Notifications
The cardiac autonomic nervous system, consisting of the sympathetic and the parasympathetic components, has been shown to play an important role in many cardiovascular diseases, such as heart failure, hypertension, atrial fibrillation, and ventricular arrhythmias 1-4. An increased sympathetic nerve activity and a reduction of vagal cardiac tone are shown to be pathogenic in most of these conditions. In light of the autonomic imbalance involved in the development and progression of cardiovascular diseases, therapeutic interventions are focused on the inhibition of the sympathetic activation such as the use of β-receptor blockers, cardiac sympathetic denervation, and renal sympathetic denervation, and the increase of the parasympathetic activity such as vagus nerve stimulation. During the last decades, dysregulation of the autonomic nervous system in cardiovascular diseases has received considerable attention and interventions aiming at autonomic rebalance are emerging as new therapeutic options for the management of cardiovascular diseases. Autonomic modulation by electrical stimulation of the autonomic nervous system has been used clinically in patients with intractable angina pectoris 5-7, epilepsy 8, and depression 9, 10. Here, we review the experimental studies and clinical trials that are focused on autonomic modulation by electrical stimulation of the parasympathetic nervous system (ES-PNS), including vagus nerve stimulation (VNS), transcutaneous auricular vagal stimulation (tragus stimulation, TS), spinal cord stimulation (SCS), and ganglionated plexi stimulation (GPS), in the treatment of heart failure (HF), atrial fibrillation (AF), and ventricular arrhythmias (VAs).
Sympathetic activation and parasympathetic withdrawal are the characteristics of autonomic imbalance in HF 2, 11, 12. A variety of experimental studies have shown that cervical VNS produced powerful antifibrillatory effect 13, significant improvement in left ventricular function 14, attenuation in systemic inflammation 15, and marked reduction in mortality with increased long-term survival rate 16 in animals with myocardial infarction or HF. Based on the contribution of experimental studies, VNS was translated to clinical application for the treatment of HF. In 2008, Schwartz et al. 17 first performed long-term VNS in patients with advanced HF in man using the CardioFit 5000 device (BioControl Medical Ltd, Yehud, Israel) in which the bipolar stimulation electrode was surgically implanted around the right vagal nerve. Six-month follow-up showed a significant improvement in New York Heart Association (NYHA) classes, Minnesota quality of life, left ventricular end-systolic volume, and a favorable trend toward reduction in end-diastolic volume. After demonstrating the feasibility and safety of VNS for HF, they expanded the study to a multicenter single-arm open-label phase II study, which enrolled a total of 32 patients with a history of chronic HF 18. One-year follow-up showed significant improvements in NYHA classes, 6-minute walk test, quality-of-life test, left ventricular ejection fraction (LVEF), and left ventricular end-systolic volume index. Similar promising results were also obtained in a recent observational study 19, the ANTHEM-HF Trial, in which either left or right VNS was performed using an open-loop stimulation system. However, another randomized controlled trial, the NECTAR-HF trial 20, in which 96 symptomatic HF patients were randomized in a 2:1 ratio to receive VNS or control for a 6-month period, showed that VNS failed to significantly improve the primary and secondary endpoint measures of cardiac remodeling and functional capacity, even though quality-of-life measures and NYHA classes were significantly improved. The discrepancies in the results among these studies can to some extent be explained by the stimulation protocol, the stimulation intensity and frequency, the following time, and the small number of patients involved in these studies. It is important to mention that recent experimental evidence has indicated that different levels of parasympathetic stimulation release different neurotransmitters, which affect cardiac pathologies 21-23. For example, Stavrakis et al. 21, 22 have shown that low-level VNS, well below that which slows the heart rate, that is, via acetylcholine release, suppresses AF associated with the anti-adrenergic action of vasostatin-1 and the nitric oxide signaling pathway. Others have shown that vaso-intestinal peptide, co-released with acetylcholine, can influence the propensity for AF 23. Although the detailed mechanisms by which VNS induces favorable effects in HF require further basic and clinical research, the efficacy of VNS in the treatment of HF is expected to be answered by an ongoing international, multicenter, and randomized clinical trial, the INOVATE-HF trial 24, which will enroll 650 HF patients with NYHA class III symptoms, an LVEF ≤40% and left ventricular end-diastolic dimensions 50–80 mm in a 3:2 ratio to either receiving VNS or not.
