Cardioneuroablation (CNA) was introduced more than 20 years ago by Pachon et al., but has recently become a new tool for the treatment of asystolic reflex syncope.1 The procedure consists of endocardial radiofrequency (RF) ablation of the areas of presumed localisation of epicardial and mid-myocardial ganglionated plexi (GPs), which contain both adrenergic fibres and parasympathetic endings of the vagal nerve. Because the latter are closer to the endocardial surface and regenerate poorly compared with adrenergic bodies, CNA causes predominantly parasympathetic denervation of the sinus node (SN), atrioventricular (AV) node (AVN) and cardiac tissue. By doing this, CNA effectively diminishes vagal input into the heart and can effectively prevent recurrences of asystolic reflex syncope – a syndrome in which hyper vagotonia plays a significant causative role.
Efficacy of Cardioneuroablation
CNA for reflex syncope was introduced into clinical practice many years ago by Pachon et al.1,2 Subsequently, several electrophysiology (EP) teams adopted and further developed this technique, leading to the increasingly widespread use of CNA in recent years.3–16 From the beginning, the efficacy of CNA looked excellent, usually exceeding 90% during 1-year follow-up.3–16 The procedure is offered to patients after failure of all non-interventional conventional therapies, including the avoidance of triggers.17 Chronic pharmacotherapy is usually difficult to continue, especially in young people, because of relatively frequent drug-related adverse effects and the patient’s unwillingness to resolve their problems.
The efficacy of CNA has been summarised in meta-analyses and the recent European Heart Rhythm Association (EHRA) statement.17–19 The first meta-analysis, published in 2022, included 465 patients from 14 reports.18 During a mean follow-up of 2 years, the overall efficacy of CNA was 91.9% and depended on the approach used. In patients in whom a bi-atrial or left-atrial (LA) approach was used, efficacy was significantly higher compared with a right-atrial (RA) approach only (92.7% and 94% versus 81.5%, respectively; p<0.0001). The second, more recent meta-analysis included 27 observational studies and one randomised controlled trial capturing 1,153 patients with a median follow-up of 21.4 months.19 The overall weighted rate of syncope recurrence after CNA was only 5.94%. In the 2024 EHRA statement, the results of 28 studies were analysed; CNA efficacy ranged from 73.2% to 100%, with the majority of studies claiming a >90% success rate.17 However, one very recent study showed a higher recurrence rate of 33.8%, which may suggest that real-life CNA efficacy might be slightly lower than previously reported. Similar findings were reported by Choubdar et al. who, using extracardiac vagal stimulation (ECVS), achieved a 77% efficacy rate during a mean follow-up close to 1 year.20,21
It must be remembered that true CNA efficacy should be evaluated on top of a possible placebo-like effect of CNA, which may be important, especially in those without an objective measure of vagal denervation, such as heart rate variability (HRV) depression. Another consideration is that spontaneous remission rates in patients with reflex syncope could be as high as 50% during 1- to 2-year follow-up.22 And, if one considers the adverse effects associated with CNA such as symptomatic inappropriate sinus tachycardia (IST) occurring – according to some studies – in up to 7% of patients, the real net benefit of CNA could be reduced to 20–30%.23 However, not all studies report such a high risk of post-CNA IST, showing that sinus rhythm usually slows down 6–12 months after the procedure.8,9,11,20 The long-term risk of attenuation of the parasympathetic protection of the heart should also be taken into account. However, so far, no data indicating such a risk have been published apart from an animal study, which may suggest increased susceptibility to ventricular arrhythmias 6 weeks following CNA in the presence of acute coronary artery occlusion.24
In summary, it seems that approximately 80% (between 70% and 100%) of patients with reflex syncope do not have recurrence of symptoms during 12–24 months of follow-up. Reported CNA efficacy varies between the studies for various reasons, such as differences in experience, the number of patients, duration of follow-up or the techniques used to perform CNA. However, there are several factors that may influence CNA efficacy, which are discussed in detail in the next paragraphs and summarised in Table 1 and Figure 1.
Factors Associated with the Outcome of Cardioneuroablation
Age
Age may be an important predictor of CNA efficacy. With ageing, the organic component of impaired SN node automaticity or AVN conduction becomes more important and – even in patients with a very good response to atropine (>30% increase in sinus rate) – it may still diminish the favourable effects of CNA. In addition, the hypotensive phenotype of reflex syncope is becoming more frequent in older people, and this is a type of reflex in which CNA is probably not effective.
