Cardioneuroablation focused on the atrioventricular node: a comparison of right and left atrial approach

Background

Radiofrequency catheter ablation of posteromedial left ganglionated plexus is a critical step to eliminate the vagal input to the atrioventricular node (AVN) for the treatment of symptomatic episodes of functional AV block. This ganglionated plexus can be effectively targeted from the coronary sinus (CS) or from the endocardial aspect of the right (RA) and left (LA) atria.

Purpose

We investigated the effect of ablation at individual sites on the suppression of parasympathetic modulation of AVN.

Methods

The study included 20 patients (age: 42±13 years, 45% males) who underwent cardioneuroablation in general anesthesia. Posteromedial left ganglionated plexus was ablated from [1] the CS (proximal 2-cm segment), [2] the RA aspect (between the fossa ovalis and inferior vena cava), and [3] the LA aspect (middle bottom part adjacent to inferior rim of fossa ovalis). Patients were randomly (1:1) assigned to CS-to-RA or RA-to-CS ablation order. LA ablation was always the last step. The response to extracardiac vagus nerve stimulation (ECVS; 50 Hz, 0.05 ms, 1 V/kg [<70V], 5 s) while atrial pacing (100 bpm) was recorded at baseline and after each ablation step. The number of non-AV-conducted beats during the ECVS was considered a measure of AV nodal denervation. Both right and left vagus nerves were sequentially stimulated and the stronger response of the AV node was taken into account.

Results

Temporal development of outcome measure with the progression of ablation is shown in Figure 1. CS ablation resulted in much stronger AV nodal denervation compared to RA ablation (P=0.02). However, RA ablation still provided some effect on top of CS ablation. The combination of CS + RA ablation resulted in complete AVN denervation in 8 (40%) patients. Subsequent LA ablation increased the number of denervated patients to 14 (70%). Two more patients were subsequently denervated by ablation elsewhere. In four patients, AVN denervation was not achieved but their responsiveness to ECVS was significantly suppressed compared to the baseline.

Conclusions

All ablation clusters targeting posteromedial ganglionated plexus convey complementary effects. Biatrial cardioneuroablation seems essential for efficacious suppression of parasympathetic modulation of AVN.

Funding Acknowledgement

Type of funding sources: None.

Importance of bilateral vagus nerve stimulation for effective atrioventricular node denervation during cardioneuroablation

Background

The progression of parasympathetic denervation of the atrioventricular node (AVN) during cardioneuroablation (CNA) can be evaluated by extracardiac vagal stimulation (ECVS). The right vagus nerve is usually used for stimulation (R-ECVS) because the right jugular vein is easily accessible. However, the AVN node is predominantly under the control of the left vagus nerve.

Purpose

To highlight the importance of left vagus stimulation (L-ECVS) for effective AVN denervation.

Methods

Both R-ECVS and L-ECVS (frequency: 50 Hz; pulse width: 0.05 ms; output 1 V / 1 kg; max. 70 V, duration 5 s) was attempted in 80 patients (age: 41±12 years, 45% men) undergoing CNA with stepwise strategy consisting of ablation of right anterior ganglionated plexus (RAGP) followed by ablation of posteromedial left ganglionated plexus (PMLGP). The study objective was the AVN response to L-ECVS (evaluated as the max. R-R interval during stimulation train) at the point when AVN non-reactivity to R-ECVS was achieved.

Results

A total of 59 patients were suitable for the analysis. Of the remaining 21 patients, left (n=14) or right (n=2) jugular veins were not accessible, AVN non-reactivity to L-ECVS was achieved before non-reactivity to R-ECVS (n=4), or AVN denervation was not achieved at all (n=1). At baseline, the AVN response was identical for R-ECVS (max. R-R median: 6.9 s, interquartile range [IQR]: 5.7–8.2 s) and L-ECVS (median: 7.1 s, IQR: 6.0–8.3 s), P=0.44. AVN non-reactivity to R-ECVS was present already at baseline (n=2); was achieved after ablation of RAGP (n=14), after ablation PMLGP (n=38), or after extensive ablation (n=5). At the point of AVN non-reactivity to R-ECVS, the response of AVN to L-ECVS was as follows: none (n=25), 2: 1 AV block (n=13) or complete AV block (n=21). The corresponding median of max. R-R interval was: 1.2 s, IQR: 0.6–4.8 s distributed as shown in Figure 1.