Besides cervical VNS, transcutaneous VNS may also be an alternative noninvasive approach for the treatment of HF. In a recent study 25, we performed chronic intermittent low-level VNS by transcutaneous electrical stimulation of auricular branch of vagus nerve (TS) in conscious dogs with healed myocardial infarction. In the TS group, each dog was received 4 h stimulation at 7–9 AM and 4–6 PM. At the end of 90-day follow-up, the TS group showed significantly reduced left atrial and left ventricular dilatation, significantly improved left ventricular contractile and diastolic function and significantly reduced infarct size compared with myocardial infarction group. The protein expression level of collagen I, collagen III, transforming growth factor β1, and matrix metallopeptidase 9 in left ventricular tissues, which reflect cardiac fibrosis, was significantly decreased. The LVEF was increased, and the plasma level of N-terminal prohormone brain natriuretic peptide (NT-proBNP) was decreased.
Another intervention approach that enhances the parasympathetic activity is SCS. Olgin et al. 26 showed that SCS at the T1 to T2 level enhanced cardiac parasympathetic activity. In a chronic HF model induced by anterior myocardial infarction and rapid pacing, Lopshire et al. 27 revealed that dogs received SCS for 5 weeks had significantly higher LVEF (47%) than that dogs received carvedilol (34%) or no therapy (28%), respectively. Similar findings were observed with SCS in a porcine animal model of ischemia and reperfusion 28. Recently, Tse et al. 29 reported the results of the SCS HEART study, which evaluated the safety and efficacy of a SCS system for the treatment of systolic HF. All the patients had a LVEF of 20–35% with NYHA class III. A significant improvement in NYHA class, Minnesota Living with Heart Failure Questionnaire, peak maximum oxygen consumption, LVEF, and left ventricular end-systolic volume was observed in the treated patients. But the NT-proBNP was not significantly changed. However, the primary results of the DEFEAT-HF study 30, the first single-blinded randomized controlled trial of SCS for HF, revealed that there was no significant difference in left ventricular end-systolic volume index, peak maximum oxygen consumption, or NT-proBNP between the SCS and control groups after 6-month follow-up. Although both of these two studies demonstrated that chronic SCS is safe and feasible, whether SCS can improve left ventricular function and structural remodeling in patients with HF requires further studies.
In recent decades, vagal overactivity has been shown to serve as a causative mechanism of AF. Thus, electrical stimulation of vagus nerve at a voltage level that slows the sinus rate or atrioventricular conduction has been utilized to induce AF in experiments. However, recent experimental and clinical studies revealed that low-level VNS, that is the voltages/currents used in VNS does not slow the sinus rate or atrioventricular conduction, induced paradoxical effects to suprathreshold VNS, and exerted an anti-arrhythmic role. Li et al. 31 first reported the anti-arrhythmic role of low-level VNS in 2009. In their study, cervical low-level VNS continued for 3 h and AF threshold was determined by high-frequency stimulation at each pulmonary vein and atrial appendages site at the end of each hour. The results showed that low-level VNS induced a progressive increase in AF threshold at all pulmonary vein and atrial appendages sites. Yu et al. 32 also demonstrated that cervical low-level VNS inhibited AF inducibility and prevented the shortening of effective refractory period (ERP) at pulmonary vein and atrial sites and the increase of ERP dispersion induced by atrial GPS. Low-level VNS also markedly suppressed the frequency and amplitude of the neural activity recorded from the major atrial GPs. Since then, a series of experimental studies indicate that either unilateral or bilateral cervical low-level VNS can induce prevention and reversal of atrial electrophysiological remodeling and autonomic remodeling induced by rapid atrial pacing as well as inhibition of AF inducibility 33-35. The anti-arrhythmic role of low-level VNS is shown to be associated with its anticholinergic and anti-adrenergic effects, which may be mediated by vasostatin-1 and nitric oxide 21, 22.
Although low-level VNS is shown to be a promising treatment of AF, cervical VNS is an invasive approach in which cervical surgery is needed to position vagal stimulation electrode. Besides, VNS may cause some adverse effects such as cough, voice alteration, dyspnea, pain, paresthesia, and headache 36. Recently, a noninvasive approach for VNS, transcutaneous electrical stimulation of the auricular branch of the vagus nerve located at the tragus (TS), was introduced by some studies. In a canine AF model induced by rapid atrial pacing, Yu et al. 37 showed that low-level TS suppressed AF and reverse acute atrial electrophysiological remodeling. Clancy et al. 38 performed electrical stimulation of tragus nerve in 48 healthy human beings and demonstrated that TS significantly decreased sympathetic nerve activity and increased heart rate variability. More recently, a randomized clinical study showed that 1-h low-level TS suppressed ERP shortening and AF inducibility as well as shortened the AF duration in paroxysmal AF patients 39. One-hour low-level TS also reduced inflammatory factors of CRP and TNF-α in this study.
