2021 Volume 85 Issue 3 Pages 275-282
Background: Pulmonary vein (PV) isolation (PVI) with balloon-based visually guided laser ablation (VGLA) is useful for treating atrial fibrillation (AF), but phrenic nerve injury (PNI) is an important complication. We investigated the predictors of developing PNI during VGLA.
Methods and Results: We included 130 consecutive patients who underwent an initial VGLA of non-valvular paroxysmal AF. Twenty patients developed PNI during the PVI. The patients with PNI had a significantly larger right superior PV ostial area (RSPVOA) than the other patients (mean [±SD] 284.7±47.0 vs. 233.1±46.4 mm2, respectively; P<0.01). Receiver operating characteristic analyses revealed that the area under the RSPVOA curve was 0.79 (95% confidence interval [CI] 0.69–0.90) with an optimal cut-off point of 238.0 mm2 (sensitivity, 0.58; specificity, 0.95). In multivariate analyses, a large RSPVOA (HR 1.02, 95% CI 1.01–1.03, P<0.01) and small balloon size (HR 0.70, 95% CI 0.50–0.99, P=0.04) were independent risk factors for PNI during VGLA. PNI remained in 13 patients after the procedure, but 12 of these patients recovered from the PNI during the follow-up period.
Conclusions: The incidence of PNI during VGLA was relatively high, but PNI improved in most cases. A large RSPVOA and small balloon size were predictors of PNI during VGLA.
Pulmonary vein isolation (PVI) is an effective treatment for atrial fibrillation (AF). PVI with balloon-based visually guided laser ablation (VGLA) is regarded as a useful tool for treating AF.1 In previous multicenter studies, VGLA exhibited non-inferiority to radiofrequency (RF) ablation in terms of efficacy and safety in treating paroxysmal and persistent AF.2,3 Moreover, a recent randomized controlled trial showed that recurrence was significantly lower following PVI with VGLA compared with RF ablation and that none of the patients in whom an adenosine provocation test was negative had recurrence of AF.4 However, phrenic nerve injury (PNI) has been an important complication. A prior randomized controlled study of PVI using either VGLA or RF ablation showed that diaphragm paralysis occurred in 3.5% of patients who underwent VGLA, which was higher than that following RF ablation.2 Although the incidence of PNI does not differ significantly between ablation procedures using a cryoballoon (CB) or laser balloon (LB), PNI is a common perioperative complication.5 Balloon dilation causes right superior pulmonary vein (RSPV) strain and reduces the anatomical distance between the balloon and right phrenic nerve, which may lead to PNI.6 In cases of CB ablation, a large RSPV maximum diameter and more distal positioning of the CB relative to the cardiac shadow are independent determinants for predicting PNI.7 Conversely, no predictors of PNI have been reported in the case of VGLA. Nagase et al reported that the compound motor action potentials (CMAPs) quickly declined during VGLA of the RSPV when PNI was provoked and concluded that it was unclear whether CMAP monitoring could prevent PNI during the VGLA.8 In the present study, we investigated the predictors of PNI during VGLA.
The study subjects were consecutive patients undergoing VGLA as an initial procedure for paroxysmal AF (PAF) between July 2018 and January 2020. The subjects were consecutive patients undergoing AF ablation for standard clinical indications.1 Written informed consent for the ablation procedures was obtained from all patients. PAF was defined according to the American Heart Association/American College of Cardiology/European Society of Cardiology guidelines as AF that self-terminated within 7 days.1 The study protocol was approved by the hospital’s institutional human ethics committee and the study complied with the tenets of the Declaration of Helsinki.
