Persistent atrial fibrillation: rate control or rhythm control

Atrial fibrillation is the commonest sustained tachyarrhythmia encountered in clinical practice.1 With an ageing population and improved survival of patients with cardiac disease its prevalence continues to rise.2 It is associated with a doubling of overall morbidity and mortality from cardiovascular disease3 and is the most common cause of embolic stroke.4 Restoring sinus rhythm holds the theoretical advantage of reducing the risk of thromboembolism and need for anticoagulation and improved haemodynamics and quality of life. However, most current anti-arrhythmic drugs have limited efficacy and several side effects. With the use of anti-arrhythmics and serial electrical cardioversion for early relapse up to 53% of patients are in sinus rhythm at one year,5 but only 25% remain so at five years.6 Considerable controversy therefore exists as to whether rhythm or rate control is the more appropriate management for most patients with persistent atrial fibrillation. Five recent trials have looked specifically at this issue.

The PIAF study was the first randomised published study to compare rate versus rhythm control in 252 patients with persistent atrial fibrillation (of < 1 year's duration).7 The investigators reported no difference in the one year primary end point of improvement in symptoms between the two groups, but higher admission rates and more frequent adverse drug effects occurred in the rhythm control group. This study included only symptomatic, younger patients (mean age 60 years) and had a short follow up period (one year).

Similar findings were reported in the equally small (205 patients) HOT CAFÉ study (mean age 61 years).8 Again more patients were admitted to hospital in the rhythm control group (P < 0.0001), despite a measurable improvement in exercise tolerance at one year follow up.

A longer follow up (mean 19.6 months) was employed in the STAF pilot study (in press), although numbers were again small (200 patients with ≤ 2 years' duration, mean age 66 years).9 This study showed that selecting patients with an increased risk of recurrence of atrial fibrillation was, not surprisingly, associated with a low rate of sinus rhythm in the rhythm control group (23% after three years), in spite of up to four cardioversions and up to four anti-arrhythmic drugs for each patient. There was no difference between the two treatment groups in terms of the combined incidence of death, stroke or transient ischaemic attacks, cardiopulmonary resuscitation, and systemic embolism. Again, except for increased admissions in the rhythm control group, there was no difference in secondary end points.

Two larger studies recruited older (and truer to “real life”) patients with longer follow up. The RACE study randomised 522 patients (mean age 68 (SD 9) years) with recurrent persistent atrial fibrillation to either rate control or rhythm control.10 After a mean follow up of 2.3 (0.6) years, no difference was seen (P=0.4) in the primary composite clinical end point of death from cardiovascular causes, heart failure, thromboembolic complications, bleeding, pacemaker implantation, and severe adverse drug effects between the rate (17.2%) and the rhythm control group (22.6%).

The landmark AFFIRM study, the largest to date, also had the longest follow up.11 It randomised 4060 patients (mean age 69.7 years) with a history of atrial fibrillation and increased risk of stroke. It is the only study large enough to have looked at mortality as a separate end point. The primary end point of overall mortality (at five years) was no different between those assigned to rhythm control (356 deaths) and those assigned to rate control (310 deaths; P=0.08), although some separation in the survival curves was evident from two years in favour of the rate control group. Furthermore, there was no difference in the composite secondary end point of death, disabling stroke, disabling anoxic encephalopathy major bleeding, and cardiac arrest. As in other studies, a rhythm control approach was associated with a higher rate of admission and more adverse drug effects, with a greater rate of treatment crossover (27.3% at three years and 37.5% at five years).

In conclusion, although these five randomised trials have compared heterogeneous groups of patients, several consistent messages have emerged for patients for whom cardiologists felt that randomisation to rate or rhythm control was an acceptable strategy. Firstly, with current anti-arrhythmics a rhythm control approach does not lead to an improvement in symptom control or quality of life or a reduction in clinical events in the short to medium term. In fact, in the longer term, mortality may increase. Secondly, maintenance of sinus rhythm remains poor, even with an aggressive strategy combining electrical cardioversion and current anti-arrhythmics. Hence we would advocate long term therapeutic anticoagulation for most patients treated with rhythm control, even if sinus rhythm is achieved in the short term. Thirdly, adverse side effects remain a problem.

The present data are thus consistent and strong enough to promote a rate control approach as the initial strategy for the vast majority of patients with persistent atrial fibrillation. For the minority who remain highly symptomatic aggressive rhythm control with consideration of invasive treatments such as pulmonary vein ablation or improved rate control with atrioventricular nodal ablation with back up ventricular pacing should be considered.