Results from a multicentre comparison of cryoballoon vs. radiofrequency ablation for paroxysmal atrial fibrillation: is cryoablation more reproducible?

Background

Atrial fibrillation (AF) is a growing epidemic, and it is expected that at least one-quarter of the population aged above 40 will develop at least one AF episode in their lifetime.¹

Anti-arrhythmic drugs lack efficacy when chosen for a rhythm control strategy in AF and have been proved to be less effective than catheter ablation,^2,3 which is now a Class I indication for symptomatic patients with drug-refractory AF.^4,5 However, radiofrequency (RF) ablation of AF (the current standard in most countries) is a technically complex procedure, with a long-learning curve, and its results seem to largely depend on Centre's experience,⁶ which restricts this technique to a small number of patients in relatively highly specialized centres.⁷

In recent years, Cryoballoon ablation of AF has been introduced as an alternative ‘single-shot’ approach for pulmonary vein isolation⁸ (the mainstay of AF ablation),⁹ and its use has increasingly grown in Europe and Northern America. The relative simplicity of this technique and its short learning curve¹⁰ could make it ideal to improve the overall success rate in all centres, especially those with less-experienced operators. In the ‘A Clinical Study of the Arctic Front Cryoablation Balloon for the Treatment of Paroxysmal Atrial Fibrillation’ (STOP AF) trial, cryoballoon ablation proved to be more effective than anti-arrhythmic drugs (69.9 vs. 7.3% freedom from arrhythmia relapse at 12-month follow-up).¹¹

Multicentre comparisons of RF vs. cryoablation are currently sparse. We proposed to set up a prospective registry to assess and compare the safety, mid-term results, and reproducibility of these two techniques in centres and operators with different annual caseloads.

Methods

Data concerning all consecutive patients undergoing a first procedure of paroxysmal AF ablation in a multicentre French ablation Survey (NCT01918670-FrenchAF) were prospectively retrieved. Approval by the local Ethics Committee was obtained in all participating institutions.

This survey gathered six referent centres for AF ablation, with different annual caseloads, and all of them are performing RF for more than 10 years. Besides to RF, Centres B and F started using Cryo in 2008 and A, C and E started using Cryo in 2010. Centres A, B, and C were Tertiary University Hospitals, and Centers D, E, and F were Private Hospitals.

Data were included from all patients older than 18 undergoing catheter ablation of paroxysmal AF refractory to at least one anti-arrhythmic drug agent from November 2010 to June 2012. Data regarding all patients within the specified time window were prospectively collected and inserted into each Centre's database in the day of the procedure. No patients were excluded from the survey as a result of acute complications. Patients with a previous AF ablation procedure, persistent AF, left atrial flutter, and/or undergoing ablation with any other single-shot technique (i.e. Ablation Frontiers, Medtronic^© or NMARQ™, Biosense Webster^©) were excluded from the current analysis.

Sample characterization

All variables at the time of the procedure were defined and categorized according to the literature or common practice. Information was collected regarding demographics, anthropometric data, baseline thromboembolic and bleeding risk, anti-arrhythmic drugs prior to the procedure, atrial dilation, left ventricular ejection fraction, and presence of structural heart disease.

Valvular cardiomyopathy was defined as the presence of one of the following: moderate or severe aortic or mitral valve regurgitation, any degree of aortic or mitral valve stenosis, and/or previous cardiac valve replacement/repair.

Ablation procedure

Data regarding the ablation procedure were recorded. The choice of pre-procedural imaging (i.e. transoesophageal echocardiogram, multidetector cardiac computed tomography and/or cardiac magnetic resonance imaging), type of anaesthesia (general anaesthesia vs. conscious sedation), choice of ablation technique, and use of a three-dimensional mapping system for every particular patient was left at the discretion of the different participating centres.

Pulmonary vein isolation was the endpoint in all procedures. Data were collected regarding procedural and fluoroscopy duration, use of 3D mapping systems, ablation technique, type of ablation catheter, and anti-arrhythmic drugs at hospital discharge.

