Pulsed-field ablation versus radiofrequency or cryoballoon thermal ablation in atrial fibrillation: a systematic review and meta-analysis

Abstract

This review investigated efficacy and safety of pulsed-field ablation (PFA) in comparison with radiofrequency ablation (RFA), cryoballoon ablation (CBA), or both combined. The Odds ratio (OR) and mean difference (MD) with 95% confidence interval (95% CI) were computed. PFA allowed shorter procedure (MD −44.27 minutes, 95% CI: −63.61; −24.93) and left atrium (LA) dwell (MD -32.71 minutes (95% CI: −58.64; −6.78) times, but with longer fluoroscopy time than RFA (MD 8.54 minutes, 95% CI: 4.03; 13.04). Post-procedural complications rate was lower with PFA than CBA (OR 0.53, 95% CI: 0.35, 0.80). Atrial arrhythmias recurrence rate within one year of follow-up was lower with PFA than RFA (OR 0.68, 95% CI; 0.53; 0.87) and CBA (OR 0.69, 95% CI: 0.48; 0.97). PFA allowed shorter procedure and LA dwell times, as well as lower atrial arrhythmia recurrence than RFA and lower post-procedural complications and atrial arrhythmias recurrence rates than CBA.

Graphical Abstract

Open in new tabDownload slide

Study characteristics and procedure outcomes. Abbreviations: CBA, cryoballoon ablation; CHA₂DS₂-VASc, congestive heart failure, hypertension, age ≥ 75 (doubled), diabetes, stroke (doubled), vascular disease, age 65 to 74 and sex category (female); hs-TnT, high sensitivity cardiac troponin-T; LA, left atrium; LVEF, left ventricular ejection fraction; MD, mean difference; OR, odds ratio; RCTs, randomized controlled trials; RFA, radiofrequency ablation.

Introduction

Atrial fibrillation (AF) is the most frequent sustained cardiac arrhythmia, which is associated with morbidity and mortality. Pulmonary vein isolation (PVI), with catheter ablation, is considered for symptomatic patients with AF refractory to antiarrhythmic drugs. [1] Moreover, catheter ablation was superior to antiarrhythmic treatment in preventing the recurrence of atrial arrhythmia and improving symptoms, exercise capacity, and quality of life without serious adverse events. [2, 3] The international guidelines recommend catheter ablation for patients with recurrent AF, paroxysmal or persistent, that is resistant to antiarrhythmic drugs. Catheter ablation is also recommended as a first-line option in symptomatic patients with recurrent paroxysmal AF or in selected patients with persistent AF. [4, 5] Repeat catheter ablation for AF after the initial catheter ablation, is reasonable in patients with recurrent AF, provided that symptomatic improvement has been achieved after initial PVI. [5] The recommendations have been extended to specified patient populations such as AF patients with left ventricular systolic dysfunction with high probability to be associated with arrhythmia-induced cardiomyopathy. [4, 5] In addition, when AF is triggered by supraventricular tachycardia, catheter ablation of supraventricular tachycardia is reasonable. [4]

Catheter ablation with thermal energy sources, namely radiofrequency ablation (RFA) or cryoballoon ablation (CBA), heats or freezes cardiac tissue to isolate pulmonary veins, respectively, which suppresses the triggers of AF. Thermal ablation demonstrated excellent acute PVI success rates and long-term freedom of AF recurrence regardless of the techniques used. [6] However, unintended injury to the adjacent atrial or extracardiac structures can result in rare but serious complications (e.g. phrenic nerve palsy, pulmonary vein stenosis, and atrio-esophageal fistula) due to the indiscriminatory nature of thermal energy. [6, 7] Pulsed-field ablation (PFA) has attracted substantial attention in recent years due to its distinct tissue-preferential non-thermal ablation technology [8] that uses microsecond-scale electric pulses that damage the cell membranes, causing cell death through electroporation (i.e. creating pores in the myocardiocyte cell membrane) in a selective and tissue-specific way. Thus, this targets myocardial fibres and reduces the risk of damaging the adjacent non-cardiac tissues such as the esophagus, pulmonary veins, and phrenic nerve. [1, 7, 8] Although it has been suggested that delivering PFA at higher doses may cause damage to the collateral tissues, no evidence has shown the occurrence of pulmonary vein stenosis, or esophageal tissue and phrenic nerve injury. The initial clinical experience and studies found that PFA modality was effective and safe in patients with symptomatic paroxysmal AF with good durability and low recurrence of atrial arrhythmia. [9–11] Most of the published studies on PFA in treating AF were single-arm studies or registries. However, the studies that compare PFA with other thermal ablation modalities have been accumulating due to the substantial progress in the safety and efficacy of catheter ablation for AF and the notable advancement in ablation technologies. Yet, many important questions remain unresolved such as the ideal setting for RFA and CBA in different left atrium (LA) regions; the ideal approach for PFA (e.g. delivery, dose, system); long-term outcomes of PFA compared with thermal energy ablation; combining PFA and thermal energy ablation to improve the efficacy and safety of AF ablation; or the ideal catheter ablation approach for persistent AF. [4] Herein, this systematic review and meta-analysis investigated the efficacy and safety of PFA in comparison with RFA, CBA, or both combined.

Methods

This systematic review was conducted in accordance with the ‘Cochrane Handbook for Systematic Reviews, [12] and its reporting was in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement. [13, 14] It was registered in the International Prospective Register of Systematic Reviews with the registration number CRD42024534599.

