Introduction

Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, estimated to affect over 37 million people worldwide, with an increasing prevalence with age [1]. AF confers significant risk for stroke, heart failure, and cardiovascular mortality, making treatment strategies aimed at controlling for AF to lower symptoms and prevent these complications a major public health priority [2, 3].

AF occurs in approximately 1–2% of younger adults, increasing to over 10% of those over 80 years old [1, 2, 4]. It often first presents with symptoms like palpitations, chest pain, fatigue, dizziness, and shortness of breath; however asymptomatic AF may exist in as many as 30% of those affected [5]. Long-standing persistent AF carries a substantial risk of tachycardia-mediated cardiomyopathy [6].

Antiarrhythmic drugs (AADs) work through various mechanisms to prevent AF recurrence but are limited by modest efficacy and risk for ventricular proarrhythmia and extracardiac toxicity [7]. Amiodarone and dofetilide are often more effective but have substantial non-cardiac adverse effects [8]. Guidelines recommend tailored AAD choice based on the presence of structural heart disease and other patient factors [9].

Catheter ablation involves using radiofrequency energy delivered to strategic locations in the left atrium to electrically isolate and ablate arrhythmogenic foci [10]. This technique aims to prevent AF triggers and maintenance [10]. Technological advancements have improved ablation success. Complications like bleeding, vascular damage and stroke remain a concern with ablation procedures [11].

Prior studies and meta-analyses suggest catheter ablation may afford greater freedom from AF recurrence compared with AADs in paroxysmal AF, but its role in persistent AF remains less defined [12, 13]. However, few studies directly compare contemporary ablation approaches to next-generation AADs [14].

Both modalities confer specific benefits and adverse effects important in treatment considerations for a given AF patient. Catheter ablation plays an important early role in management of symptomatic paroxysmal AF refractory to a single AAD. For persistent AF, decisions are more complex with both options having relatively lower efficacy [15].

In this study, we aim to systematically review randomized head-to-head trial evidence comparing outcomes of catheter ablation against antiarrhythmic drug therapy for treatment of symptomatic paroxysmal or persistent AF, providing a quantitative comparison of their efficacy and safety which can further guide clinical decision-making.

Methods

Cochrane Handbook for Systematic Reviews and Interventions [16] and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standards [17] were adhered to throughout this meta-analysis. Additionally, this systematic review and meta-analysis is registered with PROSPERO international prospective register of systematic reviews under ID of CRD42023486487 on Dec 12, 2023.

Search strategy and selection criteria

We systematically searched PubMed, SCOPUS, Cochrane Library, and Web Of Science (WOS) from database inception to November 2023. No filters were applied. The search strategy included a combination of controlled vocabulary terms and free text words for the key concepts of atrial fibrillation, antiarrhythmic medications, and catheter ablation. We also hand-searched reference lists of relevant review articles to identify additional eligible trials.

We included randomized controlled trials (RCTs) comparing antiarrhythmic drugs (AADs) versus catheter ablation, either alone or combined with AADs during the blanking period, for adult patients (≥ 18 years) with paroxysmal or persistent atrial fibrillation. Trials had to report at least one of the prespecified efficacy or safety outcomes at a minimum follow-up duration of 6 months. We excluded non-randomized studies, observational studies, case series, case reports, editorials, and conference abstracts without subsequent full publication.

Study selection and data extraction

Two investigators independently screened all retrieved titles/abstracts and potentially eligible full-text articles, determined final study inclusion, and extracted relevant data using a standardized and piloted data extraction form. Extracted information included: study characteristics (author, year, country, AF type), patient characteristics (age, gender, ), lengths of follow-up, and results for each study. Any discrepancies during study screening and data extraction were resolved via discussion and consensus between the two reviewers, consulting a third reviewer for persistent disagreements if needed.

Outcomes

The prespecified primary efficacy outcome was arrythmia recurrence. Secondary outcomes included all-cause mortality, cardiovascular hospitalizations, change in quality-of-life scores from baseline (assessed using validated instruments such as the 36-item Short Form survey [SF-36], and adverse events, including any adverse events (AEs), procedure-related AEs, stroke, vascular access complication, cardiac tamponade, pericardial effusion, and pulmonary-vein stenosis.