Cryoballoon ablation

A 14-French deflectable sheath (FlexCath Medtronic^©) was introduced into the left atrium after a single transseptal puncture. Then the Arctic Front™ (Medtronic^©) balloon was introduced in the sheath, inflated, and advanced to the ostium of each pulmonary vein and ablation of pulmonary vein antra was performed with at least two applications of 240 s per vein. The 23- or 28-mm balloon was used according to the anatomy and each practitioner's choice. Occlusion of each vein was assessed with venous angiography.

Continuous monitoring of the phrenic nerve during ablation of the PVs was systematically performed by pacing the right phrenic nerve with a quadripolar catheter in the superior vena cava.

Pulmonary vein isolation was assessed using a standard circular catheter after two applications. In procedures performed after March 2011, centres participating in the survey progressively started using the Achieve™ (Medtronic^©) catheter, which allows real-time documentation of pulmonary vein isolation.¹² If the pulmonary veins remained connected, additional applications were performed using different angulations and, finally, a 3.5- or 4-mm tip irrigated catheter was used if isolation was not possible with Cryoballoon. Pulmonary vein isolation was checked 20 min after the last ablation.

Radiofrequency ablation

A single or dual transseptal approach was used at the discretion of the operator. A left atrial geometry was collected using a circular mapping catheter guided by 3D electroanatomical mapping system (Carto 3, Biosense Webster^© or EnSite NavX, St Jude Medical^©). Wide antral circumferential ablation was performed using either a 3.5- or a 4-mm tip irrigated catheter. Power was adjusted according to the operator's preference, and suggested settings were 30 W for the anterior and 25 W for the posterior wall. Pulmonary vein isolation was checked 20 min after the last ablation.

Follow-up and outcomes

Patients could be discharged on anti-arrhythmic agents, according to centre's preference, but these were stopped after the first 3 months. The primary endpoint was AF/atrial tachycardia recurrence, defined as any symptomatic or asymptomatic atrial arrhythmia lasting >30 s after the blanking period. However, if a patient relapsed during the blanking period and remained in AF despite all attempts to restore sinus rhythm, this was also considered a procedural failure. Relapse before discharge and relapse during the 3-month blanking period were also systematically assessed.

The following monitoring protocol was proposed after discharge: a clinical appointment either at the ablating centre or with patient's local cardiologist including a 12-lead ECG and a 24-h Holter at 1, 3, 6, 9, and 12 months after the procedure. Following that, one appointment per year with an ECG and a 24-h Holter was also suggested. In the event of a patient developing symptoms suggestive of relapse, the recommendation was to perform a 12-lead ECG done as soon as possible and if this failed to document the arrhythmia, a 24-h Holter or an external loop recorder was advised. In patients with previously implanted intracardiac rhythm management devices, these were used for monitoring AF/atrial tachycardia relapse.

With regard to safety, the following complications were systematically screened: vascular complications (if requiring intervention or prolongation of admission), thromboembolism (transient ischaemic attack, stroke, and/or systemic embolism), phrenic nerve palsy persisting after the procedure, pericardial effusion (if causing haemodynamic instability and/or requiring pericardiocenthesis or prolonged monitoring), and procedure-related death.

Electrocardiograms and Holter tracings were organized and checked locally, and adjudication of relapse or procedural complications was done by the local EP investigator.

Statistical analysis

Chi-square test was used for comparison of nominal variables. The Student t-test, one-way ANOVA, or their non-parametric equivalents, Mann–Whitney and Kruskall–Wallis when appropriate, were used for comparison of continuous variables; Levene's test was used to check the homogeneity of variance. Results with P < 0.05 were regarded as significant.

Comparisons between RF and cryoablation were performed, as well as comparisons among the different participating centres. Predictors of procedural success were evaluated for the whole sample and, subsequently, for patients having ablation under any of the two techniques. Univariate Cox regression was performed. All tested variables were also used in the multivariate Cox regression (method: forward likelihood ratio, probability for stepwise = 0.05).

Data were filled into a pre-defined data introduction electronic sheet made available to all participant centres. After completion of follow-up, data from all centres were merged and analysed at the coordinating centre (Clinique Pasteur, Toulouse).

PASW-Statistics (version 18.0; SPSS Inc., Chicago, IL, USA) was used for descriptive and inferential statistical analysis.