Search strategy

A systematic literature search was independently conducted by two authors (D.A. and R.K.) using MEDLINE and EMBASE from inspection to April 22, 2024. Broad search terms “atrial fibrillation”, “ablation”, “pulsed field” and their combination were used without any limitations. Manual screening of the references of the selected studies was performed to identify more relevant trials. The search strategy is presented in Table S1.

Eligibility criteria

This systematic review included randomized controlled trials (RCTs) and cohort observational studies that recruited adult AF patients who were treated with their first ablation procedure. Eligible studies should compare the efficacy and/or safety of PFA with either RFA, CBA, or both combined (i.e. thermal energy ablation). The exclusion criteria included the studies investigating non-human subjects, undergoing redo ablation procedures, or being a single-arm design, case reports/series, or conference posters/proceedings.

Study selection and data extraction

The titles and abstracts of the search results were screened. All potential abstracts were retrieved in full text and reviewed in duplicate to confirm eligibility. Data was extracted from the original manuscripts and presented in the supplementary material by one author (R.K.), to ensure consistency, and verified by a second author (N.A.). The extracted data included study design, recruitment period, recruitment site(s), ablation modalities, selected demographics, co-morbidities, and outcomes.

The acute procedural outcomes were procedure time, LA dwell time, fluoroscopy time, and fluoroscopy dose. The definition of each outcome was according to each study. In general, procedure time was defined as the time from femoral puncture to catheter or sheath removal. LA dwell or retention time was defined as the time the ablation catheter remained in the LA (i.e. from transseptal puncture to removal of catheter from LA) following a successful ablation procedure. The main post-procedural complications outcomes were myocardial injury and all complications combined such as stroke, cardiac tamponade, phrenic nerve palsy, atrioesophageal fistula (or esophageal injury/edema), temporary ST-segment elevation (or myocardial infarction/mild chest pain), bleeding, or vascular access complications including groin hematoma. Myocardial injury was measured by high-sensitive cardiac troponin T (hs-cTnT) after the procedure. The main clinical outcomes were recurrence of AF or atrial arrhythmias and redo of ablation procedure. The analysis of a clinical outcome measure at a specific time point was conducted based on data availability.

Risk of bias and quality assessment

The revised Cochrane risk-of-bias (RoB-2) [15] and the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) risk of bias [16] tools, were used to assess methodological quality of RCTs and observational studies, respectively. RoB-2 tool has five domains of assessment and bias is judged as low risk, some concerns, or high risk. [15] ROBINS-I tool has seven domains and bias is judged as low, moderate, serious, critical risk, or no information. [16] Bias assessment for acute procedural and clinical outcomes was performed by two authors (D.A. and R.K.). Disagreement was solved by a third author.

Statistical analysis

A meta-analysis was performed using an aggregate approach. The odds ratio (OR) and the mean difference (MD) with 95% confidence interval (95% CI) were computed for categorical and continuous outcome endpoints, respectively, using a random-effects model due to the anticipated heterogeneity between studies. The values of continuous variables that were reported as median and interquartile range were converted into mean and standard deviation. A minimum of two studies were sufficient to conduct a quantitative data synthesis for each endpoint. [17] Statistical heterogeneity between the studies was examined by visual inspection of forest plots, 95% CI, Cochran’s Q statistics, and inconsistency factor (I²). The inconsistency factor (I²) value represents high heterogeneity if I² value is greater than 50%. [18] Subgroup analysis for clinical outcomes was conducted based on AF type (paroxysmal and/or persistent), the use of mapping system and intracardiac echocardiography during ablation procedure. Sensitivity analyses were performed by excluding or including studies in a meta-analysis to assess findings robustness. Publication bias was assessed by visual inspection of funnel plots. A significance level of <0.05 was used. Analyses were performed by statistical program R software (RStudio 2023.06.0 + 421 “Mountain Hydrangea” Release for windows).

Results

Search results

After excluding the duplicate records and those that did not meet the eligibility criteria (e.g. review, in-vitro study, simulation model, editorial, conference abstract, or studies on ablation techniques or mapping), 141 articles were retrieved in full texts. Thirty studies were eventually included, [19–48] of which two were RCTs [19, 20] and 28 were observational cohort studies (Fig. S1 and Tables S1, S2). [21–48]

Study characteristics

The studies enrolled patients between 2018 and 2023. Twenty and six studies were single- [19, 21, 22, 25, 27–30, 33–37, 39, 40, 43–48] and multi-centre, [20, 23, 24, 26, 31, 32] respectively. Almost all studies were conducted in Europe, [19, 21–25, 27–36, 38–48] except for a study that was conducted in China, [37] and an international multicentre trial [20] and its secondary analysis [26] that was based in the United States. Three studies [21, 29, 32] had their intervention group from other PFA initial studies. [10, 11] One of them did not report similar outcomes. [32] Two other studies were conducted in the same Swiss centre but none of their outcomes were pooled together. [24, 30] Another three studies enrolled their patients from the same Swiss registry. [25, 27, 38] One study only recruited patients with persistent AF and reported outcomes of the three ablation modalities (PFA, RFA, and CBA). [25] The second one only enrolled those with paroxysmal AF and compared PFA with any thermal ablation modality. [27] The third recruited patients with any AF type and compared PFA with CBA. [38] Common outcomes (i.e. procedure time, fluoroscopy time, and recurrence of atrial arrhythmia) between two of the three studies were not pooled together. [25, 38] Two studies that compared PFA with any thermal ablation recruited patients from the same French centre. [21, 29] However, sensitivity analysis addressed this issue. A total of 7167 patients were included in four groups (i.e. PFA: 2836, RFA: 1282, CBA: 2547, and any thermal energy source: 680). When combing patient groups in the studies that reported both RFA and CBA separately, the group of any thermal energy source would include 1911 patients. [21, 23–28] Nine studies enrolled patients with only paroxysmal AF, [20–23, 26, 27, 29, 32, 34] 18 studies with any AF type, [19, 24, 28, 30, 31, 33, 36–44, 46–48] and only one study recruited patients with persistent AF. [25] Table S3 details general study characteristics.

