Catheter ablation of atrial fibrillation in patients with hypertrophic cardiomyopathy: a European observational multicentre study

Abstract

Introduction

Hypertrophic cardiomyopathy (HCM) is the most frequent monogenic cardiovascular disease and affects 1 out of every 500 individuals.¹ Atrial fibrillation (AF) is common in patients with HCM, with an estimated prevalence of 22.5%.¹ Atrial fibrillation is associated with adverse clinical outcomes and is usually poorly tolerated, often leading to heart failure symptoms and haemodynamic instability requiring prompt treatment with direct current cardioversion.¹

Catheter ablation is an established treatment for symptomatic drug-refractory AF.² The HRS/EHRA/ECAS/APHRS/SOLAECE expert consensus statement suggests that the indications for catheter and surgical AF ablation in selected patients with HCM are the same as for other clinical scenarios.² However, data are sparse regarding patient selection and long-term efficacy in this setting, with studies based on small cohorts, with short follow-up duration, and presenting sometimes contradictory results.^1–6

Risk scores such as the APPLE score have been developed to predict AF recurrences following catheter ablation⁷ but have not yet been tested in the HCM population. Furthermore, risk stratification schemes involving cardiac structural parameters (left atrial size, outflow tract gradient, and wall thickness) are available for predicting which HCM patients are likely to develop sustained ventricular arrhythmias and should be offered primary prevention implantable cardioverter-defibrillator (ICD).^8,9 Whether these risk scores identify more severe atrial arrhythmia phenotypes which are less likely to respond to catheter ablation and can be used to select patients for ablation remains to be investigated.

Mutations in over 11 genes encoding proteins of the cardiac sarcomere are linked to HCM, with more than 1400 variants described.⁸ Whether patient genotype associates with outcomes of AF ablation remains to be determined.

Cryoballon ablation has emerged as an effective alternative to radiofrequency ablation as initial procedure in AF patients,¹⁰ having a class of recommendation IIa, level of evidence B in the current guidelines of the European Society of Cardiology (ESC) for the management of AF.¹¹ Nonetheless, limited data are available on the use of the cryoballoon technique in the HCM population.

The aim of this study was to investigate in a large cohort of HCM patients: (i) the efficacy of catheter ablation for AF beyond the initial 12 months, (ii) the available risk scores, stratification schemes and genotype as potential predictors of arrhythmia relapse post-ablation, and (iii) the impact of cryoballoon vs. radiofrequency technique on the procedural outcomes.

Methods

Non-randomized, observational study in four European centres. We included all patients aged over 18 with a definite diagnosis of HCM undergoing a catheter ablation for AF between 2006 and 2019. According to the guidelines of the European Society of Cardiology (ESC),¹¹ HCM was defined as a wall thickness ≥15 mm in one or more left ventricular myocardial segments (on either echocardiogram or cardiac magnetic resonance imaging) that is not explained solely by loading conditions. All patients provided written informed consent prior to the procedure. The study complied with the Declaration of Helsinki and the research protocol was approved by the local ethics or clinical effectiveness unit committees.

Procedures were performed under sedation or general anaesthesia, according to each institution’s protocol. Venous access was obtained via the femoral vein, with use of vascular ultrasound at operator’s discretion. In the absence of patent foramen ovale, a single or dual transseptal puncture was performed under fluoroscopic guidance. Transoesophageal echocardiography was used based on operator preference. Patients received intravenous heparin to maintain an activated clotting time of 300–350 s. Pulmonary vein isolation (PVI) was the main procedural endpoint, and was performed as a first step in all procedures. If the patient was in AF at the start of the procedure and the arrhythmia organized into an atrial tachycardia this was mapped and ablated. In patients undergoing cryoballoon ablation, if the patient remained in AF after isolation of all four pulmonary veins, direct-current cardioversion to sinus rhythm was performed and no further ablation undertaken. For repeated procedures, PVI remained the main procedural endpoint and re-isolation of the pulmonary veins was mandatory in case of reconnection. Additional ablation beyond PVI was at operator discretion for both de novo and redo procedures.

Patients were initially evaluated at 3 and 12 months after the procedure, and thereafter on an annual basis. Additional visits and further testing were allowed in case of symptoms. Information collected during follow-up included a 12-lead electrocardiogram (ECG) and either a 24-h ECG Holter monitoring or cardiac electronic implantable device interrogation at each visit. The first 3 months post-procedure were considered blanking period. Recurrence was defined as any symptomatic or asymptomatic atrial arrhythmia lasting >30 s following the 3 months blanking period. Patients with relapse during the blanking period with no response to pharmacologic or electrical cardioversion were also classified as having a relapse. Very late relapse was defined as any recurrence occurring after 12 months post-ablation, according to the international consensus statement on AF ablation.²

Irrigated-tip catheters and 3D mapping technology were used for all the radiofrequency procedures, and contact-force technology was systematically adopted since 2013. In the cryoballoon group, a second-generation device (Artic Front Advance™, Medtronic, USA) was used for all but two cases. Other single-shot ablation technologies were allowed at operator discretion but used in a very limited number of cases.