There are various age cut-off values suggested in official documents that should be used when making a choice between non-invasive and invasive treatment, which includes permanent pacing or CNA.17,25,26 According to the European Society of Cardiology guidelines on pacing, pacemaker implantation is recommended in those patients with reflex syncope who are aged >40 years.25 These guidelines do not mention at all the possibility of CNA usage. In the recent EHRA statement, the age cut-off for pacemaker implantation was set higher, at 60 years.17 Another important review suggested that in patients <40 years CNA is preferred, in those aged >60 years permanent pacing is advocated; in those aged 40–60 years an individualised approach should be used.26
The results of studies evaluating age as a risk factor for ineffective CNA are not homogeneous. Secondary analysis of the ELEGANCE study showed that during a median follow-up of 8 months there were no significant differences in the rate of patients with syncope recurrence between various age-groups (<40, 40–60 and >60 years).27 However, the Kaplan–Meier curves in fact showed major differences in CNA efficacy, ranging from 100% in those aged <40 years and 90% in those aged 41–60 years, to just below 50% in those aged >60 years.27 So in fact, this study suggested visible differences in outcomes between various age-groups but they were non-significant, probably because of the relatively small sample sizes. In addition, older patients had no heart rate acceleration during follow-up Holter ECG recordings whereas younger subjects had typical increase in heart rate, which is due to the effective vagal denervation and is associated with successful CNA.
Qin et al. showed that quality of life (QoL), measured using the SF36 questionnaire, significantly improved following CNA in patients aged <50 years, whereas it remained unchanged in those aged >50 years.13 The PIRECNA multicentre registry, which included patients with syncope undergoing CNA due to functional AV block (AVB), showed that younger people benefited more from CNA than those who were older.28 In that analysis, patients with syncope recurrence were significantly older than those without recurrence (mean 48 years [range 33–62] versus 34 years [range 27–42]; p=0.012).
Data from a study conducted at the author’s centre, which included 248 patients, showed that age matters: patients aged >60 years had higher rates of syncope recurrences and more frequently needed a pacemaker implantation after failed CNA than their younger counterparts.29 However, older age did not occur as an independent predictor of CNA failure and – in fact – the mean age of patients with recurrences was not significantly different from that of those without syncope recurrence.
Other studies did not confirm the effects of age on the efficacy of CNA. Minguiro-Carazo et al., who specifically addressed the effects of age on the outcome, did not find significant differences between subjects aged <50 versus >50 years during a median 17 months of follow-up.30 Again, the number of patients was small (17 and 33 patients, respectively). In a study by Gopinathannair et al. there was no significant association between age (HR 1.01; 95% CI [0.96–1.06]; p=0.655) and the primary outcome in univariable analyses.31 In addition, Aksu et al. failed to show the relationship between age and CNA efficacy.32
Another recent study included larger group of patients (n=77) and surprisingly showed that the recurrence rate was significantly higher in those aged ≤50 versus >50 years (48.7% versus 18.4%; p=0.032).20 The authors speculated that parasympathetic reinnervation leading to syncope recurrence may play a more important role in younger versus older patients. This is the only study to suggest older age as a good prognostic parameter following CNA; however, it was performed without the use of ECVS, thus, the real degree of vagal denervation achieved in this population is not known.
In summary, it seems that age may play a role in CNA efficacy, with patients aged >60 years having worse outcomes, although data in the literature are not consistent. The differences between studies may be because of small sample sizes in many, along with differences in patient characteristics. In addition, some older people might have been offered CNA knowing that the chance for successful procedure was not very high, but nevertheless opted for CNA as a procedure performed before scheduled pacemaker insertion with the hope that the latter procedure would not be needed. Furthermore, it is much easier to make the decision to implant a pacemaker in older rather than younger people. All these data do not mean that people aged >60 years should not be offered CNA at all. It depends on individual patient preferences, operator experience and costs. Older patients may also benefit from CNA. Simply, they should be informed at the time of consenting for the procedure that the risk of syncope recurrence may be higher than that in younger people.