Patient characteristics

The age of patients was 53–71 (most studies: 60–69) years who were predominately males (33–83%) with a body mass index of 25–30 kg/m². The most reported comorbidities were hypertension (17–97%), diabetes (3–32%), sleep apnea (8–29%), and stroke (0–28%) with CHA₂DS₂-VASc score of 0.5–3.0 (most studies: 2.0) (Table S4). The use of antiarrhythmic drugs Class II and IV ranged from 6% to 91% and drugs Class I and III ranged from 5% to 78%. In most studies, all patients received oral anticoagulants with only one study reported a rate of 84% (Table S5). The main reported echocardiographic measurements were LA diameter (33–46 mm), LA volume index (30–67 ml/m²), and left ventricular ejection fraction of ≥50% (Table S6).

Study interventions

Fourteen studies compared PFA with RFA, [19, 23, 24, 26–28, 30–37] 17 studies with CBA, [23–28, 38–48] and 10 studies with either RFA or CBA (i.e. any thermal energy source ablation). [20–29] Six studies were in common among the three analysis groups, because they reported separate results for RFA and CBA (Table S3). [23–28] Popa et al investigated standard (30–40 W) and high-power short-duration (HPSD; 70 W and 90 W) ablation techniques. [34] Pentaspline FaraWave® catheter for PFA was the main catheter utilized in the studies. [19–23, 25–34, 36, 38–41, 43–48] Only Yang et al used a disposable catheter (PFA8D18L). [37] In RFA group, the following catheters were used, ThermoCool SmartTouch®, [19, 21] ThermoCool SmartTouch® Surround Flow, [23, 25, 27, 29] NaviStar ThermoCool®, [32, 35] TactiCath®, [28, 32] and HELIOSTAR® catheters. [33] Popa et al utilized FlexAbility® SE (standard RFA and HPSD-70 W) or QDOT MICRO® (HPSD-90 W) catheters. [34] In CBA group, several cryoballoon catheters have been used such as Arctic Front Advance®, [21, 25, 27–29, 45, 47] POLARx® cryoballoon, [38, 41, 43, 44] second generation Arctic Front Advance® cryoballoon, [39, 46] and fourth generation Arctic Front Advance® catheters. [48] Della Rocca et al allowed the use of Arctic Front Advance® PRO and POLARx® cryoballoon catheters. [23] In most studies, the procedures were conducted under deep conscious sedation (e.g. fentanyl, midazolam, and propofol), [19, 27–29, 33, 34, 36, 38, 39, 42–44, 46, 48] while in six studies the procedures were conducted under general anesthesia. [22, 23, 26, 31, 32, 35] Kueffer et al adopted deep conscious sedation for all patients but those who were considered high-risk patients underwent general anesthesia. [25] In the study by Chaumont et al, CBA procedures were performed under local anesthesia and conscious sedation, while PFA procedures were conducted under general anesthesia. [40] Ablation procedures characteristics are detailed in Table S7. Five studies reported first-pass isolation in both comparison groups, [20, 23, 30, 38, 46] and one study used it in the RFA group. [34] In three studies, the use of first-pass isolation was significantly higher in PFA group (Table S8). [23, 38, 46]

Outcome measurements

Acute PVI success was achieved in all patients in most of the studies. [21, 29, 33, 34, 36–38, 43, 46, 48] Tables S8 to S10 present main study findings including ablation procedure times, acute procedural complications, and clinical outcomes at follow-up.

Procedure time

Among the observational studies, procedure time was significantly shorter with PFA than RFA [MD -44.27 minutes (95% CI: −63.61; −24.93, P < .0001; I² = 100)] and any thermal ablation modality [MD -34.08 minutes (95% CI: −45.98; −22.18, P < .0001; I² = 99%)] but only numerically shorter than CBA [MD -6.19 minutes (95% CI: −14.14; 1.76, P = .1272; I² = 100%)] with significant heterogeneity (Fig. 1). The procedure time remained significantly shorter with PFA when the three RFA power intensities (standard RFA, HPSD-70 W, and HPSD-90 W) were combined from the study by Popa et al [MD -42.36 minutes (95% CI: −61.41; −23.30, P < .0001; I² = 100%); Fig. S2]; whereas the above analysis included standard RFA alone (Fig. 1, Panel A). [34] Pooled procedure time of the two RCTs did not yield difference between PFA and any thermal modality [MD -56.44 minutes (95% CI: −133.76; 20.87, P = .152; I² = 98.9%); Fig. S3].

