- Email: [email protected]
Safety and efficacy of a second-generation cryoballoon in the ablation of paroxysmal atrial fibrillation
Safety and efficacy of a second-generation cryoballoon in the ablation of paroxysmal atrial fibrillation Raphaël P. Martins, MD,*†‡§ David Hamon, MD,*†‡§ Olivier Césari, MD,║ Albin Behaghel, MD,*†‡§ Nathalie Behar, MD,*†‡§ Jean-Marc Sellal, MD,*†‡§ Jean-Claude Daubert, MD,*†‡§ Philippe Mabo, MD,*†‡§ Dominique Pavin, MD*†‡§ From the *Service de Cardiologie et Maladies Vasculaires, CHU Rennes, Rennes, France, †Université de Rennes 1, LTSI, Rennes, France, ‡INSERM, U1099, Rennes, France, §INSERM, CIC-IT 804, Rennes, France, and ║Clinique Saint Gatien, Tours, France. BACKGROUND Compared with the first-generation Arctic Front cryoballoon (ARC-CB), the new Arctic Front Advance cryoballoon (ARC-Adv-CB) increases the efficient CB-tissue contact surface during freezing, which may increase the incidence of phrenic nerve (PN) palsy (PNP). OBJECTIVE To evaluate the safety and efficacy of paroxysmal atrial fibrillation (AF) ablation with the ARC-Adv-CB as well as the merits of a predictor of PNP. METHODS AF ablation was performed by using a “single 28-mm big CB” approach. The rate of pulmonary vein (PV) isolation with a first cryoapplication was measured. The distance between the CB and a PN pacing catheter in the superior vena cava was measured to predict PNP during freezing. RESULTS In 147 patients, PV were isolated with a single cryoapplication in 205 (81.3%) of 252 PV treated with the ARC-CB and in 280 (90.3%) of 310 PV treated with the ARC-Adv-CB (P ¼ .003). The mean time to PV isolation was 52 ⫾ 34 seconds and 40 ⫾ 25 seconds (P o .001) and the temperature at the time of isolation was 36.1 ⫾ 10.31C and 32.3 ⫾ 10.21C (P ¼ .001) in the ARC-CB and ARC-AdvCB groups, respectively. Mean procedure and fluoroscopy durations were significantly shorter in the ARC-Adv-CB group. Transient PNP was observed in 7(10.6%) and 20(24.4%) of the patients treated with
Introduction After the discovery of ectopic activity in the pulmonary veins (PVs) as a major trigger, complete PV isolation (PVI) became first-line therapy for paroxysmal atrial fibrillation (AF).1,2 Although, because of its high success rate, radiofrequency energy remains the most frequently used technique for the ablation of paroxysmal AF, it is generally agreed that pointby-point ablation around the PV ostia is highly complex, The first 2 authors contributed equally to this work. Dr Césari, Dr Daubert, Dr Mabo, and Dr Pavin have received speaker honoraria and consulting fees from Medtronic. Address reprint requests and correspondence: Dr Raphaël P. Martins, Service de Cardiologie et Maladies Vasculaires, CHU de Rennes, 2 rue Henri Le Guilloux, 35000 Rennes, France. E-mail address: [email protected].
1547-5271/$-see front matter B 2014 Heart Rhythm Society. All rights reserved.
the ARC-CB and ARC-Adv-CB, respectively (P ¼ .048). The distance between the lateral edge of the CB and a vertical line through the tip of the pacing catheter accurately predicted PNP (P o .001). CONCLUSIONS The 28-mm ARC-Adv-CB enabled more efficient ablation of paroxysmal AF and shorter procedures than did the ARC-CB. This higher performance was associated with a higher incidence of PNP, which was predicted by the distance between the CB and the PN. KEYWORDS Cryoablation; Paroxysmal atrial fibrillation; Cryoballoon ablation; Pulmonary vein; Pulmonary vein isolation; Phrenic nerve palsy ABBREVIATIONS AF ¼ atrial fibrillation; AP ¼ anterior-posterior; ARC-Adv-CB ¼ Arctic Front Advance cryoballoon; ARC-CB ¼ Arctic Front cryoballoon; CB ¼ cryoballoon; CI ¼ confidence interval; LA ¼ left atrial; LIPV ¼ left inferior pulmonary vein; PN ¼ phrenic nerve; PNP ¼ phrenic nerve palsy; PV ¼ pulmonary vein; PVI ¼ pulmonary vein isolation; RIPV ¼ right inferior pulmonary vein; RSPV ¼ right superior pulmonary vein; Se ¼ sensitivity; Spe ¼ specificity; SVC ¼ superior vena cava (Heart Rhythm 2014;11:386–393) I 2014 Heart Rhythm Society. All rights reserved.
time-consuming, and sometimes the source of major complications.3,4 Cryothermal ablation with cryoballoons (CBs) enables the PVI sometimes with a single application and is safe, straightforward, and associated with a steep learning curve.5,6 Its immediate and 1-year rates of successful PVI are similar to those of radiofrequency ablation.7–9 Phrenic nerve (PN) palsy (PNP), the main complication of CB ablation, usually occurs with the use of 23-mm CBs that can be advanced farther inside the right-sided PVs.10,11 The performance of safer, more antral ablation procedures using a “single big CB” approach was described by Chun et al.5 The first-generation Arctic Front CB (ARC-CB; Medtronic, Inc, Minneapolis, MN) was coolest at the balloon’s equator. Consequently, even with a tight occlusion, the CB had to be perfectly centered inside the PV antra to http://dx.doi.org/10.1016/j.hrthm.2014.01.002
Martins et al
Cryoballoon in Paroxysmal AF Ablation
create complete lesions. Liu et al12 underscored this limitation and hastened the development of the secondgeneration Arctic Front Advance CB (ARC-Adv-CB; Medtronic, Inc) equipped with a homogeneous refrigerant system on the distal pole of the CB. Since this new design is likely to be more adaptable to atypical anatomies or imperfect CB applications, improving its overall efficacy,13,14 one may expect an increase in the incidence of PNP because of deeper freezing and the ice cap that persists after deflation of the CB.15 This nonrandomized study was performed to compare the procedural safety and efficacy of the ARC-CB with that of the ARC-Adv-CB in the treatment of highly symptomatic paroxysmal AF. We also examined the merits of a potential predictor of PNP.
Methods This study was performed according to local institutional regulations, and all patients granted their written informed consent to participate. Between August 2011 and July 2013, consecutive patients presenting with highly symptomatic, paroxysmal AF resistant to Z1 antiarrhythmic drug(s) were included in this study. They were excluded if they presented with a history of persistent AF, left atrial (LA) ablation, or surgery or with a prosthetic heart valve, an LA thrombus, or a Z28-mm-wide or Z10-mm-long, right or left common PV trunk. Patients enrolled between August 2011 and September 2012 were treated with the ARC-CB, while patients enrolled between August 2012 and June 2013 were treated with the ARC-Adv-CB, when it became available.
387 each PV. The quality of vascular occlusion was ascertained by the injection of diluted contrast material into the PV and graded from 1 to 4 as mild, medium, subtotal, or total. Once the best occlusion was obtained, cryothermal energy was applied for 300 seconds with the ARC-CB and for 240 seconds with the new ARC-Adv-CB, as recommended by the manufacturer. An additional application of energy was systematically delivered after PVI, unless PNP was observed. Before ablation of the right-sided PVs, the quadripolar catheter was relocated to the superior vena cava (SVC) to constantly pace the right PN at a 2000-ms cycle length and 20-mA output during freezing. In the case of cessation or weakening of the right hemidiaphragmatic contraction, freezing was immediately discontinued and the CB was deflated. In a subset of 40 patients, we measured before the isolation of right-sided PVs with the ARC-Adv-CB, the distance between (1) a vertical line crossing the distal SVC PN pacing catheter and (2) the lateral edge of the CB in an anterior-posterior (AP) view, using the 28-mm equatorial diameter of the CB before the initiation of freezing as a reference measurement. We also divided the CB and its surrounding space into 4 zones to facilitate and hasten the assessment of its relationship with the PN (Figure 1; from the right to the left of the patient, respectively: zone A: right side out of CB toward right PVs; zone B1: distal pole of CB from the lateral edge to its center; zone B2: proximal pole of CB from the center to its left extremity; zone C: left side, out of CB). The CB was separated in 2 zones (B1 and B2) for
Ablation procedure Before the ablation procedure, a transesophageal echocardiogram was performed to exclude the presence of an LA thrombus and measure the LA dimensions and left ventricular ejection fraction. The computed tomography of LA was performed to examine the PV anatomy. Vitamin K antagonists or other oral anticoagulants were discontinued. All procedures were performed under conscious sedation using midazolam and fentanyl as necessary. A 6F Xtrem quadripolar catheter (Sorin SPA, Milan, Italy) was placed in the coronary sinus via the right femoral vein. A single transseptal puncture was performed under fluoroscopic and pressure guidance. Thereafter, heparin was administered intravenously to maintain an activated clotting time between 250 and 350 seconds. A single big CB approach using a 28-mm CB was performed as described previously.5 The CB catheter was introduced into the left atrium through a 12F FlexCath steerable sheath (Medtronic, Minneapolis, MN) constantly flushed with heparinized saline. Finally, an Achieve mapping catheter (Medtronic) was advanced over the CB to the PV orifice and positioned as proximally as possible inside the vessel to record the PV potentials at baseline and monitor the isolation procedure in real time. Then, the CB was inflated and advanced to the ostium of
Figure 1 Phrenic nerve-CB relationship in the anterior-posterior view. The distance between the black dotted line and the lateral edge of the CB was measured (bidirectional arrow). Four zones were defined from the right to the left of the patients: zone A: right side outside the CB toward the right pulmonary veins; zone B1: distal pole of the CB from the lateral edge to its center; zone B2: proximal pole of the CB from the center to its left border; zone C: left side outside the CB. CB ¼ cryoballoon.
