Clinical controversy: methodology and indications of cardioneuroablation for reflex syncope

Graphical Abstract

Graphical Abstract.

Observational studies and a recent randomized controlled trial have provided some evidence that cardioneuroablation (CNA) may prevent syncope recurrence in patients affected by cardioinhibitory reflex syncope during the first 2 years following the procedure. Nevertheless, many issues concerning clinical indications, methodology, and long-term results remain unresolved. These issues constitute the background of this manuscript in which pioneers in CNA and experts in syncope debated critical aspects aimed at finding agreement or, if not possible, that highlighting the controversies could be addressed in future studies.

Introduction

In 2005, Pachon et al.¹ proposed cardiac vagal denervation to treat neurally mediated syncope. Since then, a metanalysis of observational studies² and a recent randomized controlled trial (RCT)³ have provided some evidence that cardioneuroablation (CNA) is able to prevent syncope recurrence at least during the first 2 years following the procedure in patients affected by reflex syncope. In brief, the recent metanalysis² of 14 studies including a total of 465 patients reported an average freedom of syncopal recurrence in 92% of patients (95% confidence interval 88–95%) during follow-up. The only available open RCT³ reported 8% recurrence of syncope in the 24 patients randomized to CNA and 54% recurrence rate in 24 untreated controls (P = 0.0004) during 2-year follow-up. In all studies, after the ablation procedure, the patients demonstrated heart rate increase together with reduction of heart rate variability (indicating impaired parasympathetic stimulation of the heart) persisting at least for 2 years. This provides proof-of-concept for the modification of the vagal ganglia activity in the heart.

Given these facts, it is likely that CNA will become increasingly popular among physicians caring for syncope patients in the years to come. Nevertheless, many issues concerning clinical indications, methodology, and long-term results remain unresolved. These issues constitute the background of this manuscript in which leaders in CNA and experts in syncope debated critical issues and aimed to find agreement and, if not possible, to highlight the controversies that could be addressed in future studies. The authors were initially requested to give their evidence-based opinion on several predefined motions. These were merged into a manuscript draft, which was subsequently critically revised by means of two rounds of comments.

Glossary of terms

The consensus emerged on two new terms, i.e. superior paraseptal ganglionated plexus (SPSGP) and inferior paraseptal ganglionated plexus (IPSGP), that are frequently targeted during cardioneuroablation procedures. In the literature, the SPSGP is also reported as superior right atrial ganglionated plexus, or anterior right ganglionated plexus, or right superior ganglionated plexus. Even though subtle anatomical differences exist in the terminology, the authors of this document by consensus agree that all these terms should be considered synonyms of the final common pathway of the parasympathetic innervation of the sinoatrial node. Similarly, in the literature, the IPSGP is also described as posteromedial left atrial ganglionated plexus, or inferior right ganglionated plexus, or right inferior ganglionated plexus, or left inferior ganglionated plexus. Even though subtle anatomical differences exist in the terminology, the authors of this document by consensus agree that all these terms should be considered synonyms of the final common pathway of the parasympathetic innervation of the atrioventricular node.

Current indications based on the characteristics of the patients who have undergone cardioneuroablation in the literature

The mean age of the 465 patients included in the metanalysis,² was 39.8 ± 14.0 years and 53.5% of patients were female. The patients had ≥3 syncopal episodes in the previous year. A tilt test had been performed in 94% of cases. When the VASIS classification was applied, a cardioinhibitory form was present in 66%, a mixed form in 31% and a vasodepressor form in 2% of cases. In the RCT,³ the mean age of the 48 patients was 38 ± 10 years and 58% of patients were females. The patients had had 10 ± 9 spontaneous syncopal episodes during life and of 3 ± 2 episodes in the previous year. The index event was an asystolic syncope with a mean pause of 17 ± 15 s due to sinus arrest (87% of cases) and atrioventricular (AV) block (13% of cases). The inclusion criteria were: (i) at least one documented episode of spontaneous reflex syncope during the preceding 12 months or one prior syncopal event leading to injury and two presyncopal events; (ii) electrocardiogram (ECG) documentation of spontaneous asystolic syncope or symptomatic >3 s asystolic pause or bradycardia <40 b.p.m. during tilt test; and (iii) >25% increase of sinus rate after intravenous atropine injection.

