Mid-term outcomes of concomitant surgical ablation of atrial fibrillation in patients undergoing cardiac surgery for hypertrophic cardiomyopathy†

Abstract

INTRODUCTION

Atrial fibrillation (AF) is highly prevalent in patients with hypertrophic cardiomyopathy (HCM) and constitutes one of the most common reasons for hospitalization [1]. Indeed, approximately 20% of HCM patients will develop AF during the clinical course of the disease [2]. The occurrence of AF in these patients, however, is often poorly tolerated, mainly because they are unable, at the higher rates of AF, to maintain the appropriate stroke volume due to shorter time for cardiac filling. This results in a deterioration of clinical condition, worsening of symptoms and greater long-term mortality [3, 4]. As such, restoring and maintaining sinus rhythm (SR) is an important goal in this population [5]. Catheter ablation has been shown to be superior to drug therapy in the aspect of rhythm control in this population, even though it has been associated with higher risk of AF recurrence. Repeat procedures are required in approximately half of the patients undergoing catheter ablation, with left atrium (LA) size and AF duration predictive of AF recurrence [6–11]. Very few data are available on surgical ablation of concomitant AF in the setting of HCM. The underlying mechanisms of AF in these patients are variable and, therefore, the surgical ablation may have different effects. The aim of this study was to evaluate the mid-term outcome of surgical AF ablation in patients who underwent cardiac surgery due to HCM.

METHODS

Study population

The study population includes the first 31 consecutive patients (age 59 ± 6.0 years) with primary HCM and drug-refractory symptomatic AF who underwent surgical ablation with concomitant septal myectomy (77%) and/or mitral valve (MV) repair or replacement (39%) between 2002 and 2012. A total of 148 patients with HCM were surgically treated in this time frame. Out of them, 41 patients were in AF (27%, 41/148) and only 31 patients underwent concomitant surgical ablation of AF. The remaining 10 patients with HCM and AF were not ablated because they had been in permanent AF for too many years making extremely unlikely the chances of restoration of SR.

The diagnosis of HCM was based on echocardiographic evidence of a hypertrophied non-dilated left ventricle (LV) (maximum wall thickness ≥ 15 mm), in the absence of any other primary cardiac or systemic etiology, and followed the American College of Cardiology/European Society of Cardiology guidelines [12]. LV hypertrophy was assessed with 2-D echocardiography and the site and extent of maximal wall thickness were identified. Peak instantaneous left ventricle outflow tract (LVOT) gradient was estimated with continuous wave Doppler. Surgery was performed when a resting or provoked (with Valsalva or exercise) LVOT peak instantaneous gradient of ≥50 mmHg was detected [13]. The severity and mechanism of mitral regurgitation, when associated, were established by transthoracic and transoesophageal echocardiography using an integrative approach. The degree of mitral regurgitation was measured by Doppler colour flow imaging and defined as mild (1+/4+), moderate (2+/4+), moderate-to-severe (3+/4+) and severe (4+/4+).

AF was defined as paroxysmal (≥2 episodes terminating spontaneously within 7 days), persistent (lasting >7 days but <1 year, or necessitating cardioversion), or longstanding persistent (lasting >1 year) [14]. The diagnosis of AF was based on electrocardiogram (ECG) or Holter recordings (obtained either after acute onset of symptoms or during routine examination) or by an established history of paroxysmal, persistent or longstanding persistent AF. All patients had been unsuccessfully treated with class I/III antiarrhythmic drugs (AADs) and were on oral anticoagulant therapy. The absence of atrial thrombi was confirmed by transoesophageal echocardiography in all patients. For the purpose of this study, the preoperative, intraoperative, postoperative and follow-up data were prospectively entered into a dedicated database and retrospectively reviewed. The Institutional Ethic Committee approved this study (Approval number: 145/INT/2016) and waived individual consent for this retrospective analysis.

