Ablation versus medication as initial therapy for paroxysmal atrial fibrillation: An updated meta‐analysis of randomized controlled trials

Abstract Background Recent randomized controlled trials (RCTs) suggest that ablation is superior to antiarrhythmic drugs (AADs) as an initial therapy for paroxysmal atrial fibrillation (pAF) to prevent arrhythmia recurrences. We performed an updated meta‐analysis of RCTs, to include recent data from cryoballoon‐based ablation and to compare arrhythmia‐free survival and adverse events between ablation and AADs. Methods We searched MEDLINE and EMBASE from inception to December 2020. We included RCT comparing patients with pAF undergoing ablation or receiving AADs as an initial therapy. We combined data using the random‐effects model to calculate hazards ratio (HR) for arrhythmia‐free survival and odds ratio (OR) for adverse events. Results Five studies from 2005 to 2020 involving 985 patients were included (495 patients and 490 patients underwent ablation and medication as initial therapy, respectively). Patients who underwent ablation had higher freedom from atrial tachyarrhythmias (ATs) during the 12‐24 months follow‐up period (pooled HR = 0.48, 95% CI: 0.40‐0.59, P < .001). In a subgroup analysis of ablation method used, both cryoablation group (pooled HR = 0.49, 95% CI: 0.38‐0.64, P < .001) and radiofrequency ablation group (pooled HR = 0.47, 95%CI: 0.35‐0.64, P < .001) showed reduction in AT recurrence compared with AAD group. There were no differences in adverse events including cerebrovascular accident, pericardial effusion or tamponade, pulmonary vein stenosis, acute coronary syndrome, deep vein thrombosis and pulmonary embolism, and bradycardia requiring a pacemaker. Conclusion Catheter ablation (both cryoablation and radiofrequency ablation) is superior to AAD as an initial therapy for pAF in efficacy for reducing AT recurrences without a compromise in adverse events.


| INTRODUC TI ON
Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia affecting approximately 1%-2% of the worldwide population. 1 Without appropriate treatment, AF can significantly impact quality of life with risks of recurrences and/or arrhythmia progression reported to be as high as 90%. 2,3 Catheter ablation has shown to be an effective treatment with superior efficacy compared with antiarrhythmic drugs (AADs) alone for symptomatic AF. 4,5 The principal aim of AF ablation is to achieve durable circumferential pulmonary vein isolation (PVI), which electrically separates the pulmonary vein (PV) from the left atrium (LA) at the level of PV ostia/antrum.
Although AF ablation is considered relatively safe, the procedure is invasive and carries risks of devastating complications such as esophageal-related injuries, pericardial effusion with tamponade, PV stenosis, and cerebrovascular accident. 6,7 Because of this, ablation is typically used in patients who failed initial AAD therapy, and most evidence supporting the use and the superiority of ablation was derived from populations that had already received an AAD as the first-line, rhythm-based treatment. 8,9 Similarly, previously published randomized controlled trials (RCTs) suggest that PVI is superior to AAD even as initial therapy for paroxysmal atrial fibrillation (pAF). These include the Radiofrequency Ablation versus Antiarrhythmic Drugs as First-Line Treatment of Paroxysmal Atrial Fibrillation (RAAFT-2) trial, the Medical Antiarrhythmic Treatment or Radiofrequency Ablation in Paroxysmal Atrial Fibrillation (MANTRA-PAF) trial, and the Radiofrequency Ablation versus Antiarrhythmic Drugs as First-line Treatment of Symptomatic Atrial Fibrillation (RAAFT-1) trial. [10][11][12] A meta-analysis of the three RCTs by Hakalathi et al confirmed this finding but reported that ablation was associated with more serious adverse events. 13 However, all three RCTs were done using only radiofrequency ablation without any trial performing cryoablation. Moreover, the authors did not perform time-to-event analysis for the main outcome of arrhythmic recurrence. In this updated meta-analysis, we included two recently published RCTs that use cryoballoon ablation, the Early RCTs indexed in PubMed and EMBASE from inception to December 2020 using the search terms including the following: "atrial fibrillation," "ablation," and "initial" as described in Supplementary file 1.
Only articles in English were included. An additional manual search for potential additional pertinent studies was performed using the references from retrieved articles. Any conflict or discrepancy was resolved by a third author (LN).

| Inclusion criteria
The inclusion criteria were as follows: 1. RCT conducted in patients with pAF comparing ablation and AAD as an initial therapy.
2. Studies must report recurrence rates of atrial tachyarrhythmias (ATs) including AF, atrial flutter and atrial tachycardia, and adverse events following the index ablation in the ablation group and in the AAD group after randomization. Hazard ratio (HR), odds ratio (OR), or sufficient raw data to calculate effect size must be provided. 16

| Quality of included studies
Cochrane Collaboration tool for assessing risk of bias was used to evaluate the quality of each RCT by assigning a score (high, low, or unclear) for each individual element from five domains (selection, performance, attrition, reporting, and other). 17

| Data extraction
A standardized data collection form was used to obtain the following data from each study including name of the first author, year of publication, country of the study, study population, main inclusion and exclusion criteria, demographic data of participants, ablation procedure details, AAD therapy, endpoint for recurrence, recurrent rates, and reported adverse event.
To ensure accuracy, this data extraction process was independently performed by all investigators. Any data discrepancy was also resolved by referring back to the original articles.