Sex
Sex may also play a role as a risk factor for syncope recurrences. Firstly, data from studies on syncope suggest that young and middle-aged women are more affected by reflex syncope than males at a similar age. In one report, the proportion of females was significantly higher than males (58% versus 42%) among a total of 11,577 subjects with vasovagal syncope.33 Another study showed that women were younger at the time of first syncope episode and had a significantly lower systolic blood pressure compared with males (117 mmHg versus 124 mmHg; p<0.001).34 In addition, women were more likely to experience recurrent syncope.34 There are convincing data that women have lower baseline values and lower increments in systolic blood pressure compared with males until the age of 40 years, which predisposes to syncope. Hormonal changes in females, which promote vasodilatation, excessive splanchnic pooling and decreased plasma volume, may also predispose to fainting. Lastly, there are important anatomical differences between sexes. Women have not only smaller total body size but also smaller organs, including the heart and vascular beds, as well as different susceptibility to fluid pooling in the gut and adipose tissue.35
Genetic factors may also play a role. A few studies have suggested that in some cases, there is a clear family history of reflex syncope and there is evidence of sex-specific gene penetrance.36 It has also been documented that there is a sex-dependent relationship between alleles of serotonin signalling and reflex syncope, supporting the role of serotonin in the pathophysiology of vasovagal syncope.37
Of note, none of the reports suggested male sex as a risk factor. For example, Barrio-Lopez et al. showed that women had a significantly higher risk of syncope recurrence compared with men (42.8% versus 22.8%, respectively; HR; 3.3; 95% CI [1.2–8.8]; p<0.016) after adjusting for confounding factors.20 Others have also found female sex as an independent predictor of syncope recurrence.32
However, not all studies on syncope confirmed sex-dependence of fainting or syncope recurrences. For example, in the seminal paper on spontaneous syncope recurrence rate, Pournazari et al. did not find such a relationship.22 Other recent studies also do not confirm the association between sex and CNA efficacy.27,38
In summary, female sex may be associated with higher syncope recurrence rates because of lower systolic blood pressure, genetic and anatomical factors. Because CNA – by preventing asystolic reflex syncope – often unveils the presence of pre-existing vasodepression, lower systolic blood pressure in females may also play a role in syncope recurrence.
Type of Vasovagal Reflex
The cardioinhibitory type of reflex syncope is associated with better outcomes after CNA compared with other types of vasovagal reflex. This is expected because CNA aims to counteract vagally mediated pause rather than sympathetic-dependent vasodilatation. Several studies have shown that the best outcome after CNA is seen in the cardioinhibitory pattern; intermediate results can be obtained in the mixed form of reflex syncope whereas the vaso-depressive type is associated with the poorest outcomes.20,38,39 One recent study suggested a 77% efficacy of CNA in this type of vasovagal syncope.40
According to the recent scientific statement, the duration of cardiac pause should exceed 3 seconds during tilt-induced syncope or >3 seconds during symptomatic and >6 seconds during asymptomatic episodes when recorded by implantable loop recorder (ILR).17 Probably, the longer the pause, the better the candidate for CNA. In addition, cardioinhibition recorded by ILR during spontaneous syncope should more accurately identify a proper candidate for CNA than tilt-testing because the tilt-induced type of reflex response does not necessarily equal that occurring during everyday life.41 Furthermore, selection criteria for patients included into a prospective study were very strict and the mean asystole duration was 17 seconds.15 In clinical practice, when a wider range of patients undergoes CNA (not only those with asystolic reflex syncope but also very symptomatic presyncope or symptomatic functional bradycardia), the mean cardiac pause is usually shorter but always should exceed 3 or 6 seconds in patients with syncope.17,23
A few studies have suggested that CNA is effective in patients with vasodepressor response. However, the main problem with these studies is that they are small, retrospective and with patients selected for CNA based on the tilt-table testing results that showed vasodepression. However, the type of reflex reaction during tilt-testing (especially after nitroglycerin provocation) may not be the same as during everyday life and it is possible that some of these patients had also cardioinhibition during spontaneous events. In such circumstances, CNA might have appeared effective because it prevented reflex asystole. There are also some other possible explanations why CNA may work in vasodepression, for example increasing heart rate, which may improve cardiac output and counteract hypotension. However, CNA generally prevents reflex cardioinhibition rather than vasodepression because it attenuates vagal effects on the heart and has nothing or little to do with peripheral vasodilatation mediated by the adrenergic part of the autonomic nervous system.