388 technical reasons, since the refrigerated area of the CB is the distal part of the CB, corresponding to zone B1. Conversely, zone B2 is not refrigerated and considered safer. With each delivery of energy, the time to PVI, the temperature at the time of PVI, and the lowest CB temperature reached during freezing were recorded. At 20 minutes after the end of the procedure, the persistence of PVI was ascertained and the motion of the right hemidiaphragm was confirmed on fluoroscopy by using a “sniff test.”16 Intravenous heparin was continued for 24 hours after the procedure and oral anticoagulation begun thereafter and continued for Z2 months. An echocardiogram was obtained before discharge of the patient from the hospital to confirm the absence of pericardial effusion.
Study end points The primary study end point was the rate of successful initial freezes. The secondary end points of the study were as follows: (1) the overall short-term rate of successful PVI, (2) the time to PVI, (3) the lowest CB temperature and the temperature at the time of PVI, (4) the mean duration of the procedure and of exposure to fluoroscopy, and (5) the incidence of procedure-related adverse events, including PNP.
Statistical analysis Normally distributed variables were expressed as mean ⫾ SD and compared by using the Student t test or MannWhitney U test, as appropriate. Categorical variables were expressed as counts and percentages and were compared by using the χ2 test. Discrete variables were compared by using the Fisher exact test. Receiver operating characteristic curves were constructed to determine the cutoff value, sensitivity (Se), specificity (Spe), and 95% confidence interval (CI) associated with each PN-CB distance. The areas under the curves measured the overall discriminating power of a model. A P value of o.05 was considered statistically significant. The analyses were performed with the SPSS statistical package (version 11.0, SPSS Inc, Chicago, IL).
Results We enrolled 147 patients in the study, of whom 66 (44.9%) underwent procedures with the ARC-CB and 81(55.1%) with the ARC-Adv-CB. The characteristics of both groups were similar (Table 1). After the exclusion of 3 left common PVs and 3 right common PVs with long common trunks, 252 PVs were studied in the ARC-CB group, and after the exclusion of 7 left common PVs, 310 PVs were studied in the ARC-Adv-CB group.
Procedural measurements The procedural measurements are summarized in Table 2. The PVs were completely isolated with the 28-mm CB in all patients of both groups, without requiring supplemental cryo- or radiofrequency ablation. A single cryoapplication isolated 205 (81.3%) of 252 PVs treated with the ARC-CB and 280 (90.3%) of 310 PVs treated with the ARC-Adv-CB
Heart Rhythm, Vol 11, No 3, March 2014 (P ¼ .003). The rate of single successful applications was higher at each PV separately with the ARC-Adv-CB than with the ARC-CB, though the difference reached statistical significance (P ¼ .03) only at the level of the left inferior PV (LIPV; 85.7% vs 97.3%). The PVs were isolated with a mean of 1.3 ⫾ 0.7 freezing applications (range 1–6) in the ARCCB group vs 1.1 ⫾ 0.3 (range 1–3) in the ARC-Adv-CB group (P o .001). This difference was mainly due to a significantly lower mean number of freezing applications delivered in the left PVs (1.5 ⫾ 1.0 vs 1.2 ⫾ 0.5 in the left superior PV; P ¼ .021; 1.2 ⫾ 0.6 vs 1.0 ⫾ 0.2 in the LIPV; P ¼ .009). The lowest CB temperature recorded during the initial application at each PV (49.7 ⫾ 7.61C in the ARC-CB group vs 49.3 ⫾ 7.31C in the ARC-Adv-CB group; P ¼ .558) and the PV occlusion grade (3.9 ⫾ 0.4 in both groups; P ¼ .848) were similar in both groups. PVs potentials were visualized during initial freezing in 173 (68.6%) PVs with the ARC-CB and in 216 (69.7%) PVs with the ARC-AdvCB (P ¼ .712). PV potentials were recorded on 5.6 ⫾ 1.9 electrodes of the 8-pole Achieve catheter in the ARC-CB group and 5.4 ⫾ 2.0 electrodes in the ARC-Adv-CB group (P ¼ .397). The procedure lasted a mean time of 120.1 ⫾ 24.1 minutes in the ARC-CB group vs 107.4 ⫾ 24.1 minutes in the ARC-Adv-CB group (P ¼ .002), and the duration of exposure to fluoroscopy was 28.7 ⫾ 9.9 and 25.0 ⫾ 9.2 minutes, respectively (P ¼ .02).
Time to and temperature at PVI The PVI was directly visualized in 134 (53.2%) PVs in the ARC-CB group and in 187 (60.3%) PVs in the ARC-Adv-CB group (P ¼ .106; Table 3). The mean time to PVI was 52 ⫾ 34 seconds (range 10–176 seconds) with the ARC-CB and 40 ⫾ 25 seconds (range 13–185 seconds) with the ARCAdv-CB (P o .001; Figures 2A and 2B). Therefore, the temperature at the time of PVI was less cold with the new ARC-Adv-CB, that is, 32 ⫾ 101C (range 53 to þ17) vs 36 ⫾ 101C (range 57 to þ10; P ¼ .001; Figure 2A). In separate analyses of each vein, significant temperature differences were found between the left PVs and the right inferior PV (RIPV), though not between the right superior PV (RSPV).
Incidence and predictors of PNP Among the 288 right PVI procedures, 29 (10.1%) PNPs occurred, of which 7 of 126 (5.6%) occurred with the ARCCB and 22 of 162 (13.6%) with the ARC-Adv-CB (P ¼ .044) corresponding to 10.6% and 24.7% of the patients, respectively (P ¼ .048). All PNPs occurred during the first application and most of them (71% with the ARC-CB and 77% with the ARC-Adv-CB) during RSPV isolation. In 2 patients treated with the ARC-Adv-CB, who developed transient PNP during RSPV isolation, PNP recurred during treatment of the RIPV. PNP developed earlier with the ARCAdv-CB (140 ⫾ 43 seconds) than with the ARC-CB (166 ⫾ 66 seconds), though the difference (P ¼ .238) was not
Martins et al Table 1
Cryoballoon in Paroxysmal AF Ablation
389
Baseline characteristics of the study groups
Variable
ARC-CB (n ¼ 66)
ARC-Adv-CB (n ¼ 81)
P
Sex: man Age (y) CHADS2-VASc score Years with atrial fibrillation The duration of episodes o24 h Z24 h No. of unsuccessful antiarrhythmic trials Left ventricular ejection fraction (%) Left atrial dilatation None (r33 mL/m²) Mild (34–44 mL/m²) Moderate (45–59 mL/m²) Severe (Z60 mL/m²)
48 (61) 60.9 ⫾ 10.3 1.1 ⫾ 0.9 6.0 ⫾ 5.6
54 (67) 59.4 ⫾ 9.5 1.1 ⫾ 1.1 7.9 ⫾ 7.6
.540 .350 .880 .095
55 (83.3) 11 (16.7) 2.6 ⫾ 1.1 65.8 ⫾ 5.5
61 (75.3) 20 (24.7) 2.5 ⫾ 1.1 64.7 ⫾ 7.1
.326 .326 .810 .328
43 (65.1) 11 (16.7) 8 (12.1) 4 (6.1)
52 (64.2) 16 (19.7) 8 (9.9) 5 (6.2)
.952 .790 .866 .751
Values are presented as mean ⫾ SD or as n (%). ARC-Adv-CB ¼ Arctic Front Advance cryoballoon; ARC-CB ¼ Arctic Front cryoballoon.
