Introduction
According to the European Society of Cardiology, the management of atrial fibrillation (AF) can be summarised by A for anticoagulation, B for better symptom control and C for management of comorbidities and adverse lifestyles. Unhealthy lifestyles and pathological states likely to trigger AF should be managed as soon as possible to avert the development of the arrhythmia, and once the arrhythmia has been diagnosed, to slow or prevent its progression. Anticoagulation should be properly initiated in patients with the arrhythmia at risk of AF-related stroke. Symptoms may be managed by controlling the ventricular rate at a physiological level usually slightly higher than normal sinus rhythm because of the haemodynamic disadvantage associated with AF, or by converting AF to sinus rhythm and preventing its recurrence. The latter can be achieved by using antiarrhythmic drugs or the ablation of cardiac tissue to isolate the triggers of the arrhythmia or by modifying the substrate to prevent its maintenance.
Rate versus rhythm control
Between 2001 and 2009, seven so-called rate versus rhythm control randomised trials were reported. The largest of these was the AFFIRM (Atrial Fibrillation Follow-up Investigation of Rhythm Management) trial [1] and the second largest was AF-CHF (Atrial Fibrillation – Congestive Heart Failure) [2], a similar study confined to patients with heart failure (Table 1). With minor exceptions, all but one of these trials showed that rhythm control was not superior to rate control. This applied not only to the primary endpoint, including all-cause mortality, but also to an extensive list of secondary endpoints including quality of life. However, one result stood out, rhythm control required more extensive utilisation of healthcare resources.
Table 1. Atrial fibrillation rate versus rhythm control strategy trials.
Trial | No. of patients |
Mean age (years) |
Follow-up (years) |
Outcome |
AFFIRM (2002) AF + risk factor for stroke |
4060 | 69.7 | 3.5 | No difference on overall mortality or quality of life. Increased hospitalisation in rhythm control group |
PIAF 221 (2001) Persistent AF + symptoms |
252 | 61 | 1 | No difference in symptoms or quality of life; Increased hospitalisation in rhythm control group; Increased walking distance in rhythm control group |
RACE (2002) Persistent AF + flutter post cardioversion |
522 | 68 | 2.3 | No difference in symptoms or quality of life; Increased hospitalisation in rhythm control group; Increased walking distance in rhythm control group |
STAF (2001) Persistent AF + symptoms + left ventricular ejection fraction 45% |
200 | 65 | 1.6 | No difference in composite endpoints (death, cerebrovascular event, systemic embolisation, cardiopulmonary resuscitation); Increased hospitalisation in rhythm control group |
HOT-CAFE (2003) First clinically overt persistent AF |
205 | 60.8 | 1.7 | No difference in composite endpoints of death, thromboembolic events, intracranial or major haemorrhage |
AF-CHF (2008) AF + symptomatic HF + left ventricular ejection fraction <35% |
1376 | 66 | 3.1 | No difference in cardiovascular deaths |
J-RHYTHM (2009) | 823 | 65 | 1.6 | 1° composite endpoint (ACM, symptomatic cerebral infarction, systemic embolism, major bleeding, hospitalisation for HF, or physical/psychological disability requiring alteration of treatment strategy) occurred 22.0% less often in patients assigned to rhythm control (P=0.0128) |
AFFIRM: Atrial Fibrillation Follow-up Investigation of Rhythm Management; AF-CHF: Atrial Fibrillation and Congestive Heart Failure; HOT-CAFÉ: How to Treat Chronic Atrial Fibrillation; J-Rhythm: Japanese RHYTHM management trial for atrial fibrillation; PIAF: Pharmacological Intervention in Atrial Fibrillation; RACE: RAte Control versus Electrical cardioversion for persistent atrial fibrillation; STAF: Strategies of Treatment of Atrial Fibrillation
These unexpected results occurred despite sinus rhythm being successfully achieved in far more patients assigned to rhythm control than those treated with rate control. These largely consistent results led to professional society guidelines which recommended rate control as the first-line therapy for recurrent forms of AF (paroxysmal and persistent AF). Rhythm control was relegated to the management of patients whose symptoms remained troublesome despite optimal rate control. To a large extent, this continues to be the primary recommendation for rhythm management in patients with non-permanent AF [3].
