Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia with increased risks of stroke and heart failure (1). Catheter ablation (CA) has been recommended as the first-line treatment option for patients with symptomatic AF in recent decades, with pulmonary vein isolation (PVI) being the standard treatment strategy (2).

Point-by-point radiofrequency ablation (RFA) has been commonly performed to achieve PVI; however, it still remains a complex and time-consuming procedure (3). To simplify PVI procedures, several balloon-based ablation technologies have been developed, including cryoballoon ablation (CBA), hot balloon ablation (HBA) and laser balloon ablation (LBA), which have rapidly emerged as alternatives to conventional CA for AF owing to their impressive procedural advantages and feasibility in several multicenter trials (3–5).

Several studies were conducted to compare the efficacy, safety, and procedural parameters between these three balloon-based CA treatments; however, the results were conflicting. Thus, we performed pairwise and Bayesian network meta-analyses to assess the efficacy, safety, and procedural characteristics between HBA, LBA and CBA and guide the optimal selection of these balloon-based strategies as initial treatments for AF.

The present study used both pairwise and network meta-analytic methods based on 20 trials with 1,995 patients; however, there are several limitations. First, only three trials were RCTs, and most evidence was from nonrandomized trials. Although the quality of the included studies was adequate, blinding of patients and the operators was not possible. Second, a mixed population of AF (PAF and PerAF) was included, and subgroup analysis was not available due to the lack of certain data. Third, the follow-up period of some included studies was relatively short, which could have influenced the clinical outcomes. Finally, SUCRA was calculated to rank each treatment option in the NMA, whereas, SUCRA does not measure the magnitude of difference between treatments. Thus, the results should be interpreted with a combination of the estimated OR/WMD and 95% Cr/CrI.

In recent years, balloon-based CA strategies, including HBA, CBA and LBA, have rapidly emerged as alternative therapeutic approaches to conventional point-by-point RFA therapy for the initial treatment of AF (3–5). However, the best treatment option remains unclear. To the best of our knowledge, the present study is the first network meta-analysis to comprehensively compare the clinical efficacy, safety, and procedural parameters between these three balloon-based CA modalities and to rank these treatment strategies for guiding the optimal selection. The principal findings were as follows: HBA, LBA and CBA had comparable rates of AF recurrence and procedure-related complications; SUCRA suggested that HBA may be the optimal approach concerning the primary outcomes; HBA had a significantly higher incidence of TUA than LBA and CBA. LBA required more procedure time than CBA and HBA, but less fluoroscopy time than HBA.

Balloon-based CA technologies have the advantage of reducing the complexity of point-by-point RFA procedures by placing the balloon catheter at the antrum/ostium of the PV (3). Based on evidence from both pairwise and network meta-analyses, we found no significant difference in terms of AF recurrence between the three techniques. These results are consistent with the previously published pairwise meta-analysis, which demonstrated comparable AF recurrence rates during the mid-term follow-up between HBA and CBA (29).

CBA is currently the most common alternative therapy to RFA (14). Plenty of evidence from RCTs, including the large-scale Fire and ICE trial, showed the high effectiveness of CBA as first-line therapy for AF patients, especially for symptomatic PAF (3, 19). LBA and HBA were mainly applied in JAPAN, and LBA has the advantage of providing endoscopic visualization of the endocardial surface with a compliant balloon (30). A prospective, multicenter, randomized trial showed non-inferior safety and efficacy profiles of LBA to RFA (4). However, as the most recent proposed technology, little evidence has been reported regarding the direct comparisons between HBA and RFA. The only RCT that was proposed in 2016 showed the superiority of HBA compared with antiarrhythmic drug therapy for the treatment of patients with PAF (5). However, when compared with CBA and LBA, HBA demonstrated the best rank probability in terms of AF recurrence, although TUA was most frequently needed. This may suggest that the SUCRA recurrence rate preference of HBA is the result of a hybrid procedure.

Similar to the efficacy endpoint, comparable risk estimates for procedure-related complications were found between the three ablation techniques, ranging from 7.2% to 7.8%. Common major complications included PNP, cardiac tamponade and pulmonary vein stenosis (13, 14). The prevalence of major complications after CA for AF varies from 0.8% to 16.3% (31). The study by Chun et al. evaluated the complications in CA of AF in 3,000 patients and showed that cardiac tamponade occurred in 0.1% of the patients treated with balloon catheters, which was significantly lower than RFA (1.5%) (32). This study indicated a low chance of cardiac tamponade for balloon-based ablation therapy, and the overall risk for balloon-based technique-specific complications, such as PNP, was reported to be approximately 1.5% (32). PNP represents a relatively common complication for balloon-based therapy; however, most PNP cases are usually transient, and persistent PNP occurs in <1% of patients, who usually recover function after 3 months (12, 33).

