Procedural Efficiency and Clinical Outcomes of a Region‐Specific Radiofrequency Ablation Strategy for Pulmonary Vein Isolation

ABSTRACT

Background

High‐power short‐duration radiofrequency ablation improves procedural efficiency during pulmonary vein isolation (PVI), but concerns remain regarding lesion transmurality and posterior wall collateral injury. We evaluated whether a region‐specific ablation strategy was associated with improved procedural efficiency without a detected difference in acute efficacy or 12‐month arrhythmia recurrence compared with conventional ablation index‐guided PVI.

Methods

In this retrospective single‐center study, 250 patients undergoing first‐time PVI were treated with either ablation index‐guided ablation or a region‐specific strategy using very high‐power short‐duration applications on the posterior wall and ablation index‐guided lesions on the anterior wall. The primary endpoint was first‐pass isolation, analyzed per pulmonary vein using generalized estimating equations to account for within‐patient clustering. Secondary endpoints included procedural efficiency, peri‐procedural safety, and 12‐month arrhythmia recurrence.

Results

First‐pass isolation was achieved in 919 of 1 000 pulmonary veins (91.9%): 454 of 500 veins (90.8%) with the region‐specific strategy and 465 of 500 veins (93.0%) with ablation index‐guided ablation, with no significant between‐group difference (OR, 0.74; 95% CI, 0.41–1.34; p = 0.322). The region‐specific strategy was associated with shorter procedure duration, ablation time, radiofrequency time, fluoroscopy time, and lower radiation dose (all p < 0.001). Acute complications occurred in 4 patients (1.6%). At 12 months, arrhythmia recurrence occurred in 43 patients (17.2%), with no detected difference between strategies.

Conclusions

A region‐specific ablation strategy improved procedural efficiency without a detected difference in first‐pass isolation or 12‐month arrhythmia recurrence compared with ablation index‐guided PVI.

Graphical summary of a region‐specific radiofrequency ablation strategy for pulmonary vein isolation. Anatomically tailored energy delivery was associated with improved procedural efficiency, with no detected difference in first‐pass isolation or 12‐month arrhythmia recurrence compared with conventional ablation index‐guided ablation.

Abbreviations

AF

Atrial fibrillation

AI

Ablation index

CT

Computed tomography

FPI

First‐pass isolation

LA

Left atrium

LVEF

Left ventricular ejection fraction

PVI

Pulmonary vein isolation

RF

Radiofrequency

vHP‐SD

Very high‐power short‐duration

1. Introduction

Pulmonary vein isolation (PVI) has emerged as a cornerstone for the treatment of atrial fibrillation (AF) [1, 2] Despite the emergence of novel energy sources, radiofrequency (RF) ablation remains widely used for PVI [3] Conventional low‐power, long‐duration (LPLD) approaches (≈25–35 W for 20–40 s) have progressively shifted toward high‐power, short‐duration (HPSD) strategies, typically ≈50 W, to exploit greater resistive heating, shorten application time, and improve workflow efficiency while targeting equivalent lesion efficacy. A key advancement in RF ablation has been the introduction of integrative lesion‐quality metrics, such as the Ablation Index (AI), which combines power, contact force, and time into a single parameter to standardize lesion delivery and serve as a surrogate for lesion size [4]. In clinical practice, commonly used AI targets are ~400 for the posterior wall and ~500–550 for anterior segments, often within a CLOSE‐guided workflow (inter‐lesion distance < 6 mm and contact force 10–20 g) [5, 6, 7]. Application of CLOSE‐guided ablation has demonstrated high rates of durable PVI and freedom from arrhythmia, particularly in paroxysmal AF [7].

Despite these advancements, limitations remain. Application of high‐power RF energy to the posterior wall, which lies adjacent to the esophagus, increases the risk of thermal injury. Although HPSD was initially thought to reduce this risk due to less conductive heating, recent controlled data show no reduction in either the incidence or severity of esophageal lesions with HPSD compared to conventional settings [8].

To address some of these challenges, very high‐power, short‐duration (vHP‐SD) ablation has emerged as an alternative approach, using temperature‐controlled energy delivery (e.g., 90 W for 4 s). This strategy produces wide, shallow lesions with high procedural efficiency and may reduce the risk of more serious collateral damage in thin atrial tissue [9] Indeed, clinical comparisons have shown that vHP‐SD and AI‐guided HPSD both achieve high first‐pass isolation (FPI) rates and improved procedural efficiency compared to conventional ablation, without compromising acute safety outcomes [10]. Preclinical data confirm the shallower and more uniform lesion morphology of vHP‐SD compared to HPSD, especially in posterior wall applications [11] However, vHP‐SD may be limited in achieving transmural lesions in thicker atrial regions, such as the anterior wall or carina [11, 12, 13].

Despite the widespread adoption of both ablation index‐guided and very high‐power short‐duration strategies, limited data are available on approaches that tailor energy delivery according to regional atrial wall characteristics. Given the heterogeneity in atrial wall thickness and the proximity to critical structures, an anatomically guided ablation strategy may optimize lesion formation and procedural efficiency.

In this context, we evaluated a region‐specific ablation strategy that leverages the advantages of both techniques: Very high‐power short‐duration ablation is applied to the posterior wall, where shallower lesions may be desirable given the proximity of the esophagus, while ablation index‐guided ablation is used on the anterior wall, where greater lesion depth is required. By tailoring energy delivery to regional atrial characteristics, this pragmatic workflow‐oriented approach aims to improve procedural efficiency without a detected difference in acute efficacy or 12‐month arrhythmia recurrence compared with conventional ablation index‐guided ablation.