Decisions making-process regarding ablation strategy

Pulmonary vein isolation (PVI) was performed in all cases of paroxysmal AF. Left atrial set (LA set) or biatrial Maze IV were performed in presence of persistent (8 patients) or long-standing persistent AF (5 patients). In particular, the Maze IV has been carried out when concomitant tricuspid regurgitation or severe dilatation of the right atrium was present (5 cases). However, 5 patients with persistent/long-standing persistent AF underwent only PVI. The reason for that was due to the high patient’s risk profile (1 case) or the need to perform multiple surgical procedures (myectomy + MV replacement + CABG or aortic valve replacement) with long cross-clamp time (2 cases). Finally, in the remaining 2 patients, the choice to perform the PVI alone, was dictated by the fact they were considered as ‘persistent AF cases’ only because submitted to electrical cardioversion, although all their AF episodes had lasted less than 7 days.

Follow-up

All patients underwent ECG before discharge. For the 2 patients who died in the hospital, the last ECG before death was considered. Follow-ups were performed in our dedicated Institutional outpatient clinic at 3, 6, 12 months, and then yearly thereafter, with physical examination, 12-lead ECG, 24-h Holter monitoring and echocardiography. The first 3 months after surgery were considered as ‘blanking period’ and were ‘locked out’ in the outcome analysis. Follow-up was 97% complete with a median of 6.4 years [3.8–9.1]. Recurrence was considered to be any episode of AF/atrial tachyarrhythmia lasting for ≥30 s and detected due to symptoms and/or ECG/Holter recording, after the blanking period. In patients with an implanted cardioverter device (ICD), interrogation of device was used to check the recurrence. Rhythm status was defined according to Heart Rhythm Society guidelines as SR without the use of class I/III AADs [14]. Arrhythmia control was defined as maintenance of SR with or without AADs.

Statistical analysis

Statistical analyses were performed using SPSS version 22.0 (SPSS Inc., Chicago, IL, USA) for Windows (Microsoft Corp, Redmond, WA, USA). Continuous data were expressed as mean ± SD or as median [Q1–Q3]. Categorical data were reported as number and percentage. Comparison of categorical variables was performed using χ² and Fisher’s test. Pre- and post-surgery NYHA functional class were compared using Wilcoxon’s signed-rank test. Survival and freedom from events were evaluated by Kaplan–Meier analysis. For survival estimates, data are presented as mean ± standard error. Predictors associated with death and with arrhythmia recurrence during follow-up were assessed in univariate Cox proportional hazard models. A multivariable Cox regression could not be applied due to the small number of events occurred.

RESULTS

Patient characteristics

At baseline, the majority of patients had non-paroxysmal AF (18 patients, 58%) and all patients had failed treatment with ≥1 class I/III AADs and/or multiple electrical cardioversion. Amiodarone and sotalol were the most frequently used drugs. Three patients (10%) had previously been submitted to catheter ablation of AF and none had a previous cardiac surgery. At admission, 15 (48%) patients were in SR, while 16 (52%) patients were in AF. Three patients (10%) suffered a previous neurological event (transitory ischaemic attack/stroke). Significant MV regurgitation requiring mitral surgery was present in 12 patients (39%), concomitant coronary artery disease in 3 and aortic valve regurgitation in 3 cases. Most of the patients (17 patients, 55%) were in NYHA class III or IV and 8 of them (26%) had been hospitalized due to pulmonary oedema. Eleven patients (35%) complained of syncope and fainting symptoms and 3 (10%) had an implanted cardioverter device as secondary prevention. The mean left atrial diameter was 49 ± 8.5 mm and the left ventricular ejection fraction was 58 ± 2.1%. The peak LVOT gradient at rest was 56 ± 31.8 mmHg and the septum thickness was 20 ± 4 mm. In 15 patients (48%) anomalous insertion of anterolateral papillary muscle, producing mid-cavitary muscular obstruction, was detected. The baseline characteristics are reported in the Table 1.

Procedural data

All patients were operated through a conventional median sternotomy with standard cardiopulmonary bypass on moderate hypothermia and cardioplegic arrest. Most of the patients (24/31, 77%) underwent surgical septal myectomy (transaortic resection of muscle from the basal septum) to correct outflow obstruction. MV repair using the edge-to-edge technique, was performed in 5 patients (16%) and MV replacement in 7 cases (22%). Other concomitant procedures were coronary artery bypass grafting (3 patients, 10%), tricuspid annuloplasty (3 patients, 10%) and aortic valve replacement (3 patients, 10%). Isolated septal myectomy combined with ablation of AF was performed in 14 patients (45%).