The main efficacy endpoint was freedom from atrial arrhythmias following a blanking period of 3 months. Atrial fibrillation or atrial tachycardia relapse during the initial 3-month blanking period was also documented. With regard to safety, the following complications were systematically recorded: vascular complications (if requiring intervention or prolongation of admission), thromboembolism (transient ischaemic attack, stroke and/or systemic embolism during or in the first month after the procedure), phrenic nerve palsy, pericardial effusion (if causing haemodynamic instability and/or requiring pericardiocentesis or prolonged monitoring), oesophageal fistula, and procedure-related death. Other complications were reported at the discretion of the operator.

The risk of sudden cardiac death (SCD) was calculated for each patient using both the ESC HCM-risk SCD score⁸ and the stratification model proposed by the American College of Cardiology/American Heart Association (ACC/AHA) for selecting patients for primary prevention ICD.⁹

We calculated the CHA₂DS₂VASc and APPLE score [one point for age > 65 years, persistent AF, glomerular filtration rate <60 mL/min/1.73 m², left atrium (LA) diameter ≥43 mm, and left ventricular ejection fraction <50%], which had been previously validated for prediction of rhythm outcome after AF ablation.⁷

Statistical analysis

The χ² test was used for categorical and Student’s t-test for comparison of means. Levene’s test was used to check the homogeneity of variance; equivalent non-parametric tests were used when Kolmogorov–Smirnov was in favour of the absence of normal distribution. Results with P < 0.05 were regarded as significant.

Kaplan–Meier curves were traced for illustrating freedom from atrial arrhythmias, and the log-rank P test was used for assessing existing differences. Independent predictors of sinus rhythm maintenance after ablation were assessed through Cox regression (method: forward likelihood ratio, probability for stepwise 0.05). Predictive value of clinical scores for sinus rhythm maintenance after ablation was assessed using ROC curves. SPSS version 26.0 was used for descriptive and inferential statistical analysis.

Results

One hundred thirty-seven patients (mean age 55.0 ± 13.4 years, 29.1% female) underwent a total of 225 catheter ablations for AF (1.7 ± 1.0 per patient). Most patients had paroxysmal AF (57.5%) at baseline, and mean AF duration was 3.3 ± 3.1 years. Mean left atrial diameter was 47 ± 7mm. Maximal left ventricular wall thickness and left ventricular outflow tract gradient were 17 ± 4 mm and 14 ± 24 mmHg, respectively, and 46.7% of the sample had an ICD or other cardiac rhythm devices.

Based on the ESC HCM-risk SCD score, the 5-year risk of SCD was <4% in more than half of the patients, between 4% and 6% in 16.6%, and >6% in 19.7%. According to the risk stratification model of the ACC/AHA, 41% had no indication for primary prevention ICD, 33.6% should have been considered for an ICD, 17.5% might have been considered, and 8% had a definite indication for an ICD. Baseline population characteristics are reported in Table 1.

Procedural outcomes

Pulmonary vein isolation was achieved at the end of the procedure in almost all the patients (97.8%). The cryoballoon technique was utilized in 33.6% of the patients for the first procedure. Contact-force catheters were used in the 63.6% of the radiofrequency ablations.

Patients were followed up for a mean of 43.8 ± 37.0 months. Freedom from atrial arrhythmia relapse at 12 months was 48.5% after a single procedure (60.8% for paroxysmal AF vs. 31.6% for persistent AF) and 60.0% after single or multiple procedure (74.3% for paroxysmal vs. 41.0% for persistent AF).

Very-late relapses (i.e. following the initial 12-month period) were frequent, and at 24 months after the index procedure nearly all patients with persistent AF had relapsed, and only 40% of those with paroxysmal AF remained free from arrhythmia recurrence (Figure 1). At the time of the first documented arrhythmia recurrence, the underlying rhythm was AF in 69.3% of the patients and atrial tachycardia in the remaining.

Figure 1

Kaplan–Meier of arrhythmia-free survival after index procedure for paroxysmal and persistent AF. AF, atrial fibrillation.

