Pulsed field ablation (PFA) proved a better first-line treatment for persistent atrial fibrillation (AF) than did antiarrhythmic drug therapy, with less recurrence and similar adverse event risk, the AVANT GUARD randomized trial showed.

Treatment success — freedom from recurrent atrial arrhythmias, repeat ablation, or need for antiarrhythmic drugs from 90 days to 12 months after ablation or for amiodarone use at any time — occurred in 56% of the PFA group compared with 30% in the antiarrhythmic drug group. The hazard ratio for initial technical failure or 12-month treatment failure was 0.46 (95% CI, 0.33-0.65; P < .001).

Device- or procedure-related serious adverse events occurred at the same 5.1% rate in both groups, Oussama M. Wazni, MD, of the Cleveland Clinic, Cleveland, reported at the Heart Rhythm Society (HRS) 2026 in Chicago. The findings were also published in The New England Journal of Medicine.

“Everybody knows that catheter ablation is superior to antiarrhythmic drug therapy when drugs fail; that’s a given,” Wazni said at the late-breaking clinical trial session. The STOP AF First and EARLY-AF trials also showed catheter ablation was better than antiarrhythmic drugs as first-line treatment in paroxysmal AF.

While guidelines recommend a trial of antiarrhythmic drugs before catheter ablation for persistent AF, Wazni suggested the AVANT GUARD trail might change that.

“These results support consideration of PFA as an initial rhythm control strategy for selected patients with persistent atrial fibrillation,” he concluded.

Opportunity for Improvement

Indeed, catheter ablation for persistent AF has been on the rise in US practice in the past few years, such that an estimated one third of all catheter ablation procedures for AF in the US will be in the type of patient population enrolled in the AVANT GUARD trail, said Suneet Mittal, MD, of the Valley Health System in Paramus, New Jersey.

The two strategies performed as might be expected from prior separate studies in a similar population, but there’s an ongoing opportunity for improvement, said Mittal, who served as a discussant for the trial. He pointed to the 35% of ablation patients in the AVANT GUARD trail, compared with 53% in the medication group, who were left with more than a 0.1% burden of AF after the blanking period, which he highlighted as “a threshold that’s previously been demonstrated to be associated with higher use of healthcare utilization.”

The trial enrolled 440 patients with previously untreated, persistent AF. Of them, 30 were used for roll-in familiarity with the protocol and device. Another 310 were randomly assigned in a 2:1 ratio to receive PFA performed with a pentaspline catheter (Farawave) or to receive antiarrhythmic drug therapy. And 100 patients were assigned to PFA without randomization for safety assessment. All the patients got an insertable cardiac monitor.

The primary outcome measure included both short-term and long-term success. Short-term success was defined as electrical isolation of all the targeted pulmonary veins and the posterior wall of the left atrium by means of only the investigational catheter in the PFA group or freedom from ablation during the 90-day blanking period in the antiarrhythmic drug group. Long-term success was defined by no use of amiodarone at any time.

The driver of the difference between groups was recurrence of asymptomatic atrial tachyarrhythmias lasting for ≥ 1 hour (30% with ablation vs 46% with medication). Ablation also reduced median atrial arrhythmia burden at 1 year by an average of 6.5 percentage points.

The 5.1% rate of the primary safety endpoint fell within the prespecified 8.6% goal. Overall and serious adverse events were similar between groups through 12 months, with no procedure-related death, esophageal injury, pulmonary vein stenosis, atrial esophageal fistula, coronary spasm, hemolysis, or delayed onset vasospasm reported.

Lingering Questions

There were six neurologic events adjudicated as procedure-related stroke, one of which was due to inadvertent introduction of air from the deflectable sheath. While Wazni noted all the events were “clinically minor and none disabling,” the trial was paused for consultation with neurologists. It resumed with a modified protocol to exclude patients with a CHA2DS2-VASc score of ≥ 4 and to require screening for left atrial thrombus within 24 hours before the procedure, uninterrupted anticoagulation for ≥ 4 weeks before the procedure, and a minimum activated clotting time of 350 seconds before ablation. After that, no more strokes or transient ischemic attacks occurred in the ablation group.

During the pause, Wazni said the researchers realized that the patient population was sicker than in prior studies, with an average CHA2DS2-VASc score of 2.7 that subsequently dropped to 2.1 after the protocol change.

“It is possible that because of familiarity with PFA and the lower threshold to ablate, that sicker patients were enrolled in the study,” Wazni said.

However, Mittal expressed concern over the neurologic events. “ I would ask you whether we have really understood this issue completely,” he said at the session.

The ablation protocol called for both pulmonary vein isolation and posterior wall ablation. Mittal questioned whether the addition of posterior wall ablation impacted safety in this patient population.

He cited a recent publication highlighting two different patients from two different centers, one with CHA2DS2-VASc score of only 1 and the other with a score of 2.

“Both with persistent atrial fibrillation, both ablated with this technology, where you see this phenomenon, which appears to be related to posterior wall denudation of the endothelium — something that can only be seen with direct imaging with an intracardiac echo catheter in the left atrium and really not visible from right atrial imaging. So I would say that we have more questions to answer regarding the ultimate safety of these additional ablation sets,” Mittal said.

Limitations included the use of a single ablation technology and a standard lesion plan, so the outcomes may not be generalizable, Wazni said. He also noted the majority of patients had the cardiac monitor inserted at the time of initiation, so change in burden from baseline could not be evaluated. “I know that this is something that’s important for everybody, but this is how we did the study,” he said.

Three-year follow-up is ongoing.

The study was funded by Boston Scientific. Wazni disclosed having relationships with Boston Scientific, Johnson & Johnson, and Medtronic. Mittal disclosed having relationships with ScottCare, Abbott, Biosense Webster, CathVision, CVRx, Haemonetics, Impulse Dynamics, iRhythm Technologies, Philips, Viz.AI, PaceMate, B-Secur, Boston Scientific, Medtronic, BioTelemetry, and Vector Remote Care.

Crystal Phend is an award-winning medical journalist with decades of experience reporting on clinical research and healthcare developments across specialties. When not walking the halls at a medical conference, she can be found at a keyboard in upstate New York.