Conclusions

In 34/59 (58%) patients, significant AVN response to L-ECVS persists after reaching AVN non-reactivity to R-ECVS. Stimulation of both vagal nerves tightens the procedural endpoint and may increase the clinical efficacy of CNA, especially in patients with dominant AVN disorder.

Funding Acknowledgement

Type of funding sources: None.

Cardioneuroablation in patients with a prior pacemaker implant

Funding Acknowledgements

Type of funding sources: None.

Background

Cardioneuroablation (CNA) is an alternative therapeutic method for patients with functional bradyarrhythmias, who are otherwise candidates for permanent pacing. In a specific clinical scenario, CNA can also be a treatment option for patients with already implanted pacemakers.

Purpose

We investigated whether CNA could substitute permanent pacing in selected patients in whom pacemaker is associated with complications, technical failures, or decreased quality of life.

Methods

Among 160 patients who underwent CNA by radiofrequency energy between 2014 and 2022, there were 13 patients (8%) with a pacemaker in whom CNA was indicated as a "substitute" treatment. The reasons were as follows: lead failure (n = 6), recurrent decubitus of the device pocket (n = 1), infective endocarditis (n = 1), recurrent syncope even after pacemaker implantation (n = 1) and discomfort associated with the implanted device (n = 4). In 4 patients, the pacemaker was explanted shortly before CNA because of serious complications. Biatrial CNA was guided anatomically with the use of a 3D mapping system and intracardiac echocardiography. Empirical sites of ganglionated plexi were targeted. The procedural endpoint was unresponsiveness of sinus and AV nodes to extracardiac vagal nerve stimulation.

Results

Patients (n = 13, 77% male, age: 41 ± 12 years) had a pacemaker implanted 11 ± 7 years ago. They responded to atropine (2 mg IV) administration by accelerating the sinus rhythm by 55 ± 33%. CNA was technically successful and uncomplicated in all of them. During the follow-up of 29 ± 16 months (range 6 - 56 months), 2 patients (15%) had a recurrence of syncope. In both, the syncope was reproduced by tilt testing and classified as a pure vasodepressor event. In the remaining patients, the follow-up was uneventful. Elective explantation of the pacing system has so far been performed in 3 of 9 patients, while this is planned in others after reaching a sufficient length of post-procedural follow-up.

Conclusions

CNA is an effective and safe method, which is not only an alternative to the pacemaker implant but can also be used in patients with functional bradyarrhythmias, and previously implanted pacemakers who experience significant adverse events associated with the pacing system.

Heart rate acceleration during cardioneuroablation is a weak predictor of significantly reduced parasympathetic modulation of sinus node

Funding Acknowledgements

Type of funding sources: None.

Introduction

Ablation of superior parasympathetic ganglia is associated with acceleration of sinus rhythm (SR). This has been considered a favorable sign during cardioneuroablation (CNA) for the treatment of functional bradyarrhythmias.

Aim

We studied whether the relative increase in SR frequency (DeltaSR) during CNA is a reliable predictor of significantly reduced parasympathetic modulation of the sinus node (SAN).

