Ganglionic plexi (GP) ablation can be used to treat patients with various cardiac arrhythmias by selectively disrupting clusters of autonomic nerve cells, known as ganglia, on the epicardial surface of the heart. GP influence cardiac physiology through complex neural connections that can play a role in the etiology of various cardiac arrhythmias including atrial fibrillation (AF). GP are the targets of various catheter-based and surgical ablation techniques to treat arrhythmias as either stand-alone procedures or adjunctive procedures in patients undergoing cardiac surgery. , This chapter discusses the rationale for GP ablation, the techniques for the identification and ablation of the GP, and the efficacy of GP ablation in the management of arrhythmias.
Neuroanatomy of the Heart
The cardiac autonomic nervous system (ANS) consists of extrinsic (central) and intrinsic (peripheral) components ( Fig. 30.1 ). , The intrinsic cardiac ANS is a complex network of epicardial GP, interconnected neurons, and axons. , These GP are extensions of mediastinal nerves that terminate in the epicardium around the ascending aorta, pulmonary artery trunk, pulmonary veins (PVs), and venae cavae. , Multiple anatomic studies have consistently identified seven discrete areas on the epicardium where densely concentrated clusters of these ganglionic cells form. , , Two plexi innervate the right atrium (RA) and three innervate the left atrium (LA). One GP innervates the right ventricle (RV) and three innervate the left ventricle (LV).
Flow diagram of the cardiac autonomic nervous system. The anatomic structures responsible for all facets of the autonomic nervous system are the cervical and thoracic sympathetic ganglia, the vagus nerve parasympathetic fibers, and the ganglionic plexi on the surface of the atria and the proximal great vessels.
Epicardial ganglia are most densely concentrated on the posterior and posterolateral surfaces of the LA near the junctions of the PVs and LA, where about 50% of all cardiac ganglia are located. , Individual GP are located near the four PV ostia and innervate the myocardial sleeves of the PVs. , To ensure accurate description during ablation procedures, these GP near the PVs have been assigned names that reflect their association with a specific adjacent PV. The superior left ganglionic plexus (SLGP) is located on the LA roof just medial to the left superior PV ( Fig. 30.2 ). The inferior left ganglionic plexus (ILGP) is located just below the left inferior PV, extending toward the inferior aspect of the left atrial posterior wall. It is connected to the atrioventricular (AV) node and is pivotal in ventricular slowing. The anterior right ganglionic plexus (ARGP) is situated anterior to the right superior and right inferior PVs and is the only GP immediately adjacent to a PV. It delivers signals to the sinoatrial (SA) node, including integrating signals from the extrinsic ANS. The inferior right ganglionic plexus (IRGP) is located just below the right inferior PV and extends toward the inferior aspect of the left atrial posterior wall. It delivers signals to the AV node. The IRGP and ILGP are known as the “oblique sinus GP.” The ligament of Marshall ganglionic plexus (LOM GP) is positioned between the left superior PV and the left atrial appendage (LAA) and is the only GP not embedded in epicardial fat. The LOM is a vestigial remnant of the embryonic left superior vena cava (SVC) that harbors predominately parasympathetic fibers of the left vagus nerve that innervate parts of the LA. It transforms into the vein of Marshall as it connects to the coronary sinus.
Approximate locations of the ganglionic plexi (GP) in the posterior left atrium (LA). The GP near each pulmonary vein (PV) orifice have been given names to reflect their association with the corresponding PV. These GP innervate the sleeves of left atrial endocardium that extend up into the PVs. The superior left GP (SLGP) is located on the roof of the LA near the left superior PV and often extending to the medial aspect of left atrial appendage (LAA). The anterior right GP (ARGP) is located near the right superior PV and often extends inferiorly to the area near the right inferior PV. The inferior left GP (ILGP) is located just below the junction of the left inferior PV and the LA. The inferior right GP (IRGP) is located near the right inferior PV. The ligament of Marshall GP (LOM GP) is located between the left superior PV and the LAA. It is the only GP in the atrium that is not embedded in an epicardial fat pad.
