Marshaling the Autonomic Nervous System for Treatment of Atrial Fibrillation

A number of mechanisms have been proposed to explain the pathogenesis of atrial fibrillation (AF), and some are intricately linked to the autonomic nervous system (ANS) (1). A potential neural influence in AF was suggested by Coumel et al. (2) in 1978 and has since been extensively investigated. Experimentally, stimulation of autonomic nerves during the atrial refractory period has been shown to produce rapid ectopic beats from the pulmonary veins (PVs) and superior vena cava, which in turn can initiate AF (3,4). It is now well recognized that the ANS has an important contribution to the pathogenesis of AF (5,6). However, AF remains poorly understood, and the specific mechanisms underlying the relationship between the ANS and AF have yet to be fully elucidated. Despite this limited knowledge base, both cardiac and extra-cardiac autonomic structures (such as the renal arteries) have been targets of AF therapy (7).

Anatomically, the ANS has direct projections to the heart and also indirect projections via the intrinsic cardiac nervous system (ICN), distributed into ganglionated plexi (GPs). GPs are not simply “relay stations” for ANS projections to the heart; they contain sympathetic efferent, parasympathetic efferent, sensory afferent, and local circuit neurons (8,9). Extensive processing of these inputs occurs at the level of the ICN.

There is increasing evidence that the ICN plays an active role in the regulation of cardiac function. For example, the right atrial ganglionated plexus (RAGP), found near the right PV complex, mediates vagal control of the sinoatrial (SA) node. Disruption of the RAGP eliminates para-sympathetic projections to the SA node but leaves intact parasympathetic-mediated suppression of sympathetic projections to the SA node (10). GPs exert influence locally but can also influence distant regions. Stimulation of neurons in the RAGP with nicotine, for example, causes changes in unipolar wave forms not only in the atria but also surprisingly in the ventricles. Thus, spatially divergent and overlapping cardiac regions are under the influence of neurons throughout the ICN (11). This also implies a certain degree of redundancy in the network.

Imbalances in the ICN are thought to be involved in the pathogenesis of AF. Although still controversial, numerous strategies have been used to ablate GPs as adjunctive (12) or stand-alone (13) therapy in the treatment of AF. These approaches have relied on functionally identifying GP regions to target energy delivery, on the basis of autonomic responses to electrical stimulation. Other methods are anatomic, such as targeting the vein of Marshall (VOM) (14), either surgically or by catheter-based approaches (15–17).

The present study

In this issue of the Journal, Báez-Escudero et al. (18) describe the use of the VOM as a vascular route for the ablation of ICN neurons. The VOM is continuous with the ligament of Marshall (LOM), a vestigial remnant of the embryologic left cardinal vein (14). The LOM is an epicardial fibrous and neuromuscular bundle that runs from the roof of the left atrium, past the left superior PV, and inserts into the distal coronary sinus (CS) (19). Several studies have now shown that the LOM can trigger and help maintain AF. Doshi et al. (20) demonstrated that isoproterenol infusion was able to induce automatic activity in the LOM that led to AF. Furthermore, histologic evaluation of the LOM revealed muscle tracts surrounded by nerve bundles positive for tyrosine-hydroxylase (a sympathetic ANS marker), providing more evidence in support of the relationship between the ANS and ICN in the initiation of AF (20). Because the VOM has a direct connection with the CS, the VOM can be used to access the ICN neurons within the LOM. Antiarrhythmic drugs have been injected via the CS to terminate AF; presumably such drugs could interact with these neural structures (21).

In this study, the VOM was cannulated in patients who required either de novo or repeat AF ablation. Para-sympathetic responses (asystole or RR prolongation) were elicited with high-frequency stimulation, among patients who could have the VOM cannulated. Chemical ablation of the VOM with retrograde ethanol infusion eliminated parasympathetic responses and rendered AF noninducible (when tested via high frequency stimulation within the VOM) in a majority of the patients.

These findings are important on multiple fronts. Although AF ablation is improving, currently the success rate is suboptimal, with many patients requiring repeat procedures. Thus, new approaches to treat AF are critical. This study was possible due to our improved understanding of the link between the ANS and AF; targeted GP ablation might become another tool in the therapeutic armamentarium against AF.

Although the findings are intriguing, there are some limitations. A study by Leiria et al. (22) demonstrated that cryoablation of cardiac mediastinal nerves with input into the ICN resulted in a short-term reduction in AF. However, this stunning effect was time-dependent, and AF induction recovered to nearly baseline levels several months later (22). This recovery of inducibility with time illustrates the plasticity of the ICN in response to local injury. There is also evidence to suggest disruption of the ICN can have pro-arrhythmic effects on atrial tissue (23). Another limitation is that it is not possible to assess the contribution of atrial and PV myocardial junction ablation to the reduction of AF induction (24). There is also the clinical reality that the VOM is absent or too small to be cannulated in some patients, potentially increasing procedure time.

Finally, the effects of GP ablation on ventricular electrophysiology and arrhythmogenesis remain to be carefully evaluated. A recent study by He et al. (25) demonstrated that GP ablation significantly increased the risk of ventricular arrhythmias in the setting of acute myocardial ischemia (25). Thus, targeting of the ICN has the potential to create further instability in the neural network and exacerbate the risk for arrhythmias, and these patients should be carefully followed (26).

Implications

The findings of this study underscore the importance of improving our understanding of the role of the ICN in regulating cardiac function and in arrhythmogenesis. Careful evaluation of the short- and long-term effects of disrupting ICN elements on both atrial and ventricular function and electrophysiology are crucial before targeted GP ablation becomes part of clinical practice.