Selective Cardioneuroablation for Swallow-Induced Atrioventricular Block

Abstract

Background

Swallow-induced atrioventricular (AV) block is a rare mechanism of paroxysmal AV block, and no established curative treatment currently exists.

Case Summary

A 75-year-old woman without known structural heart disease presented with symptomatic paroxysmal AV block during meals. Swallowing provocation test reproducibly induced AV block at the A-H level, which was suppressed following atropine administration, suggesting a vagally mediated reflex mechanism. Despite dietary counseling, her symptoms persisted; therefore, cardioneuroablation (CNA) was performed. Among the ganglionated plexi, sites demonstrating vagal responses during high-frequency stimulation, as well as anatomically defined ganglionated plexi regions, were selectively ablated. Following the procedure, no further swallow-induced paroxysmal AV block was observed.

Discussion

In this case, combined selective and anatomical CNA approach was used, targeting sites considered to provide parasympathetic input to the AV node.

Take-Home Message

Mechanism-based selective CNA may represent a promising therapeutic option for patients with swallow-induced AV block refractory to conservative management.

Visual Summary

History of Presentation

A 75-year-old woman without known cardiac comorbidities presented with chest comfort. A Holter electrocardiogram (ECG) was performed to evaluate the cause of chest comfort, which revealed 14 times episodes of paroxysmal atrioventricular (AV) block per day, with approximately 4-second cardiac pauses predominantly during mealtimes (Figure 1). The episodes were not associated with apparent P-P and P-R interval prolongation preceding paroxysmal AV block (Figure 1). The patient was referred to our hospital and was admitted for further investigation. She had no previous history of transient loss of consciousness.

Figure 1.

Holter ECG: An Episode of Paroxysmal Atrioventricular Block With Approximately 4-Second Cardiac Pauses During Mealtimes

The episodes were not associated with apparent P-P and P-R interval prolongation preceding paroxysmal atrioventricular block. ECG = electrocardiogram.

Past Medical History

She had no past medical history other than diabetes mellitus.

Differential Diagnosis

The differential diagnosis of paroxysmal AV block included the following conditions:

• 1.Swallow-induced AV block, identified by a banana swallowing challenge test performed during the electrophysiological study (EPS) under continuous intracardiac ECG and blood pressure monitoring.
• 2.ATP (adenosine triphosphate)-sensitive AV block, excluded based on the intracardiac ECG findings obtained during the banana swallowing challenge test and the response to administration of theophylline, an adenosine receptor antagonist.
• 3.Structural abnormalities of the cardiac conduction system, which were excluded based on findings from the resting ECG and control EPS.

Investigations

Rest ECG showed sinus rhythm with a P-R interval of 159 ms and a narrow QRS with incomplete right bundle branch block, a transthoracic echocardiogram showed no structural abnormalities, and laboratory data and gastrointestinal endoscopic studies were also unremarkable. In-hospital monitoring confirmed that paroxysmal AV block occurring during meals was associated with chest discomfort.

Baseline EPS demonstrated normal AV conduction (A-H interval, 107 ms; H-V interval, 36 ms; Wenckebach point, 140 beats/min). A banana swallowing challenge reproducibly induced paroxysmal 2:1 AV block at the A-H level, accompanied by mild P-P interval prolongation (maximum 40 ms) and a significant decrease in blood pressure (Figure 2). The swallowing banana test was performed repeatedly and consistently reproduced AV block. Theophylline and atropine challenge tests were conducted on separate occasions. Atropine administration suppressed the induction of paroxysmal AV block by “banana challenge test”; however, theophylline could not suppress the induction of paroxysmal AV block. These findings suggested a swallow-induced vagally mediated reflex mechanism.

Figure 2.

A Swallowing Banana Challenge Test

While swallowing banana test under monitoring intracardiac electrocardiogram and continuous blood pressure induced paroxysmal 2:1 atrioventricular block at the A-H level with max 40 ms prolongation of P-P interval and significant blood pressure drop. For further details, please refer to the main text. EPS = electrophysiological study; His = His bundle; HRA = high right atrium; RV = right ventricle.

Management

Therapy decision-making

Following discussion with the patient, an implantable loop recorder (ILR) was implanted for close follow-up of AV block, and dietary counseling was provided to encourage thorough mastication of solid foods. At 3-month follow-up, the swallow-induced AV block persisted (with a total of 1,279 pause events in the 3 months) and the patient subsequently developed weight loss. The patient's weight loss was attributed to reduced food intake because of meal-related symptoms. Although dual-chamber pacemaker implantation was considered, the absence of syncope, normal conduction findings, relatively short pauses (maximum approximately 4 seconds), and atropine responsiveness favored cardioneuroablation (CNA) as the initial treatment strategy. Our patient consented for CNA-radiofrequency catheter ablation of ganglionic plexi (GP) aimed at modification of parasympathetic input to the AV node.

