High-power short duration vs. conventional radiofrequency ablation of atrial fibrillation: a systematic review and meta-analysis

Abstract

Aims

We sought to compare the effectiveness and safety of high-power short-duration (HPSD) radiofrequency ablation (RFA) with conventional RFA in patients with atrial fibrillation (AF).

Methods and results

MEDLINE, Cochrane, and ClinicalTrials.gov databases were searched until 15 May 2020 for relevant studies comparing HPSD vs. conventional RFA in patients undergoing initial catheter ablation for AF. A total of 15 studies involving 3718 adult patients were included in our meta-analysis (2357 in HPSD RFA and 1361 in conventional RFA). Freedom from atrial arrhythmia was higher in HPSD RFA when compared with conventional RFA [odds ratio (OR) 1.44, 95% confidence interval (CI) 1.10–1.90; P = 0.009]. Acute PV reconnection was lower (OR 0.56, P = 0.005) and first-pass isolation was higher (OR 3.58, P < 0.001) with HPSD RFA. There was no difference in total complications between the two groups (P = 0.19). Total procedure duration [mean difference (MD) −37.35 min, P < 0.001], fluoroscopy duration (MD −5.23 min, P = 0.001), and RF ablation time (MD −16.26 min, P < 0.001) were all significantly lower in HPSD RFA. High-power short-duration RFA also demonstrated higher freedom from atrial arrhythmia in the subgroup analysis of patients with paroxysmal AF (OR 1.80, 95% CI 1.29–2.50; P < 0.001), studies with ≥50 W protocol in the HPSD RFA group (OR 1.53, 95% CI 1.08–2.18; P = 0.02] and studies with contact force sensing catheter use (OR 1.65, 95% CI 1.21–2.25; P = 0.002).

Conclusion

High-power short-duration RFA was associated with better procedural effectiveness when compared with conventional RFA with comparable safety and shorter procedural duration.

Introduction

Catheter ablation for atrial fibrillation (AF) has been shown to reduce mortality, improve quality of life, and freedom from atrial arrhythmia compared with medical therapy.¹ Radiofrequency ablation (RFA) has become a cornerstone modality for catheter ablation of AF since the discovery of pulmonary vein (PV) triggers in 1998.^2,3 Technological advancements over the years such as irrigated and contact-force sensing RFA catheters have further improved effectiveness and outcomes.⁴ Successful pulmonary vein isolation (PVI) requires the creation of a contiguous, transmural lesion set avoiding collateral injury to surrounding tissues during ablation.⁵ Conventional RFA strategies have used low-to-moderate power, long-duration settings (usually 25–35 W for 30–60 s per lesion) to produce durable lesions while minimizing complications.⁶ High-power, short-duration (HPSD) RFA PVI was first reported in 2006 as an alternative strategy for improving procedural efficiency.⁷ High-power short-duration RFA strategy uses higher power (usually 45–50 W) and shorter duration (2–10 s per lesion on the posterior wall and 5–15 s at other sites in the left atrium).^7,8 High-power short-duration RFA approaches have gained in popularity with a recent study suggesting improved effectiveness and fewer complications.⁸ However, observational studies comparing HPSD RFA with conventional RFA have provided variable results regarding the effectiveness, safety, and procedural characteristics of the two strategies. At this time, no large randomized controlled studies or meta-analysis has been performed to compare the HPSD RFA with conventional RFA.

We performed a comprehensive meta-analysis of existing studies to compare the effectiveness, safety, and procedural characteristics of HPSD RFA and conventional RFA for the initial ablation of patients with paroxysmal and persistent AF.

Methods

Search strategy and study selection

The meta-analysis was performed per the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines.⁹ We searched MEDLINE, Cochrane, and ClinicalTrials.gov databases systematically from inception through 15 May 2020 for relevant published articles comparing HPSD RFA and conventional RFA in patients with AF. The search involved the following keywords: [(‘high power’ or ‘higher power’ or ‘high output’ or ‘higher output’ or ‘short duration’ or ‘shorter duration’) AND (‘ablation’ or ‘radio-frequency’ or ‘radiofrequency’ or ‘RF’ or ‘pulmonary vein’ or ‘isolation’ or ‘atrial fibrillation’ or ‘AF’)]. We also reviewed the references in the retrieved articles, relevant reviews, and prior meta-analyses for additional published data.