Results

Patients characteristics

A total of 860 patients undergoing a first ablation procedure for paroxysmal AF in the six participating centres, 467 treated with RF and 393 treated with Cryoballoon, were included in this analysis. A 3D mapping system was used in all patients in the RF group, but not in those undergoing cryoablation. The mean age was 60.0 ± 10.9 years, and 30% of patients were women. CHA₂DS₂-VASc and HAS-BLED scores were low. Left atrial dilation (>40 mm or >20 cm²) was present in 40% of the sample, but most patients had an otherwise structurally normal heart (Table 1).

Patients undergoing cryoablation had a slightly higher CHA₂DS₂-VASc score but displayed a lower prevalence of compromised left ventricular ejection fraction and hypertrophic cardiomyopathy.

Differences in centres' activity and procedure volume

Figure 1 illustrates differences in caseload and activity among centres. The annual caseload of AF ablation procedures was heterogeneous: A and B were low-volume centres (<100 AF ablations per year), C and D performed between 100 and 150 AF ablations per year, E was a moderate volume centre (>200 AF ablations per year), and Centre F was a very high-volume centre (≈500 AF ablations per year). In the setting of a first procedure of paroxysmal AF ablation, different approaches were observed: Centres A and C used cryoablation more often; Centre B used cryoablation for all procedures; Centre D used RF in all procedures; Centres E and F used RF more often.

Figure 1

The number of AF ablation procedures per year in the different participant centres. AF, atrial fibrillation; RF, radiofrequency ablation; Cryoballoon, cryoballoon ablation; A, B, C, D, E, and F, centres.

Open in new tabDownload slide

Centre A had two operators performing both cryoballoon ablation and RF cases. In Centre B, only one operator was performing cryoablation. Centre C had a single operator performing both cryoablation and RF cases. Centre D had three operators performing only RF. Centre E had three operators, two performing both cryoablation and RF and the third performing only RF cases. In Centre F, there were two operators fully dedicated to cryoballoon ablation and two other operators performed only RF. In Centres A, C, and E, where some operators were able to perform both RF and cryoablation, the decision on which method to use was based on equipment availability, patient's preference, and operators' judgment according to the findings in pre-procedural imaging.

In RF-treated patients irrigated tip non-contact force-sensing catheters were used in all patients, except in 100 from Centre F (16 Tacticath™, Endosense^©, and 84 Thermocool SmartTouch™, Biosense Webster^©). The Arctic Front™ first-generation cryoballoon was used in all cryoablation patients (23 mm in 21.4%; 28 mm in 77.3%, and both balloons in 1.3%).

Comparisons across centres revealed the presence of baseline differences among the six-centre samples regarding AF duration, CHA₂DS₂-VASc, HAS-BLED, prevalence of left atrial dilation, and compromise of left ventricular ejection fraction (Table 2 and Supplementary material online, Table S1).

Immediate procedural results and complications

The mean procedural duration was 129 ± 49 min, and fluoroscopy time was 22 ± 11 min. A minority of patients (6.6%) relapsed before discharge and almost a third was discharged under class I or class III anti-arrhythmic drugs (Table 3). A significant reduction in the procedure duration (±16 min) was observed while using cryoablation. However, patients treated with Cryoballoon were discharged more often under anti-arrhythmic agents. Differences in the procedure duration, fluoroscopy time, and immediate outcomes across the six centres are shown in Table 4.

Isolation of all pulmonary veins was achieved in 97.6% (n = 456) of RF patients and in 99.0% (n = 389) of cryoballoon ablation patients (P = 0.136). Notwithstanding, RF touch up was required in eight patients (2.0%) undergoing cryoablation in the initial phase of the survey.

Pericardial complications occurred in 1.0% (n = 9) and phrenic nerve palsy in 0.8% (n = 7) of patients. However, all patients with phrenic nerve palsy recovered during follow-up. There was a higher prevalence of pericardial effusion while using RF (1.7% RF vs. 0.3% Cryoballoon; P = 0.036; Number Needed to Harm = 71) and phrenic nerve palsy occurred more often in cryoablation patients (0% RF vs. 1.8% Cryoballoon; P = 0.004; Number Needed to Harm = 56) (Table 3). Table S2 available in Supplementary material online illustrates complications among the different participant centres.