Management Before Catheter AblationElectrocardiography and a chest X-ray were performed within 3 months before the catheter ablation. Transthoracic echocardiography was used to evaluate left atrial (LA) diameter and left ventricular ejection fraction (LVEF). All patients underwent multidetector row computed tomography (CT) within 1 week before the catheter ablation to determine the configuration of the LA cavity and rule out any thrombi in the LA appendage. The slice data of the image were reconstructed into a 3-dimensional volume rendering using ZAIO station2 (ZAIOSOFT, Tokyo, Japan). The 3-dimensional image clarified the anatomy of the pulmonary veins (PVs), consisting of the left superior PV (LSPV), the left inferior PV (LIPV), RSPV, and right inferior PV (RIPV). The isolated right middle lobe branch was regarded as the right middle PV (RMPV). The maximum diameter and ostial area of the right-side PVs were measured. Each PV ostial area was measured at the level of the intervenous carina by manually tracing the contour.
All patients received anticoagulation therapy for at least 3 weeks before the ablation procedure. Antiarrhythmic drugs were discontinued for more than 5 half-lives before the ablation procedure, and amiodarone was not administered before the ablation procedure in any patient.
Catheter Ablation ProcedureCatheter ablation was performed under general anesthesia using propofol and dexmedetomidine. A single-use supraglottic airway device (i-gel; Nihon Kohden, Tokyo, Japan) was inserted in all patients, and an esophageal temperature thermocouple catheter (Esophaster; Japan Lifeline, Tokyo, Japan) was inserted through this airway device into the esophagus for continuous monitoring. Vascular access was acquired through the right internal jugular and femoral veins. A 20-pole catheter (BeeAT; Japan Lifeline) was inserted into the coronary sinus (CS) for electrogram recording, atrial pacing, and defibrillation. For the index ablation procedure, 2 transseptal punctures were performed: 1 for the 12-Fr inner diameter deflectable sheath (CardioFocus, Marlborough, MA, USA) and VGLA catheter and the other for a circular mapping catheter. Heparin was given as an intravenous bolus followed by constant infusion to maintain an activated clotting time of >300 s during the ablation procedure. An EnSite NavX navigation system (St. Jude Medical, St. Paul, MN, USA) was used for 3-dimensional mapping.
A mapping catheter (EPstar libero; Japan Lifeline) was used to construct the geometry of the LA and PVs. The bipolar electrogram filters were set between 30 and 500 Hz. The baseline PV electrograms were recorded using fluoroscopy via the circular mapping catheter that was positioned at the PV ostium. A first-generation VGLA catheter (HeartLight; CardioFocus) was inserted into the LA and expanded to occlude each PV. The size of the LB size was increased from 1 to 9 steps and the LB was slowly inflated from Step 1 until the PV was occluded. The compliant balloon was inflated to multiple pressures to change its size so that the balloon/tissue contact was maximized regardless of the size or shape of the target PV ostium. The laser energy level was titrated according to the degree of tissue exposure between 5.5 and 12 W using LightTrackTM software (CardioFocus). Each energy application was 20–30 s. To avoid LB rupture, a dose of 5.5 W for 30 s was applied in areas that were adjacent to blood. Using the adjustable 30° aiming arc for guidance, the laser energy (5.5–16 W, 20–30 s) was delivered around the PV ostium in a contiguous manner with a 50% lesion overlap. Ablation lesions were created as proximal as possible to achieve wide isolated areas and to avoid PV stenosis and PNI. Ablation was terminated prematurely when the esophageal temperature reached 39℃.
VGLA was performed in the order of the LSPV, LIPV, RSPV, and RIPV. The RMPV was simultaneously isolated while an encircling ablation was performed for the RSPV or RIPV with VGLA. After the initial encirclement, PVI status was again assessed, and additional laser energy was delivered at electrical gap sites in cases of an unsuccessful PVI. If the PVI could not be successfully accomplished with only LB application, an irrigated RF catheter (TactiCath; Abbott, Abbott Part, IL, USA) was used for a touch-up ablation to eliminate all gap conduction sites. When PV electrograms were present, isolated PV potentials identified exit block. A successful PVI was defined as bidirectional block between the LA and inside the circumferential PVI area, which was confirmed by the electrical dissociation between the LA and PV during pacing with the electrode catheter positioned inside the PVI area and electrode catheter positioned inside the CS. Cavotricuspid isthmus ablation was also performed when atrial flutter was observed during the ablation procedure or at the discretion of the operator. Non-PV focus ablations were not performed. The RF energy was set at a maximum output of 35 W, and the catheter tip temperature was not allowed to exceed 42℃.