To many physicians, especially those specialising in cardiac rhythm management, these results were counterintuitive. Why was normal sinus rhythm not better than AF, with its haemodynamic and pathophysiological consequences? Explanations were sought for these results (Table 2). Amongst these, the toxic effects of antiarrhythmic drugs and the failure to anticoagulate appropriately those who had been successfully restored to sinus rhythm were the two most popular. The drugs that were available at the start of the century were demonstrably effective at maintaining sinus rhythm but also caused ventricular proarrhythmia and in some instances were associated with higher mortality in AF populations.
Table 2. Possible explanation for results of the “rate versus rhythm” control strategy trials.
Toxic effects of antiarrhythmic drugs | Less aggressive management of AF with available drug therapy (to reduce AF burden) |
Abandonment of oral anticoagulation after successful rhythm control | Antiarrhythmic drugs with proven value in reducing “hard” cardiovascular outcomes were not available |
Less effective treatment of underlying comorbidities | Left atrial ablation was not a therapeutic option |
Excess non-cardiovascular deaths in rhythm control arm | Relatively short duration of clinical trials |
AF: atrial fibrillation
From a practical perspective, randomised controlled trials can only be conducted over a relatively short period of time. Funding for longer studies is much more difficult to obtain. As time goes by, patients randomised to a particular therapy may often abandon the therapy or crossover to the trial alternative. In any event, the power of the trial tends to wane over time unless trial discipline can be strictly maintained. A short duration trial may also be insensitive for the detection of events that depend on the dynamic and progressive nature of AF. Registries have provided some longer-term data about mortality and stroke prevention. Some suggest that rhythm control may not be more effective than rate control over the short term but may become more effective over the long term [4]. However, registry data is no substitute for strictly controlled data acquired from randomised populations.
Clinical trials have not been designed to assess the fear and anxiety suffered by patients with AF, who are vulnerable to stroke and heart failure and to the haemorrhagic effects of anticoagulants. Quality of life instruments probably do not detect these adverse, often subconscious threats, which may develop only as age and risk increase.
Antiarrhythmic drug development
In the last decade, a range of “upstream therapies” have emerged that appear to substantially reduce the likelihood of the development or recurrence of AF. Amongst these are mineralocorticoid receptor antagonists and sodium glucose cotransporter 2 inhibitors. Few new antiarrhythmic drugs have emerged, with the exception of dronedarone, which was introduced following extensive trials of its antiarrhythmic efficacy in AF and the ATHENA trial (A placebo-controlled, double-blind, parallel arm Trial to assess the efficacy of dronedarone 400 mg b.i.d. for the prevention of cardiovascular hospitalisation or death from any cause in patiENts with Atrial fibrillation/atrial flutter), which was an important randomised, placebo-controlled trial in a high-risk AF population [5]. The antiarrhythmic efficacy of dronedarone is less than that of amiodarone but is similar to that of sotalol or flecainide. Its major advantage is that in the ATHENA trial it reduced the composite endpoint of all-cause mortality and unexpected cardiovascular hospitalisation by 26% and cardiovascular mortality by 29%. The secondary endpoints of acute coronary syndrome and stroke (post hoc) were also reduced. As yet, dronedarone is the only antiarrhythmic agent which has been shown to reduce serious cardiovascular events in the treatment of patients with recurrent AF.
Other analyses of ATHENA have looked at patients with heart failure [6]. The all-cause mortality and unexpected cardiovascular hospital admission endpoint was similarly increased in patients with heart failure with reduced or preserved left ventricular ejection fraction (LVEF) compared with patients who did not have heart failure. These data are encouraging, and the reduction in the primary endpoint in heart failure patients with preserved ejection fraction in this study constitutes the first data supporting the efficacy and safety of an antiarrhythmic drug in this population.