Left atrial dwell time

Among the observational studies, LA dwell time was significantly shorter with PFA than RFA [MD -32.71 minutes (95% CI: −58.64; −6.78, P = .013; I² = 93%); Fig. S4], with significant heterogeneity, but only numerically shorter than CBA [MD -4.61 minutes (95% CI: −14.6994; 5.4875, P = .371; I² = 98.0%); Fig. S5]. The LA dwell time remained significantly shorter with PFA when the three RFA power intensities from the study by Popa et al [34] were combined [MD -26.33 minutes (95% CI: −40.21; −12.45, P = .0002; I² = 86.0%); Fig. S6]. The LA dwell time was numerically shorter when pooled in the two RCTs [MD -51.59 minutes (95% CI: −105.78; 2.60, P = .062; I² = 98.4%); Fig. S7].

Fluoroscopy time and dose

Among the observational studies, fluoroscopy time was significantly longer with PFA than RFA [MD 8.54 minutes (95% CI: 4.03; 13.04, P = .0002; I² = 99%)] but only numerically longer than CBA [MD 1.82 minutes (95% CI: −0.27; 3.91, P = .0876; I² = 97%)] or any thermal ablation modality [MD 4.48 (95% CI: −1.12; 10.08, P = .1166; I² = 99%)] (Fig. 2). The fluoroscopy time remained significantly longer with PFA when the three RFA power intensities from Popa et al study [34] were combined [MD 8.54 minutes (95% CI: 4.03; 13.04, P = .0002; I² = 99%); Fig. S8]. The pooled fluoroscopy time was significantly longer in the two RCTs as well [MD 4.85 minutes (95% CI: 0.35; 9.36, P = .035; I² = 93.2%); Fig. S9]. The respective fluoroscopy doses were significantly higher with PFA than RFA only [(PFA versus RFA: MD 2.29 Gycm², 95% CI: 0.34; 4.23, P = .0216; I² = 99%), (PFA versus CBA: MD -0.34 Gycm², 95% I: −0.89; 0.21, P = .2271; I² = 88%), and (PFA versus any thermal ablation modality: MD 0.89 Gycm², 95% CI: −0.09; 1.87, P = .0740; I² = 96%); Figs. S10–S12].

Myocardial injury

The range of hs-cTnT levels was 8–15 pg/ml at baseline which increased to 148–7170 ng/L after ablation procedures (Table S6). In the observational studies and as a measure of myocardial injury, hs-cTnT level after ablation was numerically higher with PFA than RFA [MD 1711.85 ng/L (95% CI: −620.71; 4044.42, P = .1503; I² = 100%)] but significantly higher in comparison with CBA [MD 449.84 ng/L (95% CI: 208.29; 691.39, P = .0003; I² = 95.0%)] (Fig. S13). Similarly, the hs-cTnT level remained numerically higher with PFA when the three RFA power intensities from the study by Popa et al [34] were combined [MD 1671.12 ng/L (95% CI: −684.13; 4026.37, P = .1643; I² = 100%); Fig. S14].

Clinical endpoints

Acute procedural complications

In the observational studies, combined post-procedural complications rate was significantly lower with PFA than CBA [OR 0.53 (95% CI: 0.35; 0.80, P = .0029; I² = 0%)] without heterogeneity. There were no statistically significant differences between PFA and RFA [(Standard intensity group: OR 1.57, 95% CI: 0.67; 3.64, P = .2968; I² = 31%), and (Combined intensity groups [34]: OR 1.58, 95% CI: 0.73; 3.40, P = .2426; I² = 32%)] or any thermal ablation modality [OR 0.55 (95% CI: 0.21; 1.43, P = .2216; I² = 59%)] (Figs. 3 and S15). Overall, the rates of individual post-procedure complications were generally low. The reported complications included stroke, cardiac tamponade, phrenic nerve palsy, atrioesophageal fistula, vascular access complication, and bleeding (Table S9).

Recurrence of atrial arrhythmia

In the observational studies, atrial arrhythmia recurrence rate within one year of follow-up was significantly lower with PFA than RFA [(Standard group: OR 0.68, 95% CI; 0.53; 0.87, P = .0023; I² = 0%) and (Combined intensity groups [34]: OR 0.68, 95% CI: 0.53; 0.86, P = .0016; I² = 0%); Figs. 4 and S16]. Similarly, recurrence rate was significantly lower in PFA than CBA group [OR 0.69 (95% CI: 0.48; 0.97, P = .0341; I² = 45%); Fig. 4]. Sensitivity analysis by removing the study with two-year follow-up from the CBA group [38] yielded similar result [OR 0.67 (95% CI: 0.44; 1.0, P = .0493; I² = 53%) within one-year follow-up (Fig. S17)]. There was no difference in the recurrence rate when comparing PFA with any thermal ablation modality [OR 0.71 (95% CI: 0.42; 1.19, P = .19; I² = 66%); Fig. 4].

Redo procedure

The rate of redo ablation procedure, within one year of follow-up, did not differ between the comparison groups [(PFA versus RFA: OR 0.78, 95% CI: 0.46; 1.34, P = .3697; I² = 45%),(PFA versus CBA: OR 0.77, 95% CI: 0.41; 1.46, P = .4251; I² = 72%), and (PFA versus any thermal ablation modality: OR 0.75, 95% CI: 0.55; 1.02, P = .0693; I² = 0%); Figs. S18–S20].

