The absolute rate is still below 0.5%, in line with prior ablation studies, but the issue warrants monitoring, a researcher says.

Patients with atrial fibrillation (AF) who underwent pulsed-field ablation (PFA) were more likely to have a stroke or TIA than those treated with radiofrequency ablation at a single high-volume center, an observational study shows.

Amid low absolute rates, the risk was roughly fivefold higher with PFA (0.47% vs 0.10%), Enrico Ferro, MD (Beth Israel Deaconess Medical Center, Boston, MA), reported recently at the European Heart Rhythm Association Congress 2026 in Paris, France.

The reason is unclear, with mediation analyses indicating that patient- and operator-level differences, or a learning curve effect, could not explain the finding. In addition, more extensive ablation beyond pulmonary vein isolation (PVI), which was more common in the PFA group, was not a significant risk factor for stroke/TIA, although the low number of events didn’t allow for a definitive answer.

Ferro underscored that the absolute rate of stroke/TIA was low in both groups in the study, telling TCTMD that “the safety of the procedure is overall reassuring and is in line with what we’ve seen so far for AF ablation.”

The difference observed, however, calls for “dedicated surveillance to have informed conversations about risks and benefits of the procedure with patients,” he added, noting that feared complications like atrioesophageal fistula are less frequent with PFA than with thermal ablation modalities.

“But maybe we are trading the safety gain with a new safety signal that ought to be monitored, and stroke could be one of them. There are also reports of vasospasm and sudden death, which, again, are very rare and demand dedicated surveillance,” Ferro said, calling for studies that compare observed rates of events with benchmarks taken from the radiofrequency ablation literature or contemporary PFA performance goals.

Commenting for TCTMD, Wilber Su, MD (Banner – University Medical Center Phoenix, AZ), said he’s not too surprised by the findings because a different type of technology is expected to have a different safety profile. Though PFA carries lower risks of injury to the esophagus, phrenic nerve, and pulmonary veins compared with thermal ablation technologies, it can come with other issues related to differences in sheaths and catheters.

He noted that studies like NEMESIS-PFA have pointed to the rare occurrence of events like myocardial damage, hemolysis, renal dysfunction, and asymptomatic cerebral emboli with PFA.

“It is a trade-off sometimes,” Su said, adding that “the cerebral safety is actually my utmost concern because that’s something we notice with a lot of the PFA trials, from silent cerebral events to stroke events. We know it is not minimal, and we do know that some of them are significantly higher than other modalities, especially with silent cerebral events.”

In general, the higher rate of stroke observed in this study is not “significantly alarming,” Su said, pointing to the low overall rate. Still, he said, cerebral injury is concerning “because at the end of the day, every brain cell matters.”

PFA has been rapidly adopted in recent years, taking on a greater share of ablation procedures compared with older radiofrequency and cryoballoon technologies. This shift was driven by shorter procedure times and a perception of enhanced safety, including reduced risks of serious complications like atrioesophageal fistula, phrenic nerve paralysis, and pulmonary vein stenosis.

But, Ferro told TCTMD, “as we continued to do more of the procedures, we started to understand that perhaps the safety profile was a little bit more nuanced.” He pointed to accumulating data from registries and case reports indicating risks, though low, of events like coronary spasm, phrenic nerve injury, hemolysis, and stroke. In early 2025, he noted, ablation cases involving the Varipulse PFA system (Johnson & Johnson Medtech) were temporarily paused to investigate four neurovascular events that occurred during an external evaluation of the technology.

Ferro added that, at his center, “we also started to notice when doing our own ablations and monitoring under [intracardiac echocardiography] that there was kind of a shower of microbubbles . . . and even though they were not always converting to a neurological event, it made us worried and it made us wonder. Is this just an anecdotal experience or should we more formally study that?”

Comparison Data

For this study, Ferro and his colleagues examined data on 4,221 consecutive AF ablations performed at their center between January 2022 and January 2026—2,144 patients undergoing PFA (mean age 67.8 years; 31.7% women) and 2,077 undergoing radiofrequency ablation (mean age 66.9 years; 30.2% women). In the first half of the study period, nearly all procedures used radiofrequency catheters and in the second half, nearly all used PFA catheters.

Patients treated with radiofrequency ablation were more likely to have undergone a prior AF ablation and to have a history of atrial flutter compared with those treated with PFA. They were also more likely to have NYHA class III/IV heart failure and to be taking antiarrhythmic drugs. Overall comorbidity burden was similar in the two groups.

There were differences in procedural workflows between PFA and radiofrequency ablation. Left atrial dwell time and use of vein of Marshall alcohol infusion or left atrial appendage isolation tended to be greater with radiofrequency ablation, whereas PFA was associated with greater use of ablation beyond PVI, predominantly posterior wall isolation. These differences all may affect stroke rates, Ferro noted.

Follow-up through 30 days was conducted using electronic health records, with strokes adjudicated by two neurologists blinded to procedural data. During that time, the overall rate of stroke/TIA was low, but there were 10 events in the PFA group and two in the radiofrequency ablation group (adjusted risk difference 0.36%; 95% CI 0.03%-0.70%). The median time of stroke/TIA onset was 5.7 days. The events were spread relatively evenly across 12 operators and across PFA platforms.

Mediator analyses were performed to account for differences in procedural workflows between the PFA and radiofrequency ablation groups and did not reveal any potential explanations for the higher stroke risk associated with PFA.

Some possibilities, however, include hydrolysis and microbubble formation and the effects of more extensive ablations in the posterior wall, such as thrombus formation and endothelial disruption, Ferro said.

“The problem is that the event rates were still low enough that the resulting confidence intervals were too wide and it was not possible to make a definitive conclusion,” he said.

Ferro’s team plans on reproducing this study using a large US claims database and artificial intelligence to help adjudicate stroke outcomes. Ferro said they are also working with the US Food and Drug Administration to launch a nationwide postmarket surveillance study on stroke occurring in the context of PFA.

Su said, “the more experience we have with PFA, the more we can learn how to be safer with it. I think these [types of studies] are all good feedback for the next generations of PFA coming down the pipeline.”