History of Syncope
It has been shown that the lifetime number of syncope episodes is associated with outcomes. Tu et al. showed that disease duration was shorter in patients with non-recurrence (mean 2 years [range 1–5] than in those with recurrence (mean 10 years [range 3–22]).39 Barbic et al. found that a lifetime number of three or more syncope episodes identified patients with symptom recurrence.42 Aksu et al. reported that the pre-enrolment syncope burden was associated with the risk of syncope recurrence during follow-up (HR 1.25; 95% CI [1.04–1.51]; p=0.015).32 In our cohort of consecutive 209 patients with reflex syncope, the median total number of syncopal episodes in the past as well as during 1 year preceding CNA was significantly higher in non-responders than in responders (lifetime number: median 16.5 [range 10–50] versus median 5 [range 3–20]; p=0.002 and 1-year number: median 6.5 [range 4–19] versus median 2 [range 2–9]; p=0.002, respectively).23 However, not all studies showed such a relationship. For example, Gopinathannair et al. failed to demonstrate significant association between syncope frequency in the past year (HR 1.10; 95% CI [0.97–1.25]; p=0.122) and the primary outcome.31
A long history and high frequency of syncope episodes as risk factors for ineffective CNA may be explained in several ways. It may be because, if syncope is resistant to treatment for many years, the underlying mechanisms may be complex and less amenable to CNA. Furthermore, a low number of pre-procedural syncopal episodes in patients with a short history of fainting may result in a false-positive assessment of CNA efficacy (for example, only one or two episodes of syncope before CNA and no syncope recurrence during relatively short follow-up after procedure). It is probable that using syncope burden rather than time to first syncope recurrence as the measure of CNA efficacy is a better way to assess the outcome and usually also confirms effective ablation in patients with numerous syncopal episodes in the past in whom syncope burden is very significantly reduced.29
Syncope can occur in clusters, which represents another problem with using the number of syncopal episodes as a parameter for identifying suitable candidates for CAN and assessing procedural efficacy. To account for this phenomenon, the mean or median number of syncopal episodes during the whole life as well as a certain period before CNA (usually 1 year) is considered when discussing with the patient indications for CNA. The number of syncopal episodes required to select patients for CNA has not been universally defined and differs from patient to patient. Some patients have multiple episodes of variable intensity, whereas others have a history of only a few syncopal episodes but those are associated with injury and/or prevent them from continuing occupational or leisure activity. These patients also seek help and choose CNA as a possible effective therapy.
ECG Phenotype: Sinus Pause versus Atrio-ventricular Block
It has been demonstrated that CNA is more effective when reflex syncope is the result of sinus pause rather than AVB.43,44 The main reason for this is that effective CNA in AVB usually requires more complex bi-atrial approach and the targeting of both primarily inferior and – in some patients – also superior GPs from both atria and also from the coronary sinus. Thus, procedures performed only in the RA may be less effective when functional AVB is a cause of syncope. In a randomised prospective study, it was shown that all crossovers from the RA to the LA were due to the persistence of vagally-induced AVB, despite extensive ablation in the RA.43
Concomitant Cardiovascular Disorders
The presence of concomitant cardiovascular disorders is associated with lower CNA efficacy. Usually, these patients are older, and cardiac involvement may result in a higher degree of impairment of intrinsic SN automaticity or AVN conduction due to fibrosis, atherosclerosis, ischaemia or other factors. In the PIRECNA registry, three covariables (history of AF, hypertension, and coronary artery disease) occurred in the regression analysis as most predictive of outcomes.28 The model performed with an area under the curve of 0.74 (95% CI [0.61–0.87]) to predict syncope recurrence during follow-up. Data from a study conducted at the author’s centre also found hypertension to be an independent factor significantly associated with syncope recurrence.29
Hypertension may influence the outcome in several ways. Firstly, cardiac hypertrophy and concomitant diseases – often associated with hypertension – may cause intrinsic impairment of SN automaticity and AV conduction. Secondly, hypotensive therapy may predispose to hypotension and syncope. Lower blood pressure has been shown to be associated with a higher recurrence rate of syncope.34,35 The mean systolic blood pressure in responders was 124 ± 14 mmHg, whereas in non-responders it was 116 ± 12 mmHg (p=0.006). Of note, one study did not find a relationship between cardiomyopathy and syncope recurrence.20
Previous Treatment
It may be speculated that the more previous attempts to prevent syncope recurrences have been performed, the lower the chance for CNA to be effective. However, in all published studies, CNA was offered after all non-pharmacological and pharmacological interventions were found to be ineffective. Thus, the prognostic role of the history of previous ineffective treatment has not been established.