Subgroup analyses

Subgroup analysis was conducted to assess the impact of AF type on the clinical outcomes at follow-up. In the studies that enrolled patients with paroxysmal AF, the recurrence of atrial arrhythmias rate was significantly lower with PFA than RFA [OR 0.55 (95% CI: 0.37; 0.84, P = .005; I² = 0%); Fig. S21]. However, the rate did not differ in the studies that enrolled patients with any AF type [OR 0.78 (95% CI: 0.5083; 1.19, P = .2457; I² = 0%); Fig. S22]. The rate of redo ablation did not differ between PFA and RFA in the studies that enrolled patients with paroxysmal AF [OR 0.65 (95% CI: 0.40; 1.05, P = .079; I² = 0%); Fig. S23]. In comparison with CBA, atrial arrhythmias recurrence rate was significantly lower in PFA group in the studies that recruited patients with paroxysmal AF [OR 0.54 (95% CI: 0.35; 0.85, P = .0079; I² = 2%); Fig. S24] but not in the those enrolled any AF type [OR 0.77 (95% CI: 0.50; 1.18, P = .2223; I² = 48%); Fig. S25]. The redo procedure rate did not differ between PFA and CBA groups either in the studies that enrolled patients with paroxysmal AF [OR 0.86 (95% CI: 0.52; 1.42, P = .5602; I² = 0%); Fig. S26] or any type of AF [OR 0.75 (95% CI; 0.27; 2.13, P = .5923; I² = 86%); Fig. S27]. The comparisons with any thermal ablation modality from the studies that enrolled patients with paroxysmal AF showed lower recurrence of atrial arrhythmias with PFA [OR 0.55 (95% CI: 0.37, 0.84; P = .005, I² = 0.0%); Fig. S28] but not the rate of redo procedure [OR 0.74 (95% CI: 0.47, 1.17; P = .20, I² = 0.0%); Fig. S29] outcomes.

Another subgroup analysis was conducted for the studies that used mapping system. [27, 30, 32, 33, 35, 37, 39, 41, 48] The procedure time did not differ between the comparison groups [(PFA versus RFA: OR -32.23, 95% CI: −67.51;3.04, P = .073; I² = 99%) and (PFA versus CBA: OR 2.46, 95% CI: −13.31; 18.23, P = .76; I² = 89%); Figs. S30 and S31] but the fluoroscopy time was significantly longer with PFA [(PFA versus RFA: OR 10.61, 95% CI: 1.72; 19.51, P = .019; I² = 99%) and (PFA versus CBA: OR 5.02, 95% CI: 0.42; 9.63, P = .032; I² = 92.8%); Figs. S32 and S33]. Recurrence of atrial arrhythmias was significantly lower with PFA than CBA [(PFA versus RFA: OR 0.56, 95% CI: 0.30; 1.06, P = .075; I² = 0%) and (PFA versus CBA: OR 0.33, 95% CI: 0.13; 0.86, P = .023; I² = 0%); Figs. S34 and S35] without a difference in the acute procedure complications rate [PFA versus CBA: OR 3.52 (95% CI: 0.68; 18.24, P = .13; I² = 0%); Fig. S36]. Only two studies used intracardiac echocardiography during ablation procedure but they did not report the outcomes of interest. [32, 35] The sensitivity analyses by removing one of the two studies from the same French centre that compared PFA with any thermal ablation modalities [21, 29] were consistent with the overall results in terms of procedure and fluoroscopy times as well as acute procedural complications (Figs. 1,2,3 and S37–S42).

Risk-of-bias assessment

The overall risk of bias assessment of all studies according to ROBIN-I tool was judged as high risk for the procedure time, fluoroscopy time, and AF recurrence endpoints mainly due to confounding and participants selection domains (Tables S11–S13). The risk was judged as moderate for the redo procedure endpoint due to confounding and participants selection domains (Table S14). Sensitivity analyses were performed according to the overall risk of bias of an individual study by removing the studies with low and high risk of bias since the risk of bias of most individual studies was judged as moderate (Tables S11–S13). The procedure time results were similar to those of the overall studies in terms being significantly shorter with PFA than RFA [MD -47.46 (95% CI: −83.17; −11.74, P = .0092; I² = 99%); Fig. S43] and any thermal ablation modality [MD -43.19 minutes (95% CI: −62.53; −23.85, P < .0001; I² = 78%); Fig. S44]. Whereas, different in terms of being significantly shorter than CBA [MD -10.22 (95% CI: −18.83; −1.61, P = .0200; I² = 100%); Fig. S45]. The fluoroscopy time did not differ between any of the comparison groups (Figs. S46–S48) which was significantly longer with PFA than RFA when pooled for all studies regardless of bias judgement (Fig. 2). The recurrence of atrial arrhythmias did not differ between PFA and the comparison groups RFA and CBA (Fig. S49 and S50). In addition, redo procedure rate did not differ between PFA and CBA groups (Fig. S51). For the RCTs, the overall risk was judged as having some concerns for the procedure time, fluoroscopy time, and LA dwell time endpoints due to some concerns in the randomization process domain, according to the revised RoB-2 tool (Tables S15–S17).

Publication bias

Funnel plots of the reported outcomes showed asymmetrical patterns (i.e. presence of publication bias) in most of the findings. Some funnel plots showed less asymmetry and a few showed symmetrical pattern indicating potential or absence of publication bias, respectively (Figs. S2–S56).

