Study design and population
This retrospective observational study was conducted at the Catharina Hospital Heart Centre in Eindhoven, the Netherlands and ran from 1 January 2019 through 31 December 2022. Cryoballoon ablation was the primary treatment for paroxysmal and persistent AF until January 2022, at which point PFA became the preferred therapy for all AF patients. Consequently, patients were categorised into cryoballoon and PFA groups. Eligible participants were adults aged ≥ 18 years who had undergone either cryoballoon or PFA ablation as their initial AF treatment. Only the first PVI performed during the specified study period was included in the analysis.
Primary pulmonary vein ablation technique
During the time cryoballoon ablation was used at our hospital, groin puncture was performed blindly or under ultrasound guidance, whereas ultrasound-guided puncture became the standard approach with the implementation of PFA. Cryoballoon ablation utilised 12-French transseptal sheaths, while PFA employed 13-French sheaths. Both procedures involved inserting a second 6‑French sheath for a non-steerable diagnostic catheter, without performing arterial punctures. For cryoballoon ablation, the 27-mm Arctic FrontTM catheter (Medtronic, Fridley, MN, USA) was used, and for PFA, the FARAPULSETM catheter with a size of 31 or 35 mm (Boston Scientific, Marlborough, MA, USA) was employed at the physician’s discretion.
During cryoballoon ablation, one application was administered per PV, each lasting 240 s, unless early termination was deemed necessary due to phrenic capture loss. The application was halted if a temperature of −40 °C was not reached within 60 s, and the balloon position was adjusted accordingly. PV isolation was assessed by placing the diagnostic catheter behind the balloon and evaluating PV signals. The right PVs were ablated with concurrent right phrenic nerve stimulations via subclavian vein and manual diaphragm excursion assessments.
For PV isolation in PFA, a standard of 8 applications was used per PV: 4 in the basket position and 4 in the flower position. Variation in the number of applications was permitted at the physician’s discretion. Verification of PV isolation involved PV stimulation to confirm the absence of PV potentials and the presence of exit block.
Pre-procedural computed tomography scans were not conducted to assess PV anatomy in either group. PV anatomy assessment relied on visual inspection and contrast injection during cryoballoon ablation, with segmented ablation for suspected common ostium cases to prevent excessive depth or ostial lesions. In PFA, circumferential ablation was attempted in suspected cases.
Heparin was administered to maintain an activated clotting time > 300 s, followed by protamine administration, removal of venous sheaths and achievement of haemostasis with manual pressure and a pressure bandage. Patients had a 6-hour bed rest period and were allowed to mobilise for an additional hour before potential same-day discharge if stable and if the procedure finished before 15:00 h.
The follow-up protocol was similar for both ablation techniques. Amiodarone was discontinued immediately after the procedure, while other antiarrhythmic drugs were typically stopped after 3 months of follow-up, unless clinical indications required their continuation.
Baseline patient and procedural characteristics
Relevant data were extracted from electronic medical records. Based on the impact of outcomes after PVI according to a literature study [
17‐
19], the following baseline characteristics were included: age, gender, body mass index, renal function based on the estimated glomerular filtration rate (eGFR), left ventricular ejection fraction, left atrial volume index (LAVI) and AF type (paroxysmal or persistent).
Procedural measures include energy source/PVI method (cryoballoon or PFA), procedure duration (indicating time between arrival at and departure from procedure room) and admission duration. Baseline and procedural characteristics were defined using the definitions of the Netherlands Heart Registry [
20], which align with the European Society of Cardiology and the American College of Cardiology/American Heart Association guidelines [
1,
21].
Primary and secondary outcomes
The primary outcome encompassed procedural complications, specifically phrenic nerve palsy during admission, cardiac tamponade ≤ 30 days, thromboembolic complications ≤ 72 h, bleeding complications during admission and 30-day mortality. Phrenic nerve palsy was defined as persistent phrenic nerve failure (lasting ≥ 24 h) following intervention during admission. Phrenic nerve function was monitored by manual assessment of diaphragm excursions during procedures. Cardiac tamponade ≤ 30 days was characterised by the presence of pericardial fluid volume > 5 mm, with or without clinical or echocardiographic signs necessitating intervention. Thromboembolic complications encompassed thromboembolic events including ischaemic cerebrovascular accident, peripheral emboli and pulmonary emboli occurring ≤ 72 h post-intervention. Bleeding complications comprised groin bleeding requiring surgical intervention or transfusion during admission for the current procedure.
Secondary outcomes involved comparing the overall procedural characteristics of the PFA and cryoballoon ablation techniques, including procedure duration, admission duration, and re-do ablation rates within 6 months. Re-dos included procedures performed and scheduled within 6 months. In both ablation groups, the decision to conduct re-do procedures was primarily symptom-driven.
Statistical analysis
We assessed the dataset structure and distribution using descriptive statistics, Kolmogorov–Smirnov tests and Q‑Q plots. Continuous variables are presented as mean with standard deviation (SD) for normally distributed variables and as median with interquartile range (IQR) for non-normally distributed variables. Categorical variables are expressed as absolute frequencies (n) and relative frequencies. The Mann-Whitney U test, Student’s t-test and chi-square test were employed to identify significant differences between the treatment groups, as appropriate.
Multiple logistic regression analyses, adjusted for baseline characteristics, were used to examine the associations between short-term clinical outcomes and the PVI method used (cryoballoon or PFA). Multiple linear regression analyses, also adjusted for baseline characteristics, were utilised to investigate the associations between procedure and admission durations and the PVI methods. Baseline characteristics were included in regression models if the percentage of missing values was < 10%. Statistical significance was set at a two-tailed p-value (α < 0.05). SPSS version 29 (IBM, Chicago, IL, USA) was used to perform all data analyses.