During ablation of the RSPV or RIPV, continuous and stable right phrenic nerve pacing was performed and CMAPs were continuously recorded during the ablation procedure. A 10-pole catheter was placed in the right subclavian vein or superior vena cava to pace the right phrenic nerve. The threshold for capturing the diaphragm was measured, and pacing was performed at a rate of 40 pulses/min at an output exceeding the pacing threshold by 10%, as described previously.9 The PNI group was defined as the group in which the CMAPs decreased by more than 30% during LB application. LB application was immediately terminated when PNP developed. No further LB applications were performed around the sites where PNI developed, and additional RF ablation was performed if the RSPV could not be isolated. After successful PVI was achieved, it was again reassessed after a 15-min postablation waiting period.
During ablation of the RSPV, the position of the LB was confirmed under fluoroscopy (Infinix; CANNON, Tokyo, Japan). The fluoroscopic images were stored and analyzed using Goodnet software (GOODMAN, Nagoya, Japan). The distance between the LB apex and cardiac shadow (Figure 1) was measured in the anteroposterior position.
Positioning of the laser balloon (LB) in relation to the cardiac shadow under fluoroscopic guidance in the anteroposterior projection during right superior pulmonary vein ablation. The distance between the cardiac shadow and LB apex was measured (double-headed arrow).
All patients were routinely seen at 1 and 3 months after discharge and every 3 months thereafter in the outpatient clinic. At the follow-up visits, patients underwent a physical examination, electrocardiogram (ECG) recording, and review of their symptoms. All patients were instructed to tell the attending physician if there were any symptoms suggestive of arrhythmia recurrence. In such cases, rhythm was assessed using a 12-lead ECG. A Holter ECG examination was performed at 3 and 12 months after the VGLA procedure. Recurrence of atrial tachyarrhythmia was defined as any sustained AF or atrial tachyarrhythmia lasting >30 s that appeared >3 months after the blanking period of the catheter ablation. Clinical success, evaluated 12 months after the initial PVI procedure, was defined as freedom from any symptomatic PAF as assessed by clinical assessment, 12-lead ECG, and Holter ECG recordings.
A chest X-ray examination was also performed at 1 and 3 months after discharge. Persistent PNI was defined as an elevated right-sided diaphragm noted on post-procedural chest X-ray obtained the day after the VGLA that persisted even after the VGLA procedure. Patients with persistent PNI had a chest X-ray examination taken at each outpatient clinic visit and were evaluated until resolution of the PNI.
Statistical AnalysisContinuous variables are expressed as the mean±SD, and the significance of differences was analyzed using Student’s t-test. Categorical data are expressed as numbers and percentages, and were compared using a χ2 test or Fisher’s exact test. Univariate and multivariate logistic regression analyses were performed on candidate variables to predict the dichotomous outcome of PNI. All variables with P<0.1 in the univariate analysis were entered into the multivariate analysis. The appropriate cut-off value was determined as the sum of the highest sensitivity and specificity using a receiver operating characteristic (ROC) curve. A 95% confidence interval (CI) was shown as the area under the curve (AUC). Two-sided P<0.05 was deemed significant. Statistical analyses were conducted using EZR software (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (R Foundation for Statistical Computing, Vienna, Austria).