Methods

In patients undergoing CNA under general anesthesia, the gradual reduction in parasympathetic modulation of SAN during the procedure was assessed by extracardiac right vagal nerve stimulation (ECVS). The response to ECVS was quantified by the ratio of the maximum P-P interval (induced by ECVS) and the baseline SR cycle length (MaxPPratio). The ECVS was performed repeatedly after partial ablation steps, and therefore several pairs of DeltaSR and MaxPPratio values were obtained in the course of the single procedure. A MaxPPratio <1.5 was arbitrarily chosen as the criterion of significant attenuation of vagally induced responses. The optimum dichotomy of Delta-SR for the prediction of the MaxPPratio <1.5 was found according to the criterion of the minimum distance of the ROC curve from the point (0; 1).

Results

The study included 64 patients (mean age: 42 ± 16 years, 48% men). A total of 188 intraprocedural pairs of DeltaSR and MaxPPratio (2.9 ± 2.1 pairs per procedure) covering the wide distribution of their values (19 ± 14 bpm for DeltaSR and 2.9 ± 2.8 for MaxPPratio) were obtained. One half of ECVS tests (51%) met the criterion of MaxPPratio <1.5. In the analysis of receiver operating characteristic, DeltaSR as a predictor of significantly reduced parasympathetic modulation of SAN showed an area under the curve (AUC) of 0.69 with 95% confidence interval (CI) of 0.62 - 0.77 (Figure). The optimum cut-off (DeltaSR ≥20 bpm) had a sensitivity (SENS), specificity (SPEC), positive (PPV), and negative predictive value (NPV) of 61%, 78%, 75%, and 66%, respectively.

Conclusion

Acceleration of SR during CNA is not a relevant surrogate of significantly reduced parasympathetic modulation of SAN. Elimination or significant suppression of responses of SAN to ECVS serves as an excellent procedural endpoint that might contribute to the favorable clinical outcome of the CNA.

Acute change of cardiac autonomic regulations after thermal and non-thermal pulmonary vein ablation

Funding Acknowledgements

Type of funding sources: None.

Background

Pulmonary vein isolation (PVI) by thermal energy (radiofrequency energy or cryoenergy) results in collateral ganglionic plexi ablation. On the contrary, pulsed electric field (PEF) energy presumably spares neural tissue.

Purpose

We investigated and compared the effect of PVI on parasympathetic input into the sinus node (SAN) and AV node (AVN) when four different ablation strategies were used.

Methods

A study enrolled 49 patients who underwent PVI in general anesthesia (age: 57 ± 13 years, 71% males). In 17 patients, point-by-point radiofrequency energy delivery by the irritated-tip catheter was used for ablation while 7 patients were ablated using a second-generation cryoballoon catheter. In 7 patients, PEF energy was delivered using a single-shot Farawave catheter while 18 patients were ablated using Sphere9 lattice-tip catheter (Affera, Inc.); both subgroups with manufacturer-specific PEF settings. Before and after PVI, the responsiveness of the SAN and AVN was assessed by extracardiac vagal nerve stimulation (ECVS) via a diagnostic catheter in the right internal jugular vein. Five-second stimulation trains were delivered with a frequency of 50 Hz, pulse width of 0.05 ms, and output of 1 V/kg (<70V) both in sinus rhythm and during atrial pacing. Substantial reduction of response to ECVS was arbitrarily defined as a maximum induced pause of <1.5 seconds.

Results

At baseline, physiological response to ECVS (long sinus arrest and/or AV block) was demonstrated. After PVI, a substantial reduction of SAN response was observed in 21/24 (88%) patients after thermal PVI and 7/25 (25%) patients after non-thermal PVI (P = 0.0001). Similarly, a substantial reduction of AVN response was observed in 21/24 (88%) patients after thermal PVI and 9/25 (36%) patients after non-thermal PVI (P = 0.0003). The Figure shows on the continuous scale the post-PVI pauses in sinus rhythm (maximum P-P interval) and atrial pacing (maximum R-R interval) induced by ECVS.

Conclusion

Vagal responses of SAN and AVN are preserved in most AF patients after non-thermal PVI. This contrasts with the much stronger effect of thermal PVI. Whether this may influence the clinical outcome of AF ablation procedures remains to be investigated in future studies.