Summary

• A typical candidate undergoing CNA in the population of about 500 patients reported to date in the literature, is a patient affected by severely symptomatic recurrent reflex syncope, aged 18–60 years, with documentation of spontaneous or tilt-induced asystolic form and exclusion of intrinsic sinus node or AV node dysfunction.

• Like cardiac pacing, the fact that CNA may be effective does not mean that it is always necessary. Cardioneuroablation and cardiac pacing should be the last choices, after failure of behavioural, non-pharmacological and pharmacological therapies, and should only be considered to relieve symptoms in a small cohort of highly selected patients, i.e. those affected by severe forms of reflex syncope with frequent recurrences that alter the quality of life forcing patients to seek intervention and are associated with a high risk of injury, often due to the lack of prodrome.

Motion #1: ‘Right sided ganglia ablation should be the standard procedure’

Background

Neuroanatomical studies in cadavers of people without known cardiovascular disease at the time of death showed that most of the cholinergic fibres access the human heart through the posterior and anterior right atria.⁴ Based on the anatomical findings it seems reasonable to limit ablation to the right atrium (RA) to balance between efficacy and safety. From a functional perspective, the right vagus nerve mostly innervates the sinoatrial node, the left vagus nerve affects the AV node.⁵ Keeping the mechanisms underpinning cardioinhibition in vasovagal syncope in mind, parasympathetic fibres firing on the sinoatrial node should be the preferred target for CNA in people with treatment-resistant vasovagal syncope of cardioinhibitory type because sinus arrest is the prevailing bradycardia form in such patients. Such ablation denervates the sinoatrial node acutely almost in all patients while it also incompletely, but significantly impacts AV node innervation. Thus, ablation of parasympathetic fibres firing on the AV node should be additionally added to those patients with a vagally-induced AV block or with persistence of vagally-induced AV block after sinoatrial ablation.

The superior paraseptal ganglionated plexus (SPSGP), that is located at the junction between the interatrial septum and the superior vena cava [anterior to the left atrium (LA) and posterior to the RA], is the final common pathway of right vagus innervating the sinoatrial node. The inferior paraseptal ganglionated plexus (IPSGP), that is located near the proximal coronary sinus (and around its ostium) and the posteroseptal junction between RA and inferior vena cava adjacent to posteroinferior LA (at the so-called pyramid space), is the final common pathway of the left vagus that innervates the AV node⁶ (Figure 1). The ablation of other ganglionated plexi is needed infrequently.

Figure 1.

Drawing of the posterior view (A) and superior view (B) of the human heart and major vessels illustrating the location of the posterior atrial ganglionated plexuses. The SPSGP and the IPSGP are depicted. The SPSGP lays paraseptal close to the SVC and to the sinoatrial node (not shown). The IPSGP is in the inferoposterior part of the interatrial septum close to the AV node (not shown). Positions of the pulmonary veins (PV), superior vena cava (SVC), inferior vena cava (IVC), right ventricle (RV), and left ventricle (LV) are shown. Modified from Armour et al.,⁶ with permission. AV, atrioventricular; IPSGP, inferior paraseptal atrial ganglionated plexus; SPSGP, superior paraseptal ganglionated plexus.

