J. of Cardiovasc. Trans. Res. DOI 10.1007/s12265-014-9541-0
Left Cardiac Sympathetic Denervation in Patients with Heart Failure: a New Indication for an Old Intervention? Gaetano M. De Ferrari & Peter J. Schwartz
Received: 4 December 2013 / Accepted: 8 January 2014 # Springer Science+Business Media New York 2014
Abstract Heart failure (HF) is characterized by an autonomic imbalance with withdrawal of vagal activity and increased sympathetic activity. Novel non-pharmacological approaches to HF aimed at increasing vagal activity are being proposed. Left cardiac sympathetic denervation (LCSD) has been shown to modify favorably the outcome of several disorders characterized by life-threatening arrhythmias triggered by increased sympathetic activity. The present manuscript discusses the rationale and the limited experimental and clinical experience suggesting a potential role for LCSD in the treatment of patients with advanced heart failure. Possible future clinical applications of LCSD may include HF patients who are intolerant to β-adrenergic blockade, HF patients who have frequent Associate Editor Jozef Bartunek oversaw the review of this article G. M. De Ferrari Department of Cardiology and Cardiovascular Clinical Research Center, Fondazione IRCCS Policlinico San Matteo, Pavia, Italy P. J. Schwartz (*) Center for Cardiac Arrhythmias of Genetic Origin and Laboratory of Cardiovascular Genetics, IRCCS Istituto Auxologico Italiano, Casa di Cura San Carlo, Via Pier Lombardo, 22, 20135, Milan, Italy e-mail: [email protected] P. J. Schwartz Department of Molecular Medicine, University of Pavia, Pavia, Italy P. J. Schwartz Cardiovascular Genetics Laboratory, Hatter Institute for Cardiovascular Research in Africa, Department of Medicine, University of Cape Town, Cape Town, South Africa P. J. Schwartz Department of Internal Medicine, University of Stellenbosch, Stellenbosch, South Africa P. J. Schwartz Chair of Sudden Death, Department of Family and Community Medicine, College of Medicine, King Saud University, Riyadh, Saudi Arabia
implantable cardioverter-defibrillator shocks, and HF patients in countries where the likelihood of receiving a device is limited, but the capability to perform a one in a lifetime procedure is present. Keywords Sympathetic denervation . Autonomic nervous system . Heart failure . Life-threatening arrhythmias . Vagus nerve
Introduction Chronic heart failure remains the most prevalent chronic cardiac condition and is a major and growing public health problem. In the USA, approximately five million patients have heart failure (HF), and over 550,000 new patients/year are diagnosed with HF, which is the primary reason for 12 to 15 million office visits and 6.5 million hospital days each year [1]. Despite constant therapeutic improvements, mortality and morbidity remain high and there is an extensive unmet need for improved treatment options [2]. It is commonly accepted that heart failure is characterized by an autonomic imbalance with withdrawal of vagal activity and increased sympathetic activity. Although the increase in cardiac adrenergic drive is finalized to supporting the performance of the failing heart, long-term activation of the sympathetic nervous system exerts deleterious effects. Consequently, long-term treatment with β-adrenergic blockers [3–5] improves prognosis among patients with heart failure. New non-pharmacological approaches to heart failure have focused in the past few years in increasing parasympathetic activity by techniques such as baroreceptor stimulation [6], vagal stimulation [7], and spinal cord stimulation [8], all discussed elsewhere in this special issue of the journal. The present manuscript will examine the rationale and the potential role for left cardiac sympathetic denervation
J. of Cardiovasc. Trans. Res.
(LCSD) in the treatment of patients with advanced heart failure.
