Cardiac Neuroanatomy and Fundamentals of Neurocardiology

Introduction

The autonomic nervous system regulates all aspects of cardiac function including chronotropy, inotropy, dromotropy, and lusitropy.¹ Furthermore, autonomic dysregulation plays a major role in the development and progression of cardiovascular diseases including myocardial infarction, arrythmias, and heart failure.^1–4 While pharmacologic therapy target cardiac disease through selective neurohormonal blockade, in recent years, their therapeutic efficacy has asymptoted.^{4, 5} Leveraging the dynamic interplay between the autonomic nervous system and heart in the evolution of cardiac pathology has led to the emerging field of autonomic neuromodulation as a new therapeutic approach for cardiac disease. For such neuroscience-based approaches, it is imperative to understand the anatomy and function of the cardiac nervous system to guide development and implementation of targeted neuromodulatory therapies.⁶

The anatomy of the cardiac nervous system is complex and has been categorized into: 1) central 2) intrathoracic extracardiac and 3) intrinsic cardiac components (Figures 1 and 2).^{1, 2} There are five concepts that are central to understanding the structure- and function-based organization of the cardiac nervous system in normal and diseased states. Concept 1: Neural control of cardiac function involves a multi-tier hierarchy of interdependent reflexes.^{2, 3} Concept 2: Coordination of peripheral ganglia function depends on inter- and intra-ganglionic network interconnectivity (mediated in part via local circuit neurons), afferent projections arising from cardiac (atrial and ventricle) and extracardiac (e.g. carotid sinus) sensory neurites and descending projections from the central nervous system via parasympathetic and sympathetic-related circuits.^{2, 3, 7–9} The parasympathetic component, arising from preganglionic networks in the nucleus ambiguus and dorsal motor nucleus, projects via cranial nerve X and merges with intrathoracic sympathetic projections to form the vagosympathetic nerve trunk.^{2, 3, 10} The parasympathetic system uses acetylcholine, nitric oxide, and vasoactive intestinal peptide as neurotransmitters.^{11, 12} The sympathetic component originates in the intermediolateral cell columns of the spinal cord, projects via C7 to T6 rami to the superior cervical, middle cervical, stellate or cervicothoracic ganglia, and acts though neurotransmitters norepinephrine and neuropeptide Y.^{10, 11, 13–15} The intrinsic cardiac nervous system (ICNS), the most peripheral of these independent networks, functions with higher centers to modulate regional cardiac electrical and mechanical indices on a beat-to-beat basis.¹⁶ It contains all the neural network elements required for regional cardiac control, including afferents (sensory feedback), local circuit neurons (information processors), and efferent (motor) outputs (both sympathetic and parasympathetic).¹⁷ The ICNS is even capable of reflex control of regional cardiac function in the absence of all higher elements (e.g. after heart transplant).^{18, 19} It is the dynamic reflex interplay between all three levels of the cardiac nervous system that ultimately determines neural control of regional cardiac electrical and mechanical function.

Figure 1:

Neural regulation of cardiac function involves multiple nested feedback loops at the level of the heart, peripheral ganglia, and central nervous system. Afferent systems (blue) are mediated through the intrinsic cardiac nervous system, dorsal root ganglia, and nodose ganglia. Efferent systems (red) involve sympathetic, parasympathetic, and local circuit neurons. DRG, dorsal root ganglia. ICNS, intrinsic cardiac nervous system. Adapted from Hanna, 2018⁷⁷ with permission.

Figure 2:

Multiple interdependent feedback loops regulate regional cardiac mechanical and electrical function. At the level of the heart, sensory (blue) neurons provide input directly and indirectly through the intrinsic cardiac nervous system (ICNS), peripheral ganglia, and spinal cord with subsequent projections to higher centers. Preganglionic sympathetic and parasympathetic efferent fibers project (red, dashed) to the ICNS, including to local circuit neurons, and postganglionic sympathetic fibers (red, solid) also directly to the myocardium. The ICNS acts as the final common pathway for cardiac control, though interactions occur at all levels. Input from chemoreceptors, baroreceptors, as well as neurohormonal factors such as angiotensin II and circulating epinephrine (not displayed), also modulate cardiac function. Ang I, angiotensin I. Ang II, angiotensin II. LCN, local circuit neuron. Adapted from Hanna et al, 2018⁷⁷ with permission.