Interventions

The procedure was performed under local anesthesia using a 3-dimensional electroanatomic mapping system (CARTO3, Biosense Webster) integrated with preprocedural CT images. High-frequency stimulation (HFS) (20 Hz, 20 mV, pulse width 10 ms) was delivered endocardially at anatomically defined left atrial GP regions using a contact force-sensing ablation catheter (Thermocool SmartTouch; Biosense Webster). GP regions were defined according to Aksu et al.¹ The inferior left ganglionic plexi (ILGP) was located adjacent to the left inferior pulmonary vein. The inferoposterior left atrial ganglionic plexi (IPLA-GP) along the inferoposterior left atrial wall between the inferior pulmonary veins, and the inferior right ganglionic plexi (IRGP) adjacent to the right inferior pulmonary vein, and the posteromedial left ganglionic plexi (PMLGP) adjacent to the coronary sinus ostium, corresponding to the major parasympathetic input to the AV node. A positive HFS response was defined as either ≥50% R-R interval prolongation or a ≥20 mm Hg decrease in systolic blood pressure (Figure 3). Of 71 HFS sites, 24 demonstrated positive vagal responses, predominantly in the ILGP, IPLA-GP, and IRGP regions (Figure 4). Radiofrequency energy (30 W, 30 s) was delivered to all HFS-positive sites. Additional anatomical ablation was performed at the PMLGP from the left atrium and the coronary sinus ostium (Figure 5).

Figure 3.

Representative Positive Vagal Response Case of High-Frequency Stimulation at the GP Sites

High-frequency stimulation (20 Hz; output, 20 mV; pulse width, 10 ms) was delivered endocardially at sites of the left atrial GP. High-frequency stimulation induced atrial fibrillation accompanied by a bradycardic response with a concomitant decrease in arterial BP. Positive response was defined as either a ≥50% prolongation of the R-R interval or a ≥20 mm Hg decreases in systolic BP. ABL = ablation catheter; Ao = aortic pressure; BP = blood pressure; GP = ganglionic plexi; IRGP = inferior right gaglionated plexi; other abbreviation as in Figure 2.

Figure 4.

The Left Atrium 3-Dimensional Map and the GP Sites Undergoing High-Frequency Stimulation

A total of 71 sites underwent high-frequency stimulation, of which 24 sites (approximately one-third) demonstrated a positive vagal response. The distribution of positive sites was predominantly localized to the ILGP and IPLA-GP regions along the posterior wall of the left atrium and the IRGP. ILGP = inferior left gaglionated plexi; IPLA-GP = inferoposterior left atrial gaglionated plexi; LIPV = left inferior pulmonary vein; PA = posteroanterior view; RIPV = right inferior pulmonary vein; other abbreviations as in Figure 3.

Figure 5.

The Left Atrium 3-Dimensional Map and RF Application Sites

RF energy (30 W, 30 seconds) was applied point by point at these positive response GP sites and PMLGP, which was targeted from the left atrium and the ostium of the CS on the right side. (Left: PA view, Right: LAO view) CS = coronary sinus; LAO = left anterior oblique view; PMLGP = posteromedial left gaglionated plexi; RF = radiofrequency; other abbreviations as in Figures 3 and 4.

At the postablation procedure, the P-R interval was 135 ms (from 161 ms at preablation) and the Wenckebach point was 141 beats∖min (from 120 beats∖min at preablation).

Outcome and Follow-Up

No recurrence of AV block was observed during hospitalization. At 3 months, the patient remained asymptomatic on a normal diet, and ILR-detected pause events decreased from 1,279 before ablation to only 2 after ablation, both occurring during sleep, with complete elimination of swallow-induced AV block. Appetite and quality of life markedly improved. At 6 months, ILR showed no pause events or tachyarrhythmic complications.

Discussion

Mechanism of swallow-induced atrioventricular block

Paroxysmal AV block is broadly classified into 2 categories: structural conduction system disease and functional conduction disturbance. Functional paroxysmal AV block is typically characterized by a normal baseline 12-lead ECG, absence of structural heart disease, preserved cardiac function, abrupt onset of complete AV block on ambulatory monitoring, and negative findings on EPS.² Functional AV block includes ATP-sensitive AV block and neurally mediated AV block, which can be differentiated by ECG findings at the onset of AV block. ATP-sensitive AV block is characterized by the absence of P-P interval changes, whereas vagally mediated AV block is typically preceded by progressive prolongation of the P-P and P-R intervals. In our case, apparent P-P interval prolongation was not observed during spontaneous episodes of paroxysmal AV block. Therefore, a banana challenge test was performed under continuous intracardiac ECG and blood pressure monitoring to differentiate ATP-sensitive from vagally mediated AV block. During the test, paroxysmal AV block was accompanied by a decrease in blood pressure and subtle P-P interval prolongation. The reproducible induction of AV block by swallowing was not suppressed by theophylline, an adenosine receptor antagonist, but was successfully suppressed by atropine. Although progressive P-P interval prolongation is commonly regarded as a characteristic feature of vagally mediated AV block, vagal reflexes may preferentially affect AV nodal conduction without substantial sinus node suppression.² In swallow-induced AV block, selective vagal activation directed to the AV node has been reported,^3,4 and therefore the absence of progressive P-P interval prolongation does not necessarily exclude a neurally mediated mechanism. In addition, the marked reduction of episodes after cardioneuroablation collectively support a vagally mediated functional AV block despite the absence of preceding P-P interval prolongation.