Quality assessment

The Risk of Bias (ROB) tool, version 2, was used to assess the bias of the studies used into this meta-analysis [18]. The tool evaluates five domains: bias caused by the randomization technique, bias caused by variations from planned interventions, bias caused by missing outcome data, bias in outcome assessment, and bias in the selection of the reported result. For each domain, the risk of bias was rated as low, moderate, or high. Using ROB 2, two reviewers independently evaluated each research’s bias risk. Any differences were worked out via debate and consensus.

Data synthesis and analysis

All analyses were performed using RStudio version 2023.12.0 + 369 using the “meta” package. For outcomes reported by at least two RCTs, pooled effect estimates were calculated using fixed effect model or random-effects meta-analysis models to account for between-study heterogeneity. Dichotomous data were expressed as risk ratios (RR) and continuous data as mean differences (MD), both with 95% confidence intervals (CI). Statistical heterogeneity was assessed with the P static. If substantial heterogeneity (P > 0.1) existed, the random effect model and leave one test were used and leave one test was conducted using open meta-analyst software.

Subgroup analysis was carried out for the primary outcome based on the type of AF: paroxysmal, persistent, or studies combining both types. A subgroup analysis was also performed based on the period of follow-up: 6–12 months or > 12 months. A meta-regression was performed to provide further confirmation of the relationship between the pooled estimate and the duration of follow-up.

Results

A comprehensive literature search was conducted using PubMed, Web of Science, Cochrane Library, and SCOPUS databases to identify relevant studies comparing catheter ablation to antiarrhythmic drug therapy for atrial fibrillation. The search yielded 1842 records from PubMed, 3093 from Web of Science, 828 from Cochrane Library, and 6080 from SCOPUS. After removing 3699 duplicate records, 8144 records were screened by title and abstract. Of these, 8092 records were excluded according to predefined criteria. The remaining 52 reports underwent full-text screening, after which 40 were excluded. A total of 20 randomized controlled trials ultimately met the criteria for inclusion in the systematic review, with all 12 providing sufficient data for the meta-analysis. Fig. 1.

Fig. 1
figure 1

PRISMA Flow Chart

Full size image

Summary and baseline characteristics of the included studies

Twelve randomized controlled trials comparing catheter ablation to antiarrhythmic drug therapy for atrial fibrillation were included. These trials enrolled patients with paroxysmal (7 trials), persistent (4 trials) or chronic (1 trial) AF. Follow-up duration ranged from 9 months to 4 years. Most trials were multi-national in scope. Table 1.

Table 1 Summary of the included studies
Full size table

Baseline characteristics were well-balanced between the catheter ablation and antiarrhythmic drug arms within each included trial. The mean patient age ranged from 53 to 68 years. Most trials had 60–80% male participants. The mean duration of AF history before enrollment was 1–8 years where reported. The rate of hypertension was 40–90%. Left ventricular ejection fraction averaged between 52 and 62% across trials. Table 2.

Table 2 Baseline characteristics of the included studies
Full size table

Quality assessment

Four studies demonstrated consistent low-risk ratings across all domains, suggesting a high level of confidence in its reliability [19,20,21,22]. Conversely, five studies exhibited a high risk of bias in the deviation from intended intervention and the overall risk was high [23,24,25,26,27]. Finally, the rest three studies were judged as having overall some concerns due to few details about blinding and protocol registration [28,29,30]. Figs. 2 and 3.

Fig. 2
figure 2

Risk of bias as a percentage

Full size image
Fig. 3
figure 3

Risk of bias per protocol for individual studies

Full size image

Outcomes

Primary outcomes

Recurrent atrial arrhythmia

The pooled analysis of the 12 included studies with a total of 3977 patients showed a significantly lower AF recurrence rate in the catheter ablation group compared with AADs [RR = 0.44, 95%CI (0.33, 0.59), P ˂ 0.00001] (Fig. 4). The data were heterogenous (P ˂ 0.001, I2 = 85%), and this heterogeneity could not be resolved (Supplementary Fig. 1).

