Assessment of the effect of left atrial cryoablation enhanced by ganglionated plexi ablation in the treatment of atrial fibrillation in patients undergoing open heart surgery

The growing incidence of atrial fibrillation (AF) in the population [1] and its clinical [1] and economical [2] importance supports the progressive growth of research of pathophysiology of atrial fibrillation and factors influencing the initiation and maintenance of this most frequently observed heart arrhythmia in the last decades, which is reflected also in the frequency of updates of Guidelines for treatment of AF issued by various cardiology and heart surgery expert societies.

One of the main factors playing an important role in the initiation and duration of atrial fibrillation is the Autonomic Nervous System (ANS) [3‐5]. ANS may be divided into two main components – extrinsic and intrinsic [6]. Extrinsic ANS has a sympathetic and parasympathetic branch. Sympathetic fibres originate from the ganglia along cervical and thoracic spine, and end at the heart surface. Parasympathetic fibres originate in nucleus ambigus in medulla oblongata, and continue as a part of nervus vagus; most of them terminate in the adipose tissue (fat pad) between the superior vena cava and the aorta [7]. Intrinsic ANS is formed with a number of heart ganglia, containing approximately 200–1000 neurons, and synapses of sympathetic and parasympathetic fibres [8, 9]. Most of the ganglia are organized in ganglionated plexi (GP), embedded in epicardial fat pads on the surface of atria and ventricles. They participate upon the interconnection and modulation of interactions between extrinsic and intrinsic ANS [10], and, apart from others, influence the function of the sinoatrial and atrioventricular node [11, 12].

Already in 1978, Coumel et al. [13] in their work pointed out the fact that under certain conditions, the activity of intrinsic ANS may cause paroxysm of atrial arrhythmia. Other works looking into with a more detailed study of intrinsic ANS show the possibility of evocation of rapid trigger activity with stimulation of intrinsic ANS, with electric impulses [14], as well as administration of autonomous neurotransmitters [15‐17]. This trigger activity is most frequently observed in myocardial sleeves of the pulmonary veins [18]. The mechanism behind the initiation of trigger activity is a vegetative influencing of the intracellular level of calcium ions, and the length of duration of the action potential by causing early after depolarisations [19]. Apart from this, further changes of autonomous regulation (autonomous remodelling) may, in interaction with other AF mechanisms (with structural remodelling) lead to a formation of a substrate, which is so important for maintenance of AF [20]. Clinical trials dealing with catheterization therapy of AF present a significant improvement of the outcomes in cases of enhancement of the catheterization ablation by GP ablation [21‐23]. In accordance with the outcomes of these studies, also other outcomes of surgical treatment of lone AF via minimally invasive surgical approach are presented [24‐26]. The aim of our study was to investigate, whether enhancement of left atrial cryoablation by ablation of the autonomic nervous system of left atrium leads to influencing the outcomes of surgical treatment of atrial fibrillation in patients with structural heart disease undergoing open-heart surgery.

In our study, we refer on our early experience with GP ablation enhancing the left atrium cryoablation in the course of surgical treatment of AF in patients undergoing cardiac surgery.

As we have already mentioned in the introduction, autonomic nervous system plays an important role in the initiation and maintenance of AF. Key role is played namely with ganglionated plexi, which mediate and modulate interactions between the extrinsic and intrinsic autonomic nervous system [10, 17], and namely plexi localized in the atrial walls. These plexi are concentrated in centres with maximum incidence around orifices of pulmonary veins [8, 17, 31‐33].

Three key areas around the orifice of right-side PVs have been studied and presented in earlier publications: area between the right superior PV (RSPV) junction and LA, and superior vena cava (SVC) (dorsal right atrial subplexus) [9], medially from the right PV (RPVs) junction and LA, along interatrial groove (anterior right GP, ARGP) [8, 31] and between the inferior vena cava (IVC) orifice and the left atrium, in a close proximity of the coronary sinus orifice (the inferior area, inferior right GP, IRGP) [8, 31]. ARGP and IRGP play an important role in the regulation of SA and AV nodi [9, 12, 17, 27, 34]. Other two GP centres are located around the orifice of left-side PVs. These left-side centres manifest a significantly lower activity than the right-side GP [35, 36]. Ligament of Marshall has been reported as another independent centre, which may be the source of focal activity [37].

Although we did not keep a strict protocol regarding the frequency of activity and GP ablations in individual areas during our study, we observed, in concordance with previous studies, a significantly higher GP activity on the right side, namely in the area of RSPV junction – LA, behind SVC. Although the evidence regarding the influence of GP on the initiation and duration of AF is well known, the importance of GP ablation in the therapy of AF is not so clear, and studies, which have been performed so far, bring inconclusive evidence.