It should be noted that the estimated risks of procedure-related complications were higher than those reported by the abovementioned studies (32). A possible explanation might be that all the reported complications from the included studies were evaluated; however, the definition of procedure-related complications was relatively heterogeneous among the included studies. Complications, such as minor vascular complications, transient ischemic attacks, and gastroparesis, were also defined as procedure-related complications in several included studies.

Although the clinical outcomes in terms of efficacy and safety are comparable between the HBA, CBA and LBA techniques, HBA required a significantly higher incidence of additional TUA. This unequal phenomenon may be attributed to the inherent characteristics of HBA. It was reported that, HB had better adjustability and compliance than CB during the procedure, which could help to occlude the more distal and deeper portions of the PVs (16, 34). However, this elastic feature may lead to inadequate occlusion of the PV antrum which may not adhere well to all the tissues of the PV walls (25). Compared to the HB, the nonconforming CB stiffens the balloon during energy deliveries, creating a larger ablation area of the PV antrum (9). Several studies have demonstrated that CBA produced a wider and larger lesion width than HBA (9, 16). Therefore, the relatively smaller LA lesion areas created by HBA compared with CBA may lead to a higher incidence of TUA.

Another possible explanation could be that the surface temperature of the hot balloon may not be higher enough to cause deep tissue damage. A previous study by Hojo et al. showed that a balloon temperature of 73°C yielded a lower TUA rate than that at 70°C. The thickest anterior aspect of the left superior pulmonary vein (LSPV) was reported to be the dominant TUA area (13, 23). However, a higher balloon temperature may increase the risk of PV stenosis; thus, additional TUA is still recommended for HBA (13). Even though HBA had the highest rates of additional TUA compared with CBA and LBA, the AF recurrence rates between these techniques were comparable, suggesting benefit from the deeper lesions in the PVs. Technical developments to improve the contact between the HB and PV tissue are warranted to reduce TUA rates.

In addition, HBA may need the most fluoroscopy time according to the SUCRA, while LBA may have the least fluoroscopy time. It was reasonable to expect that LBA needed the lowest fluoroscopy time given the use of an endoscopic guide, while CBA and HBA require serial angiograms to achieve optimal PV occlusion (14). It should be noted that, the results were inconsistent between the pairwise and network meta-analyses in terms of fluoroscopy time analysis. NWM demonstrated that LBA and CBA had comparable fluoroscopy times but had less fluoroscopy time than HBA (Figure 2). However, a pairwise meta-analysis showed that, LBA had less fluoroscopy time than CBA and had comparable results compared with HBA (Supplementary Figure S5). Potential causes were investigated, and we found that when the effect quantities for pairwise meta-analysis were changed from WMD to SMD, the same results were obtained between pairwise and network meta-analyses. The relatively small sample size and the high heterogeneities between studies may be possible explanations.

Another secondary endpoint, procedural time, should also be considered, as the three balloon-based strategies had similar efficacy and safety. Both pairwise and network meta-analyses demonstrated that HBA and CBA had comparable procedural times, but LBA needed the most procedural time according to SUCRA. LBA had the advantage of providing direct PV visualization to achieve a more precise titration of ablation lesions, whereas this would consume more procedure time than the “single shot” techniques, such as HBA and CBA (35). In addition, in contrast to CBA, LBA lacks a certain mapping catheter integrated into the system for real-time PV potential recording, leading to the increased procedural time when validating PVI using an additional mapping catheter (12). New generations of LB have been introduced recently, which showed promising improvements in PV occlusion characteristics and significantly decreased procedural and total laser times (36). Thus, as newly established CA techniques for AF, balloon-based ablation strategies have provided electrophysiologists with impressive alternative options. Larger prospective multicenter randomized studies with long-term follow-up are needed to compare these three balloon-based ablation techniques.

The present NMA showed that AF recurrence and procedure-related complications after CA for AF are comparable between HBA, LBA and CBA, while HBA showed the best rank. However, HBA required significantly more TUA during the procedure than CBA and LBA. The HBA strategy ranked highest with regard to procedure time (least), while the LBA strategy ranked highest with regard to fluoroscopy time (least). This study provides decision-makers with comprehensive and comparative evidence about the efficacy and safety of different balloon-based CA strategies. Further large-scale studies are still warranted to provide an up-to-date recommendation for the superior option.