Eligibility criteria

The studies included fulfilled the following criteria: (i) patients with age ≥18, with paroxysmal and/or persistent AF undergoing initial catheter ablation; (ii) compared HPSD RFA with conventional RFA; and (iii) reported outcome data including but not limited to freedom from atrial arrhythmia, complication rates, redo-ablation procedures, procedure time, and fluoroscopy time.

The exclusion criteria were as follows: (i) conference abstracts, case reports, review articles, or non-English articles. (ii) Studies evaluating patients with prior catheter ablation for AF.

Authors (A.P., V.R., and H.D.H.) screened articles for inclusion independently, retrieved potentially relevant articles, and determined their eligibility.

Data extraction and quality assessment

Using a standardized protocol, all relevant information from each article was extracted by three authors (A.P., Q.A., and V.R.). The quality of studies used in the analysis was assessed using the Cochrane Collaboration risk of bias tool and NewCastle Ottawa scale.¹⁰ Authors (A.P., Q.A., and V.R.) independently assessed study quality. Freedom from atrial arrhythmia was defined as the absence of recurrence of AF, atrial tachycardia, or atrial flutter lasting >30 s, after the 3-month blanking period post-ablation, while not on anti-arrhythmic drug therapy. If only freedom from AF was provided in a study, then that was analysed under freedom from atrial arrhythmia. Acute PV reconnection was defined by studies as evidence of reconnection after PVI during a waiting period of 20 min or pharmacological provocation with adenosine or isoproterenol.

Statistical analysis

The principal summary effects measures were the odds ratio (OR) and corresponding 95% confidence intervals (CIs) estimated by using Mantel–Haenszel random-effects model.¹¹ Summary effects measures for continuous outcomes were the mean difference (MD) and the corresponding 95% CI estimated using the inverse-variance method. A two-sided P-value ≤0.05 was considered statistically significant. We conservatively used a priori, the random-effects model assuming substantial variability in the treatment effect size across studies.¹¹ We also planned a priori subgroup analysis based on (i) type of AF (paroxysmal, persistent), (ii) studies using ≥50 W power in the HPSD group, (iii) studies using contact force sensing ablation catheters, and (iv) studies with ablation lines in left atrium addition to PVI. Subgroup analysis of ≥50 W and contact force sensing catheters was planned as we opined that would be more reflective of contemporary real-world practice. If the studies including both paroxysmal and persistent AF patients did not provide separate outcomes based on the type of AF, then studies with at least 75% of patients with paroxysmal AF were included in the subgroup analysis of paroxysmal AF. The presence of statistical heterogeneity was evaluated by Cochran’s Q test I² statistic: when I² values were >50%, we explored individual study characteristics.¹² Sensitivity analysis was also performed based on the quality of follow-up monitoring for recurrence of AF after ablation. Publication bias was assessed using Funnel plots.¹³ Statistical analysis was performed using Review Manager (RevMan), Version 5.3, Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014.

Results

A total of 15 articles,^7,14–27 with 3718 adult patients were included in our meta-analysis as shown in Figure 1. Among these patients, 2357 underwent HPSD RFA and 1361 patients received conventional RFA. Among the 15 studies, 9 were prospective cohort studies,^{7,14,17,18,20,22,23,25,26} 2 were retrospective cohort studies,^15,19 1 was a propensity matched observation study,²⁴ and 3 studies used a historical cohort for the conventional RFA group.^16,21,27 The time of enrolment of the included studies is shown in Figure 2. Baseline characteristics among the included studies are shown in Tables 1 and 2. The mean age ranged from 51 to 69 years. Hypertension was the most common comorbidity. There were 2240 (60%) patients with paroxysmal AF and 1478 (40%) patients with persistent AF. Mean LVEF ranged from 55% to 62%.

