Effect of the stellate ganglion on atrial fibrillation and atrial electrophysiological properties and its left-right asymmetry in a canine model

Abstract

OBJECTIVE:

To investigate the effect of the stellate ganglion (SG) and its left-right asymmetry on atrial fibrillation (AF) inducibility, AF duration and atrial electrophysiological properties.

METHODS:

Sixteen adult mongrel dogs were randomly assigned to three groups. The control group (n=4) underwent 6 h rapid atrial pacing (RAP) only; the right SG (RSG) group (n=6) underwent 6 h RSG stimulation plus RAP; and the left SG (LSG) group (n=6) underwent 6 h LSG stimulation plus RAP. AF induction rate, AF duration, effective refractory period (ERP) and dispersion of ERP (dERP) were measured.

RESULTS:

In the RSG group, the induction rate of AF was significantly increased in sites in the right atrium (RA) compared with baseline (P<0.05). In the LSG group, the induction rate of AF was significantly increased (P<0.05) compared with baseline in the left atrium (LA), left superior pulmonary vein and left inferior pulmonary vein, respectively. Compared with RSG stimulation, right stellate ganglionectomy markedly decreased the AF induction rate of the RA (P<0.05). Compared with LSG stimulation, left stellate ganglionectomy markedly decreased the AF induction rate of the LA, the left superior pulmonary vein and the left inferior pulmonary vein (P<0.05). In the RSG group, the ERP was significantly shortened (P<0.05) and the dERP was significantly increased (P<0.05) in RA sites (P<0.05). The ERP was significantly shortened in the LSG group (P<0.05). The dERP was significantly increased (P<0.05) in LA and pulmonary vein sites (P<0.05).

CONCLUSIONS:

Unilateral electrical stimulation of the SG in combination with RAP can successfully establish a canine model of acute AF mediated by excessive sympathetic activity. SG stimulation facilitates AF induction and aggravates electrical remodelling in sites in the atrium and pulmonary vein. Inhibiting sympathetic nerve activation through unilateral stellate ganglionectomy can reduce AF initiation.

Atrial fibrillation (AF) is a complex arrhythmia with multiple possible mechanisms. It requires a trigger for initiation and a favourable substrate for maintenance. However, the nature of the trigger remains elusive. Many experimental and clinical studies have indicated that both the sympathetic and parasympathetic nervous systems are important in the initiation and maintenance of AF (1–5). More importantly, in recent years, direct autonomic nerve recordings in canine models have demonstrated that simultaneous sympathovagal discharges are the most common triggers of paroxysmal atrial tachycardia and paroxysmal AF (6). Some studies have shown that stellate ganglion (SG) stimulation facilitates the induction of AF, and bilateral SG ablation or stellectomy can prevent or reduce AF episodes (7). However, a bilateral SG block would result in unnecessary damage and side effects. Therefore, in the present study, we performed unilateral SG stimulation plus rapid atrial pacing (RAP) to establish a canine model of sympathetic-induced AF. The purpose was to investigate the effect of enhancement or reduction of cardiac sympathetic outflow on AF inducibility, AF duration and atrial electrophysiological properties. We also attempted to determine the left-right asymmetry of the SG on heart rate, blood pressure, AF inducibility and atrial electrophysiological properties.

METHODS

Ethics statement

The present study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee of the First Affiliated Hospital of Xinjiang Medical University, Urumqi, China (permit number IACUC-20110325009) and conformed to the guidelines of the Association for Assessment and Accreditation of Laboratory Care.

Group setting and animal preparation

Sixteen adult mongrel dogs (body weight 18 kg to 22 kg) were randomly assigned to three groups. The control group (n=4) underwent 6 h RAP only. The right SG (RSG) group (n=6) underwent 6 h RSG stimulation plus RAP. The left SG (LSG) group (n=6) underwent 6 h LSG stimulation plus RAP. All dogs were anesthetized using sodium pentobarbital (20 mg/kg) and ventilated with room air using a positive pressure respirator. Ketamine (25 mg/kg) was used as an analgesic and all efforts were made to minimize suffering. Core body temperature was maintained at a mean (± SD) temperature of 36.5±1.5°C. Standard electrocardiographic leads were continuously recorded to determine the heart rate and rhythm.

Right femoral arteries and veins were cannulated and used for catheter insertion and blood pressure recording. The standard electrocardiographic lead II and blood pressure were continuously monitored using a pressure transducer. The right external jugular vein was cannulated and used for catheter insertion into the right atrium (RA) to record right atrial potentials and to perform RAP.

