Radiofrequency Ablation versus Cryoablation in the Treatment of Paroxysmal Atrial Fibrillation: A Meta-Analysis

Background Pulmonary vein isolation is commonly performed using radiofrequency energy with cryoablation gaining acceptance. We performed a meta-analysis of randomized controlled trials which compared radiofrequency versus cryoablation for patients with atrial fibrillation. Methods A systematic search strategy identified both published and unpublished articles from inception to November 10, 2016, in multiple databases. The primary outcomes for this meta-analysis were long-term freedom from atrial fibrillation at 12-month follow-up and overall postoperative complication rates. For all included studies, the methodological quality was assessed through the Cochrane Collaboration's tool for risk of bias. Results A total of 247 articles were identified with eight being included in this review as they satisfied the prespecified inclusion criteria. Overall, there was no significant difference in freedom from atrial fibrillation at ≥12-month follow-up between those receiving cryoballoon and radiofrequency ablation, respectively (OR = 0.98, CI = 0.67–1.43, I2 = 56%, p=0.90). Additionally, the secondary outcomes of duration of ablation, fluoroscopy time, and ablation time failed to reach significance. Cryoballoon ablation had significantly greater odds of postoperative phrenic nerve injury at 12-month follow-up. Conclusions Our meta-analysis suggests that cryoballoon ablation provides comparable benefits with regard to freedom from atrial fibrillation at medium-term follow-up, fluoroscopy time, ablation time, operative duration, and overall complication rate in comparison to radiofrequency ablation.


Background
Atrial fibrillation (AF) is a common arrhythmia affecting up to 2% of the general population [1,2]. AF commonly affects the elderly and is associated with a 26% lifetime risk in males and 23% lifetime risk in females by age 80 [3]. Patients with AF have increased mortality and increased risk of stroke. In addition to that, there is an increased risk of various cardiac complications, including myocardial ischemia/infarction and congestive heart failure exacerbation and dementia [4]. e first line therapy for AF includes the use of pharmacological agents, aiming at rate control, rhythm control, and anticoagulation; however, the recurrence rate with antiarrhythmic drugs is still high and side effects of these medications require close monitoring [4,5]. e standard treatment for patients with paroxysmal drug refractory AF is catheter ablation via pulmonary vein isolation. Catheter ablation has emerged as the most effective rhythm control strategy. Catheter ablation is differentiated by the energy source used. e two current modes of energy being used are radiofrequency and cryothermal energy [4].
Radiofrequency ablation (RFA) causes tissue damage using heat while cryoablation causes tissue damage by freezing the target region. RFA has been considered to be the standard technique and is utilized more frequently than cryoablation. However, it is time consuming, requires extensive training, and is associated with an increased risk of cardiac perforation, thromboembolism, and pulmonary vein stenosis [5][6][7]. Cryoablation, using a cryoballoon catheter, is gaining acceptance as being equally as effective as RFA with potentially decreased rates of complications, although there has been a higher reported rate of phrenic nerve palsy [5,8,9].
While catheter ablation is the mainstay for treatment of AF, the optimal method of ablation is currently under debate. Prior meta-analyses on the topic have been conducted, but they were either inconclusive or included a wide array of study designs, thereby limiting their overall quality and external validity. e overall goal of this metaanalysis is to compare the long-term effectiveness, as measured by freedom from atrial fibrillation at 12-month follow-up, and the complications in treating adult patients (≥18 years old) with atrial fibrillation, using these two methods.

Criteria of Study Inclusion.
All studies that randomly allocated adult patients (≥18 years old) to receive either cryoballoon or radiofrequency ablation for atrial fibrillation were considered for eligibility. Authors of studies for which the randomization sequence or generation tool was unknown were contracted for clarification. Finally, quasi-randomized studies were also considered for eligibility if there was mention of a quasi-random method of allocation such as by date, case number, or date alteration. Nonrandomized studies were excluded to limit overall heterogeneity and improve internal validity of the results.

Search Methods for Identification Studies.
A search strategy was created to identify both published and unpublished articles from inception to May 10, 2016, in the following databases: MEDLINE, Embase, Cochrane Library, and the Database of Abstracts of Review of Effects (DARE). e full search strategy can be viewed in Table 1; however, briefly, the terms of the search related to cryoballoon, catheter ablation, radiofrequency, and cryoablation. e  2 Cardiology Research and Practice results generated by the search strategy were reviewed and screened. e initial agreement between the reviewers based on full-text eligibility was assessed through the calculation of an unweighted kappa (κ). As per the Cochrane guidelines, a κ value can quantitatively assess the agreement between two reviewers [10]. As such, values between 0.40 and 0.59 were considered to represent fair agreement, those between 0.60 and 0.74 represented good agreement, and those that were 0.75 or higher represented excellent agreement [8].