Primary pooled analysis

Freedom from atrial arrhythmia at the end of study was higher in the HPSD RFA group when compared with the conventional RFA group (OR 1.44, 95% CI 1.10–1.90; P = 0.009) (Figure 3). All ORs are reported as HPSD RFA compared with conventional RFA. Acute PV reconnection was lower (OR 0.56, 95% CI 0.38–0.85; P = 0.005) and first-pass isolation was higher (OR 3.58, 95% CI 1.93–6.61; P < 0.001) with HPSD RFA when compared with conventional RFA. There was no difference between the two groups in mortality (OR 1.28, 95% CI 0.58–2.86; P = 0.54) and total complications (OR 0.83, 95% CI 0.63–1.09; P = 0.19) (Figure 3). There was no difference between the two modalities for the complications of stroke (OR 1.13, 95% CI 0.37–3.41; P = 0.83), pericardial effusion (OR 0.35, 95% CI 0.08–1.64; P = 0.18), phrenic palsy (P = 0.69), or PV stenosis (P = 0.52).

Total procedure duration was significantly lower in the HPSD RFA group compared with conventional RFA group [MD −37.35 min (95% CI −48.3 to −26.4); P < 0.001] (Figure 4). Total fluoroscopy duration was also lower with HPSD RFA compared with conventional RFA [MD −5.23 min (95% CI −8.11 to −2.35); P < 0.001] (Figure 3). RF ablation time [MD −16.26 min (95% CI −19.22 to −13.30); P < 0.001] and left atrial time [MD −25.7 min (95% CI −32.75 to −18.65); P < 0.001] were lower in HPSD RFA. There was no difference between the two groups in average impedance drop per lesion [MD −0.92 Ω (95% CI −1.99–0.16); P = 0.10].

There was significant heterogeneity with I² > 50% for the outcomes of procedure duration (93%), fluoroscopy duration (96%), RF ablation time (91%), and left atrial dwell time (76%). All summary estimates from pooled analyses were made using a random-effects model rather than a fixed-effects model to reduce the influence of heterogeneity between studies. There was no change in heterogeneity with the sequential exclusion of studies for the above outcomes. Sensitivity analysis demonstrated the robustness of all the above results during the sequential exclusion of studies. Sensitivity analysis following exclusion of three studies^7,20,21 based on the quality of follow-up monitoring for recurrence of AF demonstrated results similar to primary pooled analysis. Three studies^7,15,19 reporting only freedom from AF were excluded, HPSD RFA still demonstrated higher freedom from atrial arrhythmia (OR 1.55, 95% CI 1.11–2.18; P = 0.01). Exclusion of three studies^16,21,27 which used a historical cohort in the convention RFA group also yielded similar results to primary pooled analysis. Funnel plots of the outcomes of the primary pooled analysis are shown in Supplementary material online, Figure S1. Newcastle Ottawa scale demonstrated that all studies included were of good quality (Supplementary material online, Figure S2).

Subgroup analysis

Paroxysmal atrial fibrillation

There were five studies^{16,17,23,25,26} which only included paroxysmal AF, one study¹⁸ which provided separate outcome data for patients with paroxysmal AF and two studies^20,22 which involved >75% patients with paroxysmal. A total of 965 patients from these eight studies (454 in HPSD RFA and 511 in conventional RFA) were included in the subgroup analysis of paroxysmal AF. Among them 941 (98%) of patients had paroxysmal AF. There was higher freedom from atrial arrhythmia in HPSD RFA compared with conventional RFA (OR 1.85, 95% CI 1.31–2.61; P < 0.001; Supplementary material online, Figure S3). Acute PV reconnection was lower in HPSD RFA (OR 0.62, 95% CI 0.39–0.99; P = 0.04). First-pass PVI was higher in HPSD RFA (OR 4.36, 95% CI 1.94–9.81; P < 0.001). There was no difference in total complications (OR 0.71, 95% CI 0.38–1.32; P = 0.28). Total procedure duration [MD −28.36 min (95% CI −35.65 to −21.07); P < 0.001] and RF ablation time [−15.45 min (95% CI −19.53 to −11.37); P < 0.001] were significantly shorter in the HPSD RFA group. There was no difference in fluoroscopy duration [MD −1.68 min (95% CI −3.58 to 0.22); P = 0.08].