| Statistical analysis
We performed meta-analysis of included studies using a randomeffects model and the generic inverse-variance method of Der Simonian and Laird to calculate pooled HR. 18 We extracted from these studies the freedom from AT rates and complications rates.

| Quality assessment of included studies
The Cochrane Collaboration tool for assessing risk of bias is shown in Table 3.

| Freedom from atrial tachyarrhythmia
The outcome of AT recurrences was available in all five studies. 10

| Adverse events
A summary of adverse events from the included studies is shown in Table 4. Studies reporting zero event rates in both ablation and AAD group were excluded from the meta-analysis for adverse events. As such, we performed a meta-analysis of adverse events with these available outcomes in at least two of the included studies. There were no differences in adverse events from studies in which data were available for meta-analysis for cerebrovascular accident, PV  summarizing adverse events are shown in Figure 3 and Table 4, respectively.

| Publication bias
We aimed to investigate potential publication bias via the funnel plot and Egger's test. However, as we only had up to five studies in the main analysis (Figures 1 and 2), the number was insufficient to reject the assumption of no funnel plot asymmetry. Thus, we did not perform a funnel plot or Egger's test. 20,21

| D ISCUSS I ON
The main finding from our updated study that includes contemporary ablation tools is that catheter ablation for PVI is more effective than AAD as initial therapy for pAF in reducing AT recurrences after the initial blanking period to 1-2 years. In addition, this benefit with ablation was the same in both cryoablation and radiofrequency ablation with similar relative reductions in risks. Despite the augmented efficacy for reduction in AF recurrences with catheter ablation, we did not observe a significant increase in any adverse events in these patients compared with AAD therapy.
In this updated meta-analysis of RCTs, we found two recently published RCTs that add to prior summary work in this area.
The new search resulted in a total of five RCTs to date. The two newly added RCTs, EARLY-AF, and STOP AF First, used cryoballoon ablation for PVI, which is different from the previous three RCTs that all used radiofrequency ablation. Evidence from the Cryoballoon or Radiofrequency Ablation for Paroxysmal Atrial Fibrillation (FIRE AND ICE) trial suggested that treatment efficacy of cryoballoon ablation was noninferior to the more traditional radiofrequency ablation in drug-refractory pAF. 22 However, the efficacy of cryoablation in treatment-naïve pAF was still unclear, especially when compared directly with AAD or radiofrequency ablation. The EARLY-AF trial and the STOP AF First trial were the first RCTs to compare cryoablation with AAD as an initial therapy for pAF. This gave us opportunity to indirectly compare the two ablation strategies as initial treatment for pAF. As shown in Figure 1, the pooled HR for the cryoablation subgroup ( Figure 1A) was similar to the pooled HR for the radiofrequency ablation subgroup (Figure 2). However, as these pooled data are derived from a retrospective comparison, further prospective RCTs that directly comparing the two techniques are needed to confirm this finding or investigate if one may be preferable over the other for treatment naïve pAF compared with AAD.
The included studies in our meta-analysis demonstrated a significant reduction in AT recurrence at 1-2 years with a similar effect size. The EARLY-AF trial and the RAAFT-2 trial reported HRs for AT recurrence. In the MANTRA-PAF trial, the RAAFT-1 trial, and the STOP-AF First trial, the authors did not report HRs for AT recurrence. Nevertheless, the authors provided sufficient raw data for us to calculate HRs as described in the methods section.
The HRs from each study are shown in Figure 1 along with the pooled HR.
The study designs are similar among the five included RCTs.
Inclusion and major exclusion criteria are shown in Table 1 Three studies used radiofrequency ablation, and two studies used cryoablation as described in Table 1. The subgroup analysis by ablation method is shown in Figure 2A,  Table 4 shows a summary of the adverse events. Overall, the adverse event rates were low, and not many specific adverse events of interest occurring in either of the ablation group or AAD group in some studies. We were able to perform a meta-analysis on several adverse events including cerebrovascular accident, PV stenosis, pericardial effusion or tamponade, bradycardia requiring pacemaker, syncope, acute coronary syndrome, DVT or PE, all of which were statistically similar between the two groups. For other adverse events including mortality, phrenic nerve injury, esophageal-related complications, and perforation, there was only one study available for a meta-analysis for each outcome. The individual ORs for these adverse events also did not demonstrate significant differences between the two groups. Nevertheless, these findings must be interpreted with caution as this could be from inadequate power from the extremely low number of events which subjects this sub-analysis to risk of a type II error.

| LI M ITATI O N S
We acknowledge certain limitations within our study. First, extracted/calculated HRs were not adjusted for confounders. Second, there were differences in follow-up times with three and two studies that followed patients up to 12 and 24 months, respectively. Also, there was a difference in the definition of arrhythmia recurrences used for the clinical endpoint as discussed above. Nevertheless, we did not observe major differences in the HR, and the random-effect model did not reveal significant heterogeneity from our analysis (I 2 = 0.0%). Third, data regarding the adverse event were limited because of the overall low event rates which subsequently limited the power of the analysis.

| CON CLUS IONS
In this updated systematic review and meta-analysis of RCTs, that now includes RCTs that specifically use cryoballoon ablation only, we found that catheter ablation is more effective than AADs as an initial therapy for pAF in reducing AT recurrences over 1-2 years following the treatment initiation. These results are the same with use of both cryoablation and radiofrequency ablation approaches. The adverse event rates are low with contemporary use of AADs and with evolved catheter ablation tools,