Open in new tabDownload slide

Class I or III anti-arrhythmic drugs post-blanking was continued in 33.6% of the patients (46) post-index procedure. On a subgroup analysis, there was no significant difference in the rate of relapse post-index procedure according to continuation vs. discontinuation of anti-arrhythmic drugs (log-rank P = 0.104; Supplementary material online, Figure S3).

Success rate of cryoballoon ablation was comparable with radiofrequency ablation in both patients with paroxysmal and persistent AF treated with contact-force catheters (log-rank P = 0.145 and 0.885, respectively; Figure 2B).

Figure 2

Trend of periprocedural complications over time (A) and success rate using different technologies. (B). CF, contact force sensing ablation catheters; Cryo, cryoballoon ablation; RF, radiofrequency.

Open in new tabDownload slide

Assessment of independent predictors of AF or atrial arrhythmia relapse is illustrated in Table 2. On multivariable Cox regression, LA area and left ventricular apical aneurysm were independent predictors of relapse post-ablation. Neither of the SCD risk stratification schemes was identified as a predictor of procedural outcomes.

The rate of peri-procedural complications was 9.3% (Table 3). There was a significant reduction in the rate of complications over time (P = 0.002), Figure 2A. There was no significant difference in the overall rate of complications between radiofrequency vs. cryoballoon ablation (4.4% vs. 6.7%, respectively; P = 0.58).

Ablation strategy during the first procedure

The first procedure was successfully completed in 132 of 137 patients (96.3%), with ablation being deferred in 5 due to onset of complications. Of the 132 acutely successful first-time ablations, a ‘PVI-only approach’ was performed in 54.5% (72) and ‘PVI plus substrate modification’ in the remaining 45.5% (60). Substrate modification overall included ablation of left atrial lines (mitral isthmus and/or roof line and/or posterior wall isolation) in 38 patients (63.3%), cavo-tricuspid isthmus line in 30 (50%), and CFAEs in 23 (38.3%). In addition, three patients (5%) underwent ablation of a microreentrant or focal left atrial tachycardia beyond PVI.

There were some differences in the baseline population characteristics between the two groups, with patients undergoing ‘PVI-only’ suffering more frequently from paroxysmal AF (69.4% vs. 43.3%, P = 0.005), and having a lower incidence of a restrictive diastolic pattern. Furthermore, subjects in the ‘PVI-only’ group had a higher left ventricular wall thickness and more commonly apical hypertrophy. These results are shown in Supplementary material online, Table S1.

On a subgroup analysis, we did not identify any significant difference in terms of freedom from arrhythmia recurrences between ‘PVI-only’ vs. ‘PVI plus substrate modification’ for both paroxysmal and persistent AF (log-rank P = 0.978; Supplementary material online, Figure S2). These results are consistent with our multivariable model, where ablation beyond PVI was not associated with improved efficacy outcomes.

Procedural outcomes in genotyped patients

Among 58 patients undergoing genetic screening, gene mutations responsible for HCM were identified in 45 patients, and the remaining were genotype negative. Nineteen patients had a mutation of the MYH7 gene, 15 of the MYBPC3, 4 of the TNNT2, and 2 of the TPM1. Other pathogenic mutations involved the genes ACTC1 (1), FHL1 (1), CSRP3 (1), FLNC (1), and LAMP-2 (1). On a Kaplan–Meier analysis of this small subset of patients, no significant difference in the ablation outcome was identified among the different genotypes (log-rank P = 0.3). These results are shown in Figure S1.

Predictive value of clinical scores for atrial arrhythmia relapses post-ablation

The APPLE score demonstrated modest discriminative capacity for predicting atrial arrhythmia relapses post-ablation (c-statistic 0.63, 95% CI 0.51–0.75; P = 0.002). The performance of HCM-risk SCD score, ACC/AHA risk stratification system, and CHA₂DS₂VASc was poor (c-statistic 0.60, 95% CI 0.49–0.71, P = 0.105; 0.56, 95% CI 0.45–0.67, P = 0.330; 0.51, 95% CI 0.38–0.64, P = 0.922; respectively). These results are shown in Figures 3 and 4.

Figure 3

Kaplan-Meyer of arrhythmia-free survival after index procedure based on APPLE score.

Open in new tabDownload slide

Figure 4

Kaplan–Meier of arrhythmia-free survival after index procedure based on HCM-risk SCD score (A) and ACC/AHA stratification model for SCD (B). ACC/AHA, American College of Cardiology/American Heart Association; HCM, hypertrophic cardiomyopathy; ICD, implantable cardioverter-defibrillator; SCD, sudden cardiac death.