Left Cardiac Sympathetic Denervation: Definition, Technique, and Side Effects Efferent sympathetic innervations of the heart include preganglionic and postganglionic fibers. The first are located within the intermediate zone of the thoracic spinal cord [9]. Axons from these neurons exit via the ventral rami of segments T1– T4, directed to the sympathetic chain, a bilateral paravertebral combination of ganglia and interconnecting nerves extending from the cervical to the lumbar levels. The synapse between preganglionic axons and postganglionic neurons directed to the heart is located within the stellate ganglion (fusion of inferior cervical and T1 ganglia) and the ganglia at levels T2–T4. The majority of the cardiac fibers originate from the stellate ganglia (left and right); additional fibers originate from T2 to T4 and, exceptionally, also from C7 to T5. They exit the ganglia via gray rami communicantes and enter the left and right cardiac nerves, which ramify en route to the heart [10]. Some nerve branches are joined by vagal fibers coursing to the heart [11]. Sympathetic nerve fibers arrive at the base of the heart and penetrate the myocardium giving off smaller branches that innervate the entire heart. Although a considerable degree of overlap exists, the left ventricle is predominantly innervated from left-sided nerves which also go posteriorly, whereas the right-sided cardiac sympathetic nerves go primarily anteriorly to the right ventricle [11, 12]. The left-sided nerves are quantitatively dominant, in terms of norepinephrine release, and this is one of the major reasons for the different effects between right and left cardiac sympathetic denervation. Two different surgical approaches have been developed and are available to perform LCSD, namely an open surgical approach and a video-assisted thoracic surgical (VATS) approach. Both approaches have been recently reviewed [13, 14] and have pros and cons. The traditional supraclavicular retropleural approach (shown in Fig. 1) is rather technically demanding and now less frequently practiced, while VATS is less invasive and is being increasingly used, also for other indications such as hyperhidrosis [15]. Recently, two small case series of patients subjected to percutaneous radiofrequency left cardiac sympathetic denervation for either hyperhidrosis or for the treatment of refractory ventricular arrhythmic storms have been reported [16, 17]. Although the term “left stellectomy” is often equated to LCSD, it should be underlined that surgical removal of the stellate ganglion results in incomplete denervation in humans, thus decreasing the efficacy of the intervention. Also, since the stellate ganglion is formed by the fusion of C7 and T1, only its lower part must be excised in order to reduce the possibility of
Fig. 1 Cartoon sequence depicting the principal surgical steps of LCSD by the supraclavicular approach. The patient lies supine, with the neck hyperextended and the head upward and turned rightward. After skin incision, the platysma and subsequently the clavicular head of the sternocleidomastoid muscle are transected (not shown). The upper cartoon shows the subsequent surgical step, consisting in the identification and cautious mobilization of the phrenic nerve and of the transection of the anterior scalene muscle. In the middle cartoon, the apical pleura is freed from its ligament, the lung is reflected caudally, and the subclavian artery is retracted upward. In the lower cartoon, the stellate ganglion and the sympathetic chain are exposed. Lidocaine 1 % is injected before excision of the nerve. Modified from [13]
a Horner syndrome. Using this caution, the patient usually experiences only transient ptosis with a minimal probability of a significant permanent Horner syndrome [13, 14, 18]; in our experience, this has occurred in less than 2 % of cases. Additional side effects include surgical pain, dry and warm left hand and forehead, compensatory sweating in other districts, and very rarely a clinically meaningful pneumothorax. To our knowledge, no case of death or major permanent disability following LCSD has never been reported. Altogether, the safety profile of this intervention appears very good, and this is in agreement with the increase for a clinical syndrome such as hyperhidrosis that is certainly not life-threatening.
J. of Cardiovasc. Trans. Res.
In 1899, Francois-Frank, who was studying the transmission of sensory information from the aorta through the cervico-thoracic sympathetic nervous system, suggested that removal of these nerves might be useful in patients with angina [19]. His suggestion was eventually accepted by the Romanian surgeon Theodore Jonnesco who had already performed cervical sympathectomy in 1896 for the treatment of epilepsy and of exophthalmic goiter [20]. Jonnesco employed for the first time in 1916 unilateral section of the last left cervical ganglion together with the first thoracic ganglion in a patient affected by invalidating angina accompanied by cardiac arrhythmias. Both angina and arrhythmias disappeared after surgery [21]. This surgical intervention was harshly criticized by Danielopolu who defined it as an anesthetic intervention since it reduced pain but not myocardial ischemia, thus depriving the patients of a useful warning signal. Furthermore, he disputed that, since the sympathetic fibers were considered to cause coronary vasodilation, their interruption could have worsened ischemia [22]. Vasoconstrictive effects of the sympathetic fibers were, on the other hand, suggested by the beneficial effects of the removal of the stellate ganglia in patients with Raynaud’s disease, a surgical approach first reported in Berlin in 1923 [23] and widely performed until the 1990s. In 1929, Leriche and Fontaine stated that the left stellate ganglion had a central role in the reflexes that initiate angina and, very important, that sympathetic nerves had a vasoconstrictive effect on the coronary arteries [24]. In spite of the long-lasting criticism and warnings of the potential deleterious effects of stellate ganglionectomy by
