A prospective multicenter study of direct comparison of feasibility and safety of pulmonary vein isolation using the minimally interrupted apixaban between second‐generation cryoballoon and radiofrequency ablation of paroxysmal atrial fibrillation: J‐HIT apixaban

Abstract Background The feasibility and safety of pulmonary vein isolation (PVI) using cryoballoon (CB) for paroxysmal atrial fibrillation (PAF) with minimally interrupted apixaban has not fully explored. Methods In this multicenter, randomized prospective study, we enrolled patients with PAF undergoing CB or radiofrequency (RF) ablation with interrupted (holding 1 dose) apixaban. The primary composite end point consisted of bleeding events, including pericardial effusion and major bleeding requiring blood transfusion, or thromboembolic events at 4 weeks after ablation; secondary end points included early recurrence of AF and procedural duration. Results A total of 250 patients underwent PVI (125 assigned to the RF ablation and 125 assigned to the CB ablation). The primary end point occurred in 1 patient in the CB ablation group (0.8%; 90% confidence interval [CI], 0.04 to 3.70) and 3 patients in the RF group (2.4%, P = .622; risk ratio, 0333; 90% CI, 0.05 to 2.20). All events were pericardial effusion, all of whom recovered after pericardiocentesis. Early recurrence of AF occurred in 4 patients (3.2%) in the RF group and in 6 patients (4.8%) in the CB group (P = .749). The procedural duration was shorter in the CB group than that in the RF group (136.5 ± 39.9 vs 179.5 ± 44.8 min, P < .001). Conclusion CB ablation with minimally interrupted apixaban was feasible and safe in patients with PAF undergoing PVI, which was equivalent to RF ablation.


| INTRODUC TI ON
Catheter ablation is a well-established therapy in patients with symptomatic drug-refractory paroxysmal atrial fibrillation (PAF), but it is one of the most complex interventional electrophysiological procedures and associated with a significant risk of complications, such as thromboembolic or bleeding events. 1 Uninterrupted warfarin during the periprocedural period has been used as a standard protocol to minimize the risk of thromboembolic events. 2 Pulmonary vein isolation (PVI) is a cornerstone of catheter ablation, particularly for PAF. Cryoballoon (CB) ablation has emerged as an alternative technique to radiofrequency (RF) ablation for the treatment of PAF. The recent largest randomized trial comparing both technologies in patients with PAF demonstrated the noninferiority of CB ablation to RF ablation in terms of efficacy and safety end points. 8 To date, there has been no prospective study to determine the feasibility and safety of CB ablation using DOACs in comparison with those of RF ablation. We hypothesized that the safety and efficacy of minimally interrupted apixaban in patients undergoing CB ablation is noninferior to those receiving RF ablation. In this multicenter, prospective study (Japan Heat and Ice Trial [J-HIT]), we sought to determine the feasibility and safety of CB ablation with minimally interrupted apixaban in patients with PAF.

| Study Design
The J-HIT was a multicenter, pilot randomized, controlled trial (RCT) to assess the feasibility and safety of CB ablation in comparison with RF ablation. The aims of the trial were 2-folds. First, we aimed to show that CB ablation with minimally interrupted apixaban was feasible and safe by comparing to the predefined threshold (noninferiority of CB ablation). Second, we aimed to compare the feasibility and safety of CB and RF ablation prospectively (exploratory comparison of CB and RF ablation). The trial was investigator initiated; the steering committee was responsible for the design, execution, and conduct of the study. The institutional review board approved the study.
This study has been registered at the UMIN Clinical Trial Registry (number UMIN000018901). A data-and safety-monitoring board reviewed interim results and monitored the safety of the patients.
The members of an end point review committee, who were unaware of the treatment-group assignments, adjudicated primary safety events. All members of the steering committee approved the statistical analysis and interpretation of the data. The trial was funded by Bristol-Myers Squibb. A contract research organization (Clinical Research Support Center, The University of Tokyo Hospital, Tokyo, Japan) collected, monitored, maintained, and analyzed the data.

| Study participants
Five Japanese centers participated in the trial, and patients aged ≥18 and ≤85 years who had symptomatic PAF refractory to class I or class III antiarrhythmic drugs or β-blockers were enrolled. Patients were mentally and linguistically able to understand the aim of the trial and to show sufficient compliance in the trial protocol and agree to these risks and benefits as stated in the patient informed consent document. All patients have signed the informed consent form for the trial and were randomly assigned, in a 1:1 ratio, to the CB ablation (CB group) or RF ablation (RF group). Patients with mechanical heart valves, advanced hepatic or renal (creatinine clearance <15 mL/min or on dialysis) dysfunction, any condition contraindicating chronic anticoagulation including hypersensitivity to apixaban or bleeding disorders, active systemic infection, pregnant or breastfeeding women, or women of childbearing potential not on adequate birth control were excluded.

