DISCUSSION

To the best of our knowledge, this is the first reported study to compare the clinical efficacy and safety of HBA and CBA performed for PerAF. The study was conducted on a propensity score-match sample, and the isolation areas were carefully evaluated by high-resolution mapping. Our two main findings were as follows: (1) The isolation areas produced by HBA and by CBA were generally wide, but those produced by HBA were significantly larger than those produced by CBA, although additional energy applications were needed when HBA was performed. (2) The procedural safety and the clinical efficacy shown at 18 months were comparable between the 2 balloon systems.

Isolation areas created by the 2 balloon systems

Recently, a single “big balloon” technique has become a standard CBA technique because the large freezing surface of the second-generation cryoballoon allows for coverage of both small and large PVs as well as creation of antral lesion sets.8 Our study findings are in line with those of a previously reported study showing that the lesion created by the cryoballoon was relatively large.9 Miyazaki et al. reported that the isolation area ratio at the LSPV antrum was significantly smaller in patients with a funnel-shaped or common left PV than in others.10 This may be attributable to the spherical nature of the cryoballoon and size mismatch of the balloon to the PV orifice, suggesting that the area covered by the cryoballoon depends highly on the patient’s LA anatomy. Anatomical factors might have affected our patients only slightly, if at all; our study included only patients with AF of relatively short duration (average: 6 months) and minimal LA remodeling (average LAD: 41 mm).
Unique to HBA is the fact that the balloon remains compliant even during energy delivery; the cryoballoon stiffens during energy delivery. However, the hot balloon often yields a relatively small ablation area because this type of balloon tends to inside the PV orifice1. Results of reported studies have indicated that larger PV antral isolation and LA posterior wall isolation areas improve clinical outcomes for patients with PerAF.11,12Further, feasibility of cryoballoon-based linear ablation along the LA roof and/or floor has been demonstrated recently.13,14On the basis of these reported findings, we employed additional upper posterior wall energy application to create a larger ablation area and so reduce the risk of recurrence. As a result, our HBA protocol yielded a wider antral isolation area than that of the standard CBA protocol. Notably, the balloon can be adjusted to fit the antral region, regardless of its size, including part of the posterior wall because it is highly compliant. Further, since HBA is performed under temperature control, it allows for stable lesion creation without complete occlusion of the PV even when the balloon is positioned near the LA antrum. Thus, the hot balloon appears be particularly suitable for creation of a wide planar antral isolation area in patients with PerAF. Additional studies are needed to further establish the validity of the approach.

Efficacy of the 2 balloon systems for PerAF

CBA-based antral PVI has been shown to be efficacious in patients with paroxysmal AF, and the 84–93% 1-year success rates have been reported after performance of PVI by HBA in such patients.1,15,16Several groups of investigators have compared antral PVI performed by CBA or by contact force-guided RF ablation in patients with paroxysmal AF 17 and those with PerAF,3 but there are no comparative data regarding clinical outcomes of PVI performed by HBA in patients with PerAF. Although recent reports have documented 1-year procedural success rates as high as 76–79%,18,19ablation for PerAF remains challenging. Our study revealed good clinical outcomes, whether by HBA or by CBA. One possible explanation for the improved outcome is that our cohort consisted of patients with relatively early-phase PerAF and a minimally remodeled LA. A wide and sufficient ablation region might have been created easily in these patients. Another possible explanation is that nearly a third of the patients were on AAD therapy even during the follow-up period. Nonetheless, the high success rates of both balloon modalities may be due to durable PVI and partial modification of the LA substrate resulting from the wide ablation areas. Randomized studies are warranted to elucidate the clinical efficacy of each of the 2 balloon systems.

Procedural safety and complications

The incidence of complications among our study patients did not differ significantly from previously reported incidences.6,14,15,20A multicenter randomized study has already revealed the safety of the hot balloon system.4Basically, myocardial tissue is ablated by conductive heating from the balloon surface via agitated fluid, which is warmed by Joule heating derived from coil electrodes within the balloon. Thus, the tissue temperature will be highest at the site were the balloon contacts the endocardial tissue surface, and the temperature will gradually decrease at deeper levels. Thus, the hot balloon poses a low risk for an abnormal rise in temperature inside the tissue, and this may prevent both thrombus formation and steam pops.
When LIPV-targeted HBA was performed, the esophageal temperature tended to rise due to placement of the balloon in close proximity to the esophagus. To avoid esophageal complications, we applied an active esophageal cooling protocol by injecting cooled saline during the energy delivery.5 This resulted in aspiration pneumonia in 1 of our patients, Although rare, care should be taken to avoid this complication during HBA. Phrenic nerve injury (PNI) is of major concern during CBA.6,21PNI occurred in 1 of our patients during CBA, and no PNI was observed during HBA. The reported incidence of PNI during HBA for paroxysmal AF is 0–3.7%,4,15and additional PV antrum application did not significantly affect the occurrence of PNI. We found additional extra PV ostial application during HBA for PerAF to be as safe as reported previously.

Study limitations

Because our study was conducted as a dual-center study that involved a relatively small number of patients with PerAF, our findings should be confirmed by further studies. Also, asymptomatic episodes might have been underestimated because the AF recurrence rate was evaluated only on the basis of spot electrocardiograms obtained during patients’ follow-up visits or by and Holter monitoring. Our study results should be interpreted cautiously because the patients were not randomized to treatment. The patients were matched on the basis of propensity scores, but unknown clinical characteristics or anatomical considerations that might have affected the outcomes could not be ruled out.