We could not perform a subgroup analysis of patients with persistent AF as there was only one study¹⁸ which provided separate outcome data for persistent AF. There were three studies^14,15,24 which each included about 50% of patients with persistent AF. An analysis of these four studies involving 2268 patients was performed. They included 1286 (57%) patients with persistent AF (982 in HPSD RFA and 304 in conventional RFA). There was no difference in freedom from atrial arrhythmia (OR 1.05, 95% CI 0.73–1.52; P = 0.78). There was no difference in total complications (OR 0.97, 95% CI 0.7–1.33; P = 0.84). Total procedure duration [MD −54.83 min (95% CI −84.94 to −24.71); P < 0.001] and RF ablation time [MD −16.2 (95% CI −20.06 to −12.34); P < 0.001] were significantly shorter in the HPSD RFA group. There was no difference in fluoroscopy duration [MD −7.13 min (95% CI −16.52–2.26); P = 0.14].

Studies with ≥50 W in high-power short-duration radiofrequency ablation group

There were 10 studies^{14,15,17,19,20,22–27} with a total of 2954 patients (1994 in HPSD RFA and 960 in conventional RFA) that involved the use of ≥50 W in the HPSD RFA group. High-power short-duration demonstrated higher freedom from atrial arrhythmia compared with conventional RFA (OR 1.53, 95% CI 1.08–2.18; P = 0.02) (Supplementary material online, Figure S4). There was no difference in total complications (OR 0.83, 95% CI 0.61–1.13; P = 0.25) as well as in the individual complications of stroke, pericardial effusion, PV stenosis, and phrenic palsy. Total procedure duration [MD −40.06 min (95% CI −54.92 to −25.2); P < 0.001], fluoroscopy duration [MD −5.5 min (95% CI −9.66 to −1.34); P = 0.01], left atrial dwell time [MD −28.42 min (95% CI −37.84 to −19); P < 0.001], and RF ablation time [MD −17.74 min (95% CI −20.39 to −15.08); P < 0.001] were significantly shorter in the HPSD RFA group.

Studies using contact force sensing ablation catheter

There were 10 studies with a total of 1198 patients (573 in HPSD RFA and 625 in conventional RFA) that used contact force sensing ablation catheters.^{16–22,25–27} HPSD RFA demonstrated higher freedom from atrial arrhythmia compared with conventional RFA [OR 1.65 (95% CI 1.21–2.25); P = 0.002] (Supplementary material online, Figure S5). Total procedure time (MD −28.84 min, P ≤ 0.001), fluoroscopy time (MD −4.09 min, P = 0.007), RF ablation time (MD −14.65 min, P < 0.001), and left atrial time (MD −25.53 min, P < 0.001) were significantly shorter in the HPSD RFA group.

Studies with ablation lesions in left atrium in addition to pulmonary vein isolation

There were 7 studies with a total of 2701 patients (1866 in HPSD RFA and 835 in conventional RFA) that included patients who underwent additional non PVI ablation.^{14,15,20,22,24,26,27} There was no difference in freedom from atrial arrhythmia between HPSD and conventional RFA (OR 1.37, 95% CI 0.93–2.01; P = 0.11; Supplementary material online, Figure S6). There was no difference in total complications between the two groups (OR 0.99, 95% CI 0.72–1.36; P = 0.93). Total procedure time (MD −52.03 min, P < 0.001) and RF ablation (MD −17.12 min, P < 0.001) were significantly shorter in HPSD RFA.

Discussion

Our study is the first meta-analysis providing a comprehensive comparison of HPSD RFA and conventional RFA in patients with AF. We made several important observations. Our results suggest that HPSD RFA may be more effective with higher freedom from atrial arrhythmia compared with conventional RFA, especially in the setting of paroxysmal AF. High-power short-duration RFA demonstrated significantly shorter procedure time, fluoroscopy time, LA dwell time, and RF ablation time compared with conventional RFA. There was no difference in safety outcomes between the two groups. In subgroup analysis, HPSD RFA demonstrated higher freedom from arrhythmia in patients with paroxysmal AF, the studies with ≥50 W protocol in the HPSD RFA group and studies with use of contact force sensing catheters with no significant difference in complications.

