubt that mechanisms of AF vary in different groups of patients. The benefit of intervention therapies may be due to different protective effects, and results cannot be directly extrapolated to specific clinical settings. Significant heterogeneity found in OR calculations in our meta-analysis may also reflect the heterogeneity of different clinical settings included in the study.
The use of statins is likely to be associated with a decreased risk of incidence or recurrence of AF in patients in sinus rhythm with a history of previous AF, in those undergoing cardiac surgery or after acute coronary syndrome. It is very possible that the antiarrhythmic effect of statins in AF is not dose-dependent or related to LDL decrease. These results provide some evidence of the benefit of statins beyond their lipid-lowering activity. However, large-scale prospective, randomised clinical trials are still needed to establish whether statins bring a similar benefit and are an appropriate therapeutic option for all patients for the management of AF.