Primary and Secondary
Outcomes. e primary outcomes for this meta-analysis were long-term freedom from atrial fibrillation at 12-month follow-up and overall postoperative complication rates. e secondary outcomes for this review were procedure time, fluoroscopy time, and ablation time.

Data Management and Extraction.
For studies that satisfied the inclusion criteria for this meta-analysis, the data was extracted as per a standardized data extraction form. e data extraction form was first created and piloted by an independent reviewer. e form was used to collect information related to the location of where the study was performed, the various follow-up times, the data reported for the prespecified outcomes, and any miscellaneous outcome information that was not intended to be analyzed by our review. Any data that was reported exclusively in graphical form was extracted using a graph-digitizing software (GraphClick, Arizona Software).

Assessment of Methodological Quality and Risk of Bias.
e Cochrane Collaboration's tool for assessing risk of bias [10] was used by two independent reviewers to evaluate the methodological quality of each included trial.
is tool included questions related to randomization, blinding, and outcome data reporting. For each question, the risk of bias was reported as low risk, unclear risk, or high risk of bias. e initial agreement between the reviewers was assessed through the calculation of an unweighted κ. Again, κ values between 0.40 and 0.59 were considered to represent fair agreement, those between 0.60 and 0.74 represented good agreement, and those that were 0.75 or higher represented excellent agreement [10]. In certain situations, the authors of included articles were contacted to elaborate on the methodology used during their research.
is was done in order to obtain additional information and to ensure an accurate quality assessment.

Statistical Analyses and Measurement of Treatment
Effect.
e primary outcomes of freedom from atrial fibrillation at 12-month follow-up and overall complications are dichotomous variables. As such, an odds ratio (OR) with a 95% confidence interval (CI) was calculated for these outcomes.
On the other hand, all secondary outcomes were measured in units of time. All time measures were first standardized to a total in minutes. In situations where a median and interquartile range (IQR) were reported, statistical conversions were made to a mean and standard deviation (SD) using the methods described by Wan et al. [11]. In situations where a mean and confidence interval (CI) were reported, statistical conversions were made to a mean and SD using the methods described by the Cochrane Collaboration [12]. Overall, a mean difference (MD) in the unit of minutes with a 95% CI was calculated for all secondary outcomes.
In situations where data could be pooled, a metaanalysis was performed using the Mantel-Haenszel random-effects model since there was expected heterogeneity between the included studies. p values less than 0.05 were considered to be significant. For continuous outcomes, an MD value less than 0 represented a decrease in value in the specific outcome (e.g., time of procedure, ablation time, or fluoroscopy time) when cryoballoon was used. On the other hand, an MD value greater than 0 represented a gain in value of the specific variable with cryoballoon use.

Assessment of Heterogeneity.
To calculate heterogeneity, an I 2 statistic test was used. e threshold for conducting subgroup analyses was an I 2 greater than 40%. As suggested by the Cochrane Handbook for Systematic Reviews, an I 2 greater than 40% suggests that heterogeneity may be present [10]. If heterogeneity was present, a priori subgroup analysis was performed on the basis of overall study quality and type of radiofrequency ablation used (Duty Cycle Radiofrequency Ablation versus Irrigated Radiofrequency Energy). Subgroups were only created if more than two studies fell into a specific subgroup. Sensitivity analysis was performed by sequentially removing studies in which a different type of radiofrequency ablation was used.

Assessment of Publication Bias.
A funnel plot was created for the primary outcomes and was visually inspected to assess for publication bias. As per the Cochrane guidelines, in the absence of bias, the plot should generally take the shape of a symmetrical, inverted funnel.
2.9. Data Management. All forest and funnel plots were generated using Review Manager software (RevMan version 5.2; Nordic Cochrane Centre, Cochrane Collaboration). Agreement between the reviewers, as assessed through the unweighted κ, was calculated using SPSS software (version 21.0; SPSS Inc.). Finally, all tests of significance were two-tailed and p values less than 0.05 were considered significant.

Study Inclusion.
A total of 247 articles were identified by the systematic literature search. After an initial title and abstract screening, a total of 17 were considered and retrieved to assess their eligibility by review of the full text. Of these articles, a total of eight satisfied the prespecified inclusion criteria and were included in this review [5,6,[13][14][15][16][17][18]. A flow diagram of study inclusion can be viewed in Figure 1. Articles were excluded if they were nonrandomized or lacked an adequate comparison group. e raw agreement between the independent reviewers for full eligibility was 88.2% and the unweighted κ was calculated to be 0.75, which represents good agreement between the two reviewers.