PRO	CONTRA
• In most recent studies, RA ablation was performed at the junction between interatrial septum and superior vena cava where the SPSGP is located and near proximal coronary sinus (and inside its ostium) and posteroseptal junction between RA and inferior vena cava, where the IPSGP is located.3,7–9	• In most recent studies, biatrial ablation was performed starting in the LA at the anterior antrum of the right superior pulmonary vein where the SPSGP is located, followed by the inferoposterior part of the interatrial septum where the IPSGP is located and finally in the RA as described above.3,10,11
• The RA approach is simple and fast. It eliminates or reduces the risks associated with transseptal puncture and manipulation in the LA, the risk of procedural and post-procedural therapy with potential bleeding complications and the risk of procedural and postprocedural cerebral and/or systemic embolization.	• CNA limited to RA ablation was associated with a significantly lower freedom from syncope than with LA ablation only (94.0%) and biatrial ablation (92.7%).2 The results of the RCT3 were achieved with the biatrial ablation. The RA approach may not be enough for complete vagal denervation. CNA from RA only may be enough for sinus node denervation, however not for AV node.12
• Freedom from syncopal recurrence after CNA performed from the RA is around 82%.2 In very symptomatic patients, at 12 months after ablation, syncope burden decreased by 95% and mean 24 h heart rate increased from 69–77 b.p.m. using right sided ablation of the SPSGP.8	• Potential complications of RA approach include right phrenic nerve and sinus node damage especially if large RA areas are targeted.
• SPSGP and, to a lesser extent, IPSGP, are closely associated with the anatomical structures of sinoatrial node and AV node respectively. Both SPSGP and IPSGP are located in the interatrial septum and hence can theoretically be ablated through a right-sided, left-sided, or bilateral approach. Any ganglionated plexus sandwiched between the left and right endocardium can potentially be heated by the radiofrequency electrode on one side and cooled contralaterally by the blood flow of the adjacent atrium.	• Albeit only a head-to-head study could compare the success rate of the techniques, the more extensive LA ablation has been claimed to be able to prevent reinnervation as it was observed several years ago in patients after cardiac transplant.13 • When the technique of extracardiac vagus nerve stimulation was used to measure the magnitude of denervation, only biatrial ablation was able to provide a complete denervation of both sinus and AV node.14,15

Summary

• The targets for CNA are the SPSGP and the IPSGP (Figure 2). While denervation to some extent can be achieved by right sided approach, complete denervation requires an approach from both the RA and the LA, especially for the ablation of the IPSGP. The ablation of SPSGP might be sufficient for denervation of the sinoatrial node, but not for the denervation of the AV node which also requires the ablation of the IPSGP.

• The final decision should weigh the risk of a particular ablation strategy and the risk of recurrent syncope with the need for a redo procedure. Head-to-head comparisons between right-sided, left-sided, or bilateral approaches in well-designed trials are required to define the best ablation strategy tailored to the patient’s individual needs.

Figure 2.

Electroanatomic map of the RA. Typical site of radiofrequency application (red bullets) of the SPSGP and of the IPSGP. An example of fragmented atrial electrograms recorded in the site of the lesion is shown. Modified from Piotrowski et al.,³ with permission. IPSGP, inferior paraseptal ganglionated plexus; RA, right atrium; SPSGP, superior paraseptal ganglionated plexus.

Motion #2: ‘CNA should be limited to otherwise healthy patients <60 years’

PRO	CONTRA
• ‘Isolated’ asystolic reflex syncope is rare in elderly adults. One needs a ‘very healthy’ vagus nerve to elicit an asystolic cardioinhibitory vasovagal response, more frequently observed in children or teenagers. With physiological aging, the vagus nerve also undergoes some degree of neurodegeneration, and the autonomic nervous system shifts towards a sympathetic predominance, which result in hypotension playing a major role in causing syncope.	• There is no strong rationale to restrict CNA to patients under the age of 60. There might be a place for CNA in older patients with reflex syncope, and absence of intrinsic sinoatrial and AV nodal disturbances who respond well to pharmacological vagal withdrawal (atropine).16
• In addition, while CNA may be effective at any age, additional risks are likely in those with comorbidities and potential explanatory mechanisms including sinus node and AV nodal dysfunction concomitant with autonomic impairment.17–19	• Anecdotal cases were described in old patients with swallowing syncope or carotid sinus syndrome.20,21
• The efficacy of CNA seems to decrease with increasing age. Qin et al.,22 in patients affected by symptomatic bradycardia, showed that the increase in heart rate after CNA was less in patients >50 years of age and lasted for a shorter time. Moreover, the Short-Form 36 Health Survey did not show obvious improvement of quality of life in patients ≥50 years of age.
• While the experience with CNA is limited to few anecdotal cases in patients over the age of 60, pacemaker therapy has proven to be effective and has gained a Class I, Level of Evidence A recommendation in the most recent 2021 ESC guidelines.23

Summary

• Intrinsic sinoatrial and AV nodal disturbances, as well as underlying structural heart disease, not age per se, are discriminators for CNA in the elderly. While there is no reason to restrict CNA to patients under age 60 who have symptoms due to vagal activation, a controlled trial should be undertaken to demonstrate that CNA is non-inferior to pacemaker implantation in relieving symptoms before this technique is applied to patients over the age of 60. Until that time, cardiac pacing is an established therapy and CNA must be considered experimental.