Although 2 nocturnal AV block episodes were detected during the first 3 months after CNA, no meal-related AV block recurred, and the overall burden of pause events was reduced from 1,279 episodes to only 2 episodes. This marked reduction strongly supports a predominantly neurally mediated mechanism underlying the paroxysmal AV block. The occurrence of a small number of residual nocturnal episodes may indicate a persistent susceptibility to functional AV block under conditions of heightened vagal tone during sleep or residual vagal influence despite CNA. Importantly, the complete elimination of meal-related episodes, together with marked symptomatic improvement, supports a swallow-induced vagal reflex as the predominant mechanism in this patient.

Cardioneuroablation

Recently, CNA has emerged as an alternative therapy for neurally mediated syncope and has been reported to reduce syncope recurrence.⁵ Two principal CNA strategies have been described: anatomical GP ablation and selective GP ablation. The anatomical approach targets the 4 GPs: the superior left GP, ILGP, anterior right GP, and IRGP, located around the pulmonary veins, whereas the selective approach identifies GP responses using high-output HFS, with radiofrequency energy delivered only to sites demonstrating positive vagal responses.

In the present case, we combined these 2 strategies. First, selective GP ablation was performed at sites showing positive vagal responses to HFS within the major GPs, including the ILGP, IPLA-GP, and IRGP (Figure 4). This approach was supported by a recent meta-analysis by Prata et al,⁶ demonstrating lower syncope recurrence rates with electroanatomic mapping-guided CNA combined with HFS. In addition, the PMLGP, considered a major parasympathetic input region to the AV node, was anatomically ablated from the left atrium and right-sided coronary sinus ostium (Figure 5).

Two previous cases of swallow-induced syncope successfully treated with CNA have been reported.^7,8 One involved sinus arrest, in which the superior right atrial GP associated with vagal input to the sinoatrial node was targeted.⁷ The other involved AV block and was treated primarily with anatomical ablation of the PMLGP.⁸ In contrast, our strategy avoided GP associated with vagal input to the sinoatrial node and selectively targeted GP related to the AV node, which may help minimize postprocedural sinus tachycardia and atrial arrhythmias.

Conclusions

This case corroborates the clinical utility of selective and anatomical approach to CNA targeting the GPs responsible for vagal input to the AV node in patients with swallow-induced AV block, highlighting the importance of elucidating the underlying mechanism of paroxysmal AV block. Further studies involving larger patient populations are warranted to determine the optional procedure strategy and to establish CNA as a standard therapy for swallow-induced AV block.

Visual Summary.

Implantable Loop Recorder Monitoring Documented 1,279 Pauses ≥3 Seconds During the 3 Months Before Ablation

Cardioneuroablation targeted high-frequency stimulation–positive sites (cyan) and the PMLGP anatomically from the LA and CS ostium; RF sites are shown in red. During the 3 months after ablation, only 2 nocturnal transient AV block events occurred, with complete elimination of swallow-induced AV block. Abl = ablation; AV = atrioventricular; CS = coronary sinus; GP = ganglionated plexus; ILGP = inferior left ganglionated plexus; IPLA-GP = inferoposterior left atrial ganglionated plexus; IRGP = inferior right ganglionated plexus; LA = left atrium; LAO = left anterior oblique; LIPV = left inferior pulmonary vein; PA = posteroanterior; PMLGP = posteromedial left ganglionated plexus; RF = radiofrequency; RIPV = right inferior pulmonary vein.

Funding Support and Author Disclosures

Boston Scientific Japan and Japan Lifeline Inc support the endowed department with which Drs Kohno, Oginosawa, and Abe are affiliated. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.

Take-Home Messages

• •Management of paroxysmal atrioventricular block requires careful differential diagnosis and consideration of the underlying mechanism rather than indiscriminate pacemaker implantation.
• •Selective and anatomical cardioneuroablation may represent an effective therapeutic option with a favorable safety profile for patients with swallow-induced atrioventricular block refractory to conservative management.