Table 2 reports in details the connecting lines accomplished in each lesion set of surgical ablation of AF. Three to five repeated applications of bipolar radiofrequency (RF) clamp were performed for each lesion to provide transmurality. Cryoprobe was used for lesions that terminate on the mitral and tricuspid valve annuli in 7 cases (22%). Left atrial appendage closure was performed in 1 patient (3%). Median cardiopulmonary bypass time and cross-clamp time were 72 [65.2–97.5] and 50 [41–73] min, respectively.

Clinical hospital outcomes

Overall hospital mortality was 6% (2/31 patients). The cause of death was low cardiac output syndrome in 1 patient and respiratory insufficiency followed by sepsis in the other one. The postoperative course was complicated by low cardiac output syndrome in 2 patients (6%), respiratory failure in 3 (10%), sepsis in 1 (3%) and re-exploration for bleeding in 2 patients (6%). One patient (3%) required a permanent pacemaker for complete heart block. Postoperative median length-of-stay was 5 [4–9] days. At hospital discharge 26 patients (84%) were in SR.

Follow-up

Clinical outcomes

At 7 years, overall survival was 87 ± 6.1%. During follow-up, 3 patients died (3/29 hospital survivors, 10%) and the cause of death was cardiac related in 1 of them (congestive heart failure). Among those 3 patients, 2 were in stable SR, while the third patient was lost to follow-up with regard to the rhythm, after 6 months (follow-up 97% complete). At univariate Cox proportional-hazard regression, age and preoperative systolic pulmonary artery pressure were identified as significant predictors of overall mortality (HR 1.2, 95% CI 1.0–1.6, P = 0.03 and HR 1.2, 95% CI 1.0–1.4, P = 0.01, respectively) (Table 3). There was no cardiac arrest, thromboembolism or major bleeding event during the follow-up period. All patients were on oral anticoagulation (27 patients with warfarin and 2 patients with rivaroxaban), even with successful ablation of AF independent of CHA₂DS₂-VASc score. No patient required a new cardiac operation except one, who was submitted to a right thoracoscopic epicardial approach by means of Cobra Fusion Ablation System (Estech), to treat recurred AF. Two patients (7%) received an implanted cardioverter device, 1 and 7 years after surgery, respectively, because of ventricular tachyarrhythmia and 2 more patients are currently scheduled for implantation. At the last follow-up NYHA class III or IV was documented in 14% of the patients (4/29) compared to 55% (17/31) at baseline (P = 0.001). Three of the 4 patients in NYHA class III–IV were in AF.

Recurrence of atrial fibrillation

AADs were not systematically used for the blanking period and anyway discontinued after that in all patients but 2 (26/29 patients, 90%). The arrhythmia-free survival off class I/III AADs was 82 ± 7.3% at 1 year and 52 ± 10.2% at 6 years (Fig. 1). The 1- and 6-year arrhythmia control (maintenance of SR with or without AADs) was 96 ± 3.5 and 80 ± 8.1%, respectively (Fig. 2). The AADs used to achieve a satisfactory rhythm control were amiodaron (6 patients), sotalol (5 patients), propafenone (2 patients) and flecainide (1 patient). The recurrent arrhythmia was AF in all patients. No predictors of AF recurrence were detected at Cox-regression analysis. The proportion of patients in stable SR off AADs at 6 months, 1, 2 and 3 years after surgery was 82% (23 of 28 patients at risk), 75% (21/28), 64% (18/28) and 60% (17/28), respectively. During follow-up, 1 patient (3%) underwent atrioventricular node ablation followed by pacemaker implantation for recurred AF, 1 patient (3%) was submitted to a successful single transcatheter ablation 2 years after surgery, and one more patient (3%) underwent multiple catheter ablation procedures and eventually a right thoracoscopic epicardial approach by means of the Cobra Fusion Ablation System (Estech). Figure 3 shows the recurrence of AF according to the type of preoperative AF and to the lesion set performed at the time of surgery. SR was maintained in 92% of the patients (12/13 patients) affected by preoperative paroxysmal AF (54% off ADDs) and in 73% (11/15 hospital survivors) of patients with non-paroxysmal AF (47% off AADs) (P = 0.3).

Figure 1

Kaplan–Meier curve showing the arrhythmia-free survival off antiarrhythmic drugs (AADs).

Open in new tabDownload slide

Figure 2

Kaplan–Meier estimate of the arrhythmia control (maintenance of SR with or without AADs).