Danielopolu [25], favorable reports on this intervention among patients with angina continued to accumulate [26–28]. With the advent of β-blockers, however, sympathectomy for angina was superseded and abandoned. Few years later, sympathectomy began to be proposed as an antiarrhythmic intervention. The anecdotal first human studies were performed by Ester and Izlar and by Zipes et al. and showed successful prevention of recurrent ventricular tachycardia [29, 30]. In 1970, Moss and McDonald performed LCSD in a long QT syndrome (LQTS) patient not protected by β-blockers [31]. Since this intervention and subsequent trials of pharmacological blockade of the left stellate ganglion failed to shorten the QT interval, an effect suggested by the canine studies by Yanowitz et al. [32], this surgical approach was not pursued further. Following the demonstration that left stellate ganglion stimulation produces in cats both QT prolongation and macroscopic T wave alternans [33], Schwartz performed in 1973 LCSD in a LQTS patient who had frequent syncope and cardiac arrest; the patient has remained free of cardiac events for the following 41 years. The initial experience with LCSD in LQTS was reported by Schwartz et al. in 1991 [34]. The same group, in 2004, reported a very high success rate in 147 patients [35]. LCSD was found highly effective also in post-infarction arrhythmias. In 1992, Schwartz et al. [36] reported that in a multicenter clinical trial in post-myocardial infarction patients at high risk for lethal arrhythmias and randomized between placebo, β-blocker therapy, and LCSD, both anti-adrenergic therapies resulted in a striking reduction of the 2-year incidence of sudden death from 22 to 3 %. Finally, in 2008, Wilde et al. reported for the first time that LCSD was extremely effective also in patients affected by catecholaminergic polymorphic ventricular tachycardia resistant to β-blockers [37].
Fig. 2 Sequential changes in left ventricular (LV) dP/dt and LV enddiastolic and end-systolic wall stresses are shown 1 day, 7 days, and 3 to 4 weeks after starting ventricular pacing at 240 beats/min in a canine model of heart failure. When ANOVA was used, the changes in LV dP/dt
and end-diastolic wall stress were significantly (P
Left Cardiac Sympathetic Denervation in Patients with Heart Failure: a New Indication for an Old Intervention? Gaetano M. De Ferrari & Peter J. Schwartz
Received: 4 December 2013 / Accepted: 8 January 2014 # Springer Science+Business Media New York 2014
Abstract Heart failure (HF) is characterized by an autonomic imbalance with withdrawal of vagal activity and increased sympathetic activity. Novel non-pharmacological approaches to HF aimed at increasing vagal activity are being proposed. Left cardiac sympathetic denervation (LCSD) has been shown to modify favorably the outcome of several disorders characterized by life-threatening arrhythmias triggered by increased sympathetic activity. The present manuscript discusses the rationale and the limited experimental and clinical experience suggesting a potential role for LCSD in the treatment of patients with advanced heart failure. Possible future clinical applications of LCSD may include HF patients who are intolerant to β-adrenergic blockade, HF patients who have frequent Associate Editor Jozef Bartunek oversaw the review of this article G. M. De Ferrari Department of Cardiology and Cardiovascular Clinical Research Center, Fondazione IRCCS Policlinico San Matteo, Pavia, Italy P. J. Schwartz (*) Center for Cardiac Arrhythmias of Genetic Origin and Laboratory of Cardiovascular Genetics, IRCCS Istituto Auxologico Italiano, Casa di Cura San Carlo, Via Pier Lombardo, 22, 20135, Milan, Italy e-mail: [email protected] P. J. Schwartz Department of Molecular Medicine, University of Pavia, Pavia, Italy P. J. Schwartz Cardiovascular Genetics Laboratory, Hatter Institute for Cardiovascular Research in Africa, Department of Medicine, University of Cape Town, Cape Town, South Africa P. J. Schwartz Department of Internal Medicine, University of Stellenbosch, Stellenbosch, South Africa P. J. Schwartz Chair of Sudden Death, Department of Family and Community Medicine, College of Medicine, King Saud University, Riyadh, Saudi Arabia
implantable cardioverter-defibrillator shocks, and HF patients in countries where the likelihood of receiving a device is limited, but the capability to perform a one in a lifetime procedure is present. Keywords Sympathetic denervation . Autonomic nervous system . Heart failure . Life-threatening arrhythmias . Vagus nerve