Antiarrhythmic interventions
There is now an alternative to using antiarrhythmic drugs for rhythm control. Left atrial catheter ablation, specifically pulmonary vein isolation, has proven to be an effective antiarrhythmic intervention. Initially, the effects were not durable, but increasingly, as the technology for left atrial ablation has improved, the procedure is now much more effective over the long term. Multiple trials have compared antiarrhythmic drug therapy with pulmonary vein isolation in drug-naïve patients. They have consistently demonstrated the superiority of catheter ablation [7]. However, the largest of the ablation studies (CABANA: Catheter ABlation vs. ANtiarrhythmic Drug Therapy for Atrial Fibrillation) did not demonstrate that ablation reduced all-cause mortality or a composite of serious adverse cardiac events when assessed using the conventional intention to treat analysis. Failure to do this is probably due to multiple crossovers from the drug treatment arm to the ablation arm of the trial since “per protocol” or “as-treated” analyses were significantly in favour of ablation treatment [8]. However, these analyses did not compare the groups according to the randomisation.
After successful catheter ablation, it is not unusual for episodes of AF to recur, but the recurrences are usually brief and often not symptomatic. The total burden of AF post-ablation is often so low (reduced by 50% or to less than 1%) that anticoagulation may not be warranted, particularly in those with relatively low thromboembolic risk. The low burden of AF may also reduce the development or worsening of heart failure.
Trials comparing left atrial ablation with antiarrhythmic drugs in patients with heart failure have demonstrated less recurrent AF in those randomised to receive ablation. This has been associated with an improvement in left ventricular ejection fraction and improved exercise tolerance. The CASTLE-AF trial (Catheter Ablation vs. Standard conventional Treatment in patients with LEft ventricular dysfunction and Atrial Fibrillation) randomised patients with heart failure to left atrial ablation or antiarrhythmic drug therapy. It demonstrated that left atrial ablation reduced all-cause mortality (hazard ratio [HR]: 0.53; 95% CI: 0.32-0.86, p=0.01) and hospitalisation for heart failure [9]. The advantage of ablation was mostly in patients with functional New York Heart Association (NYHA) Class I-II heart failure or LVEF > 25% [10]. The trial was compromised by numerous crossovers and a potential lack of generalisability because only 397 patients were randomised out of 3,013 who were screened for inclusion.
A series of five smaller trials were combined with the results of CASTLE-AF in a meta-analysis of ablation versus antiarrhythmic agents for the management of AF in heart failure patients. Altogether there were 775 patients, and all-cause mortality and heart failure hospitalisations were both significantly lower in those treated with ablation (risk ratio [RR]: 0.52 [95% CI: 0.33 to 0.81] and RR: 0.60 [CI: 0.39 to 0.93], respectively). The ejection fraction was increased on average by 7%, the 6-minute walk by 21 meters, and peak oxygen consumption by 3.17mL/kg per minute [11].
There were 778 (35%) patients with NYHA class ≥2 in the CABANA trial. The recurrence of AF was much reduced (adjusted HR [aHR]: 0.56 [95% CI: 0.42-0.74]) in the ablation-treated patients and, in an intention to treat analysis, the primary endpoint (aHR: 0.64 [95% CI: 0.41-0.99]) and all-cause mortality (aHR: 0.57 [0.33-0.96]) were also significantly reduced in this subset of patients. Much of the advantage of ablation was seen in patients with a preserved ejection fraction > 50% [12].
Early rhythm control
Using ablation for patients with AF as first-line therapy seems to be particularly effective and a meta-analysis of trials where the diagnosis-to-ablation time was known showed quite clearly that results (less recurrence of AF) were much better in those who underwent ablation less than one year after the initial diagnosis [13]. Clinical AF tends to progress over time from being paroxysmal to having more persistent forms, and the burden of AF increases substantially as the substrate matures and arrhythmia becomes more substrate- and less trigger-dependent. Although ablation may not completely prevent recurrences of AF, it certainly reduces its progression, minimises associated symptoms, and improves quality of life [14].