Discussion

The present systematic review and meta-analysis of two RCTs and 28 observational studies explored the efficacy and safety of PFA compared with thermal energy ablation modalities in AF patients. In comparison with RFA, PFA allowed shorter procedure and LA dwell times, but with longer fluoroscopy time and more myocardial injury without a difference in acute procedural complications or redo procedure at follow-up. The fluoroscopy time did not differ between the groups when pooling it from the studies of moderate risk of bias. Recurrence of atrial arrhythmia was lower with PFA within one-year of follow-up. The comparisons between PFA and CBA did not show statistical difference in terms of procedure, LA dwell, and fluoroscopy times. However, the procedure time was significantly shorter with PFA than CBA when pooling it from the studies of moderate risk of bias. The rates of post-procedural complications and recurrence of atrial arrhythmias, but not redo procedure, were significantly lower with PFA than CBA. Myocardial injury was higher with PFA than CBA (Graphical abstract). PFA resulted in shorter procedure time when compared with any thermal ablation modality. Pooled data from the two included RCTs showed only longer fluoroscopy time with PFA. However, the comparison groups differed between the RCTs. One RCT compared PFA against any thermal modality, [20] and another against RFA. [19]

Most of the published systematic reviews and meta-analyses are non-comparative meta-analyses of single proportions which pooled data from different study designs including single-arm studies. [49–54] Overall, PFA procedure has been proved safe and effective in treating AF as it showed high procedural success rate with low procedural complications and major adverse events rates. [49–51] The pooled procedural success rate was 99.7%, [51] procedure time was 83–94 minutes, fluoroscopy time 14.4–17 minutes, [50, 51] and arrhythmia recurrence rate during follow-up (i.e. after the blanking period) was 17.3%. [51] Maccioni et al computed an indirect comparison between PFA and CBA and found a more favourable safety profile with PFA than CBA in terms of lower risk of major adverse events. [52] Similarly, Aldaas et al demonstrated lower periprocedural complications rates with PFA than any thermal ablation modality without a difference in the rates of acute procedural success or AF recurrence. [53] Serban et al reported no difference in the durability-related outcome rates between four ablation techniques (PFA, RFA, CBA, and laser balloon) in terms successful ablation or PVI at baseline and maintaining durably isolated pulmonary veins during follow-up. [54] Zhang et al, in their meta-analysis of 15 studies (n = 1880) compared the efficacy and safety between PFA and CBA in AF. [55] All the 15 studies were included in the present meta-analysis except one of them because the study assessed the impact of PFA on the ganglionated plexi without comparison between catheter ablation modalities in terms of the procedural and clinical outcomes (Table S2). [56] The findings of the meta-analysis by Zhang et al were in line with those in the present meta-analysis in terms of fluoroscopy time and cardiac troponin levels but not in terms of recurrent atrial arrhythmia, periprocedural complications, and procedure time. [55] The latter discrepancy can mainly be explained the difference in the number of the included studies and the inclusion of different study designs in meta-analysis by Zhang and colleagues.

To the best of our knowledge, this is the first meta-analysis with an objective to compare PFA with either RFA or CBA. The recently published meta-analysis by Aldaas et al was the first meta-analysis of six studies (n = 1012) to compare PFA with any thermal energy ablation without a distinction between RFA and CBA modalities. [57] They pooled six studies, one RCT [20] and five non-RCTs, [21, 27, 29, 32, 39] which were also included in the present meta-analysis. In addition, while submitting the present meta-analysis, de Campos et al published their meta-analysis that compared efficacy and safety outcome between PFA and thermal ablation combined with subgroup analysis for each thermal ablation modality. [58] They identified 18 studies that are included in the present meta-analysis as well. [19, 20, 23, 27–29, 33, 34, 36–39, 41, 43, 44, 46–48] Similar to Aldaas et al, [57] de Campos et al pooled RCTs with non-RCTs but we opted not to pool variables from different study designs. Both meta-analyses found that PFA was associated with significantly shorter procedure time without a difference in acute procedural complications or rate of atrial arrhythmias recurrence when compared with any thermal ablation modality. Moreover, both meta-analyses demonstrated that PFA was associated with significantly longer fluoroscopy time than any thermal ablation modality. [57, 58] Aldaas et al attributed the increased fluoroscopy time with PFA to the short-term experience of the operator and the wide use of electro-anatomical non-fluoroscopic mapping systems with thermal ablation. They suggested that the fluoroscopy time should become shorter with more PFA utilization in practice and the incorporation of mapping systems with PFA. [57] However, our subgroup analysis according to the use of mapping system found no difference in the procedure time between comparators but longer fluoroscopy time with PFA.

The present meta-analysis showed a significantly lower acute complications rate by 53% with PFA than CBA. However, we could not analyze the acute complications separately due to the low incidence rate of each complication (e.g. stroke, cardiac tamponade, phrenic nerve palsy, atrioesophageal fistula, vascular access complication, and bleeding), and most studies reported zero event (Table S9). While PFA may avoid the thermal injury associated with RFA or CBA, certain acute complications such as pericardial effusions and tamponade require careful attention because of their consequent morbidity and mortality. The incidence of cardiac tamponade with PFA is ~1% of patients with 0.2% requiring cardiac surgery. Interestingly, the occurrence of pericardial tamponade has been associated with operator’s primary ablation technique, particularly when it was RFA. However, this can be attributed to several reasons such as the use of first generation PFA catheters (i.e. over-the-wire approach), operator learning curve, and the lack of force-sensing catheter in the initial PFA era, that might cause the application of excessive force against the myocardium. [59] In addition, there are unique complications associated with PFA such as coronary microvascular dysfunction that may occur during PFA, namely with monophasic ablation, which leads to the contraction of skeletal muscle and alterations in the systemic hemodynamic parameters. Coronary dysfunction may precipitate heart failure and the subsequent potential recurrent arrhythmias and AF. Another unique complication associated with PFA is called arcing, a phenomenon that causes gas accumulation due to the intense current density at the electrode-tissue interface, which may eventually cause myocardial damage. [60] The gaseous micro-emboli occurring during cardiac ablation were reported with thermal and non-thermal ablation. Although pulsed electric field does not employ thermal energy, microbubbles can occur. It has been reported that such microbubbles disappear after a short time without causing any cerebrovascular risk. However, emerging evidence suggested that asymptomatic cerebral embolism would occur in certain patients undergoing PFA. [60]