2. Methods

2.1. Study Design

This retrospective, observational, single‐center study included consecutive adult patients undergoing catheter ablation for atrial fibrillation at Hospital Universitario Puerta del Mar, a high‐volume electrophysiology center performing over 600 ablation procedures annually, including more than 200 PVIs. The study compared a region‐specific ablation strategy with conventional AI‐guided ablation in terms of procedural characteristics, safety, and clinical outcomes. Outcomes were additionally stratified by sex to explore potential sex‐related differences. The study adhered to the principles of the Declaration of Helsinki and received approval from the institutional ethics committee.

2.2. Study Population

Patients were eligible if they were ≥ 18 years old with symptomatic paroxysmal or persistent AF undergoing their first catheter ablation between February 2020 and August 2024. Exclusion criteria included long‐standing persistent AF, prior surgical or catheter ablation, and the initial 20 procedures with the QDOT Micro catheter to account for the initial operator learning curve.

Treatment allocation was not randomized and reflected routine clinical practice, catheter availability, and operator decision. The QDOT Micro catheter was introduced during the study period. After exclusion of the initial learning‐curve cases, consecutive patients treated with the region‐specific strategy were identified, and an AI‐guided control cohort was selected from eligible first‐time PVI procedures using 1:1 nearest‐neighbor matching with age, sex, and atrial fibrillation type included as matching covariates.

The patient‐selection process is summarized in Figure S1.

All included patients provided informed consent prior to the procedure. As part of the standard pre‐procedural workup, all patients underwent a computed tomography (CT) scan within 48 h before ablation to exclude left atrial thrombus and assist with anatomical mapping.

2.3. Ablation Technique and Study Groups

Procedures were performed under conscious sedation or general anesthesia. Vascular access was obtained via the right femoral vein. Transseptal access was achieved by a single puncture of the interatrial septum using a CARTO VIZIGO Bi‐Directional Guiding Sheath (Biosense Webster, United States) under fluoroscopic and pressure guidance and with transesophageal or intracardiac echocardiography, depending on the type of sedation.

After transseptal access, intravenous heparin was administered to maintain an activated clotting time > 300 s. Three‐dimensional left atrial mapping was performed using the CARTO 3 system (Biosense Webster) with a PentaRay NAV eco multipolar mapping catheter (Biosense Webster). Point‐by‐point RF ablation of each ipsilateral pair of pulmonary veins (PVs), including the right PVs carina, was then performed. In patients presenting in AF at the beginning of the procedure, external cardioversion was performed after completion of PVI.

Two catheter types were employed: The Thermocool SmartTouch (ST) open‐irrigated, contact‐force sensing catheter for AI‐guided ablation, and the QDOT Micro open‐irrigated catheter with real‐time temperature monitoring for vHP‐SD applications.

Patients were categorized into two treatment groups based on the ablation strategy (Figure 1 and Figure 2).

FIGURE 1.

Representative left atrial maps illustrating the differences between ablation strategies. (A) Conventional ablation index‐guided point‐by‐point ablation, showing a homogeneous lesion set applied across the PVI line. (B) Region‐specific ablation strategy, demonstrating differential energy delivery with vHP‐SD applications along the posterior wall component of the PVI line, including the adjacent superior and inferior portions of the encircling lesion set, and ablation index‐guided lesions along the anterior wall, reflecting an anatomically tailored approach to lesion formation.

FIGURE 2.

Kaplan–Meier analysis of arrhythmia‐free survival at 12 months after the blanking period, comparing the ablation index–guided strategy with the region‐specific strategy using posterior wall vHP‐SD ablation and anterior wall AI‐guided ablation. The p‐value shown corresponds to a Cox model stratified by matched pair. vHP‐SD = very high‐power short‐duration AI = ablation index.

2.3.1. AI‐Guided Group: Control Group

Conventional point‐by‐point ablation was performed using ST catheters. Radiofrequency energy was delivered at 40 W with target AI values of 350–380 for the posterior wall and 450 for the anterior wall. AI target values were selected based on established clinical practice and prior studies demonstrating their association with effective and durable lesion formation, while balancing safety considerations, particularly in posterior wall regions where tissue thickness is lower and proximity to the esophagus is relevant. The inter‐lesion distance (ILD) was maintained below 6 mm with a contact force between 10 to 20 g [6, 7, 10, 14, 15].

2.3.2. Region‐Specific Ablation Group (vHP‐SD + AI): Study Group

In this group, vHP‐SD applications were delivered along the posterior wall component of the PVI line, including the adjacent superior and inferior portions of the encircling lesion set, using QDOT Micro in QMODE+ mode (90 W, 4 s per application), with temperature‐controlled energy delivery and an automatic safety cut‐off at 55°C. The anterior wall was ablated using AI‐guided RF applications at 50 W with an AI target of 450. The ILD and contact force parameters were kept consistent with the AI‐guided group [6, 11, 14, 16].

2.4. Evaluation After Ablation

Following completion of the initial encircling ablation set, the ablation catheter was reintroduced into the pulmonary veins to assess conduction block. Conduction between the PVs and the left atrium was evaluated by pacing at 10 mA output with 2.0 ms pulse width from multiple sites within the vein ostia using the ablation catheter.

First‐pass isolation was assessed at the per‐vein level and defined as complete entrance and exit block of an individual PV after the initial encircling lesion set, without the need for additional ablation. If residual conduction was detected in any PV, additional radiofrequency applications were delivered until isolation was achieved, and the respective vein was classified as not having achieved first‐pass isolation. The final procedural endpoint was complete acute bidirectional PVI, confirmed by persistent entrance and exit block in all PVs at the end of the procedure. Adenosine testing was not performed.