statistically significant. All PNPs were transient and resolved spontaneously within 0.5–20.0 minutes. It is noteworthy that PVI was observed before the development of PNP and that all isolations persisted despite the early discontinuation of the cryoapplication. Since PNP is a serious complication jeopardizing the overall safety of CB ablation, we examined, in a subset of 80 right-sided PVI with the ARC-Adv-CB, whether the distance between the lateral edge of the CB and the vertical line crossing the tip of the SVC catheter stimulating the PN (CB-PN; Figure 1) was a predictor of the 13 PNPs that Table 2
occurred in that subgroup. The mean CB-PN distance measured in patients who developed PNP was 4.9 ⫾ 5.5 mm leftward from the lateral edge of the CB vs 8.6 ⫾ 8.9 mm rightward from the lateral edge in patients who remained free from PNP (P o .001; Figure 3A). From receiver operating characteristic analysis (Figure 3B), the highest Se and Spe values were obtained at a cutoff value of 0.0 mm (Se 92.3%; 95% CI 63.9–98.7; Spe 86.6%; 95% CI 76.0–93.7; area under the curve 0.91; 95% CI 0.82–0.96; P o .0001). When we examined the position of the SVC pacing catheter relative to the 4 zones during freezing (Figures 1 and 4),
Procedural measurements in each study group
Variable Occlusion grade: 1–4 PVP visualization All measurements LSPV LIPV RSPV RIPV No. of electrodes: 1–8 Lowest temperature (1C) All measurements LSPV LIPV RSPV RIPV Mean number of cryoapplications All measurements LSPV LIPV RSPV RIPV No. of single cryoapplications All measurements LSPV LIPV RSPV RIPV
ARC-CB (n ¼ 252 PV)
ARC-Adv-CB (n ¼ 310 PV)
P
3.9 ⫾ 0.4
3.9 ⫾ 0.4
.848
173 (68.6) 58 (92.1) 47 (74.6) 51 (80.9) 17 (27.0) 5.6 ⫾ 1.9
216 (69.7) 71 (95.9) 60 (81.1) 61 (75.3) 24 (29.6) 5.4 ⫾ 2.0
.712 .548 .362 .544 .871 .397
49.7 ⫾ 51.2 ⫾ 46.8 ⫾ 52.6 ⫾ 48.2 ⫾ 1.3 ⫾ 1.5 ⫾ 1.2 ⫾ 1.2 ⫾ 1.1 ⫾
7.6 6.9 7.3 7.2 7.6 0.7 (1–6) 1.0 0.6 0.6 0.4
205 (81.3) 43 (68.2) 54 (85.7) 52 (82.5) 56 (88.9)
49.3 ⫾ 48.4 ⫾ 47.3 ⫾ 52.0 ⫾ 49.4 ⫾ 1.1 ⫾ 1.2 ⫾ 1.0 ⫾ 1.1 ⫾ 1.1 ⫾
7.3 7.1 5.6 7.5 7.9 0.35 (1–3) 0.5 0.2 0.3 0.3
280 (90.3) 61 (82.4) 72 (97.3) 71 (87.6) 76 (93.8)
.558 .023 .670 .590 .355 o.001 .021 .009 .099 .268 .003 .083 .030 .532 .447
Values are presented as mean ⫾ SD or as n (%). ARC-Adv-CB ¼ Arctic Front Advance cryoballoon; ARC-CB ¼ Arctic Front cryoballoon; LIPV ¼ left inferior pulmonary vein; LSPV ¼ left superior pulmonary vein; PV ¼ pulmonary vein; PVP ¼ pulmonary vein potential; RIPV ¼ right inferior pulmonary vein; RSPV ¼ right superior pulmonary vein.
390 Table 3
Heart Rhythm, Vol 11, No 3, March 2014 Time and temperature at the time of PVI
Variable
ARC-CB (n ¼ 134)
ARC-CB (n ¼ 187)
Real-time PVP visualization Time to isolation (s) All veins LSPV (40/57) LIPV (40/58) RSPV (42/52) RIPV (12/20)
53.2%
60.3%
Temperature at time of isolation (1C) All veins LSPV (40/57) LIPV (40/58) RSPV (42/52) RIPV (12/20)
P .106
52 ⫾ 57 ⫾ 50 ⫾ 46 ⫾ 59 ⫾
34 35 34 32 37
40 ⫾ 44 ⫾ 34 ⫾ 41 ⫾ 48 ⫾
25 20 27 23 33
o.001 .023 .011 .326 .385
36 ⫾ 39 ⫾ 33 ⫾ 35 ⫾ 39 ⫾
10 8 10 12 6
32 ⫾ 35 ⫾ 26 ⫾ 36 ⫾ 33 ⫾
10 7 11 10 8
.001 .01 .002 .675 .038
Values are presented as mean ⫾ SD or as n (%). LIPV ¼ left inferior pulmonary vein; LSPV ¼ left superior pulmonary vein; PVP ¼ pulmonary vein potential; RIPV ¼ right inferior pulmonary vein; RSPV ¼ right superior pulmonary vein.
12 of 13 PNPs were in zone B1 and 1 was in zone A. In the 67 uncomplicated applications, the pacing catheter crossed zone B1 in 7 and another zone in 60 (Figure 3A). Thus, when the pacing catheter crossed zone B1 in the AP view, the likelihood of PN injury during right-sided PVI was high (Se 92.3%; Spe 89.9%; negative and positive predictive values 98.4% and 63.2%, respectively).
Other adverse events A small pericardial effusion was observed in the ARC-AdvCB group, and a single femoral arteriovenous fistula occurred in each group, which both resolved spontaneously within 1 month after the ablation procedure.
Discussion Main findings of our study Compared with that of the ARC-CB, the overall ability of the ARC-Adv-CB to completely isolate the PVs of patients suffering from paroxysmal AF was superior. While all PVs were ultimately isolated in all patients of both groups, fewer
freezing applications were needed to isolate the PVs and the rate of single freezes was higher when the ARC-Adv-CB was used instead of the ARC-CB, significantly shortening the mean duration of the procedure and of exposure to fluoroscopy. In most instances, the PVI was directly visualized and the time to isolation was significantly shortened, increasing the mean temperature at the time of isolation. However, the rate of transient right-sided PNP was significantly higher with the ARC-Adv-CB and strongly correlated with the estimates of procedural CB-PN proximity by using the SVC pacing catheter.
Contributions of the ARC-Adv-CB Compared with the ARC-CB, the new CB has a 2-fold greater number of injection spray jets that have been optimally located to more uniformly cool its tip. It is noteworthy that the higher performance of the ARC-AdvCB was nearly entirely confined to the left PV freezing applications while the differences in right PVs were small. This observation has several putative explanations. First, in our experience with the ARC-CB, the isolation of the
Figure 2 A: Time to disconnection and temperature at the time of PVI. B: Time-temperature relationship for all PV with visible PV potentials during cryoapplications. ARC-Adv-CB ¼ Arctic Front Advance cryoballoon; ARC-CB ¼ Arctic Front cryoballoon; PV ¼ pulmonary vein; PVI ¼ pulmonary vein isolation.
Martins et al
Cryoballoon in Paroxysmal AF Ablation
391
Figure 3 Incidence of PNP with the ARC-Adv-CB. A: Distance between the CB and the PN in patients with vs without PNP. B: Receiver operating characteristic analysis of CB-PN distance, showing an optimal cutoff value of 0.0 cm. ARC-Adv-CB ¼ Arctic Front Advance cryoballoon; CB ¼ cryoballoon; PN ¼ phrenic nerve; PNP ¼ phrenic nerve palsy.
left-sided PVs was more challenging and required a greater number of freezing applications than the right-sided PVs (Table 2). Second, as in the case of radiofrequency ablation, the ridge between the PV ostia and the LA appendage represents an anatomical obstacle for the CB, which is responsible for a high proportion of recurrent connections between the left PVs and the left atrium.17 Inferior segments are also difficult to isolate and are preferential sites of PV reconnection. Chun et al5 used the cross-talk technique in more than one-third of the procedures to isolate the left superior PV, in contrast to a “straightforward” approach for all RSPV, highlighting the difficult access to the distal electrical connection between the left PV with the ARCCB. Third, the higher ovality index of the left-sided PVs than of the right-sided PVs may also be a factor, as an inverse association was found between this index and the degree of occlusion.14
Our results18 are consistent with those of a recent study, which reported a significantly greater PVI efficacy associated with the ARC-Adv-CB than with the first-generation CB. The authors accurately described the technical improvements represented by the ARC-Adv-CB, consisting in an increase in the effective CB-tissue contact area as well as the more prominent ice formation within the PV, known as the “ice cap phenomenon.”15 They hypothesized that even with a suboptimal contact, the ice cap prevents rewarming of the blood flow and enables PVI. In the study by Fürnkranz et al,18 the overall minimum CB temperatures were lower with the ARC-Adv-CB than with the ARC-CB. Some procedural characteristics may explain this difference. Indeed, the distance between the CB surface and the proximal Achieve electrode during freezing decreased between the ARC-CB and the ARC-Adv-CB, facilitating the visualization of PV potentials. This might have promoted the contact with the PV ostia and lower freezing
Figure 4 Phrenic nerve-cryoballoon relationship in the anterior-posterior view during the right superior pulmonary vein cryoapplication. A: zone A; B: zone B1; C: zone B2.