Anatomically-guided cardioneuroablation for recurrent neurally mediated syncope

Funding Acknowledgements

Type of funding sources: None.

Background

Cardioneuroablation (CNA) has been proposed as a new therapeutic approach in selected patients with recurrent neurally mediated syncope and documented cardioinhibitory component.

Purpose

We report on procedural data and clinical outcomes of consecutive patients who underwent anatomically-guided CNA at our center in the period of 2014 - 2021.

Methods

A study investigated 145 otherwise healthy patients (age: 40 ± 14 years, 56% males) with recurrent reflex syncope and the physiological result of the atropine test. Biatrial radiofrequency (RF) ablation was performed under general anesthesia by irrigated-tip catheter at empirical sites of ganglionated plexi (GP) with the navigation by CARTO-3 system and intracardiac echocardiography. Anterior right GP and posteromedial left GP were always targeted to modulate the innervation of both sinoatrial (SAN) and atrioventricular (AVN) nodes, irrespective of clinical manifestation of the disease. The loss of responsiveness of both nodes to extracardiac vagus nerve stimulation was the procedural endpoint. Right vagus or bilateral vagus nerve stimulation was used in 86% and 54% of procedures, respectively.

Results

Enrolled patients had documented cardioinhibitory disorder of SAN (59%), AVN (30%), or both nodes (11%). CNA (duration: 157 ± 31 min; RF time: 15 ± 6 min; radiation dose: 84 ± 135 µGy.m²) resulted in sinus rate acceleration by 28 ± 12 bpm, shortening of AH interval by 15 ± 31 ms, an increase of Wenckebach point by 28 ± 33 bpm, shortening of AVN effective refractory period by 110 ± 115 ms, and sinus node recovery time by 508 ± 666 ms. During a median follow up of 26 (IQR: 12-39) months, CNA was repeated in 9 patients and is scheduled in 3 other (total 8%). Pacemaker was implanted only in 4 (3%) patients after single (n = 2) or repeated CNA (n = 2). Corresponding Kaplan-Meier curves are provided in the Figure. Any-syncope-free survival is comparable to that reported in active arms of historical and recent pacemaker studies.

Conclusions

CNA is a reasonably effective treatment option for patients with functional cardioinhibitory syncope. CNA can be performed by anatomically-guided ablation at empirical GP sites. Our study corroborates the clinical utility of CNA as a viable alternative to pacemaker implant in selected patients.

Temporary prolongation of corrected QT interval after cardioneuroablation for functional bradyarrhythmias

Background

There are controversial reports on QT interval response to ganglionic plexi ablation that are selectively targeted during cardioneuroablation (CNA) or occurs as collateral lesion during left atrial ablation procedures. Both shortening or prolongation of the heart-rate corrected QT interval (QTc) with therapeutic or safety implications were described.

Purpose

In this retrospective study, we investigated longitudinal changes of QTc after CNA.

Methods

The study included 108 patients (age: 39±12 years, 60% males) who underwent biatrial cardioneuroablation (radiofrequency time: 15.5±6.7 min) for symptomatic functional bradyarrhythmias. Surface ECG examinations were performed on the day before the CNA (N=108), 1 hour after the CNA (N=106), on the 1st post-ablation day (N=50), at the 3-month (N=99), and 1-year (N=63) follow-up visits. Automated measurements of QT interval were employed for the analysis. Four formulas (Bazett, Framingham, Fridericia, and Hodges) were used for the correction of QT interval to instant heart rate.

Results

QTc significantly prolonged immediately after the CNA with rapid return to baseline values (Figure and Table). This was particularly valid for QT correction by Framingham formula (and similarly for Fridericia and Hodges formulas). The QTc by Bazett formula, which is known to overestimate QT at higher heart rates, returned to baseline more slowly and incompletely. Several mechanisms may contribute to observed QTc dynamics: (1) direct effect of autonomic denervation with recovery phenomenon; (2) QT hysteresis with an extremely long time constant; or (3) artifact due to suboptimum QT correction to a substantial change of heart rate.