| Catheter ablation procedure and anticoagulation protocol
Patients were off antiarrhythmic drugs for at least 5 times the drugs' half-lives before the ablation. Transesophageal echocardiography was performed in all patients within 24 hours before the procedure.
All patients were treated with apixaban for at least 4 weeks before the catheter ablation. Apixaban was administered at a dose of 5 or 2.5 mg 2 times daily, and the latter dose was provided to the patients Conclusion: CB ablation with minimally interrupted apixaban was feasible and safe in patients with PAF undergoing PVI, which was equivalent to RF ablation.

K E Y W O R D S
apixaban, atrial fibrillation, catheter ablation who met 2 or more of the following criteria: age ≥ 80 years, body weight ≤ 60 kg, or serum creatinine level of ≥1.5 mg/dL. Apixaban was held at 1 dose before the procedure, and administration resumed after the procedure. All patients had a groin-entry venous- The acute procedural end point was defined as the absence of all PV potentials, as confirmed by bidirectional block using a circular mapping catheter. Use of a 3-dimensional electroanatomical mapping system, including CARTO system (Biosense Webster, Inc) or Ensite Velocity system (Abbott), was at the discretion of operators.
In all patients, the aim was to completely isolate the PVs. Therefore, touch-up ablation for complete isolation was performed if deemed necessary. No additional left atrial ablation was allowed. Patients resumed apixaban at their usual dose in the evening after the procedure and continued for 4 weeks postprocedure.

| Patient follow-up and study end points
After the procedure, a follow-up visit was scheduled for 4 weeks. At the visit, a medical history was obtained, a physical examination was performed, and a 12-lead electrocardiogram was completed. The primary end point was a composite of thromboembolic or bleeding events at 4 weeks postprocedure: cerebral stroke, transient ischemic attack, pericardial effusion requiring pericardiocentesis, hematoma or bleeding events requiring blood transfusion, and intracranial hemorrhage. The prespecified secondary end points included early recurrence of AF, minor bleeding recovered without blood transfusion, pericardial effusion not requiring pericardiocentesis, and total duration of the procedure. Early recurrence of AF was defined as the documented recurrence of AF lasting more than 30 seconds.

| Statistical analysis
In this pilot RCT, we first aimed to show single-arm, noninferiority of the CB ablation with apixaban. The reported incidence of thromboembolic or major bleeding in patients undergoing RF ablation of AF with apixaban was 2%. 9 According to the results from German ablation registry, there was no difference in the complication rate between CB and RF ablation with phenprocoumon, a derivative of coumarin. 10 In the ARISTOTLE study, the rate of stroke or systemic embolism of apixaban was not inferior to that of warfarin of 1.60% per year. In this study, we defined the noninferiority threshold of a complication rate of 5%. When the incidence and upper limit of 90% confidence interval (CI) of complications are 1.60% and 5%, respectively, we calculated that 125 patients or more would have to be enrolled. Since our second aim was to compare RF and CB ablation prospectively, we chose to randomize patients as 1:1 ratio.

| Efficacy and safety end points
The primary composite end point occurred in 1 patient in the CB group (0.8%; 90% CI, 0.04 to 3.70) and 3 patients in the RF group (2.4%), respectively (risk ratio, 0.333; 90% CI, 0.05 to 2.20; P = .622; Table 2). All 4 primary events were pericardial effusion requiring pericardiocentesis (1.6% of 250 patients). There were no cases of cerebral stroke, transient ischemic attack, hematoma or bleeding events requiring blood transfusion, or intracranial hemorrhages.
The secondary efficacy end points are also shown in Table 2. Early recurrence of AF occurred in 4 patients (3.2%) in the RF group and 6 patients All events were groin hematomas, which recovered without surgeries. The mean total procedure time was shorter in the CB group than that in the RF group (136.5 ± 39.9 vs 179.5 ± 44.8 minutes, P < .001, Figure 2).