392 temperatures.17 Conversely, in our study, though we did not measure this distance, the rates of PV potential visualization and the low temperatures were similar. The exclusion of wide and long common PV trunks may also have caused these differences since they are usually associated with lower temperatures.18
Procedural safety The greater efficacy of the ARC-Adv-CB was associated with a significant increase in transient PNP in our study. The previously published incidence of PNP has been variable: Casado-Arroyo et al19 reported a significant increase with the ARC-Adv-CB, while Fürnkranz et al18 observed no difference. This discrepancy might be explained by the definition of PNP, as some authors report only persistent PNP while others report persistent and transient PNP. The new design of the ARC-Adv-CB has considerably increased the cooling area near the tip of the CB, increasing the risk of PN injury. In this study, we found an intraprocedural, simple, and reliable predictor of PNP. The vertical projection of the SVC pacing catheter to the distal segment of the CB in the AP view (zone B1, corresponding to the cooling zone) was an excellent predictor of PNP. Of the 13 cases of PNP, a single incident occurred while the line crossed zone A in the AP view, though it may have been caused by an ice cap or by an unusual oblique course of the PN. The anatomical relationship between the right PN and the RSPV is variable and difficult to evaluate before the procedure. In a study using computed tomography scans, segments of the right pericardiophrenic artery, used to locate the PN, were identified in 20% of the patients,20 while earlier studies had reported higher rates.21 A short distance between the RSPV and the SVC was recently found to be correlated with PNP.22 However, the precise course of the PN varies among patients, and the force used to occlude the PV may distort the anatomy. The same authors found that a precipitous drop in CB temperature (ie, o411C at 30 seconds) was a strong predictor of PNP.22 However, while low temperatures are usually associated with more distal and tight PV occlusions, in this study, of 29 transient PNPs, a single case would have been predicted by such a precipitous drop in temperature.
Clinical implications The finding of a reliable predictor of PNP raises the issue of prevention when the risk of PNP is high. When the SVC pacing catheter crossed zone B1 in the AP view and as soon as a weakening of the right hemidiaphragm contraction was suspected, we immediately discontinued freezing, deflated the CB, and abstained from further treatment delivery, probably explaining the absence of persistent PNP. We believe that when the risk of PNP during RSPV treatment is high, the RIPV should be frozen first in order to allow the monitoring of PN during the remaining cryoapplications. Most importantly, Casado-Arroyo et al19 described a highly promising technique to prevent PNP, consisting, after tight wedging of the inflated CB inside the RSPV ostium, to
Heart Rhythm, Vol 11, No 3, March 2014 withdraw it until a small leak of contrast is observed, since the CB volume increases slightly at the onset of the cryoapplication. This offers the advantage of a more proximal cryoapplication. In light of our observations, their lower rate of PNP might be explained by a shift of the PN projection from zone B1 to zone A. Because of the 498% negative predictive value of our PNP predictor, we suggest the use of the CasadoArroyo et al19 technique, particularly when the vertical projection of the PN reaches the distal part of the CB (zone B1). It is noteworthy that in 2 of 80 cases of PN projection in zone B2, the withdrawal of the CB might have increased the risk of PNP since it shifted toward zone B1. Importantly, in all instances of PNP and early cessation of the cryoapplication, the right PVs were successfully isolated without short-term reconnection. Indeed, if PNP occurs, it implies that energy delivery extends to extracardiac structures. Thus, one can imagine that the PV lesion is transmural and successful PVI is achieved. Finally, the PVs were isolated more rapidly when we used the ARC-Adv-CB instead of the ARC-CB. Rapid PVIs are associated with low immediate reconnection rates,23,24 and one may hypothesize that the cryoapplication could be shortened to limit the rate of procedural complications, including PNP. However, the time to PVI has not been found to be a predictor of long-term reconnection, and further studies are needed to assess its contribution in the adjustment of the duration of applications and need for additional cryoapplications.
Study limitations Our study was not randomized, as we did not have access to both CB generations simultaneously, and the patient enrollment overlapped for a period of only 2 months. We do not believe, however, that this introduced important biases, as the same trained operators participated in both phases of the study. A learning curve effect was probably not a factor, since the occlusion grade and visualization of PV potentials were identical in both groups. Moreover, whether a faster PVI is associated with a lower long-term incidence of PV reconnection is unknown and requires further studies. The new predictor of PNP, identified on the basis of a subset of 80 isolations of right PVs, will need to be confirmed with a larger number of observations. Furthermore, only 2 patients had the PN pacing catheter crossing zone B2 and PNP did not occur during these cryoenergy applications; more observations will be needed to confirm the safety of this zone, corresponding to the nonrefrigerated part of the CB. More importantly, these observations were made with a 28-mm CB, limiting the conclusions to this size of the CB. Finally, since short-term PVI may not last indefinitely, whether this higher procedural success rate translates into a higher long-term rate of PVI and AF suppression will need to be confirmed with long-term follow-ups.
Conclusions Cryoablation of paroxysmal AF with the new 28-mm ARCAdv-CB was associated with a 490% success rate with a
Martins et al
Cryoballoon in Paroxysmal AF Ablation
single freezing application and with shorter procedures than with the first-generation CB. However, this higher efficacy was accompanied by a troubling increase in transient PNP. Consequently, we found that the distance between the CB and the PN was a reliable and straightforward intraprocedural predictor of PNP.
Acknowledgment We thank Rodolphe Ruffy for the manuscript revision.
References 1. Haissaguerre M, Jais P, Shah DC, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 1998;339: 659–666. 2. Natale A, Raviele A, Arentz T, et al. Venice Chart international consensus document on atrial fibrillation ablation. J Cardiovasc Electrophysiol 2007;18: 560–580. 3. Cappato R, Calkins H, Chen SA, et al. Updated worldwide survey on the methods, efficacy, and safety of catheter ablation for human atrial fibrillation. Circ Arrhythm Electrophysiol 2010;3:32–38. 4. Weerasooriya R, Jais P, Hocini M, et al. Balloon cryoablation for paroxysmal atrial fibrillation. Europace 2008;10:1251–1252. 5. Chun KR, Schmidt B, Metzner A, et al. The ‘single big cryoballoon’ technique for acute pulmonary vein isolation in patients with paroxysmal atrial fibrillation: a prospective observational single centre study. Eur Heart J 2009;30:699–709. 6. Khairy P, Chauvet P, Lehmann J, et al. Lower incidence of thrombus formation with cryoenergy versus radiofrequency catheter ablation. Circulation 2003;107: 2045–2050. 7. Packer DL, Kowal RC, Wheelan KR, et al. Cryoballoon ablation of pulmonary veins for paroxysmal atrial fibrillation: first results of the North American Arctic Front (STOP AF) pivotal trial. J Am Coll Cardiol 2013;61:1713–1723. 8. Van Belle Y, Janse P, Theuns D, et al. One year follow-up after cryoballoon isolation of the pulmonary veins in patients with paroxysmal atrial fibrillation. Europace 2008;10:1271–1276. 9. Vogt J, Heintze J, Gutleben KJ, et al. Long-term outcomes after cryoballoon pulmonary vein isolation: results from a prospective study in 605 patients. J Am Coll Cardiol 2013;61:1707–1712.
393 10. Van Belle Y, Janse P, Rivero-Ayerza MJ, et al. Pulmonary vein isolation using an occluding cryoballoon for circumferential ablation: feasibility, complications, and short-term outcome. Eur Heart J 2007;28:2231–2237. 11. Neumann T, Vogt J, Schumacher B, et al. Circumferential pulmonary vein isolation with the cryoballoon technique results from a prospective 3-center study. J Am Coll Cardiol 2008;52:273–278. 12. Liu CF. Pulmonary vein reconnection after cryoballoon ablation: back to the drawing board. Heart Rhythm 2010;7:191–192. 13. Kubala M, Hermida JS, Nadji G, et al. Normal pulmonary veins anatomy is associated with better AF-free survival after cryoablation as compared to atypical anatomy with common left pulmonary vein. Pacing Clin Electrophysiol 2011;34: 837–843. 14. Sorgente A, Chierchia GB, de Asmundis C, et al. Pulmonary vein ostium shape and orientation as possible predictors of occlusion in patients with drug-refractory paroxysmal AF undergoing cryoballoon ablation. Europace 2011;13:205–212. 15. Bordignon S, Furnkranz A, Schmidt B, Chun KR. Remaining ice cap on secondgeneration cryoballoon after deflation. Circ Arrhythm Electrophysiol 2012;5: e98–e99. 16. Qureshi A. Diaphragm paralysis. Semin Respir Crit Care Med 2009;30:315–320. 17. Fürnkranz A, Chun KR, Nuyens D, et al. Characterization of conduction recovery after pulmonary vein isolation using the “single big cryoballoon” technique. Heart Rhythm 2010;7:184–190. 18. Fürnkranz A, Bordignon S, Schmidt B, et al. Improved procedural efficacy of pulmonary vein isolation using the novel second-generation cryoballoon. J Cardiovasc Electrophysiol 2013;24:492–497. 19. Casado-Arroyo R, Chierchia GB, Conte G, et al. Phrenic nerve paralysis during cryoballoon ablation for atrial fibrillation: a comparison between the first- and second-generation balloon. Heart Rhythm 2013;10:1318–1324. 20. Nieto-Tolosa J, Rodriguez-Sanchez D, Hurtado-Martinez JA, et al. [Phrenic nerve identification with cardiac multidetector computed tomography]. Rev Esp Cardiol 2010;64:942–944. 21. Horton R, Di Biase L, Reddy V, et al. Locating the right phrenic nerve by imaging the right pericardiophrenic artery with computerized tomographic angiography: implications for balloon-based procedures. Heart Rhythm 2010;7:937–941. 22. Kühne M, Knecht S, Altmann D, et al. Phrenic nerve palsy during ablation of AF using a 28-mm cryoballoon catheter: predictors and prevention. J Interv Card Electrophysiol 2012;36:47–54. 23. Chun KR, Fürnkranz A, Metzner A, et al. Cryoballoon pulmonary vein isolation with real-time recordings from the pulmonary veins. J Cardiovasc Electrophysiol 2009;20:1203–1210. 24. Dorwarth U, Schmidt M, Wankerl M, Krieg J, Straube F, Hoffmann E. Pulmonary vein electrophysiology during cryoballoon ablation as a predictor for procedural success. J Interv Card Electrophysiol 2011;32:205–211.