With many of these thoughts in mind, the EAST AFNET 4 (Early Treatment of Atrial Fibrillation for Stroke Prevention Trial) trial was designed (Figure 1) [ ]. Patients at risk of cardiovascular events (CHA2DS2-VASc ≥2) with a history of AF of less than one year were enrolled and randomised to early rhythm control (ERC) with ablation, antiarrhythmic drugs, or both, or guideline-mandated rate control with rhythm control confined to those who remained symptomatic despite adequate rate control (usual care). The primary endpoint was a composite of cardiovascular death, stroke or TIA, and heart failure or acute coronary syndrome hospitalisation. After almost 2,800 patients had been randomised (median duration of AF of 36 days and 38% with their first AF episode, 30% with no AF-related symptoms, and 28% with stable heart failure), the trial was halted by the data monitoring committee because those managed by the ERC appeared to do better than those undergoing usual care. The primary endpoint was reduced by 21% (HR: 0.79, 95% CI: 0.66-0.94, p=0.004). Compared with usual care, ERC significantly reduced cardiovascular death and stroke, by 35% reduction and 28% respectively.
Figure 1. The design of the EAST-AFNET 4 trial of early rhythm control versus usual care.
ACS: acute coronary syndrome; AF: atrial fibrillation; CHA2DS2-VASc: congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, stroke, vascular disease, age 65-74 years, sex category (female); CHF: congestive heart failure; CV: cardiovascular; TIA: transient ischaemic attack
Similar advantageous results from the ERC were seen in patients with and without heart failure [16], those with and without symptoms [17], and those with paroxysmal or persistent AF. The results in women were comparable to those in men. The demonstrable advantage was only documented in the group with CHA2DS2-VASc ≥4, probably because major events are relatively rare in those with less risk, particularly in the few years following initial diagnosis. The results do, however, reinforce the advantages of the ERC guidelines in patients with more severe disease.
EAST-AFNET 4 results differ substantially from those of the rate versus rhythm control trials, probably because of the early institution of rhythm control, the careful attention to the treatment of comorbidities and thromboembolic risk, the availability of an antiarrhythmic agent effective in reducing major cardiovascular events, the introduction of highly effective left atrial ablation treatment, and allowing usual care as the comparator, which did not impair the enrolment of patients who might have needed rhythm control at a later stage.
A reanalysis of the AFFIRM trial did not find any support for the result of EAST-AFNET 4, but that trial was conducted in a bygone era. Observational data from very large administrative healthcare databases in the United States, Taiwan [18], and Korea [19] have all been very consistent with EAST-AFNET 4 results, and the raw data from a small ESC European registry were also numerically similar, though after multivariate adjustment for potential confounders they unsurprisingly lost significance [20].
Figure 2. ESC guidelines for identification of patients who may be better managed with rhythm control.
AF: atrial fibrillation; LAVI: left atrial volume index
Conclusions
Now that so much more data exist to inform the rate versus rhythm debate, it is time to revisit our guidelines [21]. The ERC significantly outperformed the usual “guideline-mandated” care. The underlying concept that symptoms should trigger the “escalation” from rate to rhythm control appears to be ill-founded, given how well the asymptomatic patients responded to the ERC. We certainly need more randomised trial data to turn our guidelines upside down or to disregard them, although many now believe that the restitution and maintenance of sinus rhythm is a worthy clinical objective. Indeed, the 2020 ESC Atrial Fibrillation guidelines [3] already concede that there is a wide range of patients, including those exercising patient choice, for whom rhythm control may be preferred (Figure 2). Existing data will inevitably colour our approach to this condition, which has potentially disastrous consequences and which causes so much anxiety and distress. Early rhythm control may eventually be offered to the many rather than the few, and rate control alone could become an end-stage strategy, only for those who cannot maintain sinus rhythm. In the meantime, practitioners should think carefully about which of their patients should be considered for early rhythm control.
Take-home messages
- Previous trials evaluating the management of AF concluded that rhythm control was not superior to rate control, and that rate control consumed fewer health care resources.
- Thereafter, guidelines recommended first-line management of recurrent episodes of AF using first line rate control.
- New approaches to rhythm control have shown that major cardiovascular outcomes can be reduced using left atrial ablation or dronedarone, neither of which were available when rate versus rhythm control strategy trials were undertaken.
- Early rhythm control been shown to be a better strategy than guideline-mandated rate control.
- There is now a change in the clinic, already anticipated in the ESC guidelines, to offer rhythm control more widely and at an early stage after initial diagnosis of AF.