Conclusions

The study suggests that CNA produces acute prolongation of QTc interval with rapid decay and virtual normalization in 3 months. CNA in otherwise healthy subjects is not likely associated with substantial long-term risk of long-QT-associated arrhythmias. In the same way, we cannot confirm earlier observations of clinically significant QTc shortening effect of ganglionated plexi ablation.

Funding Acknowledgement

Type of funding sources: None. Figure 1Table 1

Acute change in parasympathetic cardiac innervation after pulmonary vein isolation by pulse-field and radiofrequency energy

Background

In patients with atrial fibrillation (AF), pulmonary vein isolation (PVI) by radiofrequency (RF) energy is associated with a significant change of cardiac autonomic regulations due to collateral ganglionic plexi ablation. Pulse-field (PF) ablation energy presumably spares neural tissue.

Purpose

We compared the effect of PVI by PF and RF energy on cardiac autonomic function.

Methods

A study enrolled 23 patients who underwent PVI in general anaesthesia. In 12 patients, a novel lattice-tip catheter and PF energy were used for ablation while 11 patients were ablated using a conventional irrigated-tip catheter and RF energy. The response of the sinus node (SAN) and atrioventricular node (AVN) to extracardiac high-frequency vagal stimulation (ECVS) was tested before and after PVI (via right internal jugular vein; stimulation frequency of 50 Hz; pulse width of 0.05 ms; output of 1 V/kg (<70V); train duration of 5 s).

Results

At baseline, physiological massive response to ECVS (sinus arrest and/or AV block) was demonstrated in the majority of patients. After PVI, complete loss of autonomic response of the SAN in 11/11 (100%) and 3/12 (25%) patients (p=0.003), and the AVN in 9/11 (82%) and 3/12 (25%) patients (p=0.01) was observed in RF and PF groups, respectively. The figure shows the maximum duration of the pause in sinus rhythm (maximum P-P interval) and AVN block (maximum R-R interval during atrial pacing) induced by ECVS after PVI.

Conclusion

Cardiac vagal response is preserved in a considerable proportion of AF patients after PF ablation which is in contrast with a significantly stronger effect of RF energy. This may influence the clinical outcome of AF ablation procedures.

Funding Acknowledgement

Type of funding sources: Public Institution(s). Main funding source(s): Institute for Clinical and Experimental Medicine, Prague

Radiofrequency catheter denervation of sinus node: a randomized comparison of right and left atrial approach for the cardioneuroablation

Funding Acknowledgements

Type of funding sources: None.

Background

Radiofrequency catheter ablation of superior paraseptal ganglionic plexus is an important step to eliminate the vagal modulation of sinus node for the treatment of neurally-mediated syncope. The reasonable effect can be achieved by targeting this plexus from the endocardial aspect of both right (RA) and left (LA) atria.

Purpose

We investigated the efficacy of RA and LA ablation in terms of sinus nodal denervation.

Methods

The study included 24 patients (age: 42 ± 13 years, 50% males) who underwent cardioneuroablation for recurrent cardioinhibitory syncope in general anesthesia. Right atrial semicircular lesion at the posteroseptal quadrant of superior vena cava ostium was composed of 5-6 equidistantly distributed ablation sites (30 W, 30 s, 20 ml/min).  Left atrial lesion of comparable size was placed strictly contralaterally across the interatrial septum in the anterior vestibulum of a right superior pulmonary vein. Patients were randomly (1:1) assigned to RA-to-LA or LA-to-RA ablation. Sinus rate and the response to extracardiac right vagal nerve high-frequency stimulation (50 Hz, 0.05 ms, 1 V/kg [<70V], 5 s) were recorded at baseline and after each ablation cluster.