2.5. Follow‐Up

All patients underwent scheduled follow‐up with ECG and 24‐h Holter monitoring at 3–4 and 12 months after the procedure. Additional visits or extended monitoring were performed in case of symptomatic recurrence. Use of wearable ECG devices was permitted and systematically documented when available. A standard 3‐month post‐procedural blanking period was applied, in line with current recommendations. Continuous rhythm monitoring was not routinely performed.

2.6. Outcomes

The primary endpoint of the study was acute procedural efficacy, assessed by first‐pass isolation as defined above. Secondary endpoints included procedural efficiency, safety, and mid‐term effectiveness. Procedural efficiency was evaluated by total procedure time (skin‐to‐skin), total ablation time, cumulative radiofrequency application time, fluoroscopy time, and radiation dose expressed as dose‐area product (Gy·cm²). Safety outcomes comprised major and minor peri‐procedural complications. Effectiveness was assessed as freedom from documented atrial tachyarrhythmia (atrial fibrillation, atrial flutter, or atrial tachycardia) lasting ≥ 30 s within 12 months after ablation, following a 3‐month blanking period [6]. Recurrence was defined as any arrhythmia fulfilling this criterion during follow‐up.

In addition, all outcomes were analyzed separately within each sex, with comparisons performed between ablation strategies to explore potential sex‐specific differences in procedural efficiency, acute efficacy, safety, and mid‐term effectiveness.

2.7. Statistical Analysis

Continuous variables were expressed as mean ± standard deviation or median [interquartile range], depending on distribution, and categorical variables were presented as counts (percentages). To minimize baseline differences between treatment groups, patients treated with the region‐specific ablation strategy were matched 1:1 to those undergoing conventional AI‐guided ablation using nearest‐neighbor matching without replacement, with age, sex, and atrial fibrillation type (paroxysmal vs. persistent) included as matching covariates.

Baseline characteristics and procedural variables were analyzed accounting for the matched‐pair structure. Continuous variables were compared using the Wilcoxon signed‐rank test, whereas paired categorical variables were compared using the exact McNemar test when appropriate. For sex‐stratified paired analyses, only matched pairs in which both patients belonged to the same sex stratum were included. The number of matched pairs contributing to each analysis is reported in the corresponding tables.

No imputation of missing data was performed. Descriptive summaries were based on available observations, whereas paired comparisons were restricted to complete matched pairs for the variable under analysis. The number of observations and matched pairs contributing to each analysis is reported in the corresponding tables.

The primary endpoint was first‐pass isolation (FPI), analyzed at the pulmonary vein level. Because pulmonary veins within the same patient are not independent observations, between‐group comparisons of FPI were performed using generalized estimating equations logistic models with clustering at the patient level. Results are reported as odds ratios (OR) with 95% confidence intervals (CI), together with absolute between‐group differences where appropriate.

Time‐to‐event outcomes were displayed using Kaplan–Meier survival curves. Between‐group comparisons for 12‐month arrhythmia recurrence after the blanking period were assessed using Cox proportional hazards models stratified by matched pair. Sex‐stratified time‐to‐event analyses were restricted to sex‐concordant matched pairs and were performed using Cox proportional hazards models stratified by matched pair within each sex.

Exploratory predictive models for first‐pass isolation and 12‐month arrhythmia recurrence were constructed using logistic regression. In univariable analyses, candidate predictors were assessed individually. Variables were selected for multivariable modeling based on clinical relevance, prior literature, and univariable association with a p‐value < 0.10. Multivariable analyses were performed using complete cases for all variables included in each model. Final multivariable models were constructed with attention to the number of observed events to minimize overfitting. Results are reported as OR with 95% CI.

Secondary, sex‐stratified, and predictive analyses were considered exploratory. No formal adjustment for multiplicity was applied; therefore, these results should be interpreted as hypothesis‐generating. For multivariable models, the number of events per variable was calculated to assess the potential risk of overfitting.

No formal sample‐size or power calculation was performed because this was a retrospective observational study based on the available matched cohort. No pre‐specified non‐inferiority margin was defined; therefore, the study was not designed to test equivalence or non‐inferiority between ablation strategies.

Nominal two‐sided p‐values < 0.05 were considered statistically significant. All analyses were performed using R software (version 4.5.2) and SPSS Statistics (version 27; IBM Corp.).

3. Results

3.1. Study Population

A total of 250 patients undergoing first‐time PVI were included in the study, with 125 assigned to conventional ablation guided by the Ablation Index (AI group) and 125 treated using the region‐specific ablation strategy combining very high‐power short‐duration (vHP‐SD) ablation on the posterior wall and AI‐guided ablation on the anterior wall (region‐specific ablation group). General anesthesia was used in 75 procedures (30%). Baseline demographic and clinical characteristics were broadly comparable between groups (Table 1). A small difference in age was observed in the paired analysis (60 ± 9 vs. 59 ± 10 years, p = 0.045), whereas no significant differences were found in sex distribution, body mass index, atrial fibrillation type, left atrial diameter, left ventricular ejection fraction, CHA₂DS₂‐VASc score, or use of general anesthesia.

TABLE 1.

Baseline demographic and clinical characteristics of the study population.

Characteristic		Overall N = 250	AI group N = 125	Region‐specific group N = 125	p
Age		60 ± 9	60 ± 9	59 ± 10	0.045
Gender	Male	159 (64%)	82 (66%)	77 (62%)	0.062
	Female	91 (36%)	43 (34%)	48 (38%)
BMI		29 ± 5	29 ± 4	29 ± 5	0.634
AF Type	Paroxysmal	175 (70%)	87 (70%)	88 (70%)	> 0.9
	Persistent	75 (30%)	38 (30%)	37 (30%)
LA Diameter		42 ± 6	42 ± 7	42 ± 6	0.874
LVEF		60 ± 10	60 ± 11	60 ± 9	0.485
CHA2DS2‐VASc	< 3	206 (82%)	103 (82%)	103 (82%)	> 0.9
	≥ 3	44 (18%)	22 (18%)	22 (18%)
Anesthesia		75 (30%)	37 (30%)	38 (30%)	> 0.9

Note: All baseline variables were available for all 250 patients, and all comparisons included 125 matched pairs.