Introduction After the discovery of ectopic activity in the pulmonary veins (PVs) as a major trigger, complete PV isolation (PVI) became first-line therapy for paroxysmal atrial fibrillation (AF).1,2 Although, because of its high success rate, radiofrequency energy remains the most frequently used technique for the ablation of paroxysmal AF, it is generally agreed that pointby-point ablation around the PV ostia is highly complex, The first 2 authors contributed equally to this work. Dr Césari, Dr Daubert, Dr Mabo, and Dr Pavin have received speaker honoraria and consulting fees from Medtronic. Address reprint requests and correspondence: Dr Raphaël P. Martins, Service de Cardiologie et Maladies Vasculaires, CHU de Rennes, 2 rue Henri Le Guilloux, 35000 Rennes, France. E-mail address: [email protected].
1547-5271/$-see front matter B 2014 Heart Rhythm Society. All rights reserved.
the ARC-CB and ARC-Adv-CB, respectively (P ¼ .048). The distance between the lateral edge of the CB and a vertical line through the tip of the pacing catheter accurately predicted PNP (P o .001). CONCLUSIONS The 28-mm ARC-Adv-CB enabled more efficient ablation of paroxysmal AF and shorter procedures than did the ARC-CB. This higher performance was associated with a higher incidence of PNP, which was predicted by the distance between the CB and the PN. KEYWORDS Cryoablation; Paroxysmal atrial fibrillation; Cryoballoon ablation; Pulmonary vein; Pulmonary vein isolation; Phrenic nerve palsy ABBREVIATIONS AF ¼ atrial fibrillation; AP ¼ anterior-posterior; ARC-Adv-CB ¼ Arctic Front Advance cryoballoon; ARC-CB ¼ Arctic Front cryoballoon; CB ¼ cryoballoon; CI ¼ confidence interval; LA ¼ left atrial; LIPV ¼ left inferior pulmonary vein; PN ¼ phrenic nerve; PNP ¼ phrenic nerve palsy; PV ¼ pulmonary vein; PVI ¼ pulmonary vein isolation; RIPV ¼ right inferior pulmonary vein; RSPV ¼ right superior pulmonary vein; Se ¼ sensitivity; Spe ¼ specificity; SVC ¼ superior vena cava (Heart Rhythm 2014;11:386–393) I 2014 Heart Rhythm Society. All rights reserved.
time-consuming, and sometimes the source of major complications.3,4 Cryothermal ablation with cryoballoons (CBs) enables the PVI sometimes with a single application and is safe, straightforward, and associated with a steep learning curve.5,6 Its immediate and 1-year rates of successful PVI are similar to those of radiofrequency ablation.7–9 Phrenic nerve (PN) palsy (PNP), the main complication of CB ablation, usually occurs with the use of 23-mm CBs that can be advanced farther inside the right-sided PVs.10,11 The performance of safer, more antral ablation procedures using a “single big CB” approach was described by Chun et al.5 The first-generation Arctic Front CB (ARC-CB; Medtronic, Inc, Minneapolis, MN) was coolest at the balloon’s equator. Consequently, even with a tight occlusion, the CB had to be perfectly centered inside the PV antra to http://dx.doi.org/10.1016/j.hrthm.2014.01.002
Martins et al
Cryoballoon in Paroxysmal AF Ablation
create complete lesions. Liu et al12 underscored this limitation and hastened the development of the secondgeneration Arctic Front Advance CB (ARC-Adv-CB; Medtronic, Inc) equipped with a homogeneous refrigerant system on the distal pole of the CB. Since this new design is likely to be more adaptable to atypical anatomies or imperfect CB applications, improving its overall efficacy,13,14 one may expect an increase in the incidence of PNP because of deeper freezing and the ice cap that persists after deflation of the CB.15 This nonrandomized study was performed to compare the procedural safety and efficacy of the ARC-CB with that of the ARC-Adv-CB in the treatment of highly symptomatic paroxysmal AF. We also examined the merits of a potential predictor of PNP.
Methods This study was performed according to local institutional regulations, and all patients granted their written informed consent to participate. Between August 2011 and July 2013, consecutive patients presenting with highly symptomatic, paroxysmal AF resistant to Z1 antiarrhythmic drug(s) were included in this study. They were excluded if they presented with a history of persistent AF, left atrial (LA) ablation, or surgery or with a prosthetic heart valve, an LA thrombus, or a Z28-mm-wide or Z10-mm-long, right or left common PV trunk. Patients enrolled between August 2011 and September 2012 were treated with the ARC-CB, while patients enrolled between August 2012 and June 2013 were treated with the ARC-Adv-CB, when it became available.
387 each PV. The quality of vascular occlusion was ascertained by the injection of diluted contrast material into the PV and graded from 1 to 4 as mild, medium, subtotal, or total. Once the best occlusion was obtained, cryothermal energy was applied for 300 seconds with the ARC-CB and for 240 seconds with the new ARC-Adv-CB, as recommended by the manufacturer. An additional application of energy was systematically delivered after PVI, unless PNP was observed. Before ablation of the right-sided PVs, the quadripolar catheter was relocated to the superior vena cava (SVC) to constantly pace the right PN at a 2000-ms cycle length and 20-mA output during freezing. In the case of cessation or weakening of the right hemidiaphragmatic contraction, freezing was immediately discontinued and the CB was deflated. In a subset of 40 patients, we measured before the isolation of right-sided PVs with the ARC-Adv-CB, the distance between (1) a vertical line crossing the distal SVC PN pacing catheter and (2) the lateral edge of the CB in an anterior-posterior (AP) view, using the 28-mm equatorial diameter of the CB before the initiation of freezing as a reference measurement. We also divided the CB and its surrounding space into 4 zones to facilitate and hasten the assessment of its relationship with the PN (Figure 1; from the right to the left of the patient, respectively: zone A: right side out of CB toward right PVs; zone B1: distal pole of CB from the lateral edge to its center; zone B2: proximal pole of CB from the center to its left extremity; zone C: left side, out of CB). The CB was separated in 2 zones (B1 and B2) for
Ablation procedure Before the ablation procedure, a transesophageal echocardiogram was performed to exclude the presence of an LA thrombus and measure the LA dimensions and left ventricular ejection fraction. The computed tomography of LA was performed to examine the PV anatomy. Vitamin K antagonists or other oral anticoagulants were discontinued. All procedures were performed under conscious sedation using midazolam and fentanyl as necessary. A 6F Xtrem quadripolar catheter (Sorin SPA, Milan, Italy) was placed in the coronary sinus via the right femoral vein. A single transseptal puncture was performed under fluoroscopic and pressure guidance. Thereafter, heparin was administered intravenously to maintain an activated clotting time between 250 and 350 seconds. A single big CB approach using a 28-mm CB was performed as described previously.5 The CB catheter was introduced into the left atrium through a 12F FlexCath steerable sheath (Medtronic, Minneapolis, MN) constantly flushed with heparinized saline. Finally, an Achieve mapping catheter (Medtronic) was advanced over the CB to the PV orifice and positioned as proximally as possible inside the vessel to record the PV potentials at baseline and monitor the isolation procedure in real time. Then, the CB was inflated and advanced to the ostium of
Figure 1 Phrenic nerve-CB relationship in the anterior-posterior view. The distance between the black dotted line and the lateral edge of the CB was measured (bidirectional arrow). Four zones were defined from the right to the left of the patients: zone A: right side outside the CB toward the right pulmonary veins; zone B1: distal pole of the CB from the lateral edge to its center; zone B2: proximal pole of the CB from the center to its left border; zone C: left side outside the CB. CB ¼ cryoballoon.