Results

Study protocol ablations overall resulted in sinus acceleration (81 ± 13 vs. 59 ± 12 bpm, P <0.0001) and attenuation of inducible sinus arrests (maximum pause: 1.2 ± 1.4 vs. 5.5 ± 3.0 s, P <0.0001). Temporal development of outcome measures with the progression of ablation is shown in the Figure. There was no significant difference between study groups. Irrespective of ablation order, the first ablation cluster on average generated 77% of the final effect on sinus rate and 68% of the final effect on suppression of vagally-induced sinus pauses.  

Conclusions

 Neither RA nor LA approach is preferable for targeting the superior paraseptal ganglionic plexus. Both ablation clusters convey complementary and, in part, mutually independent effects. Biatrial cardioneuroablation seems essential for efficacious sinus nodal denervation. Abstract Figure.

Randomized investigation of the left atrial sweet spot for vagal denervation of atrioventricular node during the cardioneuroablation

Background

Radiofrequency catheter ablation of inferior atrial ganglionic plexi frequently results in vagal denervation of the atrioventricular (AV) node. The effective sites are, however, considerably variable.

Purpose

We prospectively sought the left atrial (LA) ablation site with the maximum effect on AV nodal modulation.

Methods

The study included 16 patients (age: 46±14 years, 56% males) who underwent cardioneuroablation for recurrent reflex cardioinhibitory syncope in general anesthesia. After targeting the superior paraseptal ganglionic plexi and achieving the denervation of the sinus node, study ablations were performed at the bottom of the LA to accomplish the AV nodal denervation (Figure). Five equidistantly distributed ablation lesions (30W / 30s / 20ml/min) were created on the virtual line connecting inferior ostium of right inferior pulmonary vein (RIPV) and inferior mitral annulus (MA). Lesions were centered symmetrically relative to the posterior mid-left-atrial line. They were numbered in ascending order from #1 (more septal, closer to the RIPV) to #5 (more lateral, closer to the MA). Patients were randomly (1:1) assigned to mutually opposite direction of ablation (from site #1 to #5 or from site #5 to #1). The response of heart rhythm to extracardiac vagal nerve high-frequency stimulation (50Hz/0.05ms/1V/kg [<70V]/5s) were recorded at baseline and after each radiofrequency energy delivery.

Results

Study protocol ablations overall resulted in elimination or attenuation of inducible AV block (maximum R-R interval: 2.9±2.8 vs. 5.2±2.4s, P<0.001). Temporal development of effect with the progression of ablation is shown in the Figure indicating that the most lateral lesion alone produces the maximum effect. The AV nodal denervation was incomplete after per-protocol ablations in 7/16 patients. In the majority of them, the final success was achieved by extension of ablation lesion toward the inferior mitral annulus either endocardially or via the proximal coronary sinus.

Conclusion

Ablation of perimitral region of the inferior LA conveyed the maximum effect in terms of AV nodal denervation.

Funding Acknowledgement

Type of funding source: None

P1097Heart rate acceleration is a poor surrogate of complete parasympathetic denervation of sinus node during cardioneuroablation

Background

Ablation of superior paraseptal ganglionic plexi is invariantly associated with the acceleration of sinus rhythm. This is considered a favourable sign during cardioneuroablation for the treatment of recurrent neurally-mediated cardioinhibitory syncope or symptomatic sinus bradycardia.

Purpose

In this retrospective study, we investigated whether the magnitude of sinus rhythm acceleration corresponds with directly assessed sinus nodal parasympathetic denervation.

Methods

The study included 48 patients (age: 39 ± 13 years, 58% males) who underwent cardioneuroablation in general anaesthesia. The procedural endpoint was non-responsiveness (i.e. loss of original cardioinhibitory response) of the sinus node to extracardiac high-frequency stimulation of the vagal nerve. The magnitude of sinus rhythm acceleration was compared between patients who reached or did not reach this endpoint.