Study Characteristics.
A detailed description of all the included studies can be viewed in Table 2. A total of eight studies encompassing 1548 patients undergoing either cryoballoon or radiofrequency ablation were included. All of the included studies were published between the years 2012 and 2015. Studies were conducted at centers across continental Europe, the United Kingdom, and Russia. Seven of the included studies had published study protocols available for review [5,6,[13][14][15][16][17]. Seven of the studies compared cryoballoon with irrigated radiofrequency energy ablation  [5,6,[13][14][15][16][17], whereas one made the comparison with phased duty-cycled radiofrequency energy ablation [18].

Risk of Bias Assessment.
e majority of studies had a low risk or unclear risk of bias for several methodological parameters as assessed through the Cochrane tool for risk of bias assessment (Figure 2). Only one study sufficiently reported information on the blinding of study participants and personnel [16]. Visual inspection of the funnel plot for all primary outcomes (freedom from atrial fibrillation and overall complication rate) did not suggest publication bias; however, there were few studies with large effect sizes (Figures 3 and 4). us, publication bias cannot be entirely ruled out.

Discussion
In the present meta-analysis, we found no difference in the freedom from atrial fibrillation at 12 months between patients randomized to cryoballoon and those randomized to radiofrequency catheter ablation of atrial fibrillation.
ese results are consistent with recently published data from the Fire and ICE study [6]. Although patients undergoing cryoballoon ablation had shorter total procedure duration when compared to RF ablation by a mean of 4 mins, this difference did not reach statistical difference in the hands of experienced operators. RF ablation requires point-by-point circumferential ablation around pulmonary veins and may be more time consuming and technically challenging for new operators. e fluoroscopy time was prolonged in cryoballoon ablation group by a mean of 1.17 minutes, but this small difference was not statistically significant. We found that there was a higher risk of postprocedure atrial flutter in the RF group, which was not significant. ere was a significantly higher incidence of phrenic nerve injury with cryoablation; however, most of these injuries resolved at 12 months. us, there was no difference in the risk of persistent phrenic nerve injury at 12 months. Phrenic nerve injury is a common concern during cryoballoon ablation of AF. Although the risk of transient phrenic nerve injury was almost ten times higher in cryoballoon group, only one patient in one study had persistent injury at 12 months of follow-up. us, the risk of permanent phrenic injury was low in the cryoballoon group as well.
In this analysis, we found no significant differences in the overall success, fluoroscopy times, total ablation times, procedure times, and any complications seen in the two groups. Cryoballoon has become increasingly used over last few years due to several potential advantages. e need of a single transseptal puncture for cryoballoon is a definite advantage, although some centers perform RF ablation using a single transseptal catheterization as well.
e learning curve for cryoablation may also be shorter.
Our analysis provides some important results, but several important questions remain unanswered. In recent years, force-sensing catheters have been used increasingly for RF ablation. ese catheters provide real-time information about the catheter contact with the tissue and thus may improve the effectiveness of delivered RF lesions. We have no data systematically comparing force-sensing RF catheters verses cryoballoon ablation. A recent study showed no differences in the long-term outcomes when contact force information was available to the operators [19]. Our analysis included studies including both first-and second-generation cryoballoon. ere was only one multicenter randomized study comparing radiofrequency ablation with contact force sense catheters with secondgeneration cryoballoon and found no difference in the outcomes following ablation using either strategy [20]. We also do not have information about patients in whom cryoballoon ablation was attempted but PVI was either not achieved or not attempted because of unsuitable anatomy. e fluoroscopy times were found to be higher in the cryoballoon group. Although not statistically significant, this difference may become more relevant in the future as more centers are utilizing advanced mapping technology coupled with intracardiac echocardiography to minimize or eliminate the use of fluoroscopy during RF ablation.
ere is a need for devising strategies to minimize the use of fluoroscopy in patients undergoing cryoballoon ablation in the future. Monitoring for recurrence of atrial fibrillation after ablation was performed either using an implantable loop recorder or a Holter monitor for two to seven days at 3, 6, and 12 months of the follow-up.

Conclusions
In conclusion, the results of this meta-analysis reveal similar overall success rates at 12 months and comparable fluoroscopy and procedural times and long-term complications between patients undergoing cryoballoon and RF ablation for AF ablation. Given similar outcomes, operators should choose AF ablation technology based on patient-specific features and preferences as well as operator experience and preference.