SCS may represent another treatment option for AF. In a canine AF model induced by tachypacing, Bernstein et al. 40 demonstrated that SCS significantly prolonged atrial ERPs by 21 ± 14 ms in the left atrium and 29 ± 12 ms in the right atrium and reduced AF burden and inducibility. Also in a canine model of rapid atrial pacing-induced AF 41, we showed that SCS attenuated the decrease in ERP, the increase in ERP dispersion, the window of vulnerability and AF inducibility, and the activation of anterior right GP and left stellate ganglion, which were induced by rapid atrial pacing. SCS also prevented the upregulation of c-fos and nerve growth factor and the downregulation of small conductance calcium-activated potassium channel type 2 induced by rapid atrial pacing in our study. These results indicate that SCS might suppress rapid atrial pacing-induced AF by inhibiting the neural remodeling.
It has been known that parasympathetic overactivity is protective for ventricular arrhythmogenesis. In animal models of acute myocardial ischemia and ischemia reperfusion 42, 43, cervical VNS has been shown to suppress the incidence of ventricular arrhythmias, especially ventricular tachycardia and ventricular fibrillation. In conscious animal with healed myocardial infarction, direct vagal stimulation also effectively prevents ventricular arrhythmias including ventricular fibrillation 13, 44. Multiple mechanisms are shown to involve in the anti-arrhythmic effect of VNS, such as bradycardiac effect 42, anti-adrenergic effects 4, prevention of the loss of phosphorylated connexin 43 proteins 45, and inhibition of the opening of the mitochondrial permeability transition pore 46. Although the cardiac protective effect of VNS is mainly mediated by muscarinic receptor, recent study indicated that the vagal antifibrillatory action can be mediated by neuronal nitric oxide released by intracardiac neurons or nerve fibers 47, which reduces the maximal slope of the ventricular action potential restitution curves 48.
SCS has also been shown to play an anti-arrhythmic role in ventricular arrhythmias. Issa et al. 49 first reported that thoracic SCS reduces the risk of ischemic ventricular arrhythmias in a postinfarction heart failure canine model in which the incidence of ventricular tachycardia/fibrillation was reduced from 59% to 23%. Similar results were reported by Lopshire et al. 27 that intermittent thoracic SCS significantly improved left ventricular function and reduced spontaneous and ischemic ventricular arrhythmia. In a more recent study 50, we demonstrated that SCS significantly prevented the incidence of ventricular arrhythmias in an acute myocardial infarction canine model, possibly by suppressing left stellate ganglion activity. One-hour SCS also significantly prolonged ventricular ERP and increased heart rate variability.
We also investigated the anti-arrhythmic role of atrial GPS. In the normal heart, GPS prolonged ventricular ERP, decreased the slope of ventricular action potential restitution curves, and delayed action potential duration alternans 51. In the ischemic heart, atrial GPS significantly inhibited the incidence of ventricular arrhythmias not only in acute myocardial ischemia 52 but also in ischemia reperfusion 53. Atrial GPS also increased heart rate variability and prevented the loss of connexin43 induced by ischemia/reperfusion 53. Transvenous endocardial atrial GPS has been safely and feasibly applied to ventricular rate control during AF 54-56, and it is reasonable that transvenous endocardial atrial GPS may also be a feasible approach for the treatment of VAs or HF 57. Further studies are warranted to clarify this issue.
Autonomic modulation by electrical stimulation of the parasympathetic nervous system, including cervical vagus nerve, auricular vagus nerve, spinal cord, and atrial ganglionated plexi, has emerged as a new therapeutic modality to treat HF, AF, and VAs. As cervical VNS, SCS, and GPS are invasive technologies, transcutaneous VNS (TS) is expected to serve as a noninvasive approach. Besides the standard neurotransmitters, other neurotransmitters such as vasostatin-1, nitric oxide, and vaso-intestinal peptide may also participate in the pathologic process of cardiovascular diseases and may serve as new modulation targets. Further basic and clinical studies are required to clarify the detailed roles of the autonomic nervous system in the development and progression of cardiovascular diseases and to provide a better understanding of the long-term consequences of device-based electrical stimulation of the autonomic nervous system.
This work was supported by grants 81300181 (BH), 81370281 (ZL), 81400254 (WH), and 81530011 (HJ) from the National Natural Science Foundation of China.