Open in new tabDownload slide

Discussion

The main findings of this multicentre study are that success rate of catheter ablation for AF in HCM is much lower than what is reported for the general population, especially for patients with persistent AF. Relapses tend to be more frequent not only in the initial 12 months post-ablation, but also longer term. Cryoballoon and radiofrequency technique performs comparably in HCM patients. Procedural complications appear to be higher compared with non-HCM series. The APPLE score can be used for predicting atrial arrhythmia recurrence, however, its discriminative performance is modest. Clinical scores for SCD are not useful in predicting success rate, however, left atrial diameter and left ventricular apical aneurysm are independently associated with relapse. Outcomes appear to be comparable among subjects with different genotype.

Previous small observational studies have evaluated the efficacy and safety of AF ablation in HCM, with conflicting results. A systematic review and meta-analysis from our group¹ showed that catheter ablation might be a valuable option in this setting, however, risk of relapse was two-fold higher compared to the general population; in addition, a sensitivity analyses suggested that the outcome in HCM patients with smaller atria and paroxysmal AF might be more comparable with the general population. Notably, the largest cohort in that systematic review included only 61 participants and the median cohort size was only 27 patients.

To the best of our knowledge, our multicentre data constitute the largest cohort of HCM patients undergoing AF ablation. In keeping with previous evidence, the 12-month success rate of ablation for paroxysmal AF appears comparable to what is observed in the general population. However, a significant proportion of patients required more than one procedure and outcomes were less favourable after the first year. On a multivariable analysis, paroxysmal AF was associated with a 57% lower rate of arrhythmia relapses.

Performance of catheter ablation for persistent AF was extremely poor, with a success rate after single or multiple procedures as low as 40.0% at the 12-month follow-up, and nearly all patients having relapsed before the end of the second year. Considering the non-negligible rate of complications, our findings rise some concerns about the role of ablation in this group of HCM patients. It is conceivable that the extent of atriopathy in HCM patients with persistent forms of AF significantly affects the success rate of the ablation.

The physio-pathological connection between HCM and AF is complex and multifactorial. Hypertrophic cardiomyopathy is associated with diastolic dysfunction, which results in an increase of the end diastolic left ventricular pressure and subsequently of the left atrial afterload. As a consequence, the LA undergoes a process of remodelling and dilatation with proarrhythmic implications. Hypertrophic cardiomyopathy patients have also a high prevalence of mitral regurgitation, which can promote a structural atrial remodelling.¹ Furthermore, atrial fibrosis is common in HCM¹² and may play a key role of substrate for AF by promoting slow conduction and inter-atrial re-entry in conjunction with myocyte disarray. Other possible mechanisms include abnormal calcium handling which could account for triggered activity, myocardial ischaemia due to microvascular dysfunction, and polymorphisms in the angiotensin receptor gene.¹

Previous studies showed that only <10% of patients undergoing AF ablation suffer from very late recurrences.^13,14 Interestingly, our data suggest that the latter may be much more frequent in HCM population; in fact, as many as 20–40% of the patients relapsed annually after the initial 12-month period in our cohort. Progression of atrial fibrosis, enlargement of LA, and adverse electrical and molecular remodelling of myocardial tissue might account for this finding.

Pulmonary vein isolation was the main procedural endpoint in our series. Whether a more extensive ablation might improve outcomes in this population has not been established. Of note, on both our subgroup analysis and multivariable model, creation of atrial lines or ablation of complex fractionated electrograms were not associated with a reduction of arrhythmia relapses. Scar homogenization¹⁵ or ablation of non-pulmonary vein triggers⁵ in the left and/or right atrium might represent alternative approaches beyond PVI.

Another relevant finding of the present study is that cryoballoon ablation appears to be effective and safe in HCM patients, showing comparable results with the radiofrequency technique. These findings are of interest, as patients with HCM were excluded in the landmark trials evaluating cryoballoon AF ablation such as the FIRE and ICE.¹⁰

In keeping with the previous literature,¹ we found that left atrial diameter was negatively associated with long-term outcomes of AF ablation. Furthermore, we identified the presence of left ventricular apical aneurysm as an independent predictor of arrhythmia relapses in our multivariate model. However, this finding should be interpreted carefully given the limited sample size. Hypertrophic cardiomyopathy patients with apical aneurysms represent a subgroup at higher risk of adverse events, including arrhythmic SCD, thromboembolism, and heart failure.¹⁶ Apical aneurysms are often but not exclusively associated with mid-ventricular obstruction,⁸ which might favour AF recurrences by promoting atrial stretch and subsequently shortening the atrial effective refractory period and increasing the dispersion of repolarization. A genetic predisposition to apical aneurysm has been suggested,¹⁶ whether this genetic background might also predispose to a more progressive substrate of AF is currently unknown.