The baseline clinical characteristics of the study patients are summarized in Table 1. In all, 130 patients (67% male, mean age 66±12 years) who underwent VGLA as an initial PVI were analyzed in this study. The mean CHA2DS2-VASc score was 2.0±1.5 and the mean LA diameter was 38.4±6.1 mm. The RIPV could not be isolated in 4 patients because the VGLA was prematurely terminated due to the sudden failure to capture the phrenic nerve, which lasted until the end of the ablation procedure. Except for these 4 patients, all 4 PVs were successfully isolated by the VGLA combined with touch-up RF ablation. An LB rupture was provoked in 3 patients, and 26 patients needed touch-up RF ablation. There were no significant differences in the age, sex, body mass index, or comorbidities such as hypertension, diabetes, and a history of congestive heart failure or stroke between those with and without PNI during the VGLA procedure. The LA diameter and LVEF were also comparable between the 2 groups. The LB size during RSPV isolation was significantly smaller in the group that developed PNI during the ablation procedure (P=0.01), but there was no significant differences in the LB size for the other PVs. Moreover, there was no significant difference in the total procedure time, energy needed to isolate each PV, or the distance between the cardiac shadow and LB apex between the 2 groups (Table 1).
| Overall (n=130) |
PNI | P value | ||
|---|---|---|---|---|
| No (n=110) | Yes (n=20) | |||
| Age (years) | 66.0±11.8 | 65.8±11.8 | 67.2±12.3 | 0.61 |
| Female sex | 43 (33.1) | 33 (30.0) | 10 (50.0) | 0.14 |
| BMI (kg/m2) | 23.8±3.7 | 24.1±3.7 | 22.6±3.6 | 0.10 |
| Hypertension | 54 (41.5) | 44 (40.0) | 10 (50.0) | 0.56 |
| Diabetes | 10 (7.7) | 8 (7.3) | 2 (10.0) | 1.0 |
| Heart failure | 8 (6.2) | 5 (4.5) | 3 (15.0) | 0.11 |
| Stroke or TIA | 11 (8.5) | 11 (10.0) | 0 (0) | 0.30 |
| CHA2DS2-VASc score | 2.0±1.5 | 2.0±1.5 | 2.3±1.3 | 0.47 |
| LA diameter (mm) | 38.4±6.1 | 38.8±6.1 | 36.1±5.6 | 0.07 |
| LVEF (%) | 65.9±9.0 | 65.9±8.4 | 65.7±12.0 | 0.91 |
| CT parameters | ||||
| RSPV diameter (mm) | 19.7±2.8 | 19.4±2.6 | 21.9±2.9 | <0.01 |
| RSPVOA (mm2) | 241.1±49.9 | 233.1±46.4 | 284.7±47.0 | <0.01 |
| RIPV diameter (mm) | 18.9±2.9 | 18.9±2.9 | 18.5±3.0 | 0.54 |
| RIPVOA (mm2) | 230.6±63.7 | 231.8±64.2 | 224.4±62.4 | 0.63 |
| Total procedure time (min) | 172.7±31.7 | 173.9±32.4 | 165.7±27.0 | 0.29 |
| Ablation time (min) | 84.8±32.5 | 86.2±34.3 | 76.4±18.6 | 0.21 |
| Fluoroscopy time (min) | 55.5±15.3 | 55.8±15.5 | 54.0±14.4 | 0.64 |
| Laser mean output (W) | ||||
| RSPV | 8.7±1.4 | 8.7±1.5 | 8.7±1.3 | 0.96 |
| RIPV | 7.9±1.4 | 7.9±1.4 | 7.7±1.6 | 0.62 |
| LSPV | 8.6±1.5 | 8.6±1.5 | 8.6±1.5 | 0.90 |
| LIPV | 7.8±1.4 | 7.7±1.4 | 7.9±1.6 | 0.58 |
| Total energy (J) | ||||
| RSPV | 6,014.9±2,667.3 | 6,164.2±2,785.2 | 5,248.1±1,752.6 | 0.16 |
| RIPV | 4,489.5±1,506.7 | 4,481.1±1,436.9 | 4,528.0±2,000.5 | 0.92 |
| LSPV | 6,768.2±3,128.5 | 6,706.1±3,000.5 | 7,106.6±3,823.8 | 0.60 |