The rates of atrial arrhythmia recurrence within one year of follow-up was significantly lower by 32% and 31% with PFA than RFA and CBA, respectively. The ablation procedure is considered efficient when it is rapid and successful without an increased complications due to gap searching and repeated ablation. PFA provides a high rate of single-shot PVI and short ablation time regardless of the usual 20-minute LA dwell time. [61] The first-generation PFA ablation system design requires, for an optimum consolidation, catheter-tissue contact or proximity and additional catheter manipulations such as rotation and different configurations. With more familiarity with the system and improved learning curve, the fluoroscopy time may be reduced even for relatively complicated cases. [61] Schmidt et al, in the EU-PORIA (EUropean Real World Outcomes with Pulsed Field AblatiOn in Patients with Symptomatic AtRIAl Fibrillation) multicentre registry, have treated the first 1233 cases with FARAPULSE® PFA system when was commercially available, [62] then, five centres from the EU-PORIA registry have participated in the MANIFEST-PF (Multi-National Survey on the Methods, Efficacy, and Safety on the Post-Approval Clinical Use of Pulsed Field Ablation) registry (n = 1568). [63] Overall, there was short procedure time (i.e. median of 58 minutes) and short learning curve without appreciable difference between the 42 operators. Operators demonstrated good performance, i.e. 75% of them did less than 25 cases in centres with volumes ranged between 78 and 347 cases. [64] Ruwald et al found that the procedure and fluoroscopy times significantly decreased over the course of 121 ablation procedures. [65] PFA has transformed from a research modality to an everyday AF ablation technique. [64] In a contemporary real-world study (n = 707) of the three ablation modalities by Calvert et al, skin-to-skin and catheter laboratory times were shorter with PFA (68 and 102 minutes) than with RFA (89 and 123 minutes) and CBA (91 and 122 minutes) (P < .001 for both times), respectively, without a difference in major complications rates which were low (1.0%) in general. Although shorter times led to lower costs (i.e. staffing and laboratory) with PFA, the overall cost was higher (£10 010, P < .001) than that with RFA (£8949) and CBA (£8106) due to significantly higher equipment cost. [66] Another study by Duxbury et al found that routine PFA using the pentaspline catheter might be as affordable as the cryoablation. [67]

Our subgroup analysis found that atrial arrhythmias recurrence rate was significantly lower with PFA than RFA in the studies that recruited patients with paroxysmal AF but did not differ in the studies that enrolled patients with any AF type. It was reported that success rates were more modest in patients with persistent AF compared to those with paroxysmal AF which may be attributed to the progressive cardiac structural and electrical remodelling in patients with persistent AF. Thus, generating a substrate to initiating and maintaining arrhythmia that requires more extensive ablation techniques. [1] Reddy et al in their single-arm study, found that PFA for both PVI and adjacent LA posterior wall (LAPW) may be a safe and durable approach since PVI alone may be inadequate for many patients with persistent AF. [68] LAPW ablation adjunctive to PVI can increase the freedom from atrial arrhythmia rates and decrease recurrence rates, [69] especially in patients with persistent AF [70] without a difference in post-procedural adverse events. [69, 70] Although, the rate of acute post-procedural complications was significantly lower with PFA than CBA, the level of hs-cTnT level after ablation, as a measure of myocardial injury, was significantly higher with PFA than CBA. It has been reported that hs-cTnT release post PFA was ~1.6 and 1.9 folds higher than RFA and CBA, respectively. [24] Intuitively, the troponin rise could be interpreted by a larger damage of cardiac tissue induced by PFA. [34] The imaging study Nakatani et al found that PFA was acutely associated with a 60% increased late gadolinium enhancement compared with thermal ablation (P < .001), but with 20% smaller edema (P = .002) and no microvascular damage or intramural haemorrhage. However, in the chronic stage, most late gadolinium enhancement lesions disappeared after PFA but maintained after thermal ablation. Thus, proposing a reparative process that involves less chronic fibrosis. [29]

Limitations

The present analysis has important limitations. It is a study-level, not a patient-level, meta-analysis. Thus, inter-study heterogeneity cannot be ruled out. The pooled analyses were mainly derived from observational studies that are inherently prone to selection bias and confounding. Thus, a conservative random-effects model was considered. There was a concern about confounding potential and randomization process in the included observational and randomized studies, respectively. The absence of randomization in observational studies does not allow a causality conclusion but a hypothesis generation. There was a clear heterogeneity among studies in terms of AF type, procedure protocols, use of antiarrhythmic drugs, and monitoring of AF at follow-up visits (i.e. use of electrocardiogram, Holter monitors, or event monitors, etc.). Moreover, funnel plots of the reported outcomes showed publication bias in most of the findings. There was not enough data to pool to allow examining the impact of first-pass isolation in the studies that reported it. [20, 23, 30, 38, 46] It has been reported that HPSD ablation led to higher rates of first-pass isolation than low-power long-duration ablation and consequently to higher rates of maintaining sinus rhythm at 12-month follow-up. [71] Furthermore, the absence of first-pass isolation was associated with worse poor PVI durability (i.e. increased spontaneous pulmonary veins reconnection), hence poor ablation outcomes. [72] Most studies reported arrhythmia recurrence with electrocardiogram or Holter monitoring rather than continuous invasive methods (e.g. implantable loop recorders), which may lead to a potential underreporting of recurrent arrhythmia (i.e. failed detection of subclinical AF). Moreover, follow-up period was limited to one year. Although the same PFA catheter was used in most studies, the comparator modalities utilized various ablation systems (i.e. sheath, mapping catheter, ablation catheter, and generator). The generalizability of results may be limited due to the stated heterogeneity and the fact that the studies were primarily conducted in Europe. However, in comparison with RCTs, real-word evidence may have a better generalizability. [73]