Fig. 4
figure 4

Forest plot of recurrent arrythmias

Full size image

Secondary outcomes

Rate of hospitalization

The pooled analysis of 5 studies with total 2781 patients showed a significant lower hospitalization rate in the catheter ablation group compared with AADs [RR = 0.44, 95%CI (0.23, 0.82), P = 0.009] (Fig. 5). The data were heterogenous (P = 0.001, I2 = 78%), and this was resolved by excluding Packer et al. 2019 (P = 0.93, I2 = 0). After resolving heterogeneity, the hospitalization rate in the catheter ablation group remained lower compared with AADs [RR = 0.34, 95%CI (0.18, 0.64), P ˂ 0.0001] (Supplementary Fig. 2).

Fig. 5
figure 5

Forrest plot of rate of hospitalization

Full size image

Adverse events

Any adverse events rate in the catheter ablation group was comparable to AADs based on the pooled analysis of 7 studies with a total of 3105 patients [RR = 1.30, 95%CI (0.83, 2.05), P = 0.24], and the data were heterogenous (P = 0.08, I2 = 46%) (Fig. 6), and this was resolved by excluding Forleo et al. 2019 (I2 = 23%). After resolving heterogeneity, the any AEs outcome in the catheter ablation group was significantly higher compared with AADs [RR = 1.51, 95%CI (1.08, 2.10), P = 0.02] (Supplementary Fig. 3).

Fig. 6
figure 6

Forest plot of adverse events

Full size image

The procedure-related AEs and cardiac tamponade in the catheter ablation group were significantly higher than those in the AADs group [RR = 15.70, 95%CI (4.53, 54.38), P < 0.0001] and [RR = 9.22, 95%CI (2.16, 39.40), P = 0.0027], respectively. The data were homogenous (P = 0.72, I2 = 0%) and heterogeneous (P = 0.88, I2 = 0%) (Fig. 7a and d). Stroke, vascular access complications, pericardial effusion, and pulmonary-vein stenosis were all similar between the two groups (Fig. 7b, c, d, e, and f, respectively). All the outcomes were homogenous.

Fig. 7
figure 7

Forrest plot of death

Full size image

All-cause mortality

There was no significant difference between the two groups in the all-cause mortality rate [RR = 0.78, 95%CI (0.58, 1.05), P = 0.10], and the data were homogenous (P = 0.40, I2 = 0%) (Fig. 8).

Fig. 8
figure 8

Forest plot of change in SF-36-Physical component

Full size image

Secondary outcomes

Quality of life

Our pooled analysis showed a significant improvement in the SF-36 physical component in the catheter ablation group compared with AADs [MD = 7.61, 95%CI (-0.70, 15.92), P = 0.07]. The data were heterogenous (P ˂ 0.0001, I2 = 98%) (Fig. 9a), and this heterogeneity could not be resolved with the sensitivity analysis (Supplementary Fig. 3). The pooled analysis of the 3 studies showed no significant difference between the catheter ablation group and AADs in the change in SF-36-Mental component [MD = 0.96, 95%CI (-3.21, 5.13), P = 0.65]. The data were heterogenous (P ˂ 0.0001, I2 = 90%) (Fig. 9b), and this heterogeneity could not be resolved (Supplementary Fig. 4).

Fig. 9
figure 9

Forest plot of change in SF-36-Mental component

Full size image

Subgroup analysis

Subgroup analysis based on the type of AF didn’t show a significant difference between those with persistent and paroxysmal AF regarding arrythmia recurrence; however, the heterogeneity was partially resolved (Supplementary Fig. 5).

Subgroup analysis based on the duration of follow-up showed that studies with a follow-up duration of 6–12 months were significantly higher than those with more than one year (P < 0.05); however, in both subgroups, ablation showed significant lower risk or AF recurrence (Supplementary Fig. 6).