| LIPV | 4,628.3±2,359.4 | 4,501.0±2,213.9 | 5,304.9±2,999.7 | 0.17 |
| LB size | ||||
| RSPV | 7±2 | 8±2 | 7±2 | 0.01 |
| RIPV | 4±1 | 4±1 | 5±1 | 0.23 |
| LSPV | 8±2 | 8±2 | 8±2 | 0.95 |
| LIPV | 5±1 | 5±1 | 4±1 | 0.76 |
| Dormant conduction | 19 (14.6) | 15 (13.7) | 4 (20) | 0.60 |
| Radiofrequency touch up | 26 (20.0) | 21 (19.1) | 5 (25.0) | 0.76 |
| Distance between cardiac shadow and LB apex (mm) |
22.2±7.0 | 22.2±7.1 | 22.5±6.5 | 0.85 |
| CTI ablation | 119 (91.5) | 101 (91.8) | 18 (90.0) | 1.0 |
Values are expressed as the mean±SD or as n (%). P values are for comparisons between the groups with and without phrenic nerve injury (PNI). BMI, body mass index; CTI, cavotricuspid isthmus; LA, left atrium; LB, laser balloon; LIPV, left inferior pulmonary vein; LSPV, left superior pulmonary vein; LVEF, left ventricular ejection fraction; RIPV, right inferior pulmonary vein; RIPVOA, right inferior pulmonary vein ostium area; RSPV, right superior pulmonary vein; RSPVOA, right superior pulmonary vein ostium area; TIA, transient ischemic attack.
PNI developed in 20 of 130 (15.4%) patients during VGLA. The PNI occurred during RSPV and RIPV isolation in 17 and 3 patients, respectively. Among these patients, the procedure was completed successfully in 14 despite sustained PNI, which manifested as a decrease in the amplitude of the CMAPs of >30% at the end of the PVI. Table 2 summarizes the details of the patients who developed PNI during the procedure. The mean time from the initiation of laser application to the development of PNI was 3.6±1.0 s (range 1.5–4.4 s) and the mean energy was 31.8±10.6 J (range 12.8–52.8 J). Figure 2 shows an example of the development of PNI during VGLA. Both stable phrenic nerve stimulation and CMAP amplitude recordings were confirmed before laser energy application to the RSPV. However, the loss of capture of the phrenic nerve during electrical stimulation occurred suddenly, within only 1.5 s. The sites where PNI was provoked were the carina of the RSPV (n=8), anterior RSPV (n=7), anterior RIPV (n=3), and the roof of the RSPV (n=2; Figure 3).
| Patient no. | PV with PNI | Laser output (W) |
Time to PNI (s) |
Energy to PNI (J) |
Amplitude of CMAP (mV) | Persistent PNI |
||
|---|---|---|---|---|---|---|---|---|
| RSPV | RIPV | Maximum | Minimum | |||||
| 1 | + | − | 8.5 | 4.4 | 37.4 | 0.74 | 0.34 | − |
| 2 | + | − | 8.5 | 4.4 | 37.4 | 0.82 | 0.28 | + |
| 3 | + | − | 8.5 | 2.9 | 24.7 | 0.81 | 0.54 | + |
| 4 | + | − | 8.5 | 4.4 | 37.4 | 0.42 | 0.15 | − |
| 5 | + | − | 8.5 | 2.9 | 24.7 | 0.47 | 0.14 | + |
| 6 | + | − | 8.5 | 2.9 | 24.7 | 0.94 | 0.39 | − |
| 7 | + | − | 8.5 | 4.4 | 24.2 | 0.46 | 0.20 | − |
| 8 | + | − | 8.5 | 4.4 | 37.4 | 0.66 | 0.30 | + |
| 9 | + | − | 8.5 | 1.5 | 12.8 | 0.64 | 0 | + |
| 10 | − | + | 8.5 | 2.9 | 24.7 | 1.0 | 0 | + |
| 11 | + | − | 10 | 4.4 | 44 | 0.39 | 0 | + |
| 12 | + | − | 10 | 2.9 | 29 | 0.34 | 0 | − |
| 13 | + | − | 12 | 1.5 | 18 | 0.70 | 0 | + |
| 14 | + | − | 10 | 4.4 | 44 | 0.46 | 0.24 | + |