While much is gained from understanding the structural and functional organization of the cardiac nervous system at rest, applying stressors to the system reveals critical aspects of how the system responds to deviations from homeostasis and diversity in patient populations with respect to their ability to respond to such challenges.^20–24 For example, disruptions in peripheral ganglia network processing are arrhythmogenic (Concept 3). To address this concept using a preclinical model of atrial fibrillation (AF), we utilized burst stimulation of mediastinal nerves to induce short periods of reproducible AF (of ~30 sec duration) with an onset latency of ~1 sec.^25–28 ICNS activity, derived from extracellular recording, can be functionally defined into afferent, efferent, or convergent (both afferent and efferent) (Figure 3).⁸ While burst stimulation of mediastinal nerves increases activity in all three functional classes of ICNS neurons, convergent neurons were the preferential target.²⁹ The neural nature of this AF model is confirmed by its extinction with pre-emptive ganglionic blockade and its mitigation by cholinergic, adrenergic, and peptidergic blockers.^{25, 27, 30, 31} This model of AF induction can be truncated by pre-emptive neuromodulation therapy including spinal cord and cervical vagal nerve stimulation (VNS), both of which target information processing within the ICNS networks (Figure 3).^{29, 31, 32} Further support for the concept of neural involvement in AF is indicated by clinical studies with auricular nerve stimulation for post-operative and paroxysmal AF^33–36 and by the increased efficacy of AF ablation when neural network processing within the ICNS is blunted by either focal ablations or pharmacological injections.^{37, 38}

Figure 3.

Vagal nerve stimulation targets local circuit neurons within the intrinsic cardiac nervous system to reduce arrhythmogenesis. (A) In response to mediastinal nerve stimulation, activity within the right atrial ganglionated plexus significantly increased and transient periods of AF were induced. (B) Increased neural activity during mediastinal nerve stimulation was primarily driven by increased activity of local circuit neurons. (C) VNS mitigated neural activity changes in response to mediastinal nerve stimulation and prevented or blunted atrial arrhythmogenesis. Adapted from Salavatian et al., 2016²⁹ with permission.

One of the fundamental unmet challenges in the field of neurocardiology is the identification of biomarkers that are predictive of adverse events in response to sudden stressors. This point is the cornerstone of Concept 4: there are inherent and acquired factors that impact the progression of cardiac disease. To this end, consider the case of acute coronary syndrome and ventricular arrhythmias. In a preclinical model of myocardial infarction, animals could be classified into three groups: super-susceptible that experience sudden cardiac death (SCD) during the initial infarction (usually within 20 min of presentation), susceptible that survive the initial heart attack but are at high risk for SCD in subsequent ischemic challenges, and resistant that survive the initial heart attack and all subsequent ischemic challenges.^39–41 Data, primarily derived from preclinical studies, indicates that those with excessive and heterogeneous sympatho-excitation to stressor, such as ischemia, are those that are at high-risk for adverse outcomes, including potential for SCD.⁴² As a corollary, if pre-emptive neuromodulation therapy can be applied to dampen this sympathetic response, individuals can be rendered resistant regardless of the inherent differences in reflex processing.^42–45 Defining the biomarkers that are predictive of future adverse events will be important in patient selection for neuromodulation therapy.

Acquired factors likewise are a major determinant of the progression of cardiac disease, involving both cardiac substrate and the autonomic reflexes that control the heart. Figure 4 depicts such induced changes following a heart attack leading to chronic myocardial infarction. The characteristics of the scar impact cardiac electrical stability; dense scars being more stable than patchy scars.^{39, 46} These heterogeneities in conduction pathways can be amplified by afferent-driven alterations in cardiac reflexes, from the ICNS, through the extracardiac intrathoracic sympathetic ganglia and their connections with the spinal cord and higher centers of the central nervous system.^{21, 47 48} Focusing on the stellate ganglia, altered afferent transmission leads to oxidative stress, lipofuscin deposition, inflammation and neutrophil infiltration, neurochemical remodeling (e.g. increased Neuropeptide Y expression), and glial activation.^{46, 49, 50} These focal changes, in addition to alterations in central peripheral neural networks, lead to a net increase in sympatho-excitation.^{1, 5} Altered afferent transmission correspondingly results in a decrease in central parasympathetic drive.⁵¹ Together, the excessive activation of sympathetic outflow with minimal counteracting central drive of parasympathetic activity translate into an increase potential for arrhythmias, including the potential for sudden cardiac death.¹ Stone and colleagues, in a series of preclinical studies, demonstrated that even in subjects that survived the initial myocardial infarction (approximately 70% survived), subsequent exercise stress with transient myocardial ischemia resulted in SCD in approximately half of the remaining animals.^{39, 41} These animals represent the susceptible group. The remaining 1/3 of animals reflect the resistant group.