| CB vs RF ablation
In this study, there were no thromboembolic events at 4 weeks postprocedure, suggesting the efficacy of minimally interrupted apixaban as an optimal anticoagulation method in both the RF and CB ablation groups. Pericardial effusion was the only bleeding event, and the incidence was 2.4% in the RF group and 0.8% in the CB group. The reported incidence of pericardial effusion in patients undergoing PVI using a contact force-sensing catheter was 2.5%, 11 whereas it was 1.5% in patients undergoing CB ablation. 12 In the FIRE AND ICE trial, a prospective randomized comparison of the efficacy between CB and RF ablation, the incidence of pericardial effusion was 1.3% in the RF ablation group and 0.3% in the CB ablation group. 8 A recent registry showed that the incidence of pericardial effusion was similar between CB and RF ablation groups, but it was slightly higher in patients undergoing RF ablation (1.3% vs 0.8%, P = .261). 13 In a metaanalysis comparing the efficacy and safety of CB vs RF ablations, the incidence of cardiac tamponade was significantly higher in the RF ablation group. 14 The relatively high incidence of cardiac tamponade in the RF group in comparison with the CB group in our study was consistent with these results. Although the results of our study were similar to those of these previous studies, the lack of previous randomized comparisons between the 2 technologies under the same anticoagulation methods using DOACs highlights the novelty of J-HIT apixaban.
The incidence of minor bleeding (3.2%) in our study was similar to that in the FIRE AND ICE trial (4.3%). 8 In our study, there was no difference in the early recurrence of AF across the 2 technologies (3.2% in the RF group and 4.8% in the CB group), which was also consistent with the results in the FIRE AND ICE trial (2.7% in the RF ablation group vs 0.8% in the cryoablation group, P = .09).
Phrenic nerve injury has been recognized as a common complication, particularly with CB ablation. 1,14 There were no phrenic nerve injuries in this study, possibly because of the operator's experience

TA B L E 2 Efficacy and safety end points
associated with the adequate hospital procedure volume among the centers. The procedure duration was shorter in the CB group, which was consistent with the results of previous reports, 8,13,[15][16][17][18][19][20] likely because of the shorter duration required for isolating each PV by a single-step circumferential lesion with CB ablation than that by a point-by-point circumferential lesion with RF ablation.

| Minimally interrupted vs uninterrupted apixaban
Uninterrupted warfarin has been a standard anticoagulation therapy during the periprocedural period of catheter ablation.
Recent studies have demonstrated that the use of uninterrupted or minimally interrupted DOACs during catheter ablation of AF was noninferior to the use of uninterrupted warfarin in terms of the thromboembolic and bleeding events. 1 Literature specific to apixaban also described favorable results with other DOACs. 4,9,[21][22][23][24][25] One prospective study showed that uninterrupted apixaban was safe and effective compared with uninterrupted warfarin. 4 The recent AEIOU trial investigated the utility of minimally interrupted apixaban in comparison with uninterrupted apixaban. 7 This prospective, multicenter trial randomized 300 patients to uninterrupted vs minimally interrupted periprocedural apixaban, and the results showed no stroke or thromboembolic events at 4 weeks after ablation and no difference in bleeding events between the 2 anticoagulation strategies. In the AEIOU study, 30% of the patients underwent CB ablation, but efficacy and safety were not compared between the different ablation technologies. In the J-HIT study, which collected data from 5 Japanese institutes, the addition of apixaban added unique data to the randomized comparison between CB and RF ablation, in terms of the outcomes of minimally interrupted apixaban.

| Limitations
First, this is a pilot RCT and sample size may be small to demonstrate the noninferiority in head-to-head comparison of these methods.
Second, we did not routinely perform brain magnetic resonance imaging to search for new brain lesions after the ablation, so asymptomatic thromboembolic events may have occurred. Third, neither left atrium dwell time nor fluoroscopic time during the procedure were assessed in this study. In addition, PV occlusion scores, time to PVI, duration of the freezing procedure, and number of freezes were not recorded in the CB ablation group. Fourth, neither the number of RF applications nor the contact force measurements were recorded in the RF ablation group.

| CON CLUS IONS
In this pilot RCT, CB ablation with minimally interrupted apixaban was feasible and safe in patients with PAF undergoing PVI by comparing to the predefined threshold, and CB and RF ablation was equally feasible and safe.

ACK N OWLED G M ENTS
We thank Dr Takuya Kawahara (Clinical Research Support Center, The University of Tokyo Hospital, Tokyo, Japan) for assistance with statistical analysis.

D I SCLOS U R E
The Comparison of the procedural duration between radiofrequency and cryoballoon ablation. Note the shorter procedural duration in patients undergoing cryoballoon ablation that they have no relationships to disclose that are relevant to the contents of this paper.