| 15 | − | + | 10 | 4.4 | 44 | 0.66 | 0.12 | − |
| 16 | + | − | 8.5 | 4.4 | 37.4 | 0.54 | 0.21 | + |
| 17 | + | − | 12 | 2.9 | 34.8 | 0.46 | 0 | + |
| 18 | − | + | 5.5 | 2.9 | 16 | 0.72 | 0 | + |
| 19 | + | − | 7 | 4.4 | 30.8 | 0.33 | 0.07 | − |
| 20 | + | − | 12 | 4.4 | 52.8 | 0.51 | 0 | + |
| All patients | 17 (85.0) | 3 (15.0) | 9.0±1.8 | 3.6±1.0 | 31.8±10.6 | 13 (65.0) | ||
Data for all patients are given as the mean±SD or as n (%). +, present; −, absent; CMAP, compound motor action potential. Other abbreviations as in Table 1.
Example of phrenic nerve injury (PNI) provoked during visually guided laser ablation. Phrenic nerve pacing was performed at a rate of 40 pulses/min at an output of 7V. The amplitude of the compound motor action potential was stably recorded as 0.65–0.70 mV. However, pacing failure of the phrenic nerve occurred within only 1.5 s after the initiation of laser application.
Distribution of phrenic nerve injury (PNI) sites. Eight sites were located at the carina of the right superior pulmonary vein (RSPV), 7 were located at the anterior RSPV, 3 were located at the anterior right inferior pulmonary vein (RIPV), and 2 were located on the roof of the RSPV.
The optimal cut-off values for RSPV diameter and RSPV ostial area (RSPVOA) for predicting PNI were investigated. The ROC AUC was used to assess the predictive power of RSPV diameter and RSPVOA for PNI. The optimal cut-off values for RSPV diameter and RSPVOA were 19.0 mm and 238.0 mm2, respectively. The AUC of the RSPV diameter and RSPVOA was 0.72 (95% CI 0.61–0.83; sensitivity, 95.0%; specificity, 37.3%) and 0.79 (95% CI 0.69–0.90; sensitivity, 95.0%; specificity, 58.2%), respectively.
Predictors of PNI During VGLAMultivariate analysis showed that a large RSPVOA was an independent predictor of PNI (hazard ratio [HR] 1.02, 95% CI 1.01–1.03, P<0.01) and that a large balloon size during RSPV ablation was an independent risk reduction factor for PNI (HR 0.70, 95% CI 0.50–0.99, P=0.04; Table 3). Moreover, a larger LA diameter was a significant predictor for avoiding PNI (HR 0.90, 95% CI 0.81–1.00, P=0.04). The distance between the cardiac shadow and LB apex was not significantly associated with the occurrence of PNI.
| Univariate analysis | Multivariate analysis | |||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Age | 1.01 (0.97–1.05) | 0.61 | ||
| Female sex | 2.33 (0.88–6.14) | 0.09 | 1.87 (0.59–5.91) | 0.29 |
| BMI | 0.88 (0.75–1.03) | 0.10 | ||
| CHA2DS2-VASc score | 1.13 (0.82–1.56) | 0.47 | ||
| LA diameter | 0.92 (0.85–1.01) | 0.07 | 0.90 (0.81–1.00) | 0.04 |
| RSPVOA | 1.02 (1.01–1.03) | <0.01 | 1.02 (1.01–1.03) | <0.01 |
| LB size in: | ||||
| RSPV | 0.71 (0.54–0.93) | 0.01 | 0.70 (0.50–0.99) | 0.04 |
| RIPV | 1.30 (0.84–2.01) | 0.24 | ||
| LSPV | 0.99 (0.74–1.32) | 0.95 | ||
| LIPV | 0.95 (0.66–1.36) | 0.76 | ||
| Distance between cardiac shadow and LB apex |
1.01 (0.94–1.08) | 0.89 | ||
CI, confidence interval; HR, hazard ratio. Other abbreviations as in Table 1.