A left-sided thoracotomy was performed at the fourth intercostal space. Multielectrode catheters were sutured to the left atrial appendage (LAA), left superior pulmonary vein (LSPV) and left inferior pulmonary vein (LIPV).

Canine model of acute AF mediated by excessive sympathetic activity

Unilateral electrical stimulation of the SG plus RAP for 6 h were used to establish a canine model of acute AF mediated by excessive sympathetic activity. Before pacing and stimulation, the baseline electrophysiological data of AF inducibility, AF duration, effective refractory period (ERP) and dispersion of ERP (dERP) were measured. AF was defined as an irregular atrial rate >500 beats/min and a duration >5 s associated with irregular atrioventricular conduction (8). After pacing every 2 h, RAP and LSG stimulation was temporarily discontinued to obtain the electrophysiological data.

RAP and electrophysiology:

Continuous RAP (600 beats/min, 0.5 ms, twice-threshold current) was administered to all dogs at the RA for 6 h. Programmed stimulation at atrial sites or pulmonary vein sleeves was performed using a programmable cardiac stimulator (Lead-2000 EP Control, Sichuan Jinjiang, China). ERP was defined as the longest S1–S2 interval that failed to produce a response. ERP was measured at an atrial pacing cycle length of 300 ms, and the S1–S2 intervals were decreased from 200 ms by increments of 5 ms until no response was observed (S1:S2=8:1, twice-threshold current, 0.5 ms in duration). dERP was defined as the coefficient of variation (SD/mean) of the ERP at all four sites (LSPV, LIPV, LAA and RA).

SG electrical stimulation:

A transverse incision was made in the supraclavicular fossa. Behind the subclavian artery and vertebral artery, but in front of the seventh cervical vertebra, the star-shaped SG was visible in the adipose tissue (Figure 1). Adipose tissue surrounding the SG was bluntly dissected using a glass dissecting needle to expose its branches and the cardiac sympathetic nerve. LSG or RSG were stimulated at a gradual level of 1 V to 20 V (20 Hz, 2 ms) for a period of 30 s. The stimulation threshold of SG is defined as the current needed to produce a rise in systolic blood pressure (SBP) or heart rate of ≥20% (6). LSG or RSG were then unilaterally and continuously stimulated (20 Hz, 2 ms) at the threshold voltage for 6 h.

Figure 1).

Stellate ganglion. L Left; R Right

After completion of the electrophysiological studies, the SG that had been stimulated was severed (unilateral stellate ganglionectomy). AF inducibility, AF duration, ERP and dERP were also measured after stellate ganglionectomy.

Statistical analyses

Measurement data are expressed as mean ± SD. Qualitative data are expressed as ratios. Values between the two groups were compared using paired t tests, and comparisons among multiple groups were made using ANOVA. ANOVA for repeated measures was used to compare the changes at different times during pacing and stimulation. The χ² test was used to compare qualitative data; P<0.05 was considered to be statistically significant.

RESULTS

Effect and asymmetry of SG on heart rate and SBP

Figure 2 illustrates the results of graded stimulation of the LSG versus the RSG. The threshold to induce a 20% rise in heart rate during RSG stimulation was significantly lower than during LSG stimulation (8.1±2.5 V versus 14.3±3.8 V; P<0.05). The threshold to induce a 20% rise in SBP during RSG stimulation was significantly higher than LSG stimulation (12.4±3.6 V versus 8.5±2.9 V; P<0.05). RSG stimulation induced a greater change in heart rate than LSG stimulation (P<0.05) at weak stimulation strengths (6 V to 18 V). However, at stronger stimulation strengths (20 V), there were no significant differences between the effects of RSG versus LSG stimulation on heart rate (Figure 2A). While LSG stimulation produced a greater change in SBP than RSG stimulation at weak stimulation strengths (6 V to 10 V; P<0.05), at stronger stimulation strengths (12 V to 20 V) there were no significant differences between the effects of RSG versus LSG stimulation on SBP (Figure 2B).

Figure 2).

Asymmetry of stellate ganglion on heart rate (HR) and systolic blood pressure (SBP). A Percentage change in heart rate. B Percentage change in SBP. *P<0.05 between right stellate ganglion (RSG) group and left stellate ganglion (LSG) group

Effect and asymmetry of SG on AF inducibility

In the RSG group, the induction rate of AF was significantly increased in RA sites compared with baseline (73.3% versus 25.0%; P<0.05). Right stellate ganglionectomy markedly decreased the AF induction rate of the RA (31.3% versus 73.3%; P<0.05). However, there were no significant changes in the LA, LSPV and LIPV. In the LSG group, the induction rate of AF was significantly increased compared with baseline in the LA (63.0% versus 27.10%; P<0.05), LSPV (70.8% versus 33.30%; P<0.05) and LIPV (47.9% versus 18.80%; P<0.05), respectively. Left stellate ganglionectomy markedly decreased the AF induction rate of the LA (35.4% versus 63.0%; P<0.05), LSPV (39.6% versus 70.8%; P<0.05) and LIPV (25.0% versus 47.9%; P<0.05). However, there were no significant changes in RA sites (Table 1).