Less accessible atrial GP, primarily studied in animal models because of limited endocardial access, include the transverse sinus or great artery GP and the SVC–aortic (SVC-Ao) or superior right GP (SRGP). Experimental canine models have identified the SVC-Ao GP as the “head station” connecting the extrinsic and intrinsic ANS. This characterization is due to the finding that the majority of vagosympathetic fibers carrying signals from the CNS to heart innervation pass through the SVC-Ao GP. There are numerous interconnections between the GP, and there is a final common pathway to the sinus node via the ARGP and to the AV node via the IRGP. , GP embedded within epicardial fat pads and in the LOM form a complex neural network that modulates autonomic innervation. They also facilitate communication between intrinsic and extrinsic components of the ANS.
The extrinsic component of the ANS is divided into two parts: sympathetic and parasympathetic ( Fig. 30.3 ). The sympathetic component originates from major autonomic ganglia located along the cervical and thoracic spinal cord, including the superior cervical ganglia (connected with C1–C3), the stellate ganglia (connected with C7–C8 to T1–T2), and the thoracic ganglia (extending down to the seventh thoracic ganglion). Most postganglionic sympathetic neurons have cell bodies in these ganglia. Their axons form the superior, middle, and inferior cardiac nerves, connecting with GP neurons. The parasympathetic component derives primarily from the nucleus ambiguus and the dorsal motor nucleus of the medulla oblongata, which houses preganglionic neurons. These neurons send their axons primarily through the vagus nerve to reach postganglionic neurons located in the GP. ,
The autonomic nervous system (ANS) of the heart. See text for further discussion.
Neurophysiology of the Heart
The ANS maintains cardiovascular homeostasis by balancing sympathetic and parasympathetic inputs to regulate heart rate, contractility, and conduction. Sympathetic stimulation, via norepinephrine release, exerts positive chronotropic, inotropic, and dromotropic effects on the heart. Activation of β-adrenergic receptors enhances SA node automaticity, increases myocardial contractility, and facilitates AV conduction, thereby augmenting cardiac output to meet increased metabolic demands during stress or exercise. Conversely, the parasympathetic activation, predominantly mediated by acetylcholine, exerts negative chronotropic, inotropic, and dromotropic effects. Stimulation of vagal nerve fibers decreases heart rate by slowing SA node depolarization, reduces AV conduction velocity, and diminishes myocardial contractility. Thus vagal input has an important role during rest and recovery.
GP are recognized as “integration centers” between the intrinsic and extrinsic ANS and therefore can play a role in the induction of AF. Evidence suggests that the extrinsic component of the ANS suppresses the intrinsic component, and loss of this inhibition may lead to the induction of AF. , Animal studies have also indicated that a hyperactive intrinsic ANS may create an environment conducive to the induction of AF.
Ganglionic Plexus Mapping and Ablation
Localization and mapping of the GP can be performed using high-frequency stimulation (HFS). , The primary goal of GP ablation for AF is to ablate the five major GP in the LA, the four associated with specific PVs and one in the LOM. , , , These GP have been targeted in both endocardial and epicardial ablation procedures because of their proximity to the PVs and left atrial posterior wall. Although electrophysiological mapping provides a general estimate of the GP locations, precise localization of the plexi remains challenging.
Epicardial electrophysiologic mapping of GP was first reported during thoracoscopic surgical ablation in 2007 by Mehall et al. Subsequent mapping and ablation were also demonstrated in both limited thoracotomy and open chest surgical approaches. , Identification of an active GP involves low-voltage, HFS while monitoring for a vagal response. An active GP is identified when HFS of the atrial epicardial surface elicits a bradycardic response, with a greater than 50% increase in the R-to-R interval, resulting in a more than 50% heart rate reduction ( Fig. 30.4 ). An example of HFS parameters for GP identification is an 18-V impulse with a 1.5-ms pulse width at 1000 pulses per minute. , In the surgical setting, HFS is performed by using a transpolar pen that can pace, sense, and ablate.
Identification of a ganglionic plexus. The surface electrocardiogram documents that high-frequency stimulation yields a vagal response with slowing of the intrinsic heart rate.
As previously discussed, GP have both sympathetic and parasympathetic innervation. Because the sympathetic response is delayed until after the immediate parasympathetic response, HFS is delivered for 2 to 5 seconds to elicit the initial parasympathetic response and avoid interference from the later sympathetic response. The same transpolar pen used to identify the GP can then be used to ablate it. HFS that fails to illicit a vagal response at the same site after ablation confirms GP denervation ( Fig. 30.5 ).
Documentation of successful ganglionic plexus ablation. The surface electrocardiogram documents that high-frequency stimulation at the same site after ablation fails to elicit the vagal response of a slowing of the intrinsic heart rate.
