Ablation for persistent atrial fibrillation: do we know who to turn away?

This editorial refers to ‘Predictors of recurrence following radiofrequency ablation for persistent atrial fibrillation’ by J.W. McCready et al., on page 355. The use of catheter-based techniques for the treatment of atrial fibrillation (AF) has grown rapidly over the last decade. Once-novel strategies, focusing principally on circumferential ablation to isolate pulmonary vein (PV) ostia, have become commonplace. As ablative therapy for AF has gained wider acceptance, the indications and target patient populations referred for AF ablation have expanded accordingly. The role of ablation for patients with drug-refractory, symptomatic paroxysmal AF is generally agreed upon. Less is known, though, about the long-term success of catheter ablation in patients with persistent (and particularly long-standing persistent) AF. Large trials which have included patients with long-standing persistent AF suggest that the rate of atrial arrhythmia recurrence in this group is high. Identifying predictors of recurrence in this patient population—particularly those that pinpoint patients for whom AF ablation is truly futile—is of obvious clinical importance. McCready et al.1 provide data on predictors of AF recurrence following ablation in patients with persistent AF. The focus of the study was long-standing persistent AF, since the large majority of the patients studied (88%) had a continuous AF duration of >1 year leading up to ablation. The authors collected data in 191 patients undergoing a total of 292 procedures. The ablation strategy used was left to the discretion of the ablating physician; typically, circumferential PV isolation was performed (100% of patients) in conjunction with a roof line (71%), and occasionally with a mitral isthmus line (44%) and complex fractionated electrogram ablation (42%). It is important to note that ablation until return of sinus rhythm was not the endpoint of the procedure; restoration of sinus rhythm during ablation was only observed in 12% of the patients. Patients were followed for an average of 13 months, with a variety of types of follow-up monitoring including clinical assessment, electrocardiograms, Holter monitoring, and device interrogation. Procedural failure was defined as recurrent AF or atrial tachycardia (AT) >30 s duration, after a 3 m blanking period on or off an antiarrhythmic medication. Sixty-seven patients underwent two or more procedures. The results presented by McCready et al. are a stark reminder of the difficulties posed by patients with long-standing persistent AF. Although PVs were isolated in 100% of the cases, operators did not succeed in terminating AF during ablation in 88% of the patients; rather, patients with AF or AT at the conclusion of the case were cardioverted. During the follow-up period, single-procedure success occurred in 32% of the patients over 13 months. Among patients who were categorized as having a successful outcome, 30% were still taking antiarrhythmic medications. Thus, the antiarrhythmic-free single-procedure success rate in this patient population can be predicted to be <32%. Overall success (i.e. after up to four ablation procedures and on or off antiarrhythmic drug therapy) was 64% over the same period. Importantly, the incidence of major complications was 6.2%, including cardiac tamponade in 2.4% (two patients requiring surgical repair), stroke in 1%, phrenic nerve paralysis in 0.7%, and one patient with restrictive pericarditis requiring a surgical pericardectomy. The authors found that left atrial (LA) size, measured by transthoracic echocardiography, was the sole univariate and multivariate predictor of single-procedure success. Left atrial size was also the principal predictor of overall success. Atrial fibrillation duration, procedural strategy, and other variables that one might expect to predict outcome did not do so in this series of patients. On the basis of these findings, the authors performed receiver operator characteristic analysis and found that an LA size of <43 mm gave greatest accuracy in predicting outcome. McCready et al. are to be congratulated for sharing their data with us for careful consideration. There are many important aspects of this study. First, the very low single-procedure success rate (<32%) is sobering. This is particularly the case when one considers that the follow-up duration in this study was relatively short (13.5 ± 9.4 months). One can only wonder what the single-procedure antiarrhythmic-free success rate might be after 3 or more years of follow-up. Secondly, the incidence of major complications is high (6.2%). Although there were no deaths, three patients in this series required surgical intervention because of a complication and two sustained a stroke. This rate of complication is consistent with our impression that catheter ablation of long-standing persistent AF, especially when it involves more aggressive ablation strategies including linear ablation and ablation of complex fractionated electrograms, is associated with a significantly higher complication rate than in patients with paroxysmal AF. And thirdly, the authors identify LA size as assessed by conventional echocardiography as an important predictor of success in this series of patients. This makes physiological sense. Remodelling of the atria during persistent AF is a time-dependent process, and typically results in enlargement of the LA chamber. It is for this reason that many studies of AF ablation have excluded patients with a markedly dilated LA, defined more commonly as an LA diameter >5 cm. A review of the literature reveals that some studies have identified LA size as the best predictor of success, whereas others have highlighted the importance of AF duration. Recently, for example, Takahashi et al.2 investigated clinical characteristics of patients with persistent AF undergoing successful ablation, in a study population that was similar (but not identical, with a shorter duration of continuous AF and slightly smaller average LA size) to the group presented by McCready et al. The Japanese group found that AF duration, but not LA size, predicted recurrence of AF and AT after ablation. The Bordeaux group, investigating clinical characteristics linked to successful termination of AF during the ablative procedure (which in turn predicts long-term success), has reported that both LA size and duration are predictors of success.3 Clearly, there are published data to support what we believe is commonly recognized by experienced clinical practitioners. Patients with large atria and/or who have been in continuous AF for very long periods have high recurrence rates of atrial arrhythmias post-ablation. At the end of the day, what do the results of this study really mean? In our mind, it would be a mistake to use LA diameter of >4.3 cm to exclude patients from undergoing AF ablation. For one, LA size as assessed with standard echocardiographic techniques is known to correlate poorly with true atrial volume. More importantly, we need to consider whether LA size provides a clinically applicable test that will significantly change current practice.4 Although the identification of clinical features that predict procedural success or failure is laudable, we believe that the patients who need identification the most are those for whom the procedure is overwhelmingly likely to be futile. In the McCready cohort, even those patients with LA size >43 mm (127/191 patients) had an AF-free survival of nearly 50% at 24 months. We suspect that these sorts of results, while certainly not ideal, would be acceptable by many patients with highly symptomatic AF refractory to multiple antiarrhythmic medications. The McCready report fails to identify a group that had strikingly poor long-term freedom from AF. Quite the opposite, the study found that there was a plateau effect in AF-free survival for patients with LA size >46 mm. What is still needed, then, is a metric (or more realistically, a risk score system) that can identify patients for whom the risks of AF ablation all too clearly outweigh the benefits. It is notable that preliminary evidence from one centre suggests that the extent of atrial fibrosis as detected with magnetic resonance imaging may be an even better predictor of outcome than LA size.5 It should be clear to all in this field that more research needs to be done on the techniques and outcomes of AF ablation in patients with long-standing persistent AF. Hopefully, our joint efforts will answer our many questions over the next several years. Conflict of interest: none declared.