TABLE 1.

Effect of right stellate ganglion and left stellate ganglion stimulation and resection on atrial fibrillation induction rate

	Group
	Right stellate ganglion (n=6)			Left stellate ganglion (n=6)
	Baseline	Stimulation	Resection	Baseline	Stimulation	Resection
Right atrium	12/48 (25.0)	33/45 (73.3)*	15/48 (31.3)†	14/48 (29.2)	19/48 (39.6)	18/46 (39.1)
Left atrium	14/48 (29.2)	20/47 (42.6)	22/47 (46.8)	13/48 (27.1)	29/46 (63.0)*‡	17/48 (35.4)†
LSPV	17/48 (35.4)	20/46 (43.5)	21/48 (43.8)	16/48 (33.3)	34/48 (70.8)*‡	19/48 (39.6)†
LIPV	10/48 (20.8)	15/48 (31.3)	16/47 (34.0)	9/48 (18.8)	23/48 (47.9)*‡	12/48 (25.0)†

Data presented as frequency of successful transduction/total frequency of stimulation (%).

^*P<0.05 compared with baseline in respective group;

^†P<0.05 compared with stimulation in respective group;

^‡P<0.05 compared with stimulation in the right stellate ganglion group. LIPV Left inferior pulmonary vein; LSPV Left superior pulmonary vein

Effect and asymmetry of SG on AF duration

In the RSG group, the AF duration was significantly prolonged in RA sites (P<0.05) compared with baseline. Right stellate ganglionectomy markedly shortened the AF duration of the RA (P<0.05). However, there were no significant changes in the LA, LSPV and LIPV. In the LSG group, the AF duration was significantly prolonged (P<0.05) compared with baseline in the LA, LSPV and LIPV, respectively. Left stellate ganglionectomy markedly shortened the AF duration of the LA, LSPV and LIPV (P<0.05). However, there were no significant changes in RA sites (Table 2).

TABLE 2.

Effect of right stellate ganglion and left stellate ganglion stimulation and resection on the duration of atrial fibrillation

	Group
	Right stellate ganglion (n=6)			Left stellate ganglion (n=6)
	Baseline	Stimulation	Resection	Baseline	Stimulation	Resection
Right atrium	20.64±1.76	76.47±2.23*	25.12±4.67	19.83±2.54	25.23±2.13	23.51±1.51
Left atrium	20.22±4.46	26.91±3.03	23.87±3.67	23.75±1.88	92.44±1.91*‡	30.47±5.25†
LSPV	28.30±2.01	35.83±1.83	30.11±3.24	20.80±3.60	81.72±3.03*‡	38.32±4.12†
LIPV	24.31±2.25	26.12±2.65	21.56±1.28	25.31±1.52	66.39±4.76*‡	33.45±3.11†

Data presented as s, mean ± SD.

^*P<0.05 compared with baseline in respective group;

^†P<0.05 compared with stimulation in respective group;

^‡P<0.05 compared with stimulation in the right stellate ganglion group. LIPV Left inferior pulmonary vein; LSPV Left superior pulmonary vein

Effect and asymmetry of the SG on ERP

In the RA:

There were no differences in baseline ERP among the three groups. In the control group, isolated RAP induced slight ERP shortening, but the difference was not statistically significant (P>0.05). RSG stimulation induced significant ERP shortening at 2 h, 4 h and 6 h in RA sites compared with baseline (P<0.05 for all). Right stellate ganglionectomy reversed these ERP changes compared with stimulation at 4 h and 6 h (P<0.05 for both). However, LSG stimulation failed to induce significant ERP shortening in the RA (Figure 3A).

Figure 3).