Motion #3: ‘There is lack of rationale for a benefit of CNA for vasodepressor forms of reflex syncope’

Background

In reflex syncope, the afferent pathways transfer information from the circulatory and visceral receptors to the brain. Haemodynamic instability (evidenced by central hypovolaemia, hypotension, and/or tachycardia), gastrointestinal symptoms, pain, and other triggers can activate the reflex. Higher brain functions such as emotional triggers can facilitate activation of the reflex or trigger it directly.

The main efferent components of the reflex are the parasympathetic/vagal branch to the heart producing bradycardia up to asystole and the sympathetic pre/post-ganglionic fibres, which, in case of phasic withdrawal, produce abrupt vasodilation of the capacitance vessels in the splanchnic region and lower limbs, with consequent hypotension. The combination of vasodepressor effects and bradycardia to varying degrees results in a vasodepressor, cardioinhibitory, or mixed type of reflex syncope.²⁴

The mechanism of action of CNA is quite similar to that of cardiac pacing since they both counteract the vagal outflow to the sinoatrial node and to the AV node. Both therapies are unable to counteract the vasodepressor reflex in the vasculature, particularly in the splanchnic region and in the vessels. The experience with cardiac pacing has shown that pacing is more effective when the cardioinhibitory response is dominant and, conversely, it is less effective or ineffective when the hypotensive component is dominant²⁵ Even in patients with documented asystolic syncope, recurrence occurs in up to 25% of patients at 3 years of follow-up due to coexistent hypotensive susceptibility.²⁶ Thus, a similar effect should be expected also for CNA.

PRO	CONTRA
• Piotrowski et al.27 studied the impact of CNA on the type of vasovagal response during tilt testing. Before CNA, 20 patients had cardioinhibitory and 1 had mixed syncope. During 1-year follow-up, no spontaneous syncopal episodes were noted. However, at follow-up tilt testing, 13 patients had syncope—12 due to vasodepressor mechanism and only 1 due to asystole. These data demonstrates that CNA profoundly affects the parasympathetic efferent branch of the reflex without affecting the sympathetic efferent branch and the brainstem integration pathway.	• While there is no rationale for a benefit of CNA, as well for cardiac pacing, in pure vasodepressor forms, and in the mixed forms bradycardia contributes to the fall in cardiac output and arterial blood pressure during Phase II and in the Phase III preceding syncope.28 Thus, increasing heart rate by CNA or by cardiac pacing could be helpful limit the effect of the vasovagal response in mixed forms. Considering the clinical characteristics of patients included in CNA studies,2 a mixed response was observed in 31.4% of patients and a vasodepressor response in 2.5% of patients.
• The phenotype of the CNA candidate should be detailed, i.e. not only documented reflex syncope of cardioinhibitory type and positive atropine test but also documentation of asystole during spontaneous syncope with an implantable loop recorder ( ILR) and exclusion of hypotensive susceptibility with a 24 h ambulatory blood pressure monitor29,30 are recommended to increase the likelihood of therapeutic success.	• The phenotype of reflex syncope is often difficult if accurate criteria are not used. For example, in patients who faint systematically because of vasodepressor reflex response diagnosed with 24 h ambulatory blood pressure monitor or with tilt testing, the cardioinhibition could finally be documented if the patient undergoes prolonged ILR monitoring or carotid sinus massage.26

Summary

• At the present time, both from a pathophysiological perspective and based on the available evidence, there are few arguments to support CNA in patients with vasodepressor syncope.

• For practical reasons, at this stage of knowledge, considering the difficulty in assessing therapeutic benefit at the time of the ablation or thereafter, it is prudent firstly to establish the efficacy of CNA in patients with cardioinhibitory forms before extending indications to mixed and vasodepressor forms of reflex syncope.