Abbreviations: AF, Atrial Fibrillation; BMI, Body mass index; LA, Left Atrium; LVEF, Left Ventricular Ejection Fraction.

3.2. Procedural Characteristics and Acute Efficacy

Procedural performance was consistently superior in the region‐specific ablation group compared with the AI strategy (Table 2). Overall procedure duration was significantly shorter with the region‐specific ablation strategy, with a median skin‐to‐skin time of 66 min versus 75 min in the AI group (p < 0.001). Fluoroscopy exposure was also markedly reduced, with median screening times of 95 versus 190 s, respectively (p < 0.001). In line with these findings, transeptal‐to‐end intervals, total ablation duration, radiofrequency delivery time, and radiation dose were all significantly lower with the region‐specific ablation strategy.

TABLE 2.

Procedural characteristics according to ablation strategy.

Variable	Overall N = 250	AI group N = 125	Region‐specific group N = 125	p
Skin to Skin [min]	70 [60–88]	75 [64–95]	66 [55–77]	< 0.001
Transeptal to end [min]	53 [43–67]	60 [47–72]	47 [39–60]	< 0.001
Total Ablation time [min]	30 [21–39]	35 [29–46]	23 [19–32]	< 0.001
RF Time [s]	668 [456–1078]	1 069 [880–1 327]	462 [400–527]	< 0.001
X‐ray Time [s]	123 [79–221]	190 [103–273]	95 [72–147]	< 0.001
X‐ray Dose [Gy·cm2]	2.26 [1.18–4.38]	3.24 [1.61–5.85]	1.62 [1.02–3.00]	< 0.001

Note: p‐values were calculated using the Wilcoxon signed‐rank test for matched pairs. All procedural variables were available for all 250 patients, and all comparisons included 125 matched pairs.

Abbreviation: RF, Radiofrequency. Median [Q1–Q3].

3.2.1. Primary Endpoint

First‐pass isolation was achieved in 919 of 1 000 pulmonary veins overall (91.9%). FPI was achieved in 465 of 500 veins (93.0%) in the AI‐guided group and in 454 of 500 veins (90.8%) in the region‐specific group. After accounting for within‐patient clustering using a generalized estimating equations model, no significant between‐group difference was observed (GEE‐derived OR, 0.74; 95% CI, 0.41–1.34; p = 0.322). The absolute between‐group difference was −2.2 percentage points (95% CI, −6.6 to 2.2). At the patient level, complete FPI of all pulmonary veins was achieved in 196 of 250 patients (78.4%), including 100 of 125 patients (80.0%) in the AI‐guided group and 96 of 125 patients (76.8%) in the region‐specific group.

3.3. Safety

Overall, four acute procedural complications were observed (1.6% of the study population). One complication occurred in the AI group (0.8%), consisting of a femoral pseudoaneurysm. In the region‐specific group, three patients (2.4%) experienced complications: One pericardial tamponade requiring pericardiocentesis, one case of transient phrenic nerve paralysis, and one left pulmonary vein stenosis. No clinically apparent esophageal complications were observed in either group. In matched‐pair analysis, the overall complication rate did not differ significantly between groups (exact McNemar test, p = 0.625).

3.4. One‐Year Outcomes

At 12‐month follow‐up, recurrent atrial tachyarrhythmia after the blanking period occurred in 43 of 250 patients (17.2%). Recurrence was observed in 25 of 125 patients (20.0%) in the AI‐guided group and in 18 of 125 patients (14.4%) in the region‐specific group. In matched‐pair analysis, recurrence did not differ significantly between groups (exact McNemar test, p = 0.337). The majority of recurrences were atrial fibrillation (18.4% in the AI‐guided group and 13.6% in the region‐specific group), whereas atrial flutter was less frequent (1.6% and 0.8%, respectively); the distribution of recurrence subtypes did not differ significantly between groups (Fisher's exact test, p = 0.50).

Kaplan–Meier analysis showed no detected difference in arrhythmia‐free survival after the blanking period between the two strategies (Figure 2). In a Cox model stratified by matched pair, the region‐specific strategy was not significantly associated with recurrence (HR, 0.78; 95% CI, 0.42–1.45; p = 0.44).

In an exploratory patient‐level analysis, 12‐month arrhythmia recurrence after the blanking period occurred in 14.3% of patients with FPI compared with 27.8% of those without FPI (Fisher's exact test, p = 0.030). This association remained significant after accounting for matched‐pair clustering using robust standard errors (OR, 0.43; 95% CI, 0.21–0.90; p = 0.026).

3.5. Predictors of First Pass Isolation

In exploratory univariable analyses, male sex, higher body mass index (BMI), persistent atrial fibrillation, and larger left atrial diameter were associated with a lower likelihood of achieving first‐pass isolation, although only BMI reached statistical significance (OR, 0.93; 95% CI, 0.88–0.99; p = 0.031). Neither age, left ventricular ejection fraction, nor obstructive sleep apnea showed a significant association with acute isolation.

In the exploratory multivariable model, no variable remained significantly associated with first‐pass isolation. Higher BMI showed a trend toward a lower likelihood of FPI (OR, 0.94; 95% CI, 0.88–1.00; p = 0.058). The ablation technique was not associated with a significantly different probability of achieving FPI compared with AI‐guided ablation (Table 3). Sex, atrial fibrillation type, and left atrial diameter were not independently associated with acute isolation success.

TABLE 3.

Univariable and multivariable predictors of first‐pass isolation.