Effect and asymmetry of stellate ganglion on the effective refractory period (ERP) in the right atrium (RA) (A), the left atrium (LA) (B), the left superior pulmonary vein (LSPV) (C) and the left inferior pulmonary vein (LIPV) (D). *P<0.05 compared with baseline in the same group; ^†P<0.05 compared with resection in the same group. LA Left atrium; LSG Left stellate ganglion; RSG Right stellate ganglion; Sti Stimulation

In the LA, LSPV and LIPV:

There were no differences in baseline ERP among the three groups. In the control group, isolated RAP induced slight ERP shortening, although this was not statistically significant (P>0.05). LSG stimulation induced significant ERP shortening at 2 h, 4 h and 6 h in the LA, LSPV and LIPV sites compared with baseline, respectively (P<0.05 for all). Left stellate ganglionectomy reversed these ERP changes compared with stimulation at 4 h and 6 h (P<0.05 for both). However, RSG stimulation failed to induce significant ERP shortening in the LA (Figure 3B), LSPV (Figure 3C) and LIPV (Figure 3D) sites.

Effect and asymmetry of SG on dERP

In the RSG group, RSG stimulation significantly increased dERP at 2 h, 4 h and 6 h compared with baseline (P<0.05 for all). Right stellate ganglionectomy markedly reduced the dERP compared with stimulation at 2 h, 4 h and 6 h (P<0.05, respectively). In the LSG group, LSG stimulation significantly increased dERP at 2 h, 4 h and 6 h compared with baseline (P<0.05 for all). Left stellate ganglionectomy markedly reduced the dERP compared with stimulation at 2 h, 4 h and 6 h (P<0.05, respectively). At 2 h, the dERP reached a maximum, and decreased slightly at 4 h and 6 h. Otherwise, there were no differences in baseline dERP among three groups (P>0.05), whereas the dERP at 2 h, 4 h and 6 h in the RSG group and LSG group were significantly larger than the control group (P<0.05, respectively). There were no statistically significant differences in dERP at 2 h, 4 h and 6 h between the RSG group and the LSG group (P>0.05) (Figure 4).

Figure 4).

Effect and asymmetry of the stellate ganglion on the dispersion of the effective refractory period (dERP). *P<0.05 compared with baseline in the same group; ^†P<0.05 compared with control group in the same intervention status. LSG Left stellate ganglion; RSG Right stellate ganglion; Sti Stimulation

DISCUSSION

Major findings

In the present study, we found that unilateral SG stimulation facilitated AF induction and aggravated atrial electrical remodelling in the first 2 h to 4 h in the atrium and pulmonary sites. The inhibition of sympathetic nerve activation through unilateral stellate ganglionectomy can reduce the AF induction and reverse the process of atrial electrical remodelling. We also demonstrated that both RSG and LSG stimulation induced changes in heart rate and SBP: low-level stimulation of RSG mainly induced changes in heart rate, and low-level stimulation of LSG mainly induced changes in SBP.

Stimulation of SG in atrial arrhythmia, AF and atrial electrophysiological properties

Tan et al (6) observed that thoracic veins become the most common sites of initiation of ectopic atrial arrhythmias with electrical stimulation of LSG and RSG following the elimination of a dominant sinus pacemaker influence. Swissa et al (9) reported that sympathetic hyperinnervation induced either by nerve growth factor infusion or subthreshold electrical stimulation of the LSG in dogs with complete atrioventricular block and myocardial infarctions can induce even more atrial nerve sprouting and a significantly higher incidence of paroxysmal AF and paroxysmal atrial tachycardia compared with normal dogs. The present study showed that enhancement of sympathetic nerve activation by electrical stimulation of the SG can facilitate AF induction and aggravate electrical remodelling (shortening the ERP, increasing the dERP) in atrial and pulmonary vein sites. The mechanisms by which AF was initiated and sustained were usually attributed to atrial electrical remodelling, as indicated by a shortening of the atrial ERP, increased dispersion of the atrial ERP, increased incidence of the atrial ERP maladaptation and a decreased atrial conduction velocity (10). Electrical remodelling is an important electrophysiological mechanism of ‘AF begetting AF’. Our findings confirmed an association between the external cardiac sympathetic nervous system (the SG) and AF, as suggested by other investigators.

Inhibition of SG in atrial arrhythmia and AF

In 1978, Tanaka et al (11) reported an unusual type of atrial tachycardia in a 52-year-old woman. Propranolol, atropine sulfate, carotid sinus massage and ocular compression were not effective in terminating or preventing the arrhythmia. LSG block had stopped the tachycardia for at least two years. Atrial tachycardia in this patient was believed to be caused by mechanical stimulation of the LSG, resulting in pacemaker shifting. Multiple studies have demonstrated that continuous low-level vagus nerve stimulation suppressed AF inducibility by directly suppressing the neural activity of major ganglionated plexi within the intrinsic cardiac autonomic nervous system and the sympathetic nerve activity, which were assessed by continuously recording the neural activity of the SG and ganglionated plexi (12–15).