388 technical reasons, since the refrigerated area of the CB is the distal part of the CB, corresponding to zone B1. Conversely, zone B2 is not refrigerated and considered safer. With each delivery of energy, the time to PVI, the temperature at the time of PVI, and the lowest CB temperature reached during freezing were recorded. At 20 minutes after the end of the procedure, the persistence of PVI was ascertained and the motion of the right hemidiaphragm was confirmed on fluoroscopy by using a “sniff test.”16 Intravenous heparin was continued for 24 hours after the procedure and oral anticoagulation begun thereafter and continued for Z2 months. An echocardiogram was obtained before discharge of the patient from the hospital to confirm the absence of pericardial effusion.
Study end points The primary study end point was the rate of successful initial freezes. The secondary end points of the study were as follows: (1) the overall short-term rate of successful PVI, (2) the time to PVI, (3) the lowest CB temperature and the temperature at the time of PVI, (4) the mean duration of the procedure and of exposure to fluoroscopy, and (5) the incidence of procedure-related adverse events, including PNP.
Statistical analysis Normally distributed variables were expressed as mean ⫾ SD and compared by using the Student t test or MannWhitney U test, as appropriate. Categorical variables were expressed as counts and percentages and were compared by using the χ2 test. Discrete variables were compared by using the Fisher exact test. Receiver operating characteristic curves were constructed to determine the cutoff value, sensitivity (Se), specificity (Spe), and 95% confidence interval (CI) associated with each PN-CB distance. The areas under the curves measured the overall discriminating power of a model. A P value of o.05 was considered statistically significant. The analyses were performed with the SPSS statistical package (version 11.0, SPSS Inc, Chicago, IL).
Results We enrolled 147 patients in the study, of whom 66 (44.9%) underwent procedures with the ARC-CB and 81(55.1%) with the ARC-Adv-CB. The characteristics of both groups were similar (Table 1). After the exclusion of 3 left common PVs and 3 right common PVs with long common trunks, 252 PVs were studied in the ARC-CB group, and after the exclusion of 7 left common PVs, 310 PVs were studied in the ARC-Adv-CB group.
Procedural measurements The procedural measurements are summarized in Table 2. The PVs were completely isolated with the 28-mm CB in all patients of both groups, without requiring supplemental cryo- or radiofrequency ablation. A single cryoapplication isolated 205 (81.3%) of 252 PVs treated with the ARC-CB and 280 (90.3%) of 310 PVs treated with the ARC-Adv-CB
Heart Rhythm, Vol 11, No 3, March 2014 (P ¼ .003). The rate of single successful applications was higher at each PV separately with the ARC-Adv-CB than with the ARC-CB, though the difference reached statistical significance (P ¼ .03) only at the level of the left inferior PV (LIPV; 85.7% vs 97.3%). The PVs were isolated with a mean of 1.3 ⫾ 0.7 freezing applications (range 1–6) in the ARCCB group vs 1.1 ⫾ 0.3 (range 1–3) in the ARC-Adv-CB group (P o .001). This difference was mainly due to a significantly lower mean number of freezing applications delivered in the left PVs (1.5 ⫾ 1.0 vs 1.2 ⫾ 0.5 in the left superior PV; P ¼ .021; 1.2 ⫾ 0.6 vs 1.0 ⫾ 0.2 in the LIPV; P ¼ .009). The lowest CB temperature recorded during the initial application at each PV (49.7 ⫾ 7.61C in the ARC-CB group vs 49.3 ⫾ 7.31C in the ARC-Adv-CB group; P ¼ .558) and the PV occlusion grade (3.9 ⫾ 0.4 in both groups; P ¼ .848) were similar in both groups. PVs potentials were visualized during initial freezing in 173 (68.6%) PVs with the ARC-CB and in 216 (69.7%) PVs with the ARC-AdvCB (P ¼ .712). PV potentials were recorded on 5.6 ⫾ 1.9 electrodes of the 8-pole Achieve catheter in the ARC-CB group and 5.4 ⫾ 2.0 electrodes in the ARC-Adv-CB group (P ¼ .397). The procedure lasted a mean time of 120.1 ⫾ 24.1 minutes in the ARC-CB group vs 107.4 ⫾ 24.1 minutes in the ARC-Adv-CB group (P ¼ .002), and the duration of exposure to fluoroscopy was 28.7 ⫾ 9.9 and 25.0 ⫾ 9.2 minutes, respectively (P ¼ .02).
Time to and temperature at PVI The PVI was directly visualized in 134 (53.2%) PVs in the ARC-CB group and in 187 (60.3%) PVs in the ARC-Adv-CB group (P ¼ .106; Table 3). The mean time to PVI was 52 ⫾ 34 seconds (range 10–176 seconds) with the ARC-CB and 40 ⫾ 25 seconds (range 13–185 seconds) with the ARCAdv-CB (P o .001; Figures 2A and 2B). Therefore, the temperature at the time of PVI was less cold with the new ARC-Adv-CB, that is, 32 ⫾ 101C (range 53 to þ17) vs 36 ⫾ 101C (range 57 to þ10; P ¼ .001; Figure 2A). In separate analyses of each vein, significant temperature differences were found between the left PVs and the right inferior PV (RIPV), though not between the right superior PV (RSPV).
Incidence and predictors of PNP Among the 288 right PVI procedures, 29 (10.1%) PNPs occurred, of which 7 of 126 (5.6%) occurred with the ARCCB and 22 of 162 (13.6%) with the ARC-Adv-CB (P ¼ .044) corresponding to 10.6% and 24.7% of the patients, respectively (P ¼ .048). All PNPs occurred during the first application and most of them (71% with the ARC-CB and 77% with the ARC-Adv-CB) during RSPV isolation. In 2 patients treated with the ARC-Adv-CB, who developed transient PNP during RSPV isolation, PNP recurred during treatment of the RIPV. PNP developed earlier with the ARCAdv-CB (140 ⫾ 43 seconds) than with the ARC-CB (166 ⫾ 66 seconds), though the difference (P ¼ .238) was not
Martins et al Table 1
Cryoballoon in Paroxysmal AF Ablation
389
Baseline characteristics of the study groups
Variable
ARC-CB (n ¼ 66)
ARC-Adv-CB (n ¼ 81)
P
Sex: man Age (y) CHADS2-VASc score Years with atrial fibrillation The duration of episodes o24 h Z24 h No. of unsuccessful antiarrhythmic trials Left ventricular ejection fraction (%) Left atrial dilatation None (r33 mL/m²) Mild (34–44 mL/m²) Moderate (45–59 mL/m²) Severe (Z60 mL/m²)
48 (61) 60.9 ⫾ 10.3 1.1 ⫾ 0.9 6.0 ⫾ 5.6
54 (67) 59.4 ⫾ 9.5 1.1 ⫾ 1.1 7.9 ⫾ 7.6
.540 .350 .880 .095
55 (83.3) 11 (16.7) 2.6 ⫾ 1.1 65.8 ⫾ 5.5
61 (75.3) 20 (24.7) 2.5 ⫾ 1.1 64.7 ⫾ 7.1
.326 .326 .810 .328
43 (65.1) 11 (16.7) 8 (12.1) 4 (6.1)
52 (64.2) 16 (19.7) 8 (9.9) 5 (6.2)
.952 .790 .866 .751
Values are presented as mean ⫾ SD or as n (%). ARC-Adv-CB ¼ Arctic Front Advance cryoballoon; ARC-CB ¼ Arctic Front cryoballoon.