	Univariable OR (95% CI)	p	Multivariable OR (95% CI)	p
Gender (Male vs. Female)	0.61 (0.31–1.16)	0.139	0.59 (0.29–1.14)	0.126
Age (per year)	1.02 (0.99–1.05)	0.279
BMI (per kg/m2)	0.93 (0.88–0.99)	0.031	0.94 (0.88–1.00)	0.058
LVEF (%)	0.99 (0.96–1.02)	0.551
LA diameter (per mm)	0.96 (0.91–1.01)	0.086	0.97 (0.92–1.03)	0.330
OSA (Yes vs. No)	0.76 (0.30–2.20)	0.584
AF type (Persistent vs. Parox)	0.66 (0.35–1.26)	0.204	0.81 (0.41–1.61)	0.536
Technique (Region‐Specific vs. AI)	0.83 (0.45–1.51)	0.539	0.82 (0.44–1.52)	0.523

Note: These analyses were exploratory. No adjustment for multiplicity was performed. The multivariable model included 250 patients, 54 non‐complete FPI events, and 5 predictor variables, corresponding to 10.8 events per variable. Values in bold indicate statistical significance (p < 0.05).

Abbreviations: BMI, body mass index; CI, confidence interval; LA, left atrium; LVEF, left ventricular ejection fraction; OR, odds ratio; OSA, obstructive sleep apnea.

The multivariable model included 250 patients, 54 non‐complete FPI events, and 5 predictor variables, corresponding to 10.8 events per variable.

3.6. Predictors of Recurrence

In exploratory univariable analyses, larger left atrial diameter, persistent atrial fibrillation, and the absence of first‐pass isolation were associated with an increased risk of recurrence at 12 months after the blanking period. Left atrial diameter showed a strong association with recurrence (OR, 1.11 per mm; 95% CI, 1.05–1.18; p < 0.001), persistent atrial fibrillation was associated with higher recurrence risk compared with paroxysmal atrial fibrillation (OR, 2.40; 95% CI, 1.22–4.72; p = 0.011), and first‐pass isolation was associated with a lower risk of recurrence (OR, 0.43; 95% CI, 0.21–0.90; p = 0.022). No significant associations were observed for sex, age, body mass index, left ventricular ejection fraction, obstructive sleep apnea, or ablation technique in univariable analysis.

In the exploratory multivariable model, larger left atrial diameter remained independently associated with a higher likelihood of arrhythmia recurrence at 12 months (OR, 1.08 per mm; 95% CI, 1.02–1.16; p = 0.009). Atrial fibrillation type, first‐pass isolation, ablation technique, sex, body mass index, and left ventricular ejection fraction were not independently associated with recurrence (Table 4).

TABLE 4.

Univariable and multivariable predictors of 12‐month recurrence.

	Univariable OR (95% CI)	p	Multivariable OR (95% CI)	p
Gender (Male vs. Female)	1.40 (0.70–2.92)	0.357	1.12 (0.53–2.42)	0.776
Age (per year)	0.99 (0.95–1.02)	0.408
BMI (per kg/m2)	1.04 (0.97–1.12)	0.216	1.02 (0.94–1.10)	0.632
LVEF (%)	0.97 (0.95–1.00)	0.076	0.99 (0.96–1.02)	0.587
LA diameter (per mm)	1.11 (1.05–1.18)	< 0.001	1.08 (1.02–1.16)	0.009
OSA (Yes vs. No)	0.43 (0.07–1.55)	0.269
AF type (Persistent vs. Parox)	2.40 (1.22–4.72)	0.011	1.50 (0.69–3.16)	0.296
Technique (Region‐Specific vs. AI)	0.67 (0.34–1.30)	0.243	0.67 (0.33–1.34)	0.261
First‐pass isolation (Yes vs. No)	0.43 (0.21–0.90)	0.022	0.49 (0.23–1.09)	0.072

Note: These analyses were exploratory. No adjustment for multiplicity was performed. The multivariable model included 250 patients, 43 recurrence events, and 7 predictor variables, corresponding to 6.1 events per variable. Values in bold indicate statistical significance (p < 0.05).

Abbreviations: BMI, Body mass index; CI, Confidence Interval; LA, Left Atrium; LVEF, Left Ventricular Ejection Fraction; OR, Odds Ratio; OSA, Obstructive Sleep Apnea.

The multivariable model for recurrence included 43 recurrence events and 7 predictor variables, corresponding to 6.1 events per variable.

3.7. Sex‐Stratified Analysis

In exploratory sex‐stratified analyses, the region‐specific strategy was associated with consistent differences in procedural efficiency, whereas acute efficacy, safety, and 12‐month clinical outcomes were not significantly different between strategies in either men or women.

At baseline, women presented with higher CHA₂DS₂‐VASc scores than men, while age, atrial fibrillation type, left atrial diameter, and left ventricular ejection fraction were similar between sexes.

Within‐sex comparisons showed that the region‐specific ablation strategy was associated with improved procedural efficiency in both men and women. In men, the region‐specific strategy was associated with significant reductions in total procedure duration, left atrial dwell time, total ablation time, radiofrequency application time, fluoroscopy time, and radiation dose compared with ablation index‐guided ablation. In women, the region‐specific strategy was associated with significant reductions in total procedure duration, left atrial dwell time, total ablation time, and radiofrequency application time, whereas fluoroscopy time and radiation dose were not significantly different between strategies (Table 5).

TABLE 5.

Procedural characteristics stratified by sex and ablation strategy.