Experimental Studies on Ganglionic Plexus Ablation for Atrial Fibrillation
Numerous animal studies have examined the long-term effects of autonomic denervation after RF GP ablation. Permanently damaging nerve cells requires destruction of the bodies of the cells within the GP. However, GP consist not only of cell bodies but also of interconnecting axons. Because of neuroplasticity, they can regenerate and even increase innervation through nerve sprouting in response to axonal injury. Thus neuroplasticity suggests the possibility of GP reinnervation after ablation, with heterogeneous reinnervation and proarrhythmic potential. Damiano et al. examined the acute and chronic effects of surgical GP ablation alone or with additional surgical ablation (pulmonary vein isolation [PVI], PVI plus box lesion) in canines. To ensure lesion transmurality, they dissected epicardial fat pads. Postoperatively, vagosympathetic trunk stimulation elicited significantly reduced responses from the atrial conduction system and myocardium, confirming initial atrial denervation. However, evidence of parasympathetic reinnervation was observed 4 weeks after surgery, revealing transient and nonuniform denervation, potentially increasing vulnerability to the development of a new arrhythmogenic substrate.
Mao et al. observed similar findings when they studied the electrophysiologic outcomes of GP ablation in dogs. They divided the dogs into three groups: group 1, pre- and postablation of four major GP and LOM GP; group 2, 8 weeks after ablation of four major GP and LOM GP; and group 3, sham-operated controls at 8 weeks. Immediately after GP ablation, group 1 exhibited significantly prolonged atrial refractory periods compared with preablation levels. However, at 8 weeks, group 2 showed a significantly shortened atrial refractory periods compared with the group 3 control subjects and the preablation levels in group 1. Moreover, the refractory periods were significantly shorter in the LA compared with the RA in group 2, which was not observed in group 3 control subjects. In group 1, atrial tachyarrhythmia (ATA) inducibility decreased immediately after GP ablation compared with before the ablation. However, after 8 weeks, dogs in group 2 showed significantly higher ATA inducibility. In group 2, ATA inducibility was higher in the LA than in the RA, unlike in the group 3 control subjects. Furthermore, immunohistochemical staining showed a significantly higher nerve density in GP-ablated dogs at 8 weeks (group 2) than controls (group 3). The authors suggested that incomplete denervation, along with factors such as axonal regeneration, nerve sprouting in response to RF energy nerve injury, and repeated HFS of GP, could have contributed to the observed changes in atrial refractory periods and ATA inducibility at 8 weeks after ablation.
Wang et al. evaluated the longer-term effectiveness of GP ablation in dogs at 6 and 12 months after ablation. Similar to Mao et al., they focused solely on ablating the RIGP and ARGP without performing PV or posterior wall isolation. Although there was an initial refractory period prolongation 1 month after GP ablation, refractory periods returned to preablation values at 6 and 12 months. AF was not inducible immediately before or after GP ablation. However, the ability to induce AF improved 1 month after the procedure and remained elevated compared with control subjects at 6 and 12 months. Immunohistochemical staining revealed reduced nerve density 1 month after GP ablation, but nerve density was similar to control subjects by 6 and 12 months.
As previously mentioned, the GP at the junction of the SVC, right pulmonary artery, and aorta is the primary relay between the extrinsic and intrinsic ANS. Po et al. ablated this “headway station” GP in five dogs and assessed atrial refractory periods acutely and after 10 weeks. A control group of five dogs underwent refractory period measurements without GP ablation. Although the refractory periods were prolonged initially after GP ablation, they were significantly shortened after 10 weeks, which contrasted with no change in the refractory periods in the control group. Furthermore, significant AF and ATA burden increased in the ablated group but not in the control group.
Clinical Outcomes of Ganglionic Plexus Ablation for Atrial Fibrillation
Surgical GP ablation has the advantage of direct access to the GP embedded in fat. In contrast, endocardial ablation requires penetrating through the myocardium to reach the GP, leading to difficulty in achieving complete GP ablation. During surgical access to the LAA, management often includes dissecting or ablating the LOM GP. Avazzadeh et al. provided a comprehensive review of surgical and catheter-based GP ablation studies in 2020. The following sections discuss selected observational and randomized studies.
Nonrandomized Studies of Surgical Ganglionic Plexus Ablation for Atrial Fibrillation
Nonrandomized studies of GP ablation during surgical ablation procedures have yielded mixed results regarding efficacy when added to traditional lesion sets.