High-power short-duration RFA approach demonstrated a 9% higher freedom from atrial arrhythmia when compared with conventional RFA. This benefit was also demonstrated in the subgroup analysis of studies using ≥50 W protocols for HPSD RFA. The observable injury to the myocardium with RFA starts at 45°C, being partially reversible with transient stunning below 50°C, and definitive durable necrosis above 50°C.²⁸ Thermal injury induced by an irrigated-RFA catheter tip comprises resistive and conductive phases. Most of the RF energy delivered at high-power settings is absorbed within the first 1–1.5 mm of tissue from the ablation catheter’s electrode tip.^29,30 By increasing the rim size of resistive heating >50°C, HPSD RFA favours the creation of durable lesions,²⁷ with dimensions (larger size, lesser depth) that might be particularly suitable for PVI, as antral thickness is usually below 4 mm.^29,30 This can lead to better PVI with contiguous ablation lesions and lower risk of extracardiac injury due to lower dependence upon the conductive phase of heating during HPSD ablation. This is consistent with our observation of lower acute PV reconnection and higher first-pass isolation with HPSD RFA. Additionally, the issue of catheter stability caused by the respiratory movement of the heart leading to suboptimal lesions from tissue oedema may be less of a problem given the shorter duration of lesion delivery.²³ Along the posterior wall in particular, more durable lesions may be achieved with HPSD RFA in comparison with conventional RFA, where temperature rise with oesophageal monitoring and concern for oesophageal injury may limit sufficient energy delivery to obtain continuous, transmural lesions.

There was variation in the definition of freedom from arrhythmia in individual studies and the use of AAD. However, sensitivity analysis with exclusion of these studies still demonstrated HPSD had higher freedom from atrial arrhythmia. Our overall results favour the use of HPSD RFA strategy over conventional RFA for acute procedural success and freedom from subsequent atrial arrhythmia.

In the subgroup analysis of patients with paroxysmal AF, HPSD RFA demonstrated higher freedom from arrhythmia, shorter procedural time, and no difference in complications compared with conventional RFA. Although the subgroup analysis of patients with persistent AF demonstrated no difference between the two strategies of RFA in terms of freedom from atrial arrhythmia, it must be noted that only 57% of these patients had persistent AF. Based on this observation, it appears that the benefit of higher freedom from arrhythmia with HPSD RFA was probably driven by patients with paroxysmal AF. Whether HPSD RFA demonstrates similar benefits in patients with persistent AF needs to be evaluated further. In addition, the subgroup analysis of patients with additional ablation lesions to PVI did not demonstrate a difference in freedom from atrial arrhythmia. The effectiveness of HPSD RFA in the creation of non-PV ablation lesions, which may be needed in patients with persistent AF has been questioned, with insufficient lesion depth a concern when ablating on thicker atrial tissue, such as ablation of a mitral isthmus line.^23,30 In both subgroups, procedural durations still favoured HPSD RFA and there was no difference in safety outcomes.

High-power short-duration RFA demonstrated a clear advantage compared with conventional RFA in terms of shorter procedure time, RF ablation time, left atrial dwell time, and fluoroscopy duration. The reduction in procedure time and left atrial time limits patient exposure to anaesthetic agents, intravenous fluids, intravenous heparin as well as significantly improving procedural efficiency and laboratory throughput. Reduction in fluoroscopy time benefits the patient, operator, and the support staff. This benefit was initially proposed by Nilsson et al.⁷ when using HPSD to reduce the procedural duration and has been consistently demonstrated in multiple prior studies.^14–27