Atropine Test Before Cardioneuroablation
Theoretically, the greater the magnitude of sinus rate acceleration following the atropine injection, the greater the role of the vagal nerve in syncope mechanism may be anticipated. Because CNA decreases vagal activity, subjects with good responses to atropine should also respond to CNA. However, there are no reports suggesting that the degree of sinus rate increase following pre-procedural atropine test may have prognostic value. This may be in part because only patients with good response to atropine (sinus rate increase >25% or >30%) are selected for CNA, which causes preselection, and those with lower baseline vagal activity are not available for analysis. In addition, atropine test is not a perfect tool to assess sympatho-parasympathetic balance and other tests – such as full autonomic blockade with atropine and propranolol – might be more appropriate in this setting.
Methods for the Identification of Presumed Areas of Ganglionated Plexi
The methods currently used for the identification of GP areas – the target for CNA – are not accurate because none of them directly identify the GP location. These methods include anatomical GP-related area identification using electro-anatomical mapping systems, intracardiac echocardiography or CT; responses to high-frequency stimulation; or identification of endocardial areas of fractionated atrial electrograms. Data do not consistently show any significant advantage of one over another. Aksu et al. showed that the outcome of patients undergoing electro-anatomical mapping-guided CNA versus those having combined approach with high-frequency stimulation and spectral analysis was similar.45 Sun et al. found no differences between high-frequency stimulation versus anatomically guided CNA.9 A meta-analysis by Vanderberg et al. showed no difference in outcomes after CNA based on various GP-mapping techniques.18 Xu et al. showed, in 108 patients, that both high-frequency stimulation mapping and anatomical ablation can effectively improve the symptoms of patients with vasovagal syncope.46 However, the most recent meta-analysis showed a higher prevalence of syncope recurrence in the electro-anatomical mapping subgroup (overall weighted rate of syncope recurrence 6.21%; 95% CI [2.93–10.28%]; I2 5.0%) compared with spectral analysis of fragmented atrial potentials or high frequency stimulation.19
In summary, there is no consensus on which technique for the identification of GP areas is the best in terms of CNA efficacy. It is difficult and somewhat artificial to compare different approaches, because in clinical practice they are often used concomitantly. For example, the signal-based and high-frequency stimulation-based approaches always use anatomical data coming from the fluoroscopy, electro-anatomical mapping and/or intracardiac echocardiography. In addition, many of these techniques have significant limitations. For example, high frequency stimulation HFS provides a very localised assessment, is fundamentally a trial-and-error technique, is ineffective in identifying deeply localised GPs or those with high activation thresholds and frequently induces AF. Analysis of fragmented atrial potentials – although very promising and used by several operators – is not perfect and has unknown reproducibility and accuracy.2,45
It must be kept in mind that neither technique is capable of directly imaging the GPs. Even new CT systems for epicardial fat imaging are not perfect and there is no strong correlation between CT and other methods for identifying GP areas.47 In addition, it has been assumed that the higher the density of the epicardial fat, the more the GPs are localised in this area, which is a speculation rather than fact. Furthermore, anatomical and histological data suggest that epicardial fat is not the only place where GPs are located.48 Thus, we need more accurate methods to guide CNA and it is not surprising that none of the currently used tools is clearly the best to predict CNA efficacy.