	Male (77 matched pairs)			Female (43 matched pairs)
	AI group	Region‐specific group	p	AI group	Region‐specific group	p
Skin to Skin	75 [65–95]	69 [60–80]	< 0.001	75 [60–93]	65 [53–74]	0.038
LA dwell time [min]	61 [47–72]	51 [41–62]	< 0.001	55 [48–68]	45 [38–54]	< 0.001
Total Ablation time [min]	35 [30–47]	25 [21–33]	< 0.001	35 [27–42]	21 [19–30]	< 0.001
RF Time [s]	1 098 [900–1 334]	475 [407–563]	< 0.001	1 014 [823–1 310]	435 [376–518]	< 0.001
X‐ray Time [s]	198 [115–273]	101 [77–151]	< 0.001	146 [82–240]	89 [58–120]	0.073
X‐ray Dose [Gy·cm2]	3.71 [2.00–6.46]	1.91 [1.28–3.84]	< 0.001	2.00 [0.92–4.11]	1.13 [0.75–2.44]	0.069

Note: Values are presented as median [Q1–Q3]. p‐values were calculated using the Wilcoxon signed‐rank test for matched pairs within each sex. All procedural variables were available for all patients. Sex‐stratified analyses included 77 male matched pairs and 43 female matched pairs. Five sex‐discordant matched pairs were excluded from the sex‐stratified paired analyses.

Abbreviations: LA, Left Atrium; RF, Radiofrequency.

FPI rates were similar across sexes and ablation strategies. Among men, FPI was achieved in 91.2% of veins with AI‐guided ablation and in 89.6% with the region‐specific strategy, with no significant difference after accounting for within‐patient clustering using GEE (OR, 0.84; 95% CI, 0.42–1.68; p = 0.615). Among women, FPI rates were 96.5% and 92.7%, respectively, with no significant between‐strategy difference after clustering correction (OR, 0.46; 95% CI, 0.16–1.31; p = 0.145).

Kaplan–Meier survival analyses stratified by sex showed no detected difference in arrhythmia‐free survival between ablation strategies in either women or men during 12‐month follow‐up. Sex‐stratified analyses included sex‐concordant matched pairs only, and comparisons were based on Cox models stratified by matched pair (Figure 3).

FIGURE 3.

Kaplan–Meier estimates of arrhythmia‐free survival at 12 months after the blanking period, stratified by sex. (A) Women and (B) men undergoing pulmonary vein isolation with either ablation index–guided ablation or the region‐specific strategy, consisting of posterior wall vHP‐SD ablation and anterior wall AI‐guided ablation. Sex‐stratified analyses included sex‐concordant matched pairs only. The p‐values shown correspond to Cox models stratified by matched pair. vHP‐SD = very high‐power short‐duration; AI = ablation index.

Sex‐specific exploratory univariable and multivariable analyses for predictors of first‐pass isolation are provided in the (Tables S1 and S2).

4. Discussion

4.1. Main Findings

This study evaluated whether an anatomically guided, region‐specific radiofrequency ablation strategy could improve procedural efficiency during pulmonary vein isolation without a detected difference in acute efficacy or 12‐month arrhythmia recurrence compared with conventional AI‐guided ablation. Three main findings emerge from our analysis. First, the region‐specific strategy significantly improved procedural efficiency, with shorter procedure, ablation, and fluoroscopy times, as well as reduced radiation exposure. Second, first‐pass isolation rates were high in both groups, with no statistically significant between‐group difference after accounting for within‐patient clustering. Third, no statistically significant difference in 12‐month arrhythmia recurrence was detected in Cox models stratified by matched pair. Safety findings should be interpreted cautiously because the number of complications was low and the study was not powered to detect differences in infrequent adverse events.

The primary aim of this study was not to demonstrate superiority in long‐term clinical outcomes, but rather to determine whether procedural workflow could be optimized through anatomically tailored energy delivery. The concept of tailoring ablation energy to regional atrial characteristics has been explored previously using hybrid, vHP‐SD‐only, and wall‐thickness‐guided approaches. Unlike individualized wall‐thickness‐guided strategies, the present region‐specific approach provides a simplified anatomical workflow applicable when wall‐thickness maps are not available [17, 18, 19], using vHP‐SD ablation along the posterior wall component of the PVI line and AI‐guided ablation along the anterior wall, where deeper lesions may be required. Within this framework, the absence of a detected difference in 12‐month arrhythmia recurrence in a relatively large real‐world cohort of 250 patients suggests that procedural efficiency can be substantially improved using a pragmatic, anatomy‐based workflow. Taken together, these findings suggest that the region‐specific ablation strategy was associated with high acute procedural efficacy in this cohort while providing a significant procedural efficiency benefit through a simplified, anatomically guided workflow.

4.2. Primary Endpoint: Acute Procedural Efficacy

In our cohort, first‐pass pulmonary vein isolation was achieved in nearly 92% of individual pulmonary veins. This high rate should be interpreted in the context of the per‐vein definition used in the present study. When FPI was considered at the patient level, complete isolation of all pulmonary veins was achieved in 78.4% of patients. These values are consistent with prior reports of AI‐guided or optimized radiofrequency ablation protocols in which FPI rates typically range between 70%–85% depending on the population and methodology used [14, 20] Importantly, after accounting for within‐patient clustering using generalized estimating equations, no significant difference in per‐vein FPI was observed between conventional AI‐guided ablation and the region‐specific ablation strategy.

FPI rates achieved with the region‐specific ablation strategy were also consistent with those reported in recent vHP‐SD ablation studies, although differences in endpoint definition should be considered. The FAST AND FURIOUS study reported bilateral FPI in 77% of cases [13] while Peller et al. observed complete patient‐level first‐pass isolation in approximately one‐third of patients using vHP‐SD, though regional differences in reconnection sites were noted [14] Likewise, in the randomized POWER PLUS trial, first‐pass isolation showed a nonsignificant trend to be lower with 90 W/4 s than with conventional AI‐guided ablation (83.9% vs. 90%; p = 0.085), without impact on clinical outcomes [21] Taken together, these findings suggest that the region‐specific ablation strategy achieves high acute procedural efficacy while offering a significant procedural efficiency benefit.