Future directions

Table 1 summarizes the comparison between the current ablation modalities. [61, 74] From the clinical aspect of catheter ablation, the ADVENT trial [20] was the first and currently the only well-powered RCT to compare between PFA and thermal ablation. Other powered RCTs, ideally sham-controlled, are needed to investigate longer term outcomes and technical aspects of ablation systems such as catheter design, ablation dosing, and procedure workflow. Two single-arm studies are ongoing and their results are awaited to inform about the long-term clinical outcomes related to PFA modality using the FARAPULSE® (NCT05443594) and VARIPULSE® (NCT05293639) PFA systems. Another important clinical aspect focuses on patient-tailored approach in management. The recurrence of AF remains a significant challenge for patients undergoing catheter ablation, i.e. recurrence rate of ~35% at 12 months post catheter ablation. Thus, the prediction of AF recurrence after catheter ablation is of importance in the selection of patients and their management. Conventional prediction methods involve using the univariate predictors and scoring systems. [75] For example, increased LA volume and diameter are factors for an increased recurrence rate. LA strain and LA ejection fraction may indicate the progression of the condition. [75, 76] Prediction models generated by artificial intelligence and cardiac imaging can also improve the selection of patients and their management. [75] From the technical aspect, there are several potential directions to further optimize the PFA technology including the parameters of the pulsed electric field, PFA system (i.e. ablation catheter, pre-ablation mapping system), and ablation position (i.e. endothelium, epithelium). [60] For the parameters of the pulsed electric field, PFA may have certain thermal effect that may damage cells and tissues during the procedure, therefore, the use of low frequency pulsed electric field is considered the main method to decrease such thermal damage. [60] The integration of PFA in a 3-D mapping systems may decrease or eliminate fluoroscopy while deliver an accurate point-by-point energy. [74] Several PFA catheter ablation systems are being investigated at the preclinical and clinical stages. [77] Lastly, deciding on the optimal ablation position may necessitate finding supplemental targets for ablation because PFA, in comparison with thermal ablation has a more limited tissue penetration, which would question its thoroughness and persistence. Moreover, since persistent AF has an underlying complex mechanism, simple PVI is insufficient in achieving persistent therapeutic effect; thus, identifying new targets (e.g. LAPW) for ablation can improve prognosis after ablation for persistent AF. [60]

Conculsion

Pulsed-field ablation showed more favorable outcomes in terms shorter procedure and LA dwell times, as well as lower atrial arrhythmia recurrence rates than RFA. It had lower rates of acute procedural complications and recurrence of atrial arrhythmias than CBA. However, PFA was associated with longer fluoroscopy time and more myocardial injury.

Acknowledgements

Thanks to all the peer reviewers and editors for their time in the first place and for their valuable opinions and suggestions which helped refine and improve the manuscript.

Author Contributions

All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.

Conflict of interest: The authors declare no conflict of interest.

Funding

None declared.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate

Not applicable.

Main messages

1. Catheter ablation through pulmonary vein isolation is an effective approach for symptomatic patients with atrial fibrillation.

2. Thermal ablation demonstrated excellent acute pulmonary vein isolation success rates and long-term freedom of atrial fibrillation recurrence regardless of the techniques used, but may be associated with unintended injury to the adjacent atrial or extracardiac structures.

3. Pulsed-field ablation had more favourable outcomes in terms shorter procedure time and lower rate of atrial arrhythmia recurrence when compared with any thermal ablation modality.

Current research questions

1. What is the value of diagnostic cardioversion for persistent atrial fibrillation in steering management of atrial fibrillation?

2. Does the decision on continuing oral anticoagulation is based on stroke risk scores and irrespective of having atrial fibrillation episodes for patients undergoing successful catheter ablation?

3. What are the optimal catheter ablation techniques and strategy for patients after the first failed catheter ablation for paroxysmal atrial fibrillation or for those with persistent atrial fibrillation?

Self-assessment questions

True or false?

1. Catheter ablation with radiofrequency ablation freezes tissue to isolate pulmonary veins which suppress the triggers of atrial fibrillation.

2. Pulsed-field ablation targets myocardial fibres and reduces the risk of damaging the adjacent non-cardiac tissues such as the esophagus, pulmonary veins, and phrenic nerve.

3. Pulsed-field ablation procedure has proved safe and effective in treating AF as it showed high procedural success rate with high procedural complications and major adverse events rates.

4. Atrial fibrillation recurrence rate after catheter ablation may differ according to atrial fibrillation type, i.e. paroxysmal or persistent atrial fibrillation.

5. The two ongoing studies that will inform about the long-term clinical outcomes related to pulsed-field ablation using FARAPULSE® and VARIPULSE® systems are sham-controlled intervention studies.

Answers

1. False

2. True

3. False

4. True

5. False

References