Importantly, we did not observe any difference in terms of total complications and the individual complications between HPSD and conventional RFA approaches. High-power short-duration RFA has been demonstrated in multiple animal studies to create lesions with a wider diameter and lesser depth, avoiding distant conductive heating, thereby theoretically reducing the risk of collateral damage to extracardiac structures.^30–32 This may theoretically reduce the risk of oesophageal complications as observed in a study that retrospectively evaluated the use of HPSD and conventional RFA on the posterior wall of the left atrium during AF ablation.⁸ Another study which evaluated performed endoscopy within 48 h of the ablation in 96% of the patients also demonstrated lower risk of oesophageal injury.²¹ However, we did not observe any difference between the two groups in our primary pooled analysis. Our results are consistent with the low complication rates observed with HPSD RFA in a large retrospective, multicentre study evaluating HPSD RFA.⁸ In addition, there was no difference in complications between the two groups in the subgroup analysis of studies using ≥50 W in the HPSD RFA strategy. In an animal study by Bhaskaran et al.,³² 50 and 60 W ablations for 5 s were more effective with no steam pops when compared with 40 W for 30 s. However, the same study suggested an upper limit to the power to be 80 W for 5 s as it resulted in an 8% occurrence of steam pops.³² Hence, utilization of strict ablation settings, contact force monitoring, and follow-up of parameters such as impedance, ablation index can reduce the risk of collateral damage with HPSD RFA.⁸ Our results demonstrate that HPSD RFA appears to be as safe as conventional RFA.

Our results of this pooled analysis favour the use of the HPSD RFA strategy over conventional RFA. However, these results remain hypothesis-generating and need to be evaluated further in RCTs comparing HPSD RFA with conventional RFA.

Limitations

We acknowledge several limitations. All the studies in our analysis were non-RCT, hence our pooled analysis needs to be interpreted with caution. However, all included studies were of good quality based on NewCastle Ottawa scale, reflect a real-world experience, and thus have the advantage of being generalizable to contemporary practice. Three of the included studies used a historical cohort for the conventional RFA group. However, sensitivity analysis with exclusion of these studies yielded similar results to primary pooled analysis. The definition of freedom from atrial arrhythmia varied in some studies where atrial flutter, atrial tachycardia was not included. We do not have individual patient data from the studies as this is a study data-level meta-analysis. There was significant heterogeneity between studies, but this should have been limited with the usage of a random-effects model for analyses. There was variation in the strategy of HPSD ablation among the various studies in terms of exact parameters for maximum power, duration, and type of catheter used which could have biased our results. The study by Winkle et al. compared 50 W for 3–10 s with 40–45 W for 20–45 s, which was higher than the power used in conventional RFA in rest of the studies. Sensitivity analysis did not change outcomes in primary pooled analysis or subgroup analysis. In most of the studies, monitoring was based on clinic visits and Holter monitoring, however, some studies used additional methods such as implantable loop recorders.

Conclusions

High-power short-duration RFA was associated with better procedural effectiveness and higher freedom from atrial arrhythmia with comparable safety compared with conventional RFA. The total procedure duration with HPSD RFA was shorter and required lesser fluoroscopy time and RF ablation time compared with conventional RFA. In the subgroup analysis of patients with paroxysmal AF, the use of ≥50 W protocol, and contact force sensing catheters, HPSD RFA demonstrated better effectiveness with comparable safety to conventional RFA.

Supplementary material

Supplementary material is available at Europace online.

Conflict of interest: K.K. serves as a consultant to Abbott/St. Jude Medical, Cardiva, and Zoll and research funding from Abbott/St. Jude Medical. P.S.S. has been a speaker for Medtronic and has been a consultant for Abbott, Boston Scientific, and Biotronik. R.G.T. reports serving as an advisor to Boston Scientific/Guidant; receiving research grants from Boston Scientific/Guidant, Medtronic Inc., and St. Jude Medical (Abbott); serving as a consultant for St. Jude Medical (Abbott); and receiving speakers fees or honoraria from Boston Scientific/Guidant CRM, Medtronic Inc., and St. Jude Medical (Abbott). H.D.H. reports serving as a consultant for Cardiofocus and receiving research grants from Medtronic. The remaining authors declare that there is no conflict of interest.

Data availability

The data underlying this article are available in the article and in its online supplementary material.

References