statistically significant. All PNPs were transient and resolved spontaneously within 0.5–20.0 minutes. It is noteworthy that PVI was observed before the development of PNP and that all isolations persisted despite the early discontinuation of the cryoapplication. Since PNP is a serious complication jeopardizing the overall safety of CB ablation, we examined, in a subset of 80 right-sided PVI with the ARC-Adv-CB, whether the distance between the lateral edge of the CB and the vertical line crossing the tip of the SVC catheter stimulating the PN (CB-PN; Figure 1) was a predictor of the 13 PNPs that Table 2
occurred in that subgroup. The mean CB-PN distance measured in patients who developed PNP was 4.9 ⫾ 5.5 mm leftward from the lateral edge of the CB vs 8.6 ⫾ 8.9 mm rightward from the lateral edge in patients who remained free from PNP (P o .001; Figure 3A). From receiver operating characteristic analysis (Figure 3B), the highest Se and Spe values were obtained at a cutoff value of 0.0 mm (Se 92.3%; 95% CI 63.9–98.7; Spe 86.6%; 95% CI 76.0–93.7; area under the curve 0.91; 95% CI 0.82–0.96; P o .0001). When we examined the position of the SVC pacing catheter relative to the 4 zones during freezing (Figures 1 and 4),
Procedural measurements in each study group
Variable Occlusion grade: 1–4 PVP visualization All measurements LSPV LIPV RSPV RIPV No. of electrodes: 1–8 Lowest temperature (1C) All measurements LSPV LIPV RSPV RIPV Mean number of cryoapplications All measurements LSPV LIPV RSPV RIPV No. of single cryoapplications All measurements LSPV LIPV RSPV RIPV
ARC-CB (n ¼ 252 PV)
ARC-Adv-CB (n ¼ 310 PV)
P
3.9 ⫾ 0.4
3.9 ⫾ 0.4
.848
173 (68.6) 58 (92.1) 47 (74.6) 51 (80.9) 17 (27.0) 5.6 ⫾ 1.9
216 (69.7) 71 (95.9) 60 (81.1) 61 (75.3) 24 (29.6) 5.4 ⫾ 2.0
.712 .548 .362 .544 .871 .397
49.7 ⫾ 51.2 ⫾ 46.8 ⫾ 52.6 ⫾ 48.2 ⫾ 1.3 ⫾ 1.5 ⫾ 1.2 ⫾ 1.2 ⫾ 1.1 ⫾
7.6 6.9 7.3 7.2 7.6 0.7 (1–6) 1.0 0.6 0.6 0.4
205 (81.3) 43 (68.2) 54 (85.7) 52 (82.5) 56 (88.9)
49.3 ⫾ 48.4 ⫾ 47.3 ⫾ 52.0 ⫾ 49.4 ⫾ 1.1 ⫾ 1.2 ⫾ 1.0 ⫾ 1.1 ⫾ 1.1 ⫾
7.3 7.1 5.6 7.5 7.9 0.35 (1–3) 0.5 0.2 0.3 0.3
280 (90.3) 61 (82.4) 72 (97.3) 71 (87.6) 76 (93.8)
.558 .023 .670 .590 .355 o.001 .021 .009 .099 .268 .003 .083 .030 .532 .447
Values are presented as mean ⫾ SD or as n (%). ARC-Adv-CB ¼ Arctic Front Advance cryoballoon; ARC-CB ¼ Arctic Front cryoballoon; LIPV ¼ left inferior pulmonary vein; LSPV ¼ left superior pulmonary vein; PV ¼ pulmonary vein; PVP ¼ pulmonary vein potential; RIPV ¼ right inferior pulmonary vein; RSPV ¼ right superior pulmonary vein.
390 Table 3
Heart Rhythm, Vol 11, No 3, March 2014 Time and temperature at the time of PVI
Variable
ARC-CB (n ¼ 134)
ARC-CB (n ¼ 187)
Real-time PVP visualization Time to isolation (s) All veins LSPV (40/57) LIPV (40/58) RSPV (42/52) RIPV (12/20)
53.2%
60.3%
Temperature at time of isolation (1C) All veins LSPV (40/57) LIPV (40/58) RSPV (42/52) RIPV (12/20)
P .106
52 ⫾ 57 ⫾ 50 ⫾ 46 ⫾ 59 ⫾
34 35 34 32 37
40 ⫾ 44 ⫾ 34 ⫾ 41 ⫾ 48 ⫾
25 20 27 23 33
o.001 .023 .011 .326 .385
36 ⫾ 39 ⫾ 33 ⫾ 35 ⫾ 39 ⫾
10 8 10 12 6
32 ⫾ 35 ⫾ 26 ⫾ 36 ⫾ 33 ⫾
10 7 11 10 8
.001 .01 .002 .675 .038
Values are presented as mean ⫾ SD or as n (%). LIPV ¼ left inferior pulmonary vein; LSPV ¼ left superior pulmonary vein; PVP ¼ pulmonary vein potential; RIPV ¼ right inferior pulmonary vein; RSPV ¼ right superior pulmonary vein.
12 of 13 PNPs were in zone B1 and 1 was in zone A. In the 67 uncomplicated applications, the pacing catheter crossed zone B1 in 7 and another zone in 60 (Figure 3A). Thus, when the pacing catheter crossed zone B1 in the AP view, the likelihood of PN injury during right-sided PVI was high (Se 92.3%; Spe 89.9%; negative and positive predictive values 98.4% and 63.2%, respectively).
Other adverse events A small pericardial effusion was observed in the ARC-AdvCB group, and a single femoral arteriovenous fistula occurred in each group, which both resolved spontaneously within 1 month after the ablation procedure.
Discussion Main findings of our study Compared with that of the ARC-CB, the overall ability of the ARC-Adv-CB to completely isolate the PVs of patients suffering from paroxysmal AF was superior. While all PVs were ultimately isolated in all patients of both groups, fewer
freezing applications were needed to isolate the PVs and the rate of single freezes was higher when the ARC-Adv-CB was used instead of the ARC-CB, significantly shortening the mean duration of the procedure and of exposure to fluoroscopy. In most instances, the PVI was directly visualized and the time to isolation was significantly shortened, increasing the mean temperature at the time of isolation. However, the rate of transient right-sided PNP was significantly higher with the ARC-Adv-CB and strongly correlated with the estimates of procedural CB-PN proximity by using the SVC pacing catheter.
Contributions of the ARC-Adv-CB Compared with the ARC-CB, the new CB has a 2-fold greater number of injection spray jets that have been optimally located to more uniformly cool its tip. It is noteworthy that the higher performance of the ARC-AdvCB was nearly entirely confined to the left PV freezing applications while the differences in right PVs were small. This observation has several putative explanations. First, in our experience with the ARC-CB, the isolation of the
Figure 2 A: Time to disconnection and temperature at the time of PVI. B: Time-temperature relationship for all PV with visible PV potentials during cryoapplications. ARC-Adv-CB ¼ Arctic Front Advance cryoballoon; ARC-CB ¼ Arctic Front cryoballoon; PV ¼ pulmonary vein; PVI ¼ pulmonary vein isolation.
Martins et al
Cryoballoon in Paroxysmal AF Ablation
391
Figure 3 Incidence of PNP with the ARC-Adv-CB. A: Distance between the CB and the PN in patients with vs without PNP. B: Receiver operating characteristic analysis of CB-PN distance, showing an optimal cutoff value of 0.0 cm. ARC-Adv-CB ¼ Arctic Front Advance cryoballoon; CB ¼ cryoballoon; PN ¼ phrenic nerve; PNP ¼ phrenic nerve palsy.
left-sided PVs was more challenging and required a greater number of freezing applications than the right-sided PVs (Table 2). Second, as in the case of radiofrequency ablation, the ridge between the PV ostia and the LA appendage represents an anatomical obstacle for the CB, which is responsible for a high proportion of recurrent connections between the left PVs and the left atrium.17 Inferior segments are also difficult to isolate and are preferential sites of PV reconnection. Chun et al5 used the cross-talk technique in more than one-third of the procedures to isolate the left superior PV, in contrast to a “straightforward” approach for all RSPV, highlighting the difficult access to the distal electrical connection between the left PV with the ARCCB. Third, the higher ovality index of the left-sided PVs than of the right-sided PVs may also be a factor, as an inverse association was found between this index and the degree of occlusion.14
Our results18 are consistent with those of a recent study, which reported a significantly greater PVI efficacy associated with the ARC-Adv-CB than with the first-generation CB. The authors accurately described the technical improvements represented by the ARC-Adv-CB, consisting in an increase in the effective CB-tissue contact area as well as the more prominent ice formation within the PV, known as the “ice cap phenomenon.”15 They hypothesized that even with a suboptimal contact, the ice cap prevents rewarming of the blood flow and enables PVI. In the study by Fürnkranz et al,18 the overall minimum CB temperatures were lower with the ARC-Adv-CB than with the ARC-CB. Some procedural characteristics may explain this difference. Indeed, the distance between the CB surface and the proximal Achieve electrode during freezing decreased between the ARC-CB and the ARC-Adv-CB, facilitating the visualization of PV potentials. This might have promoted the contact with the PV ostia and lower freezing
Figure 4 Phrenic nerve-cryoballoon relationship in the anterior-posterior view during the right superior pulmonary vein cryoapplication. A: zone A; B: zone B1; C: zone B2.
392 temperatures.17 Conversely, in our study, though we did not measure this distance, the rates of PV potential visualization and the low temperatures were similar. The exclusion of wide and long common PV trunks may also have caused these differences since they are usually associated with lower temperatures.18
Procedural safety The greater efficacy of the ARC-Adv-CB was associated with a significant increase in transient PNP in our study. The previously published incidence of PNP has been variable: Casado-Arroyo et al19 reported a significant increase with the ARC-Adv-CB, while Fürnkranz et al18 observed no difference. This discrepancy might be explained by the definition of PNP, as some authors report only persistent PNP while others report persistent and transient PNP. The new design of the ARC-Adv-CB has considerably increased the cooling area near the tip of the CB, increasing the risk of PN injury. In this study, we found an intraprocedural, simple, and reliable predictor of PNP. The vertical projection of the SVC pacing catheter to the distal segment of the CB in the AP view (zone B1, corresponding to the cooling zone) was an excellent predictor of PNP. Of the 13 cases of PNP, a single incident occurred while the line crossed zone A in the AP view, though it may have been caused by an ice cap or by an unusual oblique course of the PN. The anatomical relationship between the right PN and the RSPV is variable and difficult to evaluate before the procedure. In a study using computed tomography scans, segments of the right pericardiophrenic artery, used to locate the PN, were identified in 20% of the patients,20 while earlier studies had reported higher rates.21 A short distance between the RSPV and the SVC was recently found to be correlated with PNP.22 However, the precise course of the PN varies among patients, and the force used to occlude the PV may distort the anatomy. The same authors found that a precipitous drop in CB temperature (ie, o411C at 30 seconds) was a strong predictor of PNP.22 However, while low temperatures are usually associated with more distal and tight PV occlusions, in this study, of 29 transient PNPs, a single case would have been predicted by such a precipitous drop in temperature.