Meta-regression

A meta-regression analysis revealed a statistically significant positive correlation between the duration of follow-up and the pooled estimate for both arrhythmia recurrence (r = 0.02, P = 0.018).

Discussion

In this systematic review and meta-analysis of 12 RCTs including 3977 patients with paroxysmal or persistent atrial fibrillation, catheter ablation was associated with significantly higher free from atrial arrhythmia recurrence between 6 months post-procedure compared to antiarrhythmic drug therapy. The pooled analysis found a 44% relative risk reduction for AF recurrence with ablation. Additionally, rates of hospitalization were substantially lower with ablation. However, these improved efficacy outcomes came at the cost of an increased risk of procedural-related complications and cardiac tamponade. No difference in all-cause mortality was noted between the treatment strategies.

Catheter ablation also demonstrated benefits to quality of life based on SF-36 scores. The physical health composite score improved by approximately 5 points more with ablation than with medications. However, the difference in mental health component score change was not statistically significant between groups. This data warrants further trials with larger sample size.

Our efficacy findings confirm an advantage of ablation aligned with previous meta-analyses in paroxysmal AF patients [31,32,33]. However, few prior reviews assessed outcomes separately for persistent AF or incorporated the latest-generation ablation tools and techniques. Our study is among the first to pool trial data across both paroxysmal and persistent AF. It provides updated estimates following major evolutions in both pharmacological and ablative treatment options.

While ablation offered efficacy benefits, it did confer moderately higher risk of complications like vascular injury and pericardial effusions. Yet, the 1.5 times greater adverse event rate aligns with expectations for a complex, invasive procedure relative to oral medications. The spectrum of complications was generally manageable without excess mortality. Careful patient selection, operator experience, and performing procedures in higher volume centers can help mitigate procedural risks. Future technological advances may also enhance the safety profile of AF ablation [34].

There remains debate around the appropriate timing for ablation in AF management pathways for stroke prevention. Most patients in these trials continued anticoagulation per guidelines. The open question of whether successful ablation can allow stopping anticoagulants after a several-month blanking periods requires further study through longitudinal cohort follow-up [35].

An important limitation in assessing ablation’s definitive role relates to discrepancies in healthcare systems and reimbursement models internationally. Due to high upfront facility costs, ablation is estimated to have unfavorable short-term economic profiles compared to generic AADs in European nationalized systems [36, 37]. Yet models suggest cost equivalence or even superiority over longer horizons for ablation, considering savings from downstream cardiovascular care [36]. Value-based research is needed to clarify clinical and financial trade-offs globally.

This meta-analysis synthesized only RCT data, supporting internal validity and causality for the treatment effects. However, gaps remain in real-world, generalizable evidence. Participants enrolled in trials may not reflect heterogeneous AF populations or community practice patterns. There was heterogeneity in some pooled analyses, attributable to differences in ablation methods, antiarrhythmic drug choice/dosing, AF type, and monitoring protocols. Another aspect could be the difference in sample size, such as Packer et al. 2019, which created significant heterogeneity but had a significantly higher sample size (2,204). However, the main effects remained consistent after sensitivity analyses. More high-quality head-to-head trials are warranted, particularly focusing on persistent AF cohorts.

This systematic review substantiates an overall beneficial efficacy and safety profile for catheter ablation over antiarrhythmic drug therapy for rhythm control in patients with symptomatic paroxysmal or persistent AF refractory to at least one medication attempt. Subgroup differences based on arrhythmia persistence, patient comorbidities, or other modifiers remain less clear. While evaluating long-term outcomes and ideal timing for ablation are needed, our results support catheter ablation as a recommendable treatment option in patients failing initial antiarrhythmic medications. Ablation is likely underutilized currently and could be offered more widely rather than pursuing multiple unsuccessful drug trials in AF patients. Shared decision-making discussions should weigh upfront risks against reduced arrhythmia burden, hospitalizations, and improved quality of life after ablation. Additional study is justified to enhance patient selection criteria, procedural technique, management protocols and longitudinal monitoring to optimize the risk-benefit ratio with AF ablation.