Results

All patients had positive atropine test (baseline heart rate: 65 ± 14 bpm; post-atropine: 109 ± 22 bpm). Complete sinus nodal denervation as assessed by vagal nerve stimulation was achieved in 44/48 (92%) patients. Intraprocedurally, heart rate accelerated from 54 ± 11 to 85 ± 14 bpm (difference: 31 ± 10; median 29; interquartile range: 24–40; total range: 13–61 bpm). This change did not correlate with age and was not related to pre-procedural post-atropine sinus rhythm acceleration. There was no difference in heart rate acceleration between the patient with and without sinus nodal denervation (Figure).

Conclusions

Sinus rhythm acceleration is not reliable endpoint for cardioneuroablation. Guidance by extracardiac vagal nerve stimulation may help to tailor the procedures to increase the clinical success rate and, at the same time, to avoid patient overtreatment.

Abstract Figure.

P3752Ablation of the superior left ganglionic plexus is not necessary for effective denervation of the sinoatrial and atrioventricular nodes

Background

Cardioneuroablation by targeting of atrial ganglionic plexi (GP) has been proposed as a new therapeutic option in selected patients with reflex syncope. Contribution of individual GPs to cardiac autonomic regulations is not fully established.

Purpose

Because consistent vagal responses have been observed during left superior pulmonary vein isolation in patients undergoing ablation for atrial fibrillation, we investigated whether standalone ablation of the superior left GP modifies the vagal input into the sinoatrial (SAN) and atrioventricular node (AVN).

Methods

Study hypothesis was investigated in otherwise healthy patients undergoing cardioneuroablation for symptomatic bradyarrhythmias. All had preprocedural atropine test suggesting functional disorder. Anatomically-navigated (CARTO-3) radiofrequency (RF) ablation (25–30 W/30 s/20 ml/min) at empirical GP sites was performed in general anaesthesia. Extracardiac high-frequency vagal nerve stimulation (25–60 V/30–50 Hz/0.05–0.1 ms) via right jugular vein was performed at baseline, after initial superior left GP ablation, and after the ablation of remaining septal and inferior GPs. High-frequency vagal nerve stimulation was always done in both sinus rhythm and atrial pacing. The elimination of all stimulation-induced vagal responses was the endpoint of the procedure.

Results

A study included 8 patients (34±8 years; 5 males). Six of them had recurrent syncope with cardioinhibitory response at the SAN (n=4), AVN (n=1) or both nodes (n=1); and 2 patients had symptomatic sinus bradycardia. At baseline, high-frequency vagal nerve stimulation induced long episodes of sinus arrest and advanced AV block in all patients. Cluster ablation at the superior left GP (RF time: 192±28 s) did not change the sinus rate (59±14 vs 60±15, NS), PQ interval (174±37 vs 173±37, NS), and did not induce any tangible change in SAN/AVN response to high-frequency vagal nerve stimulation. Subsequently, anterior right GP was targeted from the aspect of right atrium (n=7), from the anterior antrum of right pulmonary veins (n=6), and inferior GPs were targeted from the aspect of left atrium (n=7). This lesion set finally resulted in complete non-responsiveness of SAN and AVN to high-frequency vagal nerve stimulation in all patients. Ablation procedure overall (duration: 172±15 min; RF time: 988±306 s; radiation dose: 70±34 μGy·m2) led to sinus rate acceleration by a median of 29 (IQR: 18–38) bpm, increase of Wenckebach point by 21 (IQR: 9–28) bpm, and shortening of AVN effective refractory period by 40 (IQR: 15–73) ms.

Conclusions

Cardioneuroablation guided by extracardiac high-frequency vagal nerve stimulation can achieve complete denervation of SAN and AVN by ablation of postero(para)septal and inferior GPs only. Ablation of the superior left GP appears unnecessary and can be eliminated from the lesion set design.