Clinical implications The finding of a reliable predictor of PNP raises the issue of prevention when the risk of PNP is high. When the SVC pacing catheter crossed zone B1 in the AP view and as soon as a weakening of the right hemidiaphragm contraction was suspected, we immediately discontinued freezing, deflated the CB, and abstained from further treatment delivery, probably explaining the absence of persistent PNP. We believe that when the risk of PNP during RSPV treatment is high, the RIPV should be frozen first in order to allow the monitoring of PN during the remaining cryoapplications. Most importantly, Casado-Arroyo et al19 described a highly promising technique to prevent PNP, consisting, after tight wedging of the inflated CB inside the RSPV ostium, to
Heart Rhythm, Vol 11, No 3, March 2014 withdraw it until a small leak of contrast is observed, since the CB volume increases slightly at the onset of the cryoapplication. This offers the advantage of a more proximal cryoapplication. In light of our observations, their lower rate of PNP might be explained by a shift of the PN projection from zone B1 to zone A. Because of the 498% negative predictive value of our PNP predictor, we suggest the use of the CasadoArroyo et al19 technique, particularly when the vertical projection of the PN reaches the distal part of the CB (zone B1). It is noteworthy that in 2 of 80 cases of PN projection in zone B2, the withdrawal of the CB might have increased the risk of PNP since it shifted toward zone B1. Importantly, in all instances of PNP and early cessation of the cryoapplication, the right PVs were successfully isolated without short-term reconnection. Indeed, if PNP occurs, it implies that energy delivery extends to extracardiac structures. Thus, one can imagine that the PV lesion is transmural and successful PVI is achieved. Finally, the PVs were isolated more rapidly when we used the ARC-Adv-CB instead of the ARC-CB. Rapid PVIs are associated with low immediate reconnection rates,23,24 and one may hypothesize that the cryoapplication could be shortened to limit the rate of procedural complications, including PNP. However, the time to PVI has not been found to be a predictor of long-term reconnection, and further studies are needed to assess its contribution in the adjustment of the duration of applications and need for additional cryoapplications.
Study limitations Our study was not randomized, as we did not have access to both CB generations simultaneously, and the patient enrollment overlapped for a period of only 2 months. We do not believe, however, that this introduced important biases, as the same trained operators participated in both phases of the study. A learning curve effect was probably not a factor, since the occlusion grade and visualization of PV potentials were identical in both groups. Moreover, whether a faster PVI is associated with a lower long-term incidence of PV reconnection is unknown and requires further studies. The new predictor of PNP, identified on the basis of a subset of 80 isolations of right PVs, will need to be confirmed with a larger number of observations. Furthermore, only 2 patients had the PN pacing catheter crossing zone B2 and PNP did not occur during these cryoenergy applications; more observations will be needed to confirm the safety of this zone, corresponding to the nonrefrigerated part of the CB. More importantly, these observations were made with a 28-mm CB, limiting the conclusions to this size of the CB. Finally, since short-term PVI may not last indefinitely, whether this higher procedural success rate translates into a higher long-term rate of PVI and AF suppression will need to be confirmed with long-term follow-ups.
Conclusions Cryoablation of paroxysmal AF with the new 28-mm ARCAdv-CB was associated with a 490% success rate with a
Martins et al
Cryoballoon in Paroxysmal AF Ablation
single freezing application and with shorter procedures than with the first-generation CB. However, this higher efficacy was accompanied by a troubling increase in transient PNP. Consequently, we found that the distance between the CB and the PN was a reliable and straightforward intraprocedural predictor of PNP.
Acknowledgment We thank Rodolphe Ruffy for the manuscript revision.
References 1. Haissaguerre M, Jais P, Shah DC, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 1998;339: 659–666. 2. Natale A, Raviele A, Arentz T, et al. Venice Chart international consensus document on atrial fibrillation ablation. J Cardiovasc Electrophysiol 2007;18: 560–580. 3. Cappato R, Calkins H, Chen SA, et al. Updated worldwide survey on the methods, efficacy, and safety of catheter ablation for human atrial fibrillation. Circ Arrhythm Electrophysiol 2010;3:32–38. 4. Weerasooriya R, Jais P, Hocini M, et al. Balloon cryoablation for paroxysmal atrial fibrillation. Europace 2008;10:1251–1252. 5. Chun KR, Schmidt B, Metzner A, et al. The ‘single big cryoballoon’ technique for acute pulmonary vein isolation in patients with paroxysmal atrial fibrillation: a prospective observational single centre study. Eur Heart J 2009;30:699–709. 6. Khairy P, Chauvet P, Lehmann J, et al. Lower incidence of thrombus formation with cryoenergy versus radiofrequency catheter ablation. Circulation 2003;107: 2045–2050. 7. Packer DL, Kowal RC, Wheelan KR, et al. Cryoballoon ablation of pulmonary veins for paroxysmal atrial fibrillation: first results of the North American Arctic Front (STOP AF) pivotal trial. J Am Coll Cardiol 2013;61:1713–1723. 8. Van Belle Y, Janse P, Theuns D, et al. One year follow-up after cryoballoon isolation of the pulmonary veins in patients with paroxysmal atrial fibrillation. Europace 2008;10:1271–1276. 9. Vogt J, Heintze J, Gutleben KJ, et al. Long-term outcomes after cryoballoon pulmonary vein isolation: results from a prospective study in 605 patients. J Am Coll Cardiol 2013;61:1707–1712.
393 10. Van Belle Y, Janse P, Rivero-Ayerza MJ, et al. Pulmonary vein isolation using an occluding cryoballoon for circumferential ablation: feasibility, complications, and short-term outcome. Eur Heart J 2007;28:2231–2237. 11. Neumann T, Vogt J, Schumacher B, et al. Circumferential pulmonary vein isolation with the cryoballoon technique results from a prospective 3-center study. J Am Coll Cardiol 2008;52:273–278. 12. Liu CF. Pulmonary vein reconnection after cryoballoon ablation: back to the drawing board. Heart Rhythm 2010;7:191–192. 13. Kubala M, Hermida JS, Nadji G, et al. Normal pulmonary veins anatomy is associated with better AF-free survival after cryoablation as compared to atypical anatomy with common left pulmonary vein. Pacing Clin Electrophysiol 2011;34: 837–843. 14. Sorgente A, Chierchia GB, de Asmundis C, et al. Pulmonary vein ostium shape and orientation as possible predictors of occlusion in patients with drug-refractory paroxysmal AF undergoing cryoballoon ablation. Europace 2011;13:205–212. 15. Bordignon S, Furnkranz A, Schmidt B, Chun KR. Remaining ice cap on secondgeneration cryoballoon after deflation. Circ Arrhythm Electrophysiol 2012;5: e98–e99. 16. Qureshi A. Diaphragm paralysis. Semin Respir Crit Care Med 2009;30:315–320. 17. Fürnkranz A, Chun KR, Nuyens D, et al. Characterization of conduction recovery after pulmonary vein isolation using the “single big cryoballoon” technique. Heart Rhythm 2010;7:184–190. 18. Fürnkranz A, Bordignon S, Schmidt B, et al. Improved procedural efficacy of pulmonary vein isolation using the novel second-generation cryoballoon. J Cardiovasc Electrophysiol 2013;24:492–497. 19. Casado-Arroyo R, Chierchia GB, Conte G, et al. Phrenic nerve paralysis during cryoballoon ablation for atrial fibrillation: a comparison between the first- and second-generation balloon. Heart Rhythm 2013;10:1318–1324. 20. Nieto-Tolosa J, Rodriguez-Sanchez D, Hurtado-Martinez JA, et al. [Phrenic nerve identification with cardiac multidetector computed tomography]. Rev Esp Cardiol 2010;64:942–944. 21. Horton R, Di Biase L, Reddy V, et al. Locating the right phrenic nerve by imaging the right pericardiophrenic artery with computerized tomographic angiography: implications for balloon-based procedures. Heart Rhythm 2010;7:937–941. 22. Kühne M, Knecht S, Altmann D, et al. Phrenic nerve palsy during ablation of AF using a 28-mm cryoballoon catheter: predictors and prevention. J Interv Card Electrophysiol 2012;36:47–54. 23. Chun KR, Fürnkranz A, Metzner A, et al. Cryoballoon pulmonary vein isolation with real-time recordings from the pulmonary veins. J Cardiovasc Electrophysiol 2009;20:1203–1210. 24. Dorwarth U, Schmidt M, Wankerl M, Krieg J, Straube F, Hoffmann E. Pulmonary vein electrophysiology during cryoballoon ablation as a predictor for procedural success. J Interv Card Electrophysiol 2011;32:205–211.