Figure 4.

Potential pathways driving adverse remodeling within stellate ganglia and impact of efferent sympathetic neurotransmission. Potential afferent-driven alterations in parasympathetic outflow also depicted. Right panel summaries primary targets for neuroscience-based neuromodulation of the cardiac nervous system in setting of chronic heart disease. Adapted from Ajijola et al, 2017⁴⁹ with permission.

Treatment of arrhythmias in the patient with ischemic heart disease has primarily involved pharmacological, surgical, device-based (e.g., implantable cardioverter-defibrillator [ICD]) and/or ablation approaches.¹ While showing some efficacy, there are significant shortcomings in their long-term effectiveness and often carry with them substantial off-target effects. Neuroscience-based neuromodulation therapies offer the opportunity to intervene at earlier time points and to thereby alter the trajectory of cardiac disease and render a state of overall cardioprotection. Concept 5: Neuromodulation-based therapies impact multiple levels of control to exert cardioprotection. There are three primary targets for such neuromodulation (Figure 4): aberrant afferent signaling, restoration of parasympathetic tone, and mitigation of sympatho-excitation. With regards to aberrant afferent signaling, the cardiac nervous system can be catastrophically overwhelmed by the activation of cardiac nociceptors.^{20, 21} This can be mitigated by disruption of the primary sensory inputs, preclinically for now, by the injection of resiniferatoxin (RTX), a selective transient potential vanilloid 1 (TRPV1) receptor agonist, that blunts cardiac sympathetic afferent reflexes with long-duration efficacy.^{52, 53} Parasympathetic tone can be bioelectrically restored using either transcutaneous (auricular)^{34, 35, 54–56} or direct cervical vagal nerve stimulation (VNS).^57–59 In patients, auricular nerve stimulation is effective in mitigating post-operative and paroxysmal AF and could potentially serve as a screening tool prior to implant of cervical VNS devices.^{33, 36} In preclinical models, reactive cervical VNS reduced ventricular arrhythmias (Figure 5), mitigated adverse neural cardiac remodeling post-myocardial infarction, and restored sympathetic control to “normal” levels. Bioelectric interventions, applied to spinal cord and/or cervical vagus nerve, similarly mitigate aberrant afferent signaling from cardiac nociceptors.^{48, 60, 61} Excessive sympatho-excitation itself can be selectively targeted using axonal modulation therapy (Kilohertz AC or charge-balance DC partial block) applied to the T1-T2 paravertebral chain, a scalable, reversible, and on-demand methodology that has the potential to eventually replace the ICD.^62–64 As is evident from the above discussion, the neuroscience-based targets are interdependent and often work together to exert cardioprotection. To date, these therapies are delivered in open-loop fashion. Future efforts are being directed towards transitioning these approaches to closed-loop systems, with the realization that each will have to be optimized for the specific disease process being targeted.

Figure 5.

Antiarrhythmic effects of chronic vagal nerve stimulation. (A) Representative left ventricular pressure tracing, and unipolar and bipolar electrograms during programmed electrical stimulation in myocardial infarction (MI) and MI + cervical vagal nerve stimulation (cVNS) animal. (B) Chronic VNS significantly reduced the inducibility of sustained ventricular tachycardia/fibrillation (VT/VF) regardless of whether PES was performed with or without active VNS. (C) MI animals demonstrated significantly greater ease of VT/VF induction compared to MI + cVNS pigs. Ctrl, control; EGM; electrogram; LV, left ventricular; NSVT, non-sustained ventricular tachycardia; PES, programmed electrical stimulation; Data presented as count (B) or mean ± SEM (C) and analyzed using chi-squared test (B) or Kruskal-Wallis and Wilcoxon rank-sum test (C). *p<0.05, **p<0.01, ***p<0.001. Adapted from Hadaya et al, 2023⁴⁶.