Experimental studies have proven influence of AF initiation in vagal denervation [38, 39], or chemical GP ablation [14]. In clinical practice, isolated GP ablation in the treatment of AF has been shown as insufficient [40, 41], and is not recommended [42]. Nevertheless, some authors support GP ablation as a part of a more complex catheterization [21, 22] or surgical ablation [24, 43‐45]. In this relation, work of Pappone et al. [21] is frequently cited; the authors achieved 99% success rate (freedom from AF) in the course of 12-month follow-up in patients with paroxysmal AF, in whom catheterization RF ablation was performed – circulation of PVs with a complete vagal denervation (= interruption of all vagal reflexes around PVs orifice).

In surgery, GP ablation is most frequently used as an enhancement of ablation therapy of AF alone performed from mini-invasive aproache, with or without the use of extracorporeal circulation, from bilateral thoracotomy [45, 46], or via fully thoracoscopic approach [24, 43, 44]. These studies present very good results achieved in the treatment of paroxysmal, persistent and long-standing persistent AF. However, most of these studies are retrospective, non-randomized clinical trials. There also exist works speaking against the use of GP ablation, presenting good results of AF treatment with ablation therapy only, without the use of GP ablation [47, 48]. In these cases, certain role may have been played by the already mentioned GP topography. Most of the plexi are located in areas of ablation lines; that is why it is possible to presume that they will be destroyed (eliminated) already in the course of a standard ablation, without a targeted GP mapping. This theory is also supported by the findings of McClelland et al. [46]. The authors claim that following circumferential isolation of pulmonary veins with bipolar RF, the GP activity was limited by 79%. Gelsomino et al. [49] use this theory to explain the failure to demonstrate the benefit of GP ablation in their study. This study was published last year, and similarly to our study, deals with the importance of GP ablation in patients with concomitant atrial fibrillation in the course of open-heart surgery. The authors compared groups of patients with persistent and long-standing persistent AF, in whom radiofrequency ablation was performed, including (Group 1), or not including (Group 2) GP ablation. The average follow-up period was 36.7 months. When comparing both groups, no difference was observed in the percentage of patients with NSR on-AAD (anti-arrhythmic drugs) (86.4% vs. 79.7%, p = 0.07), or in the percentage of patients with NSR off-AAD at the end of the follow-up period (75.5% vs. 67.8%, p = 0.08).The performed risk analysis showed identical incidence of AF recurrence in both groups. Also the group of Alex et al. [50] was unsuccessful in their attempt to influence the incidence of AF in patients after CABG (coronary revascularization) procedure with heart denervation.

Other reasons standing in opposition of the performance of targeted GP ablation are the findings in the area of restoration of the heart autonomic system. The sympathetic re-innervation was described by Kaye et al. [51] already in 1993, in patients following heart transplantation. It is possible to object that in cases of heart transplantation, neuronal fibres are only interrupted, and direct destruction of neurons of the heart neural system is not performed. However, newer experimental studies bring clear evidence regarding re-innervation also following radiofrequency parasympathetic denervation (vagal denervation). Tests were also performed on canine hearts, the samples were monitored for the period of four weeks following ablation on average [52, 53]. Similar results of neural function restitution have also been presented regarding experimental demonstration of administration of cryoenergy on the neural tissue. After administration of cryoenergy, the function of sensory and motor nerves is restored within several months [54‐56]. This work cites only selected publications regarding GP ablation. Nevertheless, inconclusiveness of results and conclusions presented in these studies points towards the need of further research in the form of randomized, prospective clinical trials.

The issue of long-term efficacy of GP ablation has been discussed on numerous occasions and conclusion is not clear. On the basis of our early experiences and of recent experiences of other colleagues from cardiac centres we decided disrupt GP ablation performance and we are waiting for results of following studies. Although our team was educated by Atricure promotor, we performed GP ablation in only 35 patients and the learning curve may play role determinig our negative result.

A frequent topic of discussions regarding the surgical treatment of AF is the extent of the lesion set. This topic was analysed in the meta-analysis performed by Barnett et al. [57]; the results of this analysis show that surgical biatrial ablation (range, 92.0%-87.1%) in the analysed studies was more effective in the treatment of AF than the procedure limited to the left atrium (range, 86.1-73.4%). Also the results of comparing various extent of ablation in the above cited work of the Italian authors [49] are in accordance with this finding; the authors concluded that biatrial ablation, as well as compete left atrial lesion set are suitable for improving rhythm outcomes, however, on more detailed analysis, a significantly lower incidence of AF recurrence was observed in patients with biatrial ablation when compared to the patients with left atrium ablation (18.5 vs. 36.5%, p = 0.001), and absence of RA lesions (right atrium ablation) was determined as one of the predictors of AF recurrence.