Open in new tabDownload slide

Figure 3

Flow chart depicting recurrence of AF according to the type of preoperative AF and to the lesion sets performed for the ablation. AF: atrial fibrillation; PVI: pulmonary vein isolation; LA: left atrial set; AADs: antiarrhythmic drugs; AV: atrioventricular; TCA: transcatheter ablation.

Open in new tabDownload slide

DISCUSSION

The present study reports the results of concomitant surgical ablation of AF in a series of patients with HCM who underwent septal myectomy and/or MV repair or replacement, with the longest follow-up to date. The major finding is that patients with HCM undergoing concomitant surgical ablation of AF have a satisfactory mid-term arrhythmia control (maintenance of SR with or without AADs) although the arrhythmia-free survival off class I/III AADs remains lower than that reported in the general AF population.

AF is a well-recognized comorbidity in patients with HCM, occurring in about 20% of patients [2, 15]. Several mechanisms may be responsible for the increased vulnerability of the LA myocardium to initiate and sustain AF in HCM. MV regurgitation due to systolic anterior motion [16], LV outflow tract obstruction and particularly LV diastolic dysfunction, all lead to increased LA pressure and consequently to chronic atrial stretch and remodelling, which finally results in a diffuse atrial fibrosis that plays a crucial role in the development and maintenance of AF [17]. Besides these issues, additional factors such as a primary sarcomeric atrial myopathy, polymorphisms in the angiotensin receptor gene [11, 18–20] and specific HCM-mutations [18] have been implicated in the development of AF in those patients.

The occurrence of AF represents a turning point in the clinical course of the HCM, and it is generally associated with deterioration of clinical status, functional capacity, quality of life and long-term outcome [3, 7, 18, 21]. The loss of atrial kick can reduce preload and exacerbate or unmask LV outflow obstruction, and rapid ventricular rate can decrease cardiac output by reducing LV filling time in the setting of diastolic dysfunction and reduced LV compliance [18]. As such, restoring and maintaining SR is an important goal in this population. AADs, however, are frequently not effective enough for a rhythm control strategy in those patients [11]. Catheter ablation can be considered in patients with symptomatic AF refractory to medical therapy, but restoration of SR and suppression of AF recurrence can be achieved in approximately two thirds of HCM patients over 1–2 years and typically decline with longer follow-up [7, 21, 22]. Bunch et al. [23] reported a success of 64% that decreased to 47% during 3-year follow-up. Repeat procedures are needed in approximately half of the patients and maintenance of chronic AADs is frequently required [7, 24]. The lower success rate of catheter ablation in HCM compared to the general AF population may likely be due to the extensive atrial scarring, dilation and remodelling, which frequently characterize HCM patients. Those alterations, due to haemodynamics or sarcomere protein gene mutations [3, 9], might provide in such patients the substrate for multiple arrhythmogenic areas beyond the PVs. Non-PV triggers may be potentially responsible for the high rates of AF recurrence in HCM patients [8–9]. Hypertrophy of atrial myocytes, as well as the LA wall thickening, may also contribute to the poorer results by limiting the ability of RF to create permanent transmural lesions to effectively isolate the pulmonary veins [24]. This suggestion has been supported by the high incidence of PV conduction recovery identified on repeat ablations [24].

As the surgical management of AF in HCM is concerned, limited data are available. Surgical septal myectomy reduces gradients and systolic anterior motion-related MV regurgitation in over 90% of HCM cases, improving exercise capacity and symptoms [13], but it has not been well established if it decreases AF [25]. Therefore, for symptomatic AF patients with outflow obstruction, combining septal myectomy with surgical ablation of AF has been suggested. Chen et al. [26] described a series of 10 patients submitted to surgical myectomy combined with the cut-and-saw Maze III procedure. At a mean follow-up of 15 months they reported a success rate of 60%. The only other study published by Bassiouny et al. [24] reported a surgical series of 68 HCM patients mainly with paroxysmal and persistent AF, who underwent Cox-Maze III (28%), modified Maze using cryothermia/RF energy (19%) or limited PVI by applying a bipolar RF clamp (53%). The surgical ablation was associated in most of the cases with septal myectomy. The reported AF-free survival was 81% at 6 months and decreased to 74% at 1 year. At last follow-up of 39 months, only 51% of the patients maintained SR, 44% of whom off AADs [24]. These data demonstrate that, due to the peculiar nature of the HCM disease process, the efficacy of surgical ablation of AF in maintaining SR over time remains uncertain.

