All patients undergoing RF-balloon-based PVI between September 2021 and November 2022 at the University Hospital of Cologne were prospectively included in this single-center registry. Inclusion criteria were symptomatic paroxysmal AF, age > 18 years, and prior written informed consent. Consecutive patients were included, and there were no predefined criteria for RF or cryoballoon use. Patients with prior left atrial ablation, persistent AF, or history of atrial tachycardia (AT) or atrial flutter were excluded. Procedural data and outcome were assessed and compared with data from consecutive patients undergoing initial PVI for paroxysmal AF with the cryoballoon from our specifically designed database (RedCap Database, Nashville, Tennessee, USA).

Alongside the 28 mm compliant balloon with 10 irrigated gold‐plated surface electrodes (Fig. 1), the RF-balloon ablation system comprises a 15/20 mm circular mapping catheter (Lassostar™, Biosense Webster, Diamond Bar, CA, USA), a 13.5/14Fr steerable sheath (Guidestar™, Biosense Webster, Diamond Bar, CA, USA), and a multichannel RF generator (nGEN™, Biosense Webster, Diamond Bar, CA, USA).

A Showing the 2nd generation Heliostar™ 28 mm diameter RF-balloon. B Fluoroscopic position of the RF-balloon at the antrum of the left superior pulmonary vein. C Visualization of the RF-balloon in the CARTO® 3D map indicating parameters for optimal balloon positioning. The electrode numbers highlighted in white fulfill impedance and temperature criteria for optimal balloon position

The first-generation RF-balloon system used a decapolar, non‐nav mapping catheter, Lassostar™ as guidewire and for EGM recording during PVI. For initial 3D map acquisition and post-PVI re-map, an additional circumferential mapping catheter (LassoNav™ Biosense Webster, Diamond Bar, CA, USA) was necessary.

The second-generation RF-balloon including the Lassostar™NAV (Biosense Webster, Diamond Bar, CA, USA) circumferential mapping catheter allowed direct 3D map acquisition via the RF-balloon making intraprocedural catheter and sheath exchange for map and re-map acquisition dispensable.

Parameters for optimal balloon positioning (Fig. 1) and ablation parameters have been previously described in detail [8, 10]. Briefly, the RF-balloon is inflated by increased irrigation flow. After balloon placement at the respective PV antrum, the ablation electrode positions can be identified in the 3D map. Anterior and posterior electrodes are defined before ablation (Fig. 1). Afterwards, energy is delivered in a unipolar mode at 15 W for 60 s anterior and 20 s posterior. All ten electrodes can be selected individually to deliver circumferential or segmental ablations up to 60 s duration. Every electrode provides real-time impedance, temperature, and electrogram data [8].

All RF-balloon procedures were performed by the same two experienced operators, and as we report our initial clinical experience, no cases had been performed prior to the present cohort.

For cryoballoon PVI, a 28‐mm cryoballoon (Arctic Front Advance Pro™, Medtronic, Dublin, Ireland or POLARx™, Boston Scientific, Marlborough, MA, USA) was applied. The workflow of a cryoballoon PVI has been previously described in detail [4]. Briefly, the balloon was inserted through a steerable sheath (15Fr FlexCathAdvance™, Medtronic or 15.9Fr POLARSHEATH™, Boston Scientific). The cryoballoon was placed in each PV antrum under fluoroscopy guidance, and PV occlusion was verified via contrast agent application. PV potentials were recorded using a 20‐mm circular inner lumen mapping catheter with 8 electrodes (Achieve™ and Achieve Advance™ Mapping Catheters, Medtronic or POLARMAP™, Boston Scientific). The duration of cryoablation was determined by either the observed time-to-isolation (TTI; in case of visible PV signals) or achieved nadir temperature [11].

Oral anticoagulation was discontinued the morning of the procedure and continued immediately after the procedure. All procedures were performed under deep propofol sedation amended by midazolam and fentanyl. An esophageal temperature probe (CIRCA S-CATH™, Circa Scientific, Englewood, CO, USA) was inserted transorally, and the position was confirmed with an X‐ray. An alarm was set at 38/20 °C followed by a maximal/minimal cut‐off at 39/19 °C, respectively, at which point RF delivery/cryoablation was discontinued.

