Anatomical Science International

, Volume 86, Issue 1, pp 30–49

Anatomy of the cardiac nervous system with clinical and comparative morphological implications

Authors

    • Department of Anatomy, School of MedicineTokyo Women’s Medical University
Review Article

DOI: 10.1007/s12565-010-0096-0

Cite this article as:
Kawashima, T. Anat Sci Int (2011) 86: 30. doi:10.1007/s12565-010-0096-0

Abstract

Unlike autonomic nervous preservation in other surgeries for improving patient quality of life, autonomic cardiac nervous system (ACNS) preservation has been neglected in cardiovascular surgery because of technical difficulties and other unsolved issues. Because such ACNS preservation in cardiovascular surgery is anticipated in the future, detailed anatomical investigation of the human ACNS is required. Therefore, we have conducted morphological studies of the ACNS from macroscopic, clinical, and evolutionary anatomical viewpoints. In this study, I review detailed anatomical studies of the human ACNS together with their clinical implications. In addition, the evolutionary comparative anatomical significance of primate ACNS is also summarized to help understand and translate the findings of functional experiments to humans. These integrated findings will be the subject of a future study unifying molecular embryological and anatomical findings to clarify cardiac functions based on functional animal experiments, clinical applications such as improving surgery techniques and individual order-made surgery in cardiac surgery, and for future evaluation in regenerative medicine.

Keywords

Autonomic cardiac nervous systemClinical anatomyComparative anatomyEvolutionPrimateHuman heartCardiac nerveCardiac branch

Abbreviations

AA

Ascending aorta

AI

Superior interventricular artery (anterior interventricular/descending branch of the coronary artery)

aMG

Accessory middle cervical ganglion

Ao

Aorta

Az

Azygos vein

AV

Anterior cardiac vein

BC

Brachiocephalic trunk

BP

Brachial plexus

Br-A

Bronchial artery

CB

Circumflex branch of the coronary artery

CC

Common carotid artery

CT

Cervicothoracic (stellate) ganglion

C1-8

First to eighth cervical nerves

Es

Esophagus

GV

Great cardiac vein

IB

Inferior (vagal) cardiac branch

IG

Inferior cervical ganglion

IN

Inferior cervical cardiac nerve

IVC

Inferior caval vein

L

Nerve branches to the lung

LA

Left atrium

LC

Left common carotid artery

LCA

Left coronary artery

LS

Left subclavian artery

LSVC

Left superior vena cava

LV

Left ventricle

MG

Middle cervical ganglion

MN

Middle cardiac nerve

NX

Vagus nerve

OV

Oblique vein of the left atrium

P

Pectoral nerve

PA

Pulmonary artery

PC

Pericardium

Ph

Phrenic nerve

PT

Pulmonary trunk

PV

Pulmonary vein

PVi

Inferior pulmonary vein

PVs

Superior pulmonary vein

RA

Right atrium

RC

Right common carotid artery

RCA

Right coronary artery

RERSA

Retroesophageal right subclavian artery

RL

Recurrent laryngeal nerve of the vagus nerve

RS

Right subclavian artery

RSVC

Right superior vena cava

RV

Right ventricle

SAN-B

Sinuatrial nodal branch of the coronary artery

SB

Superior (vagal) cardiac branch

SbC

Nerve to the subclavian muscle

SG

Superior cervical ganglion

SN

Superior cardiac nerve

SVC

Superior caval vein

TB

Thoracic (vagal) cardiac branch

TN

Thoracic cardiac nerve

T1-6

First to sixth thoracic nerves

VA

Vertebral artery

VG

Vertebral ganglion

VN

Vertebral nerve

X

Vagus nerve

XI

Accessory nerve

XII

Hypogastric nerve

2TG–6TG

Second to sixth thoracic ganglia

Introduction

Because the autonomic cardiac nervous system (ACNS) controls life-maintaining functions such as pulse, blood pressure, blood flow and cardiac pain delivery, many gross and clinical anatomical studies of the ACNS have been conducted over the long course of medical history. However, most of the anatomical research on ACNS has been incomplete because of the long course from the origin to the target and the existence of two systems, i.e., sympathetic and parasympathetic. Thus, most gross anatomical studies on the ACNS have been limited to either the sympathetic or parasympathetic nervous system. Furthermore, studies on the sympathetic nervous system have been divided into the following two types: research on the proximal part from the spinal nerves to the sympathetic ganglia (Axford 1927–1928; Becker and Grunt 1957; Harman 1900; Hoffman 1957; Jamieson et al. 1952; Kuntz 1946; Mitchell 1953; Pick and Sheehan 1946; White and Smithwick 1952; Wrete 1959), and research on the distal part from the sympathetic ganglia to the visceral nerves (Ellison and Williams 1962; Fukuyama 1982; Hausman 1956; Janes et al. 1986; Kuntz and Morehouse 1930; Mizeres 1963; Mitchell 1953; Pick 1970). Although a few excellent anatomical studies on the ACNS have been conducted recently (Batulevicius et al. 2003, 2005; Pauza et al. 2000, 2002; Saburkina et al. 2010; Tanaka et al. 1998, 2007), numerous points remain to be clarified regarding future human applications.