Bi-atrial versus Left-atrial versus Right-atrial Approaches
The majority of studies confirmed that a bi-atrial approach was associated with better outcomes than the RA-only approach, especially when AVB is the cause of syncope.43,49 Francia et al. showed that bi-atrial approach was more effective than LA- or RA-only approaches and that starting CNA from the LA was associated with the greatest attenuation of vagal nerve activity. Also, both meta-analyses confirmed these findings.50 Recently, Tu et al. showed in a randomised trial that CNA performed in the LA only was equally effective as the bi-atrial approach; however, the latter was associated with longer procedure duration and fluoroscopy time as well as higher number of RF applications.51 Surprisingly, the PIRECNA registry, which included only patients with functional AVB, did not show such a relationship.28 Similarly, Barrio-Lopez et al. did not confirm a difference between single versus bi-atrial approaches.20
These data suggest that, although bi-atrial approach may be more effective and particularly required in patients with AVB or redo procedures, the RA-only approach may be regarded as the first, less invasive step in many patients – especially those with sinus pauses – because it is quite effective in experienced hands, with efficacy exceeding 80%.5,6,11,12
Extent of Cardioneuroablation
It may be speculated that the greater the area ablated and the higher the number of RF applications and cumulative energy used, the greater the efficacy of CNA because more GPs are ablated. Barrio-Lopez et al. found that the number of RF applications was significantly higher in responders than in non-responders (>30 versus >30, respectively; HR 0.4; 95% CI [0.2–0.9]; p=0.042), whereas there was no difference in the number of GPs treated.20 One meta-analysis failed to show that the number of GPs treated was associated with outcomes.18 In the PIRECNA registry the number of ablation points was similar in responders versus non-responders for all GP areas, with right superior GP showing a trend towards higher efficacy with more RF applications performed (mean 8 versus 10; p=0.054).28
The extent of CNA may be associated with procedural efficacy in different ways. On the one hand, it may be speculated that in patients in whom all GPs areas were ablated and a high number of RF applications and/or high cumulative energy was used, efficacy would be enhanced because greater damage to the GPs was done. On the other hand, a high number of RF applications and large ablated area may suggest that there was a problem with achieving full vagal denervation because of factors such as difficult anatomy or unusual distribution of GPs, meaning that the procedure might have been less effective.
The Intra-procedural Use of Extracardiac Vagal Stimulation
The method was developed by Professor Pachon and his team several years ago.52 Currently, the main clinical application of the ECVS is the acute intraprocedural assessment of the effects of CNA on the SN and AVN. At baseline, ECVS causes sinus arrest and AVN conduction block, whereas after CNA, these effects disappear, confirming complete CNA-induced vagal denervation. ECVS is the only currently available reliable tool to acutely evaluate CNA efficacy. Other methods, which include standard EP measurements of SN recovery time and AVN conduction, can also be used for this purpose but are probably less accurate and more prone to be influenced by other factors such as actual sympathetic and parasympathetic balance or medication used during general anaesthesia, muscle relaxation or mild sedation.
The question of whether the use of ECVS increases CNA efficacy remains open. Data in literature on the use of ECVS versus conventional EP parameters as an endpoint for CNA are scarce. There is no randomised prospective study to address this issue; only retrospective observational studies are available. There are two published studies and one abstract dealing with the use of ECVS.28,53,54 All reports showed a tendency towards better outcomes when ECVS was used. When combined, the use of ECVS seems to be associated with better outcome (91% versus 82% of patients free from syncope recurrence), but with a p-value of a borderline significance.
Published data suggest that ECVS is particularly important in patients with reflex syncope due to functional AVB in whom ablation is often more extensive, requires bi-atrial approach, including inferior septal GPs, coronary sinus and areas close to the left pulmonary veins, compared with patients with sinus pauses only.43,49
Type of Anaesthesia
The type of anaesthesia can influence the outcome because medications used for general anaesthesia and muscle relaxation change the sympatho-parasympathetic balance and make assessment of standard EP parameters less reliable. Even conscious sedation with midazolam versus deep sedation with midazolam and propofol differ in their effects on heart rate.55 In such situations, ECVS seems to be the best tool to assess intra-procedurally vagal denervation. Data in the literature are not consistent, with the PIRECNA registry showing no difference in outcomes and the study by Barrio-Lopez et al. suggesting more effective CNA when performed under general anaesthesia.20,28
Sinus Rate Increase after Cardioneuroablation
Sinus rate increase after CNA is a well-known phenomenon, confirmed in all studies dealing with this technique.1–21 Whether the extent of heart rate acceleration is associated with the outcome is less clear. The ELEGANCE study showed that sinus rate increase was significant in patients aged <60 years but small and non-significant in older patients.27 It has been shown that the mean heart rate, calculated from the 30-minute ECG recordings performed 1, 3 and 6 months after CNA, was significantly faster in responders than in non-responders at each of these time-points. However, Barrio-Lopez et al. failed to confirm these findings.20 In their cohort of 77 patients, heart rate increase after CNA did not predict outcomes. The usual pattern of heart rate behaviour after the procedure is a steep increase immediately after CNA, with gradual slowing during the following year, but usually remaining faster than the pre-procedural values.28 It has been shown that in those in whom sinus rate significantly slowed or returned to normal, suggesting prompt early parasympathetic reinnervation, the risk of syncope recurrence was increased.56,57
To sum up, the prognostic role of the magnitude of sinus rate increase immediately after CNA is debatable; however, return to the pre-procedural values at any stage of follow-up, detected by ECG ambulatory monitoring or implantable loop recorder, predicts syncope recurrence.