Motion #4: ‘The effect of CNA decreases with time due to re-innervation of the sinus and AV nodes’

Background

The literature on histology of the vagus nerve indicates that the post-ganglionic parasympathetic neurons are located in the epicardial ganglionated plexi or they are directly embedded in the cardiac wall, where the efferent vagal fibres are certainly located.⁴ The depth of the lesions caused by radiofrequency lesions is crucial for this issue. The general concept is that if the soma/nucleus of the neuron is damaged, the cell will not regenerate. If only the (transmural) fibres are damaged, then axonal regeneration and new synaptic sprouting will occur, and neighbouring spared neurons will reinnervate some myocytes which were previously innervated by ablated neurons. At the end re-innervation will occur, even though with a ‘patchy’ pattern.

In patients undergoing heart transplant, which represents a different group, since denervation is at the pre-ganglionic level, reinnervation has been reported to occur after >1 year.^13,31–33 Thus, data from longer follow-up time after CNA is necessary.

PRO	CONTRA
• In the CNA literature, there is some evidence of heart rate decrease and of recovery of heart rate variability (i.e. vagal reinnervation) toward pre-ablation values after 2 or 3 years.2,7,34 which was greater in patients who had recurrence of syncope or positive tilt testing after ablation.34,35 It appears that the rate of reinnervation/recovery is highest acutely in early days/weeks with non-linear decay during subsequent months, so that little change if any, continues 1 year after the index procedure. It depends on the study design whether this phenomenon is detected or not. Indices of heart rate variability are relatively insensitive. Post-procedural atropine re-testing seems a reproducible approach.	• Some studies showed that the effect of reinnervation is insignificant at 2 years.3,36 Even if there is some evidence of reinnervation in other studies,2,7,34,35 this might be partial and there is not sufficient data indicating that the magnitude of reinnervation/recovery is associated with the risk of syncope recurrence. The magnitude of reinnervation/recovery is highly variable and logically depends on the procedure-related factors (ablation technology, extent of ablation, and proper targeting of ganglionic sites) and patient-related factors (thickness of myocardium as well as epicardial fat at critical regions and, hypothetically, intra-individually variable capacity of neural regeneration).
	• The ultimate absolute level of vagal modulation should correspond to the risk of recurrent cardioinhibitory syncope. In addition, the clinically efficacious degree of neuromodulation is unknown.
	• A recovery process is not always unwanted. In some patients, the recovery may help to attenuate the post-procedural sinus tachycardia if this is symptomatic. It is known that the natural history of vasovagal syncope in young patients is characterized by gradual spontaneous reduction of syncopal attacks with advancing age. A ‘bridge to pacemaker’ ablation strategy might be the best solution for these patients while preserving them from potentially adverse effect of a persistent vagal denervation with persistent increased resting heart rate on one hand and from and too early pacemaker implantation on the other hand.37

Summary

• There is a need for studies with long-term follow-up aimed at assessing the magnitude of reinnervation and its relationship with clinical efficacy and the potential long-term adverse effect of increased resting heart rate.

Motion #5: ‘Future randomized trials should be double blind’

PRO	CONTRA
• In theory, everybody agrees that future RCT should be double blind, preferably with a sham-procedure. The SIMPLICITY-3 HTN trial38 showed that a sham procedure is feasible in a condition similar to that of CNA, i.e. renal artery denervation in patients with refractory hypertension.	• There are several practical obstacles to the execution of a double-blind trial, especially if it includes a sham procedure: - Ethical issues. - Unlikely to be well accepted by most patients and clinicians. - Recruitment would only be possible in a handful of ‘selected’ patients who are carefully identified and agree, limiting the generalizability of the results. - The sham procedure cannot guarantee blindness in follow-up. - Who should be blinded? The physicians performing the ablation cannot be blind. - Many established procedures are currently accepted in clinical practice and recommended in the guidelines without having a sham study. - The study of the placebo component of CNA has lower priority than other studies, for example: (i) CNA efficacy should be compared with conventional therapies, as per guidelines, i.e. general measures and drugs in the young group and pacemakers in those >40 years; and (ii) the bradycardia-episode burden assessed by implantable loop recorder before and after CAN, is currently being investigated in an Italian-CNA observational study.
• In a recently published study, Vandenberk et al.39 performed a Research Electronic Data Capture survey that included international physicians treating patients with vasovagal syncope on their opinion about a RCT design for CNA. Performing a sham procedure in the control arm was supported by 56% of respondents, providing equipoise in RCT design.
• A study of CNA comparing sham procedure with RA approach only is ongoing (CardNMH3, NCT04755101).