Because accurate anatomical knowledge of autonomic innervation is essential for function-preserving operations aimed at improving patient quality of life (QOL), anatomical reexamination of the autonomic nervous system in the abdominal and pelvic regions has been performed. In cardiovascular surgery, however, operations to preserve the autonomic nervous system have never been performed because of technical difficulties and other unsolved issues. Furthermore, autonomic regeneration may help to accelerate nonpreserving operation of ACNS in complicated cardiac surgery. However, it is well known that patients cannot perceive anginal pain caused by denervation after undergoing cardiac transplantation (Raczak et al. 1999; Bengel et al. 2001, 2002). In addition, denervated patients constantly experience orthostatic hypotension and ineffectiveness of antiarrhythmic agents such as atropine and foxgloves, which act on autonomic nerves. Because this type of cardiovascular surgery is anticipated in the future, detailed anatomical investigation of the human ACNS is necessary.

To date, several mammals, such as rats, mice, and dogs, have been used as model animals for medical and developmental studies of the ACNS to understand its function and morphogenesis. The morphology of the ACNS in these experimental animals shows considerable interspecific differences in line with those shown in the evolutionary divergence timeline. In addition, to completely understand the morphological principles of the ACNS, it is essential to examine the ACNS in primates—positioned phyletically between humans and rodents—with regard to whether ACNS morphology reflects its phylogeny.

In this review, I discuss the detailed human ACNS anatomy, the clinical significance of topographical changes in branchial arterial anomalies of the ACNS, and the evolutionary morphology of the primate ACNS based on previous studies by our group and others.

Anatomy of human cardiac innervation and its clinical significance

Terms

Because many technical terms regarding ACNS anatomy have been introduced arbitrarily without clear definitions, several authors have separately coined new terms, leading to great difficulties in comparing results among previous studies (Nonidez 1939; Mitchell 1953; Hoffman 1957; Pick 1970; Fukuyama 1982; Tanaka et al. 1998). Our definitions of the cardiac nerve/branch follow those coined by Fukuyama (1982), with modifications. The cardiac nerves connecting to the cardiac plexus or distributing to the heart are named according to their origins as follows. Superior cardiac nerve (SN) is the cardiac nerve originating from the superior cervical ganglion (SG) or sympathetic trunk (ST) between SG and middle cervical ganglion (MG); middle cardiac nerve (MN) is the cardiac nerve originating from MG, vertebral ganglion (VG), or ST between MG and inferior cervical ganglion (IG) (or the cervicothoracic ganglion, CT), including the subclavian ansa; inferior cardiac nerve (IN) is the cardiac nerve originating from IG or CT; thoracic cardiac nerve (TN) is the cardiac nerve originating from the thoracic ganglia (TG) or the thoracic ST below IG or CT. Furthermore, each cardiac branch is named according to its origin as follows. Superior cardiac branch (SB) is the cardiac branch originating from the vagus nerve proximal to the branching site of the recurrent laryngeal nerve (RL); inferior cardiac branch (IB) is the cardiac branch originating from any part of the RL; thoracic cardiac branch (TB) is the cardiac branch originating from the vagus nerve distal to the branching site of the RL.

Furthermore, we use some modified terms for the human heart and its surrounding structures such as superior (anterior) and inferior (posterior) interventricular arteries and caval veins (vena cava), based on their correct orientations provided by Anderson et al. (2004) and Anderson and Loukas (2009). Most of the abbreviations used follow the accepted conventions, facilitating easy identification.

Sympathetic ganglia

Based on these definitions of ACNS, we analyzed the detailed morphology of the human ACNS via surgical stereomicroscopy (Kawashima 2005; Kawashima and Sasaki 2007; Figs. 1, 2). Our observations on the relationship between the sympathetic ganglia and spinal nerves are as follows. SG was observed on all 36 sides (100.0%), and exhibited communicating branches with C1 (36/36 sides, 100.0%), C2 (36/36 sides, 100.0%), C3 (26/36 sides, 72.2%), and C4 (1/36 side, 2.8%) (Kawashima 2005).
https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig1_HTML.jpg
Fig. 1

Photographs showing the autonomic cardiac nervous system (ACNS) in humans. a Right cervical and upper thoracic sympathetic ganglia. b Regional enlargement of a. The single inferior cervical ganglion (IG) was observed without formation of the cervicothoracic ganglion. c Typical morphology of the human cervicothoracic ganglion (CT) comprising IG and first thoracic ganglion (1TG). d Reflection of the pericardium in the arterial part of the cardiac hilum and its nerve entrances are clearly observed viewed from the ventral aspect. e ACNS accompanying the great arteries viewed from the dorsal aspect. f The cardiac plexus viewed from the dorsal aspect

https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig2_HTML.gif
Fig. 2

Diagrams showing the human ACNS with great arterial bifurcations viewed from the ventral (a) and dorsal (b) aspects. The sympathetic, parasympathetic vagal, and mixed nerves are colored orange, green, and purple, respectively. Black stars, squares, and circles indicate the inlet/outlet of the cardiac nerve/branches from the arterial part of the cardiac hilum itself, and venous parts of the cardiac hilum

MG was observed on 33 sides (33/36 sides, 91.7%), and it communicated with C3 (11/33 sides, 33.3%), C4 (29/33 sides, 87.9%), C5 (20/33 sides, 60.6%), and C6 (5/33 sides, 15.2%).