Doll et al. conducted a small prospective study involving GP testing and ablation in 12 patients with paroxysmal AF (PAF) or persistent AF (4.5 ± 1.5 years’ duration) undergoing concomitant left atrial surgical ablation. They reported a sinus rhythm rate of 83% after 12 months of follow-up, concluding that GP ablation was feasible.
Onorati et al. performed a prospective study on 75 patients undergoing mitral valve surgery, in whom biatrial ablation procedures were conducted with or without additional GP ablation. The “mini-Maze” procedure, conducted using bipolar and monopolar RF ablation, involved PVI, mitral isthmus, and left superior PV to LAA ablation on the left side and SVC, IVC, cavotricuspid isthmus, right atrial appendage, and terminal crest ablation on the right side. Septal isthmus ablation was also performed if a patient had a large RA. They observed improved sinus rhythm rates with the addition of GP ablation. They also extended resection of the epicardial fat pads along Waterston’s groove, the LOM, and the left PVs for histological evaluation of GP post-operatively. Ware et al. compared 20 patients who received biatrial or left atrial surgical ablation with GP mapping and ablation during different valve replacement procedures to case-matched control subjects who had concomitant surgical ablation without GP ablation. They reported higher sinus rhythm rates at 12 months of follow-up with the addition of GP ablation. In a much larger retrospective study of 519 patients with nonparoxysmal AF (PAF) who underwent Cox Maze-IV ablation, the incorporation of GP ablation did not improve the maintenance of sinus rhythm or reduce AF recurrence.
Randomized Clinical Trials of Surgical Ganglionic Plexus Ablation for Atrial Fibrillation
The two randomized clinical trials that have evaluated GP ablation for AF were performed during open chest and minimally invasive thoracoscopic procedures, but the results did not favor GP ablation. The Japan Ganglionated Plexi Ablation Trial (JGPAT) was a multicenter randomized trial comparing 69 patients who underwent surgical Maze procedures with and without GP ablation during concomitant open chest structural heart procedures. Approximately 80% of patients had persistent or long-standing persistent AF (LSpAF). In the Maze plus GP ablation arm, ablation was performed at known anatomic GP sites and at 28 additional sites identified by HFS stimulation. The mean number of active GPs identified was 4.5. Postoperative pacemaker implantation rates and complications were similar with and without GP ablation. With a mean follow-up period of 16.3 months, there was no significant difference in freedom from AF or ATA between the two arms, with rates of 86.8% after Maze plus GP ablation and 91.4% after the Maze without GP ablation. The investigators speculated that in the control arm (Maze without GP ablation), the Maze procedure itself may have disrupted the GP by cardiac dissection and epicardial fat ablation, leading to denervation despite the lack of targeted GP testing and ablation.
The Atrial Fibrillation Ablation and Autonomic Modulation via Thoracoscopic Surgery (AFACT) trial was a single-center, randomized trial of 240 patients with advanced PAF and persistent AF who underwent thoracoscopic RF ablation and LAA exclusion with and without GP ablation. Thoracoscopic RF ablation involved PVI with a bipolar RF clamp for all patients and additional superior connecting and trigone lines for those with persistent AF. In the GP ablation arm, the GP were identified using either HFS or an anatomic approach, resulting in the ablation of all four major GP and the LOM GP. After the procedure, vagal responses were present in 0% of the GP ablation arm and 87% of the no GP ablation arm. Major adverse events were significantly higher in the GP ablation arm compared with the no GP ablation arm (19% versus 8%; P =.022), mainly because of bleeding, SA node dysfunction, and pacemaker implantation. At 1-year follow-up, freedom from ATA was comparable between the GP ablation arm and no GP ablation arm (70.9% versus 68.4%; P =.696), as were freedom from AF rates when analyzed by AF type. However, ATA recurrences were significantly more frequent in the GP ablation arm than the no GP ablation arm (78.1% versus 51%; P =.026). At 2-year follow-up, no additional procedure-related complications occurred, and rates of freedom from ATA remained similar between arms (55.6% in the GP ablation arm and 56.1% in the no GP ablation arm). As a result of these findings, the investigators advised against thoracoscopic GP ablation in this patient population. As a result of both trials, adjunctive GP ablation has largely been abandoned in most centers during either stand-alone, hybrid, or concomitant surgical ablation.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