Early and mid-term procedural success

Only 6.9% (n = 59) of patients were lost during a median follow-up of 14 months (interquartile range 8–23). Relapse during the blanking period occurred in 19.3% (n = 139) of patients, but at a similar rate in Cryoballoon and RF-treated patients (20.4 vs. 18.0%; P = 0.424).

Patients treated in the five centres performing cryoballoon ablation displayed similar rates of freedom from atrial tachycardia/AF over time off anti-arrhythmic drugs (Figure 2, left panel) (freedom from relapse at 18 months: Centre A = 68%, Centre B = 77%, Centre C = 72%, Centre E = 74%, Centre F = 80%). Therefore, mid-term efficacy at 18 months was comparable, even though the centre performing more procedures (Centre F) had a numerically higher (P = NS) success rate than Centre A (lower number of cryoballoon ablation procedures).

Figure 2

An overview of procedural results and freedom from AF in patients treated with Cryoballoon (left) and RF (right). RF, radiofrequency ablation; Cryoballoon, Cryoballoon ablation; A, B, C, D, E, and F, centres.

Open in new tabDownload slide

Among centres using RF, patients in Centre F (performing 349 RF ablations of AF per year) fared better (higher rate of sinus rhythm maintenance) than those from Centre C (47 RF ablations of AF per year), Centre D (134 RF ablations of AF per year), and Centre E (149 RF ablations of AF per year). Centre A performed only seven RF AF ablations during the study and displayed a numerically lower rate of relapse than other low-to-moderate volume centres (Figure 2, right panel) (freedom from relapse at 18 months: Centre A = 72%, Centre C = 48%, Centre D = 46%, Centre E = 66%, and Centre F = 79%). Centre B was excluded from this specific analysis due to the absence of RF ablation in first procedures of paroxysmal AF ablation.

Among the global sample, a centre's annual AF ablation caseload (HR = 0.87 per 100 procedures per year, 95% CI 0.80–0.96, P = 0.003), use of the cryoballoon (HR = 0.47, 95% CI 0.35–0.65, P < 0.001), use of contact-force sensing catheters (HR = 0.52, 95% CI 0.29–0.90, P = 0.020), and relapse during the blanking period (HR = 4.52, 95% CI 3.35–6.10, P < 0.001) were independent predictors of relapse on multivariate Cox regression (Table 5).

Sensitivity analysis for predictors of mid-term procedural success according to the ablation method identified annual AF ablation caseload (HR = 0.77 per every 100 AF ablation procedures per year; 95%CI 0.68–0.87, P = 0.001) and relapse during blanking (HR = 5.28, 95% CI 3.58–7.78, P < 0.001) as independent predictors of relapse in patients treated with RF. However, in patients treated with cryoablation, annual caseload did not have an impact on post-procedural freedom of AF, and the presence of valvular cardiomyopathy (HR = 3.66, 95% CI 1.45–9.25, P = 0.006) and relapse during blanking (HR = 3.75, 95% CI 2.30–6.10, P < 0.001) were the only independent predictors of relapse (Supplementary material online, Table S3). In other words, annual caseload seems to have greater impact on the mid-term outcomes of RF.

The comparison of both ablation techniques pooling data from the six centres showed better mid-term results with the cryoballoon (adjusted HR = 0.47, 95% CI 0.35–0.65, P < 0.001), with early separation of curves shortly after the blanking period and persisting during the remainder of the follow-up (Figure 3). The analysis of this comparison in the different centres performing the two techniques for a first procedure of paroxysmal AF ablation revealed that patients fared better with cryoablation in Centre C and had a numeric trend for better cryoballoon ablation results in Centre E. However, similar results were displayed with cryoablation and RF in Centres A and F (Figure 4). In centres which were performing cryoballoon ablation for only a few months before the beginning of the FrenchAF survey, cryoballoon ablation performed better (Centre C) or was non-inferior to RF (Centres A and E), which had been in use for 10 years in those centres. The estimated number of patients who needed to be treated with cryoablation rather than RF in Centre C to prevent one relapse in the first 18 months was 4.

Figure 3

Freedom from AF—comparison of Cryo vs. RF, pooled data from all centres. RF, radiofrequency ablation; Cryoballoon, Cryoballoon ablation.