Heart Rate Variability Decrease after Cardioneuroablation
The data regarding HRV are similar to those for sinus rhythm behaviour after CNA. There is a prompt decrease of the standard-time domain HRV indices (standard deviation of NN intervals is most frequently assessed) after the procedure, with gradual albeit not complete return towards baseline values by 1 year of follow-up and no further significant reinnervation after 2 years of follow-up.7,9,10,12,28,44,58,59 The fluctuation in frequency-domain HRV parameters is similar.6
The prognostic value of the post-CNA changes in HRV has not been clearly established. Calo et al., who used the RA approach, showed that patients with effective CNA had greater reduction of the frequency-domain HRV parameters compared with those who experienced syncope recurrence.6 In another report, return of the HRV indices to the pre-procedural values during follow-up heralded syncope recurrence.57
Deceleration Capacity
Another non-invasive parameter that reflects parasympathetic drive to the heart and may identify patients with vasovagal syncope is deceleration capacity. It has been shown that patients with recurrent syncope/pre-syncope after CNA have lower baseline deceleration capacity value than those without symptom recurrences (8.9 ± 3.2 ms versus 11.3 ± 3.7 ms; p<0.001).38 This implies that the greater the baseline vagal activity, the target for CNA, the better the procedural results. Accordingly, some centres use deceleration capacity to select patients for CNA.38,60,61 Whether CNA-induced changes in deceleration capacity have prognostic role remains to be established; one case report suggests such a possibility.61
Post-Cardioneuroablation Atropine Test to Measure Residual Vagal Activity
Theoretically, the greater the residual sinus rate response to atropine administered at the end of the procedure, the less vagal denervation is achieved, which may be associated with poorer outcomes. Barrio-Lopez et al. failed to show that the magnitude of post-CNA response to atropine (≤5 BPM versus >5 BPM) predicted the outcome.20 A study conducted by the authors also did not find such a relationship when the patients were divided according to ≤10% versus >10% increase in sinus rate after post-procedural atropine.29 In the ELEGANCE study, Franchia et al. found a high negative predictive value for the post-procedural atropine test (90.7%).27 However, positive predictive accuracy was low, achieving 20%. In a seminal study, Pachon et al. showed that post-atropine percentage increase in heart rate was significantly lower than at baseline (4% versus 79.4%, respectively; p<0.01), confirming effective vagal denervation.2 However, no specific comment was made on whether this change was different between 40 patients with effective CNA versus three subjects with syncope recurrence. In summary, data in the literature on the predictive value of post-procedural atropine test are sparse. It seems that this test has a low positive predictive value in identifying future non-responders, whereas no change or minimal sinus rate acceleration after atropine accurately identifies patients with good outcome.
Centre Experience
Operator experience usually plays a significant role in the efficacy and safety of any invasive procedure. However, such a relationship has not yet been found for CNA. In the PIRECNA registry there was only a non-significant trend for better outcome in the group with ablation conducted by an experienced operator [HR 0.52; 95% CI [0.19–1.41]; p=0.19], whereas in the multi-centre study by Aksu et al. there was virtually no difference between procedure-naïve operators and experienced ones.28,62
These results may be explained by the fact that, from a technical point of view, CNA is not a difficult procedure for an operator who is experienced in performing complex ablation procedures, including ablation for AF. Procedure-naïve operators who took part in the Aksu et al. study were probably not operators inexperienced in ablation just starting their EP training, but instead mature operators performing other types of ablations.62
Follow-up Duration
It is clear that the longer the follow-up, the higher the likelihood of syncope recurrence. In one of their studies, Aksu et al. showed that median follow-up duration was significantly shorter in responders than in non-responders (20 months versus 43 months; p=0.019, respectively).32 Another study also showed such a trend.29
These results also suggest that in studies on CNA efficacy, the outcome data using fixed follow-up duration should be presented instead of only median follow-up duration values with large inter-patient and inter-group differences in the follow-up duration.