Translational Implications of Neurocardiology

Expanding our understanding of structural and functional cardiac neuroanatomy promotes the development of neuromodulation-based therapies for coronary artery disease, heart failure, and arrhythmias. Since cardiac sympathetic denervation to ameliorate cardiac chest pain was initially described over a century ago,⁶⁵ surgical and percutaneous coronary revascularization strategies have come to the fore. However, for cases of coronary vasospasm refractory to medical therapy, stellate ganglionectomy has been pursued with good effect.⁶⁶ The use of cardiac sympathetic denervation has grown to include targeting long QT syndrome,⁶⁷ catecholaminergic polymorphic ventricular tachycardia,⁶⁸ and refractory ventricular tachycardia.⁴⁵ Bolstered by supporting evidence, cardiac sympathetic denervation is now a Class IIb recommendation in patients with refractory VT/VF storm per the 2017 AHA/ACC/HRS Guidelines for Management of Patients with Ventricular Arrhythmias.⁶⁹ Pre-clinical studies using scalable, on-demand, and reversible axonal blockade of the stellate ganglia offer promise of a bioelectric alternative to surgical sympathectomy.^{63, 64} Similarly, the impact of autonomic tone on atrial fibrillation has been recognized clinically for decades,^{70, 71} and potential interventions at the cardiac neuraxis are now being studied in patients. Freedom from AF recurrence post-catheter ablation has correlated to amount of neural injury.⁷² However, studies of neuro-interventions for post-operative AF have yielded mixed results with botulinum toxin type A injection into epicardial fat pads at time of cardiac surgery unable to reduce AF burden while low-level electrical VNS was shown to suppress post-operative AF.^{33, 73} More promisingly, low-level tragus (auricular) stimulation has been shown to reduce AF burden in a randomized, controlled trial of patients with paroxysmal AF.³⁵ Neuromodulation may counteract the autonomic remodeling that occurs in AF to improve upon the success rates of catheter ablation. Neuroablative approaches that target the ICNS for bradyarrhythmias and syncope to obviate the need for permanent pacemaker implantation in younger patients has been pursued with great interest.^{74, 75} Future studies should hone in on ideal nexus points within the cardiac nervous system and mode of targeting. We favor a neuromodulatory over a neuroablative approach, as cardioneuroablation has not proved to reduce AF burden thus far and a case report demonstrated recurrence when applied to syncope.^{38, 76} Harnessing the exquisite control of the cardiac nervous system over cardiac function in disease promises to yield effective and targeted neuromodulatory therapies.

Synopsis.

Cardiac control is mediated via nested-feedback reflex control networks involving the intrinsic cardiac ganglia, intra-thoracic extra-cardiac ganglia, spinal cord, brainstem, and higher centers. Each of the neural networks contains afferent, efferent, and local circuit neurons that interact locally and in an interdependent fashion with the other levels of the neuroaxis to coordinate regional cardiac electrical and mechanical indices on a beat-to-beat basis. This control system is optimized to respond to normal physiological stressors (standing, exercise, temperature); however, it can be catastrophically disrupted by pathological events such as myocardial ischemia. In fact, it is now recognized that cardiac disease progression reflects the dynamic interplay between adverse remodeling of the cardiac substrate coupled with autonomic dysregulation. With advances in understanding of this network dynamic in normal and pathological states, neuroscience-based neuromodulation therapies can be devised for management of acute and chronic cardiac pathologies. Such targeted autonomic regulation therapy has the potential to mitigate the progression of pump failure and arrhythmia potential while at the same time preserving sympathetic and parasympathetic reflex control of the heart.

Keypoints:

1. Cardiac control involves nested-feedback reflex networks, spanning intrinsic cardiac ganglia, spinal cord, brainstem, and higher centers.

2. Neural networks coordinate cardiac indices beat-to-beat, responding to physiological stressors but susceptible to disruption by events like myocardial ischemia.

3. Cardiac disease progression results from adverse remodeling and autonomic dysregulation, emphasizing the dynamic interplay between these factors.

4. Understanding cardiac neural network dynamics enables neuroscience-based therapies for acute and chronic pathologies, aiming to preserve autonomic reflex control.

5. Advances in neuromodulation offer potential in managing pump failure and arrhythmia while maintaining sympathetic and parasympathetic control of the heart.