Open in new tabDownload slide

Figure 4

Freedom from AF—comparison of Cryo vs. RF in the four centres performing the two techniques for a first procedure of paroxysmal AF ablation. Note: Centres A, C, E, and F have performed 51, 64, 189, and 349 RF ablations of AF per year, respectively, at the time of data collection. RF, radiofrequency ablation; Cryoballoon, Cryoballoon ablation; B, C, D, and F, centres; AF, atrial fibrillation.

Open in new tabDownload slide

Figures S1 and S2 available as Supplementary material online illustrate procedural success according to operator caseload (Figure S1 shows data with operators grouped by caseload and Figure S2 shows individual operator data). A wide variation in mid-term success rates among RF operators was observed, contrary to what was seen with cryoballoon ablation. Impact of annual caseload on mid-term outcomes also seems to be more pronounced in RF ablation.

Discussion

This multicentre, real-world survey confirms the safety and comparable efficacy of cryoballoon ablation compared with standard RF ablation in the setting of a first procedure of paroxysmal AF ablation. Furthermore, our data suggest that, unlike RF, where a marked centre and operator dependence was observed in mid-term results, cryoballoon ablation seems to perform equally well, with a less-pronounced impact of centre or operator experience. This resulted in non-inferior results vs. RF in three of the centres performing both techniques and favourable results in one of the low-volume centres. Knowing that these centres have been performing RF procedures for much longer than cryoballoon ablation, this confirms the shorter learning curve and higher reproducibility of this technique. To the best of our knowledge, the differential reproducibility of cryoballoon ablation compared with RF had not been previously demonstrated.

However, even though data have been adjusted for possible confounders, these results should be interpreted with caution, as they result from an observational non-randomized comparison.

Efficacy of cryoballoon ablation vs. RF

Freedom from AF recurrence at 18 months in 75–80% of patients following cryoballoon ablation of paroxysmal AF using the first-generation balloon has been previously reported.^13,14

Minor differences (lower prevalence of compromised left ventricular systolic dysfunction and hypertrophic cardiomyopathy and higher CHA₂DS₂-VASc score in Cryoballoon ablation patients) were observed in our sample as regards baseline variables, when comparing RF vs. cryoballoon ablation. However, after adjustment, patients treated with Cryoballoon had better mid-term results, remaining free from arrhythmia relapse more often. These differences in baseline parameters can be possibly explained by operator-driven selection bias in low-volume centres, selecting structurally normal hearts more frequently for cryoablation.

The Kaplan–Meier curves illustrating mid-term outcomes according to operator annual caseload (Supplementary material online) suggest that even though caseload plays a role in RF procedural results, it does not seem to be the only driver for different success rates in operators performing RF, as some operators performing less procedures, perform as well, or even better, than some of those performing medium caseloads. Therefore, besides mere chance, other confounding factors such as operator skill with RF may be involved in explaining these findings.

Several observational single-centre comparisons of cryoballoon vs. RF ablation in patients with paroxysmal AF have confirmed the non-inferiority of cryoballoon ablation with regard to efficacy.^15–17

A similar efficacy also seems to occur while using the novel generation RF (allowing contact force assessment) and cryoballoon (second-generation) catheters according to data from single-centre^18,19 and multicentre observational studies.²⁰

Results of the first randomized controlled study comparing cryoballoon vs. RF ablation, the single-centre ‘A Comparison of Isolating the Pulmonary Veins With the Cryoballoon Catheter Versus Radiofrequency Segmental Isolation: a Randomized Controlled Prospective Non-inferiority Trial’ (FreezeAF) have been recently published. FreezeAF included 322 patients with paroxysmal AF who were randomized 1:1 to the first-generation cryoballoon vs. RF. After a single ablation procedure, 65% of RF-treated patients and 68% of those treated with the cryoballoon remained in sinus rhythm at 12 months (P < 0.001 for non-inferiority).²¹

A North American multicentre (five centres) observational study compared the success rate of a single procedure of AF ablation using the second-generation cryoballoon vs. point-by-point non-contact force-sensing RF. In the group of patients with paroxysmal AF (593 treated with cryoballoon ablation and 320 with RF), a significantly higher number of patients remained free from AF without anti-arrhythmic drugs after cryoballoon ablation (78.4% Cryoballoon vs. 60.8% RF; log-rank P < 0.001).²²