In our present, as well as previous study [58], we managed to achieve similar good outcomes with performance of left atrium ablation in achievement and duration of NSR in the group of patients with paroxysmal and persistent AF. However, in the group of patients with long-standing persistent, or permanent AF, we observed a significantly higher AF recurrence, which increases with the period of follow-up.

These results and conclusions made us re-evaluate the extent of the ablation procedure performed at our centre, namely in patients with persistent and long-standing persistent AF, and enhancing the procedure with ablation of RA lesions (Cox-MAZE IV procedure), the performance of which is aimed at the period of reperfusion, and thus should not have any influence on the cross-clamp time, the pump time, or the overall time of surgery [59]. Currently the Cox MAZE IV without GP ablation is preferred approach to treatment of concomitant AF in patients with persistent AF and mitral valve diseases at our centre.

Permanent pacemaker implantation is a known complication following heart surgery procedures. Its incidence reported in the literature reaches 1–11% [60‐65], and is comparable for heart surgeries with and without ablation. Elahi et al. [62] reported in their study of patients undergoing heart valve surgery the frequency of permanent cardiac stimulation of 6%, and the identified predictors for implantation include reoperation, cross-clamp time, combined procedure on heart valves and absence of NSR prior to the surgical procedure.

Low values were also reported by Merin et al. [63] (1.4%), with a higher incidence in patients following aortic valve replacement (5.7%), and the team of Gelsomino et al. [49] (0.6% in the control group of patients undergoing cardiac surgery, with a various extent of ablation lesions set due to AF). On the other hand, there exist publications reporting the frequency of permanent cardiac stimulation around 20% [66, 67]. In case of the study presented by Prasad et al. [66] the study subjects were patients after cardiac surgery with associated ablation due to AF, in most of whom sick sinus syndrome was diagnosed already prior to the surgery. In the group of patients with a normal function of the sinus node, the reported rate of implantation reached 8%. In our patient group, a total of eight patients required permanent implantation (8%) in the postoperative period (until discharge). At the end of the 12-month follow-up, the overall rate of implantation reached 10%.

Left atrium appendage (LAA) elimination in patients with AF is performed due to the risk of thromboembolic complications. Dawson et al. [68] presented the results of studies dealing with this topic in his summary work. Although the benefit of LAA elimination was not confirmed, we perform this procedure in all patients with AF undergoing cardiac surgery. The procedure may be performed safely, without the need of prolongation of the overall time of surgery and increasing of the risk of perioperative bleeding [69]. As far as the surgical technique is concerned, use of instruments designed specifically for appendage occlusion, rather than use of staplers or the cut-and-sew technique is recommended [70]. Ligation or suture of the appendage is not recommended due to a high incidence of recanalization [71, 72]. Very good results have been published regarding the use of AtriClip device (Atricure, Inc., Westchester, OH, USA); use of this device leads not only to an ideal LAA occlusion and prevention of stroke but also to achieving electrical isolation of LAA and reduced AF recurrence [73].

At present, we use at our centre the technique cut-and-sew for removal of LAA. The incidence of thromboembolic complications (stroke) among the patient population in our present study reaches approx. 2.2% at the end of one-year follow-up. In the previous study [58] performed at our centre, the incidence of thromboembolic events was 4.6%, or 5.9%, at the end of one-year and two-year follow-up, respectively. This corresponds with the frequency of thromboembolic events in patients after surgical ablation presented in published studies and results summarized in the meta-analysis by Gray et al. from 2011 [74].

The size of the left atrium is one of the factors influencing the AF recurrence. Gillinov et al. [61], in their study present a significantly higher AF recurrence in patients with the LA diameter of 60 mm (15%), in comparison with the LA size >40 mm. Damiano et al. [75] present up to 50% AF recurrence in LA diameter > 80 mm et LA reduction procedures presented in this study were not successful in the prevention of recurrence.

Upon analysis of our patient group, we did not observe any significant dependency of AF recurrence on the LA diameter in patients with LA > 50 mm. This result may have been influenced by the relatively small number of study subjects and the fact that our patient file did not include any patients with LA > 60 mm undergoing surgery; however, one patient in the GP Group had LA diameter of 63 mm.

We have not proven an improvement of results of left atrial cryoablation enhanced with a targeted RF ablation of GP in the AF treatment in patients undergoing cardiac surgery. Mapping and targeted GP ablation is not performed at our centre in other patients at the moment, as we are waiting for results of randomized controlled trials.

Cryoablation of the left atrium is a reasonable treatment of choice in patients with paroxysmal and persistent AF. In case of patients with long-standing persistent AF, higher recurrence of AF has been observed.