Summary

• There is consensus on the need for well controlled randomized efficacy trials.

• Real double blindness and sham procedure remain controversial issues.

Motion #6: ‘Future randomized trials should be controlled against the current standard of care of cardioinhibitory reflex syncope, i.e. cardiac pacing’

PRO	CONTRA
• CNA should be compared with the current standard of care. The definition of the comparator is crucial. According to ESC guidelines23 and the 2021 ESC guidelines on cardiac pacing, cardiac pacing is established therapy ‘in patients aged >40 years with severe, unpredictable, recurrent syncope who have: spontaneous documented symptomatic asystolic pause/s >3 s or asymptomatic pause/s >6 s due to sinus arrest or AV block; or cardioinhibitory carotid sinus syndrome; or asystolic syncope during tilt testing’ (Class I Level of Evidence A). The age category 40–60 years is poorly represented in pacing trials. Pacing has been mostly used in patients aged >60 years.	• Since CNA is controversial in patients >60-year-old and, conversely, cardiac pacing is controversial in patients <40 years, perhaps the age range of 40–60 years is the best age criteria for inclusion in a controlled trial of CNA vs. cardiac pacing.
• There is no current standard of care for severe, unpredictable, recurrent cardioinhibitory reflex syncope in patients aged <40 years. In the general population, education and life-style measures (Class I), modification or discontinuation of hypotensive drug regimen (Class IIa), isometric counterpressure manoeuvres (Class IIa), fludrocortisone (Class IIb) and midodrine (Class IIb) are evidence-based effective therapies.21	• The best way to identify patients and to demonstrate therapeutic efficacy remains uncertain. However, when considering the use of atrial fibrillation ablation techniques, the lack of well controlled clinical trials has not held back the use of atrial fibrillation ablation since the inception of this technique.
• Nevertheless, although pacing may not be indicated in patients under <40 years, based on guidelines, there are no data from any controlled clinical trials that clearly support the use of CNA at this point. Therefore, it would seem reasonable to consider a trial comparing pacing to CNA even for younger individuals.

Summary

• CNA should be compared in a randomized trial with the best treatment currently available, which consists of combined behavioural, non-pharmacological and pharmacological measures in young patients and pacemakers in older patients.

• In young patients, CNA should be proven to be superior to conventional treatment. In older patients, CNA should be proven to be non-inferior to pacemaker therapy in relieving symptoms and (eventually) superior to cardiac pacing for the combined endpoint of symptoms, complications, quality of life and costs.

• A randomized trial of CNA vs. pacemaker in patients <40-year-old is controversial.

Additional controversial issues (gap in knowledge)

Ablation technique

The optimal ablation technique is not yet established. A variety of methods (listed in a table in the Supplementary material online, Appendix) for the electroanatomical localization and targeting of the ganglionated plexi has been developed, but no one has been proven to be superior to the other: fluoroscopy and spectral analysis of atrial electrograms,^1,40 3-dimensional endocardial electroanatomical mapping systems^10,22 and imaging by computed tomography,^8,41 assessment of high amplitude fractionated electrograms,²² and endocardial high frequency stimulation of the atria.^10,42,43 Finally, extracardiac vagus nerve stimulation³⁶ has been used to assess the periprocedural effect of CNA. The post-ablation absence of provoked vagal responses of sinus and AV nodes is a highly sensitive and specific measure of their denervation.¹² However, the utility of extracardiac vagus nerve stimulation for guiding the CNA has not been formally investigated in CNA studies with clinical endpoints.

Other methods for the intra-procedural assessment of efficacious vagal denervation are controversial: the cut-off values of persistent increase in sinus rate,⁴⁴ baseline and after atropine infusion, persistent increase in Wenckebach point during high-rate atrial stimulation, elimination of responses to high frequency endocardial stimulation, elimination of atrial fractionated potentials is largely empirical and not validated by trials. Ad hoc comparison studies are warranted.