The ongoing multicentre European ‘A Controlled, Prospective, Non-Inferiority, Parallel-Group, Randomised, Interventional, Open, Blinded Outcome Assessment (PROBE-Design), Multi-centre Trial, Comparing Efficacy and Safety of Isolation of the PVs With a Cryoballoon Catheter vs a Radiofrequency Ablation With a ThermoCool Catheter in Patients With PAF’ (FIRE AND ICE) trial has an estimated enrolment of 768 patients randomized in a 1:1.matrix. This study is expected to assess the non-inferiority of cryoballoon vs. RF ablation during a 18-month time interval and will most likely provide stronger evidence and, very likely, the definite answer regarding this comparison.⁷

Safety of cryoballoon ablation vs. radiofrequency ablation

In our sample, we have confirmed the low and comparable incidence of procedural complications with both ablations methods. However, we have observed a higher incidence of transient phrenic nerve palsy and lower incidence of pericardial complications in patients treated with cryoballoon ablation.

A similar overall safety profile of the two techniques has been suggested by several single-centre observational studies.^15–17

In the single-centre, randomized, FreezeAF trial, temporary phrenic nerve palsy and vascular complications occurred more frequently in cryoballoon ablation (5.8 vs. 0 and 5.1 vs. 1.9%, respectively).²¹

In a North American Multicentre observational study, a higher rate of complications was also observed with cryoballoon ablation, driven by an increase in phrenic nerve palsy (7.6 vs. 0%; P < 0.001), which was persistent in 1.2% of patients. On the other hand, a trend for more pericardial complications was observed with RF (1.7 vs. 0.6%; P = 0.09).²² This trend for less pericardial effusions in patients undergoing cryoablation has also been observed in a meta-analysis by Cheng et al.²³ (2.1 vs. 5.5%, OR = 0.58, 95%CI 0.30–1.06, P = 0.08) and is concordance with the lower incidence of pericardial complications we have observed in our sample. A possible explanation for this may be the more uniform distribution of pressure across a higher area of tissue with the cryoballoon compared with the much smaller and punctiform ablating surface of the tip of the RF catheter which, in the presence of sudden variations and high levels of contact-force, may be more prone to microperforation.

Limitations

This investigation has several limitations that should be highlighted. First, this was a survey and, therefore, not a randomized, controlled trial. Secondly, all cryoballoon ablation procedures were performed using the Arctic Front™ balloon. The current standard is now the more effective Arctic Front Advance™ (Medtronic^©), which could have improved the overall performance of cryoballoon ablation, but it only became available in France after June 2012 (end of our survey's enrolment period). On the other hand, in the higher volume Centre F, 100 patients had their RF ablations performed using contact-force sensing catheters, which are associated with better long-term results.²⁴ We can postulate that the results of Centres A, C, D, and E could have improved if they had used contact-force sensing catheters. A minority of patients (6.9%) were lost to follow-up. This can be explained by the fact that patients were frequently referred for ablation by their local cardiologists and were transferred to their care after discharge. Furthermore, some patients had been referred by foreign cardiologists from Asia and Africa and returned to their home countries soon after the procedure, making follow-up difficult or impossible in some cases. Lastly, even though ablation currently aims to address patient's symptoms, we believe that if patients had been systematically followed with 7-day Holter or implantable loop recorders, more asymptomatic AF episodes might have been recorded and freedom from AF relapse might be 10% lower.²⁵ This would be in line with the current recommendations,²⁶ which suggest more prolonged monitoring of patients after the ablation procedure. However, this was not current practice when the inclusion period of the survey started.

Conclusion

These real-world data suggest that cryoballoon ablation seems to be less operator-dependent and more reproducible than RF in the setting of paroxysmal AF ablation.

Supplementary material

Supplementary material is available at Europace online.

Funding

P.D.L. is funded by the National Institute of Health Research Biomedical Research Centre at University College of London Hospitals NHS Foundation Trust and University College of London.

Conflict of interest: R.P. has received training grant from Boston Scientific, and Sorin Group and a Research Grant from Medtronic; S.B. has received consulting fees from Medtronic, Boston Scientific, and Sorin Group. J.-P.A. has received consultant fees from St Jude Medical and Biosense Webster. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.

References