After obtaining double vascular access through the right femoral vein, a decapolar catheter (Dynamic XT™, large curve, Boston Scientific, Marlborough, MA, USA) was placed in the coronary sinus (CS). Consequently, a fluoroscopy-guided single transseptal puncture using TSX™ Fixed Curve Transseptal Sheath and TSX™ Transseptal Needle (Boston Scientific, Marlborough, MA, USA) was performed. Immediately after transseptal puncture, a weight-adjusted heparin bolus was administered, with repeat activated clotting time (ACT) guided boluses every 20 min targeting an ACT of > 300 s.

For all first-generation RF-balloon procedures, the 3D map was obtained via TSX™ Fixed Curve Transseptal Sheath (Boston Scientific, Marlborough, MA, USA) and an additional 15 mm LassoNav™ (Biosense Webster, Diamond Bar, CA, USA). After map acquisition, the transseptal sheath was exchanged for the 13.5/14Fr steerable sheath (Guidestar™, Biosense Webster, Diamond Bar, CA, USA). For second-generation RF-balloon procedures, transseptal sheaths were exchanged immediately after transseptal puncture, just as for all cryoballoon procedures, and the 3D map was obtained directly via the RF-balloon and Lassostar™NAV (Biosense Webster, Diamond Bar, CA, USA).

PV signals were monitored in real‐time through the respective circumferential mapping catheter with the endpoint of an entrance block of all PVs. Finally, all PVs were re-checked for acute reconnection, and in all RF-balloon patients, a post-PVI remap was acquired confirming PV isolation via voltage map.

Phrenic nerve pacing was performed in all patients (pacing at maximal output with a cycle length of 800 ms) during ablation of the right‐sided PVs using the decapolar CS-catheter. Capture was confirmed by palpation of the diaphragmatic movement and compound motor action potential (CMAP). After sheath removal, a Z‐suture was applied for hemostasis at the femoral puncture site.

Procedure duration was defined as skin-to-skin time (groin puncture to groin suture after sheath removal), and ablation time was defined as net ablation time (start of the first ablation to end of last RF/cryoablation application). Directly, just as 2 and 24 h after the procedure, pericardial effusion was excluded using transthoracic echocardiography.

For all patients, a 3- and 12-month post-ablation follow-up visit was scheduled in our outpatient clinic. Furthermore, a 6-month telephone-based follow-up was conducted. For detection of atrial arrhythmia recurrence, a 24-h Holter electrocardiogram (ECG) was performed at 3- and 12-month in-person visit and patients were requested to perform an additional 24-h Holter ECG 6 months after PVI at the attending practitioner, which was included in the telephone-based 6-month follow-up. If patients reported symptoms, an additional visit in our outpatient clinic was scheduled, and a 12-lead ECG or additional Holter ECGs were obtained. During every follow-up visit, a 12-lead ECG was recorded, and the patient’s history was taken. Arrhythmia recurrence was defined as the occurrence of any arrhythmia longer than 30 s (AF, AT, atrial flutter) after a 90-day blanking period.

Study endpoints were differences in procedural characteristics (procedure duration, ablation time, fluoroscopy use, details on RF/cryoablation applications, and isolation times) just as procedure-related complications and rates of any atrial arrhythmia recurrence (> 30 s) after a 90-day blanking period. Furthermore, a subgroup analysis within the RF-balloon group was conducted, separately comparing procedural characteristics of the second-generation RF-balloon and cryoballoon.

Data analysis was performed using SPSS statistical software (version 26) and GraphPad Prism9. Data are shown in absolute values, percentages, and means with standard deviation. Variables were tested for normal distribution by the Shapiro–Wilk test. For the comparison of continuous variables, the Student’s t-test or Mann–Whitney U test were used. Categorical variables were compared using contingency tables and the application of the chi-square test or Fisher’s exact test. P-values < 0.05 were considered as statistically significant.