In our series all patients were severely symptomatic and 25% of them had been hospitalized due to pulmonary oedema. In most of the patients, concomitant surgical procedures, particularly on the MV, were performed besides myectomy which explains the relatively high hospital mortality. The rather low rate of perioperative complications is likely due to the fact that we did not use a cut-and-saw Maze technique which is more technically challenging and carries a relatively high operative risk. Of note the fact that only 1 patient required a permanent pacemaker for complete heart block and this rate (3%) is even lower than that reported for isolated septal myectomy [13]. In addition, unlike other studies [24, 26], we have excluded the LA appendage in only 1 patient with a previous stroke and TIA episodes while on oral anticoagulation. The choice of leaving the LA appendage was made in order to possibly maintain its role in the passive and active function of the left atrium whose contribution to the ventricular filling may be particularly important in the HCM setting. Whether this policy is worthy or not remains to be established [27].

Late survival was satisfactory and, unlike other studies [28], age and preoperative systolic pulmonary artery pressure did have an impact on mortality at follow-up. The symptomatic improvement was remarkable and restoration of SR most likely played an important role in the achievement of this clinical outcome in many patients. Our data expand the current limited knowledge in this field, by confirming with mid-term results that surgical ablation in HCM patients may result in acceptable late maintenance of SR, although with frequent need for antiarrhythmic medications. Arrhythmia control (maintenance of SR with or without AADs) was achieved in 80% of the patients at 6 years. In particular SR was maintained in a higher percentage of patients with paroxysmal AF (92%, 54% off ADDs) as compared to those with non-paroxysmal AF (73%, 47% off AADs). Although this difference was not statistically significant, this might simply be due to the small sample sizes and we cannot exclude a better outcome in the paroxysmal setting.

As it concerns the arrhythmia-free survival off class I/III AADs, our findings are similar and possibly better than those reported in the literature in this setting [24–26], with a freedom from AF off AADs of 52 ± 10.2% at 6 years. The proportion of patients in stable SR off AADs at 3 years was 60% supporting the concept that HCM patients with AF represent a specific population, in whom, as with percutaneous catheter ablation, the disease substrate may limit the effectiveness of bipolar RF ± cryotherapy to create transmural lesions, even with a surgical approach. The recurrent arrhythmia was AF in all patients. The only 2 patients submitted to subsequent catheter ablations were treated in other hospitals and therefore we could not identify if the recurrence was due to conduction recovery.

In conclusion, the likelihood of remaining free of AF off AADs after concomitant surgical ablation of AF appears lower in HCM, with recurrence rates likely tied to the severity of the underlying substrate. Antiarrhythmic medications may still be needed, but joint strategies can often improve outcome for extended periods. Even with these limitations, since AF is often debilitating for patients with HCM and restoration of SR is highly desirable, concomitant arrhythmia surgery should be considered as a treatment option for drug refractory AF in patients with HCM undergoing surgical myectomy and/or MV surgery. However, those patients should be advised regarding the lower long-term arrhythmia-free survival off AADs, the frequent need for chronic AADs and the possible necessity for subsequent catheter ablations compared to non-HCM patients with AF.

Study limitation

Although data were obtained from a prospectively collected database, this was a retrospective, single-centre study and therefore subject to the limitations of every retrospective data analysis. The sample size is small and the inclusion period long, which probably prevented the identification of predictors of failure. Another limitation is represented by the different ablation strategy performed according to the operator and the individual patient’s characteristics and risk profile. Future studies, with large number of patients, assessing the efficacy of different strategies, such as PVI-only, LA set, biatrial Cox-Maze IV and the type of energy source used (only RF versus combined RF + cryoablation), may help to identify the optimal strategy in this challenging patient population.

In addition, we did not have a control group of patients affected by HCM who underwent surgery for HCM without treating AF. The size of LA at follow-up, due to the retrospective nature of the study, was not recorded systematically and therefore could not be analysed. Finally, we did not use event recorder devices to detect AF recurrence.

Conflict of interest: none declared.

REFERENCES

Author notes