Periprocedural complications and long-term risk of vagal denervation

The periprocedural complications of CNA are probably underreported. Injuries of the sinoatrial node and of the AV node (and their arteries) have been seldom reported.⁴⁵ Phrenic nerve palsy after the RA ablation and coronary artery damage (specifically, occlusion of the distal circumflex artery after the proximal coronary sinus ablation) were observed but are rare (D. Wichterle, personal communication). This issue is particularly important in the setting of a benign disease such as reflex syncope and other vagally-dependent syndromes. In absence of solid data from large registries, it is reasonable to assume the same risk as for the other ablation procedures performed in the RA and LA. For example, the complication rate of ablation of focal atrial tachycardia and cavotricuspid-dependent atrial flutter has been reported to be 1.4–2%⁴⁶ and that of atrial fibrillation has been reported to be 2–3%.⁴⁷

There is a potential for long-term harm by ablating the vagus and increasing the heart rate. The ideal heart rate at rest may vary from individual to individual but, when mean heart rates exceed 70 b.p.m., concern is raised particularly when an underlying disease is present.³⁷

Higher resting heart rate and potentially faster rates with exercise could affect functionality, symptoms and even long-term prognosis. Only long-term follow-up studies will provide an answer to this issue.

Extrinsic (functional) sinus bradycardia and atrioventricular block

While in patients with cardioinhibitory reflex syncope bradycardia is restricted to the time of occurrence of syncopal episodes and the rhythm is normal outside these episodes, some other patients have frequent episodes of symptomatic sinus bradycardia or AV block (usually second-degree Type 1 or 2:1 AV block) even in absence of syncope. There is absence of structural heart disease. (Pre)syncope is rare. However, these patients suffer from symptoms of fatigue, irritability, lassitude, inability to concentrate, lack of interest, forgetfulness, and dizziness (due to the consequent reduction in cerebral and peripheral perfusion) that sometime alter greatly the quality of life of the patients. Nevertheless, a cause-effect clear correlation between symptom and bradycardia is often difficult to prove. Its extrinsic (likely vagally-dependent) nature is suspected because bradycardia and AV block are intermittent. Also, symptoms occur at rest and during sleep, and disappear during exercise. Conversely, chronotropic incompetence or persistence of AV block during physical activity argue in favour of an intrinsic disturbance of sinoatrial function and AV conduction properties.^22,48 A blunted response of heart rate to atropine injection (0.02 mg/kg) or, better, to autonomic blockade with propranolol (0.2 mg/kg intravenously) followed by atropine (0.04 mg/kg) may be useful to identify patients with chronotropic incompetence and differentiate intrinsic from extrinsic forms of sinus node dysfunction.^48,49 However, in the elderly, a subtle decrease of the intrinsic properties of the sinoatrial node automaticity and of the AV node conduction due to a degenerative aging process is difficult to exclude.⁵⁰ The patients with extrinsic (functional) sinus bradycardia or AV block have been included in some observational studies of CNA; improvement of sinus function and AV conduction has been observed during the medium-term follow-up.^1,8,10,51 There may be risk in choosing CNA as a first-line procedure in these older individuals who have had paroxysmal episodes of AV block or with sinus bradycardia. Qin et al.²² studied the long-term effects of CNA on heart rate and quality of life effect in patients with symptomatic sinus bradycardia. The increase in heart rate was small in patients >50 years of age and there was no improvement in quality of life. Thus, the clinical benefit is uncertain. The fact that CNA may be effective does not mean that it is always necessary. A randomized trial specifically limited to extrinsic sinus bradycardia and AV block is warranted before CNA can be applied in clinical practice to this subset of patients.

How to perform follow-up after cardioneuroablation

The only clinical reason for performing CNA in patients with reflex syncope is to avoid syncopal recurrences. Thus, syncopal recurrence, assessed either as time to first recurrence or decrease in the syncope burden, should be the primary endpoint of the trials and the unique goal of clinical practice to assess the efficacy of the therapy. The possibility of late regression of the effect of CNA due to vagal reinnervation should be assessed with a follow-up period of several years, much longer than in the present studies which is limited to 1–3 years. Potential side effects, e.g. symptomatic tachycardia, and changes in overall quality of life should also be carefully assessed during the follow-up.

The success of the therapy cannot be inferred by surrogate endpoints currently used in the literature such as average heart rate on prolonged ECG monitoring, parameters of heart rate variability, or repeated tilt tests, etc. These observations, however, are useful for to understand changes in the cardiovascular physiology induced by ablation.

Supplementary material

Supplementary material is available at Europace online.