Post-Cardioneuroablation Tilt-testing
While tilt-testing before CNA is a well-established tool to identify proper candidates for the procedure – in particular those who have no ECG documentation of cardioinhibition during a spontaneous event – the role of post-CNA tilt-testing to predict outcomes is less clear. Some studies suggest that tilt-testing may be useful for this purpose. In a study that included 51 patients, eight had syncope recurrence.56 Tilt-testing performed 1 month after CNA was normal in seven and vasodepression was demonstrated in the remainder. It is therefore difficult to state that the test had prognostic value. However, tilt-testing performed 6 months after CNA was normal in three patients, whereas it provoked syncope due to vasodepression in four patients and cardioinhibition in one.56 Francia et al. showed that post-procedural tilt-testing was a valuable tool to predict future outcome, identifying responders with a 91.7% negative predictive value.27 The authors did not specify when post-CNA tilt-testing was performed – immediately after the procedure or later on during follow-up. Our experience is similar. Of 20 patients, 14 had a positive tilt-test despite no spontaneous syncope recurrence during follow-up. Of note, in all but one with abnormal test result, only vasodepression was demonstrated.63
These results suggest a high negative predictive value of the test and also the possibility for identification of those in whom early parasympathetic reinnervation occurs. In a significant proportion of patients post-CNA, tilt testing reveals vasodepression, which is almost always present in reflex syncope but is not detected before CNA due to the presence of sudden severe cardioinhibition. The problem is that routine post-CNA tilt-testing, potentiated by nitroglycerin, in all consecutive patients may not be desirable because it can quite frequently cause syncope due to vasodilatation in patients who do not have spontaneous syncope recurrences.2,63 This may alter QoL and unnecessarily raise a patient’s concerns about the efficacy of CNA. In addition, one would like to identify those who may have syncope recurrence rather than those who do well after CNA. However, the positive predictive value of tilt-testing in this setting is low.
Definition of Endpoints During Follow-up
The definition of endpoints surely would affect the perception of CNA efficacy. The most widely used endpoint – first syncope recurrence or the time to the first syncope recurrence – may not be ideal in many patients. For example, in someone who has had multiple syncopal episodes in the past, just one or two syncope recurrences during long-term follow-up usually means a significant improvement, yet it would be defined as CNA failure. This situation is similar to that in patients after ablation for AF. Currently, AF burden rather than the time to first AF recurrence is advocated as an endpoint.64 A similar approach could probably be used in patients undergoing CNA.
Another endpoint, which should be strongly advocated when assessing treatment effects in patients with benign conditions, is QoL. Several studies using various questionnaires consistently show a significant improvement in QoL measures after CNA in those without symptom recurrences.16,30,65 Thus, perhaps this may be the best parameter for the assessment of CNA efficacy?
Conclusion
Demographic and clinical predictors of favourable outcomes after CNA include younger age, male sex, cardioinhibitory type of reflex, lack of concomitant cardiovascular disorders, low number of syncopal episodes in the past and possibly others that are not yet well established (Table 1 and Figure 1). There are currently more unknowns than well-established facts in the field of CNA, as recently comprehensively summarised.66,67 It must also be acknowledged that the quality of the majority of published data on CNA is not very high. Of approximately 150 publications in PubMed, there is only one small unblinded prospective randomised study, several retrospective observational studies (of which many are underpowered) and a large number of case reports or case series. Several studies have reported the results of procedures performed during a learning phase and, in the majority of studies, ECVS was not used, making the assessment of the real degree of vagal denervation impossible.
We really need better quality data, including sham trials that are on-going (NCT04755101), but ideally studies should include patients undergoing identical procedures (including equivalent ablation) in both arms, to establish the real role of CNA in the treatment of reflex syncope.
Clinical Perspective
- Cardioneuroablation is a promising new tool for the treatment of asystolic reflex syncope.
- There are many unresolved issues concerning cardioneuroablation, including the identification of optimal candidates and predictors of success.
- Data in the literature suggest that the cardioinhibitory form of syncope, frequent and very symptomatic episodes, lack of concomitant cardiovascular disorders, younger age and perhaps male sex characterise the best candidates for cardioneuroablation.
- Several parameters, such as heart rate, heart rate variability, deceleration capacity, use of extracardiac vagal stimulation, bi-atrial approaches or post-procedural atropine testing, may help to predict the outcome after cardioneuroablation.