After a mean follow-up of 331±121 days (range 90–572 days), atrial tachyarrhythmias recurred in 20 patients (15.4%). Persistent PNI was provoked in 13 patients during VGLA and persisted after discharge. The magnitude of the diaphragm elevation was <1 vertebral body in 6 patients, >2 vertebral bodies in 4 patients and >3 vertebral bodies in 3 patients. Three patients complained of dyspnea on exertion, which may have been a symptom associated with persistent PNI. However, complete recovery from PNI was observed in 12 patients during the follow-up period. The mean time to recovery from persistent PNI was 260±101 days (range 127–484 days).
Fifteen patients underwent a redo catheter ablation, 3 of whom had developed PNI during the first PVI session. All 3 of these patients developed PNI during the RSPV isolation with VGLA, and there was no further VGLA application or RF touch-up because the RSPVs were successfully isolated. However, RSPV reconnections were found in all 3 of these patients during the redo ablation session.
In the present retrospective single-center study, 130 patients who underwent VGLA using a first-generation LB were analyzed. The main findings of the present study are that: (1) the time from the initiation of the laser application to the occurrence of PNI was very short; (2) a large RSPV and a small LB size were predictors of PNI; and (3) the PNI during LB ablation was reversible in most patients.
Characteristics of PNI During VGLAContinuous monitoring of CMAPs during PVI is a useful technique for early detection of PNI.10,11 Moreover, a reduction in the amplitude of CMAPs of >30% during VGLA has been reported as a risk factor for the development of phrenic nerve palsy.12,13 In the present study, the reduction in CMAP amplitude occurred very quickly during the VGLA, with a mean time to the development of PNI of only 3.6±1.0s. The time to PNI occurrence was very short during the VGLA, therefore continuous CMAP monitoring cannot be regarded as a useful tool for preventing PNI during VGLA. In the present study, 17 patients developed PNI during RSPV ablation and 3 developed PNI during RIPV ablation. This result is compatible with that of PVI using a CB, and is considered to be associated with the anatomical course of the phrenic nerve.14
Previous studies have demonstrated that laser energy can create deeper and more discrete lesions in the LA antrum-PV region than RF ablation.15 Moreover, laser energy is absorbed by the water in cardiac tissue, and the maximum heat initially reaches beyond the endocardial level,16 which may quickly cause PNI.
Tohoku et al reported that most PNIs during CB ablation are transient, whereas most PNIs during VGLA are persistent.17 These authors discussed that the difference in the temperature between the blood and heated tissue of the phrenic nerve is smaller with VGLA than with a CB, which makes it difficult to cool heated tissue in the phrenic nerve, and so persistent PNI occurs frequently in VGLA.17
Predictors of PNI During VGLAIt has been reported that, in an in vitro model, lesion depth, lesion volume, and maximum lesion diameter increase not only in accordance with laser output, but also the total laser energy delivered.18,19 However, in the present study, there was no significant difference in the mean output and total energy delivered between the groups with and without PNI. Moreover, the distance between the LB and cardiac shadow on the fluoroscopic image was not significantly related to the incidence of PNI.
