RF Ablation, Antiarrhythmic Meds Comparable as Initial Therapy of Paroxysmal AF

October 31, 2012

BOSTON — Radiofrequency (RF) ablation as first-line therapy for paroxysmal atrial fibrillation (AF) was no more effective than antiarrhythmic therapy in a two-year trial conducted in Scandinavia [1].

Although the cumulative AF burden was similar with the two management strategies, as was the burden at seven-day Holter monitoring at most of the regular follow-ups, it was significantly lower in the ablation group at the final 24-month Holter period (p=0.007). Ablation also had a significant edge for some secondary end points, such as freedom from symptomatic AF at 24 months (p=0.01). In addition, more than a third of drug-treated patients required crossover to ablation during the follow-up, usually in the first year.

Still, "given the risk of complications with ablation, our data support the current guidelines recommending antiarrhythmic drugs as first-line treatment in most patients with paroxysmal atrial fibrillation," write the authors, led by Dr Jens Cosedis Nielsen (Aarhus University Hospital, Denmark).

Their report on the Medical Antiarrhythmic Treatment or Radiofrequency Ablation in Paroxysmal Atrial Fibrillation (MANTRA-PAF) trial, which randomized about 300 patients, is published in the October 25, 2012 issue of the New England Journal of Medicine.

The group acknowledges that their results apply to patients like those in the trial, "relatively young, symptomatic patients with no major coexisting conditions, but [they] should not be extrapolated to elderly patients, to those with persistent or permanent atrial fibrillation, or to those with severe heart disease."

In an accompanying editorial, Drs William G Stevenson and Christine M Albert (Brigham and Women’s Hospital and Harvard Medical School, Boston, MA) agree [2]. Patients like those in MANTRA-PAF "are good candidates for ablation but represent fewer than a quarter of the patients with atrial fibrillation who are encountered in practice. These results should not be extrapolated to the broader population of patients with atrial fibrillation, most of whom are elderly with coexisting conditions."

"They Set a Very High Bar"

As previously reported by heartwire , a recent international consensus document concluded that ablation is an acceptable first option for patients with symptomatic paroxysmal AF. "The present trial supports this recommendation but did not show the superiority of ablation over drug therapy, and many patients did well with drug therapy. When maintenance of sinus rhythm is desired, either strategy is reasonable," according to the editorialists.

The cochair of that consensus document, Dr Hugh Calkins (Johns Hopkins University School of Medicine, Baltimore, MD), said the current study's results aren't surprising. "But this was a much better-designed study than many," he told heartwire . "They set a very high bar with the seven-day Holter monitoring, and at each of the time points they were able to look at AF burden in every single patient."

In supporting AF ablation as a first-line therapy, the findings indeed support contemporary guidelines on both sides of the Atlantic, Calkins said, with the caveat that "for most patients, a trial of antiarrhythmic drug therapy is a reasonable first step."

On the other hand, said Dr Vivek Y Reddy (Mount Sinai School of Medicine, New York, NY), "I find these results disappointing. They are from a time when the techniques were not fully evolved, and the ablation strategy was not fully evolved to what most of us have been doing for the past four or five years."

In particular, ablation success in the trial was defined by outdated electrographic criteria, he observed. "Nowadays, we always try to achieve pulmonary-vein isolation." Also, MANTRA-PAF allowed linear lesions, whereas today they are avoided. Had contemporary techniques been used, Reddy said, the results of AF ablation would have been better.

Indeed, he noted, RAAFT-2 randomized similar patients to RF ablation or drug therapy but in the former group required pulmonary vein isolation as the procedural end point. In that trial, which has yet to be published, ablation suppressed AF significantly better than antiarrhythmic therapy.

"It is possible that, if ablation had been performed using the current procedural goal of electrical isolation, the outcomes would have been better," agreed editorialists Stevenson and Albert. But, they add, "it is also possible that this approach would have led to more aggressive ablation procedures with a higher risk of complications."

Calkins also observed that the ablation technique in MANTRA-PAF was outmoded. "So the case is not closed, and with better techniques and better approaches we'll keep reexamining this question. But for now, this study is consistent with all the very recent guidelines and consensus documents that have weighed in on this issue."

Primary and Secondary End Points at Odds?

MANTRA-PAF randomized 294 patients younger than 70 with symptomatic paroxysmal AF with no history of antiarrhythmic use who were considered candidates for rhythm-control drug therapy. They received RF ablation or treatment with antiarrhythmic agents of either class IC or class III as initial AF therapy.

In both groups, AF burden--that is, the percentage of time in AF on seven-day Holter monitoring--was significantly improved at all follow-up visits compared with baseline (p<0.001 for all comparisons).

On an intention-to-treat basis, there was no significant difference in the primary end point of cumulative AF burden--that is, the total AF burden at month 3, 6, 9, 12, 18, and 24 combined. With AF burden at each individual Holter session as another primary end point, the groups didn't differ significantly except at month 24, which showed an advantage for RF ablation.

Serious adverse events numbered 20 in the ablation group and 16 in the drug-therapy group and included death in six and four patients, respectively.

Primary and Secondary End Points in MANTRA-PAF

End points

RF ablation, n=146 (%)

Antiarrhythmic drugs, n=148 (%)




AF burden at 24 mo




90th percentile for cumulative burden




Secondary at 24 mo


Symptomatic AF burden




Freedom from any AF




Freedom from symptomatic AF




*Primary end points also included AF burden at 3, 6, 12, and 18 months, for which there were no significant differences between the groups

From the study's intention-to-treat perspective, Reddy observed, the two treatment methods produce relatively similar results. Ablation showed some advantages over drug therapy, "and those are statistically significant differences, but the magnitude of the differences is not so striking. So their clinical significance is questionable."

On the other hand, the crossover rate from antiarrhythmics to ablation was 36%, he noted. "So clearly there are a significant percentage of patients who, even though you're trying to keep them controlled just with drugs, end up switching over to ablation because of an increase in symptoms. So really, you have to look at MANTRA-PAF as a strategy study."

Calkins agreed. "The question isn't whether they get an AF ablation. The question is whether they get an ablation now or whether they get put on a drug to have their AF controlled but get an ablation down the road if indicated."

And that decision about first-line therapy "is very much patient-dependent," he said. "This study helps inform patients' decisions. For those who are reluctant to have a procedure, the study basically says it's very reasonable to go on a drug; there's no need to have the procedure now. Certainly for those who hate drugs, this says it's also reasonable to have an ablation now. But for most patients, given that some who get AF ablation will end up with a terrible complication, it's generally prudent to try a drug first." And that, in his experience, he said, is what most patients choose.

The trial was partly supported by Biosense Webster; Cosedis Nielsen reports serving as an advisory-board member of Sanofi and receiving lecture fees from Biotronik, Medtronic, and St Jude Medical. Disclosures for the coauthors are listed in the paper. Neither Stevenson nor Alpert had relevant disclosures.