Body mass index was associated with a lower likelihood of achieving FPI in univariable analysis and showed a borderline association after multivariable adjustment. These findings are in line with data from the REAL‐AF registry, which highlighted BMI and atrial remodeling as relevant, but not uniformly dominant, determinants of acute procedural efficacy [20] Clinically, these observations are consistent with the importance of integrating structured weight management into AF care; importantly, newer anti‐obesity pharmacotherapies are making substantial and sustained weight loss more achievable and may become a key adjunct to rhythm‐control strategies, potentially improving both procedural and long‐term outcomes [22].

4.3. Procedural Efficiency

The region‐specific ablation strategy demonstrated a clear advantage in terms of procedural efficiency. Compared with conventional AI‐guided ablation, the region‐specific ablation strategy significantly reduced skin‐to‐skin duration (66 vs. 75 min, p < 0.001), total ablation time (23 vs. 35 min, p < 0.001), and RF application time (462 vs. 1 069 s, p < 0.001). Radiation exposure was also substantially lower, with both fluoroscopy time and dose reduced by approximately half. These efficiency gains should be interpreted as reflecting the overall region‐specific workflow rather than the isolated effect of posterior wall vHP‐SD ablation alone, since anterior wall ablation also differed between groups in terms of delivered power. Reduced fluoroscopy time in the region‐specific group likely reflects the shorter left atrial ablation phase. Faster lesion deployment with very high‐power short‐duration applications reduced RF and ablation time, thereby decreasing the need for intermittent fluoroscopic confirmation during catheter repositioning, sheath support, and lesion delivery. Pulsed field ablation (PFA) has emerged as an alternative approach for pulmonary vein isolation, offering the potential for rapid procedures and myocardial selectivity. However, procedural characteristics such as procedure duration, fluoroscopy exposure, and workflow efficiency may vary depending on the specific technology and operator experience [23, 24] In this context, the present findings highlight that further optimization of radiofrequency‐based ablation workflows remains feasible, even in the era of emerging ablation technologies.

From a practical perspective, this strategy provides a simple and reproducible approach that can be readily implemented in routine clinical practice without requiring additional technologies, supporting its potential applicability across a wide range of electrophysiology centers.

Our observations are consistent with previous studies evaluating very high‐power short‐duration protocols. Sallo et al. reported a reduction in total procedure time to 70 min with vHP‐SD compared with 85 min for HPSD and 101 min for conventional low‐power ablation [10] Similarly, the randomized POWER PLUS trial by O'Neill et al. showed a modest but significant reduction in procedure time with 90 W ablation compared with 35/50 W ablation (70 vs. 75 min, p = 0.009) [21] Baumgartner and colleagues, in a direct comparison of vHP‐SD and a region‐specific ablation strategy, found that vHP‐SD alone yielded shorter ablation and fluoroscopy times, with a numerical trend toward higher FPI [25] This observation suggests that vHP‐SD‐only workflows may be highly efficient and, in selected settings, may be as efficient as or more efficient than hybrid approaches. However, preclinical data confirm that vHP‐SD lesions are significantly shallower than HPSD lesions, which may limit transmurality in thicker atrial regions such as the anterior wall or the carina [11]. Clinical studies have reinforced this concern, with Halbfass et al. reporting low FPI rates and frequent residual conduction at the carina [12] Similarly, the FAST AND FURIOUS trial showed that achieving durable isolation with vHP‐SD requires a very‐close lesion set, thereby attenuating the apparent efficiency benefit [13]. The recognition that left atrial wall thickness varies significantly across different anatomical regions, with the posterior wall typically measuring 1–2 mm and the anterior wall, septum, and carina often exceeding 3–4 mm, has prompted the development of personalized ablation strategies. Teres et al., in the ‘Ablate by‐LAW’ pilot study, demonstrated that adapting AI to LAWT (categorized into 1 mm layers) achieved high first‐pass isolation rates (93%–97%) with minimal RF time and excellent 16‐month outcomes [19]. This approach was subsequently validated in the randomized PeAF‐by‐LAWT trial, which confirmed non‐inferiority to the CLOSE protocol while significantly improving procedural efficiency [18] Our region‐specific ablation strategy was designed with these anatomical considerations in mind. By selectively applying vHP‐SD to the posterior wall, where shallower lesions may be sufficient and proximity to the esophagus is most relevant, and reserving AI‐guided ablation for the anterior wall, where greater lesion depth may be required, we aimed to match ablation energy to regional anatomical characteristics. This anatomically tailored approach optimizes the balance between lesion depth and procedural efficiency, potentially explaining the high FPI rates observed in our study despite the use of vHP‐SD in selected regions.

From a clinical perspective, this approach may be particularly relevant in settings where radiofrequency ablation remains the primary technology or where access to emerging energy sources is limited. Rather than competing with newer technologies, a region‐specific strategy may complement current practice by optimizing conventional radiofrequency ablation workflows.

4.4. Periprocedural Safety

The overall complication rate in our cohort was low (1.6%). No statistically significant difference in complications was detected between the AI‐guided and region‐specific ablation strategies; however, this finding should be interpreted cautiously given the very low number of events and the numerically higher number of clinically relevant complications in the region‐specific group. The low overall complication rate is consistent with previous reports of high‐power and very high‐power ablation, such as Q‐FFICIENCY (3.6%) [26] POWER PLUS (< 2%) [21], and other observational series confirming the low incidence of acute complications with contemporary high‐power strategies [10] Of note, Baumgartner et al. reported a similar safety profile when comparing vHP‐SD with a region‐specific ablation protocol, further supporting the feasibility of this approach [25] Silent cerebral embolism was not systematically assessed by brain MRI in this study; although initially a concern with vHP‐SD due to early reports of higher incidence, subsequent series suggested lower rates, and in the present cohort, using a single transseptal access with catheter exchange, no cases of stroke or transient ischemic attack were observed [12, 27].