VG was observed on 34 sides (34/36 sides, 94.7%), and it communicated with C4 (1/34 side, 2.9%), C5 (2/34 sides, 5.9%), C6 (3/34 sides, 8.8%), and C7 (6/34 sides, 17.6%).

A single IG without fusion to TG was present on five sides (5/36 sides, 13.9%), whereas most cases formed a CT (31/36 sides, 86.1%) and communicated with C5 (2/36, 5.6%), C6 (4/36, 11.1%), C7 (27/36, 75.0%), C8 (100.0%), T1 (100.0%), T2 (25/36, 69.4%), and T3 (1/36, 2.8%). The ganglion communicating with T3 was a CT comprising IG, 1TG, and 2TG (Fig. 1c).

Furthermore, these data on communication between the sympathetic ganglia and spinal nerves do not include the vertebral nerve (VN) consisting of the nerve following the vertebral artery and entering into the foramen transversarium. We observed VN originating from VG on 26/34 sides (76.5%) and IG or CT on 34/36 sides (94.4%). Based on further examination of VN after demineralization and removing the bony elements on six sides, VN communicated with C5 (3/6 sides, 50.0%), C6 (6/6 sides, 100.0%), and C7 (6/6 sides, 100.0%). Therefore, VN appears to be a complex of communicating branches to the spinal cervical nerve through the foramen transversarium and it was termed deep ST by Wrete (1959).

Although differences in frequencies of communicating branches between the sympathetic ganglia and spinal nerves appear to be due partially to ethnicity, most results have been consistent and are in agreement with previous reports (Axford 1927–1928; Becker and Grunt 1957; Harman 1900; Hoffman 1957; Jamieson et al. 1952; Kuntz 1946; Mitchell 1953; Pick and Sheehan 1946; White and Smithwick 1952; Wrete 1959). Only the appearances of MG and VG are controversial. Hoffman (1957) and Pick (1970) mentioned that these two ganglia were generally small and their ganglion cells could be observed throughout the ST between SG and IG, even if they could not be readily observed. Although this histological finding is important with regard to their morphological significance, it is also significant for clinicians that the frequency of VG is higher than that of MG and both the ganglia are not always macroscopically evident.

Cardiac nerves/branches

Our findings on origins and appearances of the sympathetic cardiac nerves and parasympathetic vagal cardiac nerves are as follows (Kawashima and Sasaki 2007; Fig. 2). SN originating from SG and ST was observed on 46/52 (88.5%) and 37/52 sides (71.2%), respectively. MN originating from MG, VG, and ST was observed on 43/49 (87.8%), 43/50 (86.0%), and 40/52 sides (76.9%), respectively. IN and TN were observed on 43/50 (86.0%) and 35/52 sides (67.3%), respectively. On the other hand, all parasympathetic vagal cardiac branches, which included SB, IB, and TB, were further consistently observed on 52/52 (100.0%), 52/52 (100.0%), and 43/52 sides (82.7%), respectively (Kawashima and Sasaki 2007).

Although opinions that MN originating from MG and VG, and all parasympathetic cardiac branches are main contributors, in the human ACNS are in agreement with ours (Mitchell 1953; Ellison and Williams 1962; Fukuyama 1982), limits of cardiac nerve origin, namely the upper limit of SN and the lower limit of TN, differ among authors (Ellison and Williams 1962; Fukuyama 1982; Hausman 1956; Janes et al. 1986; Kuntz and Morehouse 1930; Mizeres 1963; Mitchell 1953; Pick 1970). In humans, based on our detailed anatomy, the upper limit of SN was clearly consistent from SG. On the other hand, the lower limit of TN is difficult to correctly identify because TN is quite slender and has a complicated course along the vessel, appearing irregularly throughout the posterior mediastinum. Tanaka et al. (2007) described the lowest origin of TN originating from 9TG in shrew (Suncus murinus) and concluded that it was similar to that in humans. Based on our comparative anatomy of ACNS (discussed later), we disagree with their conclusions that the shrew shares primitive morphologies with humans, based on numerous and significant anatomical differences between humans and other mammals including primates as well as their evolutionary divergence.

Kuntz and Morehouse (1930) reported the course of the human lower TN as running transversely through the thoracic vertebra, irrespective of the course of the intercostal vessels. Moreover, Mitchell (1953) and Saccomanno (1943) described the lower origin (5TG–7TG) as ascending across the intercostal vessels to reach the cardiac entrances at a position higher than the origin of TN (Fig. 3a). However, neither the ascending TN nor the transversal TN was observed in our human cases. According to our findings, although the right TN originating from relatively higher origins such as the 2TG and 3TG can distribute to the heart from the right venous porta via the azygos arch and/or intercostal and bronchial arteries (Fig. 3b), the actual course