Introduction Chronic heart failure remains the most prevalent chronic cardiac condition and is a major and growing public health problem. In the USA, approximately five million patients have heart failure (HF), and over 550,000 new patients/year are diagnosed with HF, which is the primary reason for 12 to 15 million office visits and 6.5 million hospital days each year [1]. Despite constant therapeutic improvements, mortality and morbidity remain high and there is an extensive unmet need for improved treatment options [2]. It is commonly accepted that heart failure is characterized by an autonomic imbalance with withdrawal of vagal activity and increased sympathetic activity. Although the increase in cardiac adrenergic drive is finalized to supporting the performance of the failing heart, long-term activation of the sympathetic nervous system exerts deleterious effects. Consequently, long-term treatment with β-adrenergic blockers [3–5] improves prognosis among patients with heart failure. New non-pharmacological approaches to heart failure have focused in the past few years in increasing parasympathetic activity by techniques such as baroreceptor stimulation [6], vagal stimulation [7], and spinal cord stimulation [8], all discussed elsewhere in this special issue of the journal. The present manuscript will examine the rationale and the potential role for left cardiac sympathetic denervation
J. of Cardiovasc. Trans. Res.
(LCSD) in the treatment of patients with advanced heart failure.
Left Cardiac Sympathetic Denervation: Definition, Technique, and Side Effects Efferent sympathetic innervations of the heart include preganglionic and postganglionic fibers. The first are located within the intermediate zone of the thoracic spinal cord [9]. Axons from these neurons exit via the ventral rami of segments T1– T4, directed to the sympathetic chain, a bilateral paravertebral combination of ganglia and interconnecting nerves extending from the cervical to the lumbar levels. The synapse between preganglionic axons and postganglionic neurons directed to the heart is located within the stellate ganglion (fusion of inferior cervical and T1 ganglia) and the ganglia at levels T2–T4. The majority of the cardiac fibers originate from the stellate ganglia (left and right); additional fibers originate from T2 to T4 and, exceptionally, also from C7 to T5. They exit the ganglia via gray rami communicantes and enter the left and right cardiac nerves, which ramify en route to the heart [10]. Some nerve branches are joined by vagal fibers coursing to the heart [11]. Sympathetic nerve fibers arrive at the base of the heart and penetrate the myocardium giving off smaller branches that innervate the entire heart. Although a considerable degree of overlap exists, the left ventricle is predominantly innervated from left-sided nerves which also go posteriorly, whereas the right-sided cardiac sympathetic nerves go primarily anteriorly to the right ventricle [11, 12]. The left-sided nerves are quantitatively dominant, in terms of norepinephrine release, and this is one of the major reasons for the different effects between right and left cardiac sympathetic denervation. Two different surgical approaches have been developed and are available to perform LCSD, namely an open surgical approach and a video-assisted thoracic surgical (VATS) approach. Both approaches have been recently reviewed [13, 14] and have pros and cons. The traditional supraclavicular retropleural approach (shown in Fig. 1) is rather technically demanding and now less frequently practiced, while VATS is less invasive and is being increasingly used, also for other indications such as hyperhidrosis [15]. Recently, two small case series of patients subjected to percutaneous radiofrequency left cardiac sympathetic denervation for either hyperhidrosis or for the treatment of refractory ventricular arrhythmic storms have been reported [16, 17]. Although the term “left stellectomy” is often equated to LCSD, it should be underlined that surgical removal of the stellate ganglion results in incomplete denervation in humans, thus decreasing the efficacy of the intervention. Also, since the stellate ganglion is formed by the fusion of C7 and T1, only its lower part must be excised in order to reduce the possibility of
Fig. 1 Cartoon sequence depicting the principal surgical steps of LCSD by the supraclavicular approach. The patient lies supine, with the neck hyperextended and the head upward and turned rightward. After skin incision, the platysma and subsequently the clavicular head of the sternocleidomastoid muscle are transected (not shown). The upper cartoon shows the subsequent surgical step, consisting in the identification and cautious mobilization of the phrenic nerve and of the transection of the anterior scalene muscle. In the middle cartoon, the apical pleura is freed from its ligament, the lung is reflected caudally, and the subclavian artery is retracted upward. In the lower cartoon, the stellate ganglion and the sympathetic chain are exposed. Lidocaine 1 % is injected before excision of the nerve. Modified from [13]
a Horner syndrome. Using this caution, the patient usually experiences only transient ptosis with a minimal probability of a significant permanent Horner syndrome [13, 14, 18]; in our experience, this has occurred in less than 2 % of cases. Additional side effects include surgical pain, dry and warm left hand and forehead, compensatory sweating in other districts, and very rarely a clinically meaningful pneumothorax. To our knowledge, no case of death or major permanent disability following LCSD has never been reported. Altogether, the safety profile of this intervention appears very good, and this is in agreement with the increase for a clinical syndrome such as hyperhidrosis that is certainly not life-threatening.