4.5. Outcomes

At 12‐month follow‐up, arrhythmia recurrence was observed in 17% of patients, with no significant differences between the AI‐guided and region‐specific ablation strategies. The majority of recurrences were atrial fibrillation, while atrial flutter and atrial tachycardia were uncommon. Recurrence rates are in line with contemporary data for both AI‐guided and very high‐power protocols. The Q‐FFICIENCY trial reported freedom from atrial tachyarrhythmia of 82% at one year using a predominantly vHP‐SD approach [26]. Baumgartner and colleagues, in a direct comparison of vHP‐SD and region‐specific ablation strategies, found no significant difference in arrhythmia‐free survival at 6 months [25].

First‐pass isolation was associated with a lower rate of arrhythmia recurrence at one year in unadjusted analyses; however, this association did not remain significant after multivariable adjustment. In contrast, larger left atrial diameter remained independently associated with a higher likelihood of 12‐month arrhythmia recurrence after multivariable adjustment.

Although FPI has been proposed as a marker associated with improved long‐term outcomes [14] in our cohort this association did not remain significant after multivariable adjustment. This may reflect differences in patient selection, limited event numbers, and the fact that residual conduction after the initial lesion set was treated with additional applications until complete acute bidirectional PVI was achieved. Accordingly, FPI in the present study should be interpreted primarily as an acute procedural marker of lesion‐set quality and workflow efficiency, rather than as definitive evidence of long‐term pulmonary vein isolation.

In exploratory sex‐stratified analyses, the region‐specific ablation strategy was associated with improved procedural efficiency in both men and women, while no detected difference in first‐pass isolation or 12‐month arrhythmia recurrence was observed between ablation strategies in either sex. These findings are consistent with contemporary ablation cohorts reporting no meaningful sex‐related differences in mid‐term outcomes with modern technologies [28, 29].

5. Limitations

Several methodological considerations should be noted. First, this was a non‐randomized observational analysis with a relatively limited sample size. Although patients were matched using age, sex, and atrial fibrillation type as matching covariates and baseline characteristics were generally balanced between groups, residual confounding due to unmeasured variables cannot be fully excluded. In addition, no formal sample‐size or power calculation was performed, and no pre‐specified non‐inferiority margin was defined. Therefore, the study was not designed or powered to demonstrate equivalence or non‐inferiority between ablation strategies. Accordingly, the absence of statistically significant differences in first‐pass isolation, arrhythmia recurrence, or safety outcomes should not be interpreted as proof of comparable efficacy or safety.

Second, adenosine or isoproterenol testing was not routinely performed, which may have led to an underestimation of acute pulmonary vein reconnection. Third, while initial learning‐curve cases with the QDOT Micro catheter were excluded, procedural experience and workflow refinements over time may still have influenced procedural efficiency and outcomes. Although both strategies overlapped after the introduction of QDOT Micro, AI‐guided procedures were more frequent during the earlier part of the study period, and region‐specific procedures were performed later. Therefore, residual temporal confounding related to operator experience, catheter availability, and workflow maturation cannot be excluded. Moreover, longer follow‐up beyond 12 months would be desirable to more comprehensively assess the long‐term effectiveness of the region‐specific ablation strategy.

In addition, left atrial remodeling was evaluated using anteroposterior diameter rather than volumetric CT or MRI measurements, which may have limited the precision of atrial size as a predictor of outcomes. Rhythm monitoring was based on scheduled ECG and 24‐h Holter recordings, supplemented by symptom‐driven evaluations and wearable devices when available. As continuous monitoring was not systematically implemented, asymptomatic recurrences may have been underdetected, potentially leading to an underestimation of the true arrhythmia recurrence rate.

Finally, the study was underpowered to detect differences in infrequent safety outcomes. Although no statistically significant difference in complications was observed, the numerically higher number of clinically relevant complications in the region‐specific group warrants cautious interpretation and further evaluation in larger prospective cohorts. Furthermore, while no clinically apparent esophageal complications were observed, systematic post‐ablation endoscopy and luminal esophageal temperature monitoring were not performed. Therefore, asymptomatic or subclinical esophageal lesions cannot be excluded.

6. Conclusion

A region‐specific radiofrequency ablation strategy for pulmonary vein isolation was associated with improved procedural efficiency without a detected difference in acute efficacy or 12‐month arrhythmia recurrence compared with ablation index‐guided ablation. These findings suggest that tailoring energy delivery to regional atrial characteristics may represent a practical approach to optimize procedural workflow, with similar observed clinical effectiveness in this cohort.

Funding

This study was supported by this study received partial funding from the Consejería de Salud y Consumo of the Junta de Andalucía (HOR‐2023‐02) through the ERA4Health partnership (Agreement No. 101095426, EU Horizon Europe Research and Innovation Programme).

Ethics Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of HUPM (code: TFG‐AVPFA‐2023).

Consent

Informed consent was obtained from all individual participants included in the study.

Conflicts of Interest

J.F.‐A. reports consulting fees from Biosense Webster. The remaining authors declare no conflicts of interest.

Supporting information

Figure S1: Patient‐flow diagram.

Table S1: Univariable and multivariable predictors of first‐pass isolation in men.

Table S2: Univariable and multivariable predictors of first‐pass isolation in women.

Data Availability Statement

The anonymized dataset underlying this study is available in Zenodo at: https://doi.org/10.5281/zenodo.21859681.