By Joshua D. Moss, MD

Associate Professor of Clinical Medicine, Cardiac Electrophysiology, Division of Cardiology, University of California, San Francisco

Dr. Moss reports no financial relationships relevant to this field of study.

SYNOPSIS: Catheter ablation for atrial fibrillation in relatively young men with cardiomyopathy and heart failure is associated with a substantial reduction in arrhythmia burden, improvement in ejection fraction, and reduction in heart failure hospitalizations and mortality compared to medical therapy.

SOURCE: Marrouche NF, Brachmann J, Andresen D, et al. Catheter ablation for atrial fibrillation with heart failure. N Engl J Med 2018;378:417-427.

(Editor’s Note: Late-breaking results of the landmark CASTLE-AF trial were presented at the European Society of Cardiology Congress in August 2017, with full details published in February in the New England Journal of Medicine.)

This prospective, randomized, open-label study at 33 sites in Europe, Australia, and the United States enrolled 398 patients with left ventricular (LV) ejection fraction (EF) ≤ 35%, heart failure, paroxysmal or persistent atrial fibrillation (AF), and “absence of response to, unacceptable side effects from, or unwillingness to take antiarrhythmic drugs.” After a five-week run-in phase to adjust administration of heart failure medications according to guidelines, 363 patients who remained in the trial were randomized to catheter ablation for AF (n = 179) or medical therapy (n = 184), with a composite primary endpoint of death or hospitalization for worsening heart failure.

Catheter ablation was performed by experienced operators and consisted of pulmonary vein isolation (PVI) alone in 48% of patients or PVI plus additional lesions (at the discretion of the operator) in 52% of patients during the initial procedure. Some patients (15.6%) did not undergo any ablation for several reasons and crossed over to medical therapy. After a standard 12-week post-ablation blanking period, 22.5% of patients underwent one additional ablation procedure and 2% underwent two additional procedures. Twenty-three patients were lost to follow-up for the primary endpoint.

Medical therapy was administered based on the 2006 American College of Cardiology/American Heart Association/European Society of Cardiology guidelines. Although rhythm control was encouraged, about 70% of patients were treated using a rate-control strategy, targeting goal heart rates of 60-80 beats per minute at rest and 90-115 beats during moderate exercise. Eighteen patients crossed over to catheter ablation, and 10 were lost to follow-up.

Median age at enrollment was 64 years and median EF was about 32%, with nearly all patients (86%) male and the majority exhibiting New York Heart Association (NYHA) Class II or III heart failure symptoms. Most patients had an implantable cardioverter defibrillator (ICD) implanted before enrollment, and arrhythmia burden was assessed via ICD remote monitoring. Researchers ended the trial early (after 133 primary endpoint events had occurred) when it became apparent that the target of 195 events would not be reached in a reasonable time frame. A modified intention-to-treat analysis was performed, excluding patients who died or were withdrawn during the run-in period, endpoint events during the run-in period, and non-death events during the 12-week post-ablation blanking period (and an equivalent 12-week period after baseline in the medical-therapy group).

Death or hospitalization for heart failure occurred in 28.5% of ablation patients vs. 44.6% in the medical therapy group. Also significantly reduced in the ablation group were death from any cause (13.4% vs. 25.0%), cardiovascular death (11.2% vs. 22.3%), heart failure hospitalization (20.7% vs. 35.9%), and cardiovascular hospitalization (35.8% vs. 48.4%). There was a trend toward reduction in stroke (2.8% vs. 6.0%; P = 0.15). The burden of AF based on memory from the implanted devices was substantially lower in the ablation group; 63.1% of those patients were in sinus rhythm at the five-year follow-up visit with no recurrence since the four-year follow-up visit compared with 21.7% in the medical therapy group. Additionally, median absolute increase in EF was 8.0% in the ablation group compared with 0.2% in the medical therapy group. Serious complication rates were low in the ablation group. Three patients experienced pericardial effusion, one of whom required pericardiocentesis, and three demonstrated bleeding severe enough to require transfusion. There were no phrenic nerve injuries, strokes, or deaths related to the procedure. The authors concluded that catheter ablation for AF should be strongly considered early in the management of patients with heart failure.


The CASTLE-AF trial has quickly generated debate in the cardiology and electrophysiology community. Some point to the impressive improvements in outcomes after ablation (including a 50% relative reduction in death) as confirmation of anecdotal experience; some see an expected extension of results from other trials of catheter ablation for AF in heart failure (including CAMERA-MRI and AATAC); and some consider it a call to arms. Others have urged caution based on the relatively small number of patients, the statistical methods employed and unblinded nature of the trial, a differential loss to follow-up between the groups, and results that simply seem “too good to be true.”

Regardless of the effect size (which was largely unchanged with several alternate analyses presented in the supplementary appendix), it is difficult to argue convincingly that this population of patients did not benefit from the very plausible reduction in AF burden afforded by catheter ablation. At no time did the authors contend that catheter ablation for AF is a “cure,” and their arrhythmia results are in line with clinical experience and prior trials of ablation for AF. After all, of 151 patients in the ablation group who underwent ablation, 75 experienced an adjudicated recurrence of AF during up to five years of follow-up. However, mean burden per patient was 27% of the time at five years in the ablation group vs. 64% of the time in the medical therapy group (median burden 0% vs. 99% of the time). Improvements in heart failure and mortality clearly are plausible biologically with such a significant reduction in arrhythmia burden when not achieved with additional toxic antiarrhythmic drugs. The results also are difficult to “blame” on differential effects of antiarrhythmic drugs, since at the last follow-up with documented medication use, 25% of patients in the ablation group were taking amiodarone compared with 31% in the medical therapy group. The CASTLE-AF trial is not an invitation to sign up every patient with heart failure and AF for an ablation procedure. These results alone cannot necessarily be extrapolated to older patients or even to women. Additionally, it is worth noting that there were more patients with non-ischemic cardiomyopathy in the ablation group (59%) than in the medical therapy group (49%), and a subgroup analysis suggested no benefit of ablation in more severe cardiomyopathy (EF < 25%) or heart failure (NYHA Class III). However, there also was no subgroup in which medical therapy was superior, and rates of serious complications from catheter ablation were low. Thus, the potential benefits of ablation for AF should be strongly considered early in the management of patients with heart failure.