J. of Cardiovasc. Trans. Res.
In 1899, Francois-Frank, who was studying the transmission of sensory information from the aorta through the cervico-thoracic sympathetic nervous system, suggested that removal of these nerves might be useful in patients with angina [19]. His suggestion was eventually accepted by the Romanian surgeon Theodore Jonnesco who had already performed cervical sympathectomy in 1896 for the treatment of epilepsy and of exophthalmic goiter [20]. Jonnesco employed for the first time in 1916 unilateral section of the last left cervical ganglion together with the first thoracic ganglion in a patient affected by invalidating angina accompanied by cardiac arrhythmias. Both angina and arrhythmias disappeared after surgery [21]. This surgical intervention was harshly criticized by Danielopolu who defined it as an anesthetic intervention since it reduced pain but not myocardial ischemia, thus depriving the patients of a useful warning signal. Furthermore, he disputed that, since the sympathetic fibers were considered to cause coronary vasodilation, their interruption could have worsened ischemia [22]. Vasoconstrictive effects of the sympathetic fibers were, on the other hand, suggested by the beneficial effects of the removal of the stellate ganglia in patients with Raynaud’s disease, a surgical approach first reported in Berlin in 1923 [23] and widely performed until the 1990s. In 1929, Leriche and Fontaine stated that the left stellate ganglion had a central role in the reflexes that initiate angina and, very important, that sympathetic nerves had a vasoconstrictive effect on the coronary arteries [24]. In spite of the long-lasting criticism and warnings of the potential deleterious effects of stellate ganglionectomy by
Danielopolu [25], favorable reports on this intervention among patients with angina continued to accumulate [26–28]. With the advent of β-blockers, however, sympathectomy for angina was superseded and abandoned. Few years later, sympathectomy began to be proposed as an antiarrhythmic intervention. The anecdotal first human studies were performed by Ester and Izlar and by Zipes et al. and showed successful prevention of recurrent ventricular tachycardia [29, 30]. In 1970, Moss and McDonald performed LCSD in a long QT syndrome (LQTS) patient not protected by β-blockers [31]. Since this intervention and subsequent trials of pharmacological blockade of the left stellate ganglion failed to shorten the QT interval, an effect suggested by the canine studies by Yanowitz et al. [32], this surgical approach was not pursued further. Following the demonstration that left stellate ganglion stimulation produces in cats both QT prolongation and macroscopic T wave alternans [33], Schwartz performed in 1973 LCSD in a LQTS patient who had frequent syncope and cardiac arrest; the patient has remained free of cardiac events for the following 41 years. The initial experience with LCSD in LQTS was reported by Schwartz et al. in 1991 [34]. The same group, in 2004, reported a very high success rate in 147 patients [35]. LCSD was found highly effective also in post-infarction arrhythmias. In 1992, Schwartz et al. [36] reported that in a multicenter clinical trial in post-myocardial infarction patients at high risk for lethal arrhythmias and randomized between placebo, β-blocker therapy, and LCSD, both anti-adrenergic therapies resulted in a striking reduction of the 2-year incidence of sudden death from 22 to 3 %. Finally, in 2008, Wilde et al. reported for the first time that LCSD was extremely effective also in patients affected by catecholaminergic polymorphic ventricular tachycardia resistant to β-blockers [37].
Fig. 2 Sequential changes in left ventricular (LV) dP/dt and LV enddiastolic and end-systolic wall stresses are shown 1 day, 7 days, and 3 to 4 weeks after starting ventricular pacing at 240 beats/min in a canine model of heart failure. When ANOVA was used, the changes in LV dP/dt
and end-diastolic wall stress were significantly (P