Because CMAP monitoring was not useful for preventing PNI because of the extremely rapid reduction in CMAP amplitude during VGLA, it is important to identify other predictors of PNI. The present study showed that a larger ostium of the RSPV and a smaller LB size were predictors of PNI. Therefore, the maneuver to make the LB size the largest relative to the RSPVOA turned out to be the most effective strategy to prevent PNI, which encouraged us to ablate the RSPV with a large-sized LB as much as possible. Recently, contrast CT has been performed before ablation in order to construct the geometry of the heart chambers merged with a 3-dimensional mapping system or to confirm a thrombus in the LA. We must pay attention to PNI in patients with a large RSPV when a relatively large RSPV is recognized on CT images. Sano et al reported that pacing at each PV ostium before the CB ablation could be effective in predicting PNI.20 Therefore, when VGLA is to be performed in patients with a relatively large RSPVOA, pacing inside the PVs may be recommended to avoid PNI.
A multivariate study demonstrated that a larger LA diameter was also a significant predictor for avoiding PNI.7 Of interest, there was no significant difference in the LA diameter between the groups with and without PNI in the cure rate of PAF in cases in which the CB was used (41.1±6.3 vs. 40.3±7.6, respectively; P=0.70).7
Comparisons With Previous ReportsIn the present study, persistent PNI, which persisted even after discharge, occurred in 13 patients (10.0%). Previous studies on VGLA reported that the incidence of PNI ranged from 0% to 8.6%.2,3,21–25 A meta-analysis reported that PNI was the most common procedural complication following VGLA, observed in 32 of 1,162 (2.6%) procedures across 16 studies.26 The incidence of PNI was higher in the present study than in prior studies. There may be several reasons for this observation. First, the LB size in the present study may have been relatively smaller than that in previous studies. The present study demonstrated that LB size was an independent risk reduction factor for provoking PNI. Accordingly, a larger LB was associated with a lower incidence of PNI (HR 0.71, 95% CI 0.54–0.93, P=0.03). To the best to our knowledge, no previous study has reported the relationship between LB size and the incidence of PNI. Second, there were significant differences with regard to the definition of persistent PNI. In the present study, persistent PNI was assigned to even subtle differences in the diaphragm in the post- compared with preprocedural chest X-ray. None of the patients with persistent PNI underwent spirometry or a sniffing test, and so we failed to systematically assess the study patients in terms of investigating the genuine incidence of persistent PNI. However, the PNI provoked by VGLA in the present study was reversible in most patients during the follow-up period. Thus, the final incidence of persistent PNI was almost comparable to that reported in recently published studies on VGLA.2,3,21–25
Study LimitationsThis study has some limitations. First, the study was a single-center retrospective analysis and the study sample size was relatively small. A larger study is needed to verify our results. Second, the pacing cycle length was longer than in a previous study.12 In that study, the authors reported that a 30% reduction in the diaphragmatic CMAP amplitude was a risk factor for provoking PNI for the first time.12 They performed pacing of the right-sided PN at a frequency of 60 pulses/min. We may have been able to reduce the incidence of PNI, which occurred within a few seconds, in most cases if we had paced the phrenic nerve with a higher pacing rate than the one used in the present study. Third, the actual laser application sites may differ between patients with and without PNI even though the balloon appeared to be in the same position; in addition, the laser application sites depend on the operator, regardless of the balloon position when using the LB system. However, in the patients with and without PNI 20.0% and 19.1%, respectively, of RSPVs (P=1.0) and 25.0% and 37.3%, respectively, of RIPVs (P=0.41) required laser applications more distal than the white center line marker. We tried to deliver the ablation lesions as proximal as possible, but there was no significant relationship between the LB application site and the occurrence of PNI.
The incidence of PNI during the VGLA procedure was relatively high in the present study, but the PNI improved in most cases. A larger ostium of the RSPV and smaller LB size were reliable predictors of PNI. A strategy to avoid PNI may be to make an effort to expand the LB as much as possible relative to the size of the RSPV ostium.
There was no financial support associated with this study.
The authors declare no conflict of interest for this article.
This study was approved by the Japan Red Cross Yokohama City Bay Hospital (Reference no. 2020-9).
Deidentified participant data will not be shared.