In addition, Yano et al (16) observed that bilateral stellectomy can significantly decrease the incidence of electrically-induced AF in hypokalemic dogs. Ogawa et al (17) reported that reduction of cardiac sympathetic outflow through cryoablation of bilateral SG and T2–T4 thoracic ganglia can effectively eliminate paroxysmal atrial tachyarrhythmia in dogs with pacing-induced heart failure. Jayachandran et al (7) investigated the effect of bilateral stellectomy on AF and electrical remodelling in a rapid atrial-paced canine model. They found that 80% of dogs in the bilateral stellectomy group remained in sinus rhythm for six weeks despite RAP, whereas 100% of dogs in the control group developed sustained AF after four weeks of RAP. This result indicated that bilateral stellectomy can prevent sustained AF. Xie et al (18) demonstrated that cardiac sympathetic denervation by bilateral stellate ganglionectomy shortened the action potential duration (APD) and flattened the APD restitution curve, but did not significantly affect the spatial dispersion of the APD. They also demonstrated that bilateral stellate ganglionectomy downregulated β₁-adrenergic receptors. The authors did not compare the effect of left/right/bilateral stellate ganglionectomy on electrophysiological properties and arrhythmias. However, bilateral stellate ganglionectomy causes unnecessary damage and side effects such as Horner syndrome. In the present study, we investigated the effect of unilateral (left and right) stellate ganglionectomy on AF inducibility and atrial electrophysiological properties. The results showed that AF inducibility was reduced and the process of atrial electrical remodelling (shortening of ERP, increasing of ERP spatial dispersion) was reversed by unilateral stellate ganglionectomy. Additional studies are necessary to fully understand the APD and ionic mechanisms of unilateral stellate ganglionectomy.

Effect of left-right asymmetry of SG on atrium

Stimulation of the RSG and its branch produced localized refractory period shortening on the RA, particularly at the sinus node area, and invariably induced sinus tachycardia. Stimulation of the LSG and its branch affected only the refractory period of LA sites and induced atrioventricular junctional rhythm (19). Tan et al (6) demonstrated that at weak stimulation strengths (10 mA to 15 mA), RSG stimulation induced a greater change in heart rate than LSG stimulation, whereas LSG stimulation produced a greater change in SBP than RSG stimulation. At stronger stimulation strengths (20 mA to 35 mA), there were no significant differences between the effects of LSG versus RSG stimulation on either heart rate or SBP. In the present study, we found that RSG stimulation induced a greater change in heart rate than LSG stimulation (at 6 V to 18 V), whereas LSG stimulation produced a greater change in SBP than RSG stimulation (at 6 V to 10 V). At stronger stimulation strengths (20 V), there were no significant differences between the effects of LSG versus RSG stimulation on either heart rate or SBP. This finding was consistent with the study by Tan et al (6).

Crampton (20) reported that excessive or unopposed activity of the LSG, or subnormal activity of the RSG, account for the patho-physiological manifestations of long QT syndrome. Schwartz et al (21) reported that LSG ablation was an effective treatment for ventricular arrhythmias in patients with long QT syndrome. These data suggest that there is functional asymmetry between the effects of LSG and RSG stimulation in terms of ventricular arrhythmogenesis. However, less is known about the effects of LSG versus RSG activation on atrial electrophysiology and atrial arrhythmias. In the present study, we found that RSG stimulation mainly increased AF inducibility and promoted electrical remodelling (including ERP shortening and dERP increase) in the RA, and LSG stimulation mainly increased AF inducibility and promoted electrical remodelling in the LA, LSPV and LIPV. These findings suggest that LSG or its cardiac nerve branches may be an effective target for sympathetic-induced AF treatment.

Study limitations

Although we showed evidence that SG stimulation or resection can activate or suppress sympathetic activity by indirectly observing changes in heart rate and blood pressure, direct neural firing from the SG or the vagus trunk was not recorded in the present study. The complex interplay between sympathetic and vagal activity and the interaction between extrinsic and intrinsic cardiac sympathetic nervous systems were not explored. Future studies involving simultaneous recording of the SG, vagus trunk and ganglionated plexi will provide more definitive answers to this question.

CONCLUSIONS

Unilateral SG electrical stimulation plus RAP for 6 h can successfully establish a canine model of acute sympathetic-induced AF. SG stimulation facilitates AF induction and aggravates electrical remodelling in atrial and pulmonary vein sites. Inhibition of sympathetic nerve activation through unilateral stellate ganglionectomy can reduce AF initiation.