Among the clinical scores tested, only the APPLE score⁷ demonstrated a discriminative capacity for relapses post-ablation, however, its performance was modest. Our study suggests that the APPLE score might help selecting the HCM patients who are more likely to benefit from AF ablation.

This is the first study to report outcomes of AF ablation in a cohort of genotyped HCM patients with a broad spectrum of gene mutations. Among the several genotypes causing HCM, mutations in the MYH7 and MYBPC3 genes are the most frequent.⁸ It has been hypothesized that specific genotypes might be directly associated to atrial myopathy, resulting in a direct predisposition to AF.¹⁷ Recent evidences suggest that subjects with pathogenic variation in the MYH7 gene may be at greater risk of AF.¹⁷ Di Donna et al.⁶ showed that efficacy of AF ablation was similar in a small group of genotype-negative individuals compared with 11 genotyped-positive patients, almost exclusively with mutation on the MYBPC3 gene. Similarly to that findings, we have identified no difference in the rate of relapses post-AF ablation among the different genotypes or in genotype-negative vs. genotype-positive subjects. Further studies are required to clarify the impact of specific genetic mutations in the pathogenesis of AF and in the outcomes of ablation.

Finally, we found a higher rate of major complications, in particular cardiac tamponade, compared with the rate observed in the general population in our centres.¹⁸ Tamponade rates were 4 times those seen in non-HCM series reflecting the challenges of the anatomy on transeptal puncture and the fact that with diastolic dysfunction there is low haemodynamic tolerance to small pericardial effusions during the procedure. Based on our experience, use of intraprocedural transoesophageal or intracardiac echocardiogram should be encouraged in HCM patients to minimize risks of transeptal puncture as the heart is often very rotated resulting in misalignment of the transeptal needle on X-ray screening. Use of non-vitamin K oral anticoagulants compared with vitamin K antagonists might be associated with a lower rate of bleeding.¹⁹

Limitations

Several limitations should be acknowledged. Arrhythmia burden has been recently suggested as a useful alternative endpoint for assessing AF ablation efficacy,²⁰ however, no data on arrhythmia burden pre- and post-ablation were available. Another limitation of the present study is its observational non-randomized design. Furthermore, our cohort included patients undergoing catheter ablation over a long timeframe and in a proportion of them using technologies which are currently outdated and indeed associated with poorer outcomes, such as non-contact force ablation catheters. This should be accounted in the interpretation of our results; however, we have not identified any significant difference in the success rate on a sub-analysis investigating outcomes using different technologies, which represent the major technical developments in the field of AF ablation over the past decades. Nonetheless, procedural outcomes in terms of safety are likely nowadays to be significantly improved, as indeed demonstrated by the reduction in the rate of complication over time in our series. Left atrial fibrosis has been shown to be a predictor of AF ablation outcomes and could also influence the ablation strategy; however, no data on the amount of left atrial fibrosis were available for the patients included in the present series. The proportion of mutation positive subjects was higher in our cohort compared to the expected yield of positive genetic results in the HCM population,⁸ therefore a selection bias might be possible. Furthermore, our cohort included a large number of patients with ICD and/or previous history of VT therefore our results might not be generalizable to lower risk HCM population. There were only a limited number of patients with left ventricular aneurysm, this should be accounted for the interpretation of the results.

Conclusion

The present multicentre observational data show that very late relapses post-AF ablation is common in HCM patients, namely in those with persistent AF. However, as technology progressed since the start of the inclusion period of our cohort, further long-term follow-up studies utilizing contemporary ablation tools are required to confirm these findings. Left atrial size, left ventricular apical aneurysm, and use of the APPLE score might contribute to identify subjects at higher risk of relapse post-ablation. Cryoballoon ablation in patients with HCM appears as safe and effective as radiofrequency ablation.

Supplementary material

Supplementary material is available at Europace online.

Funding

P.D.L. was supported by UCL/UCLH Biomedicine NIHR & has received educational grants from Medtronic and Boston Scientific.

Conflict of interest: R.J.S. has had research agreements and speaker fees from Abbott, Medtronic, Boston Scientific, and Biosense Webster; M.F. has received speaker fees from Biotronik and Medtronic and is a founder and shareholder of Epicardio Ltd, Echopoint Medical Ltd, and Rhythm AI Ltd. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.

Data availability

The data underlying this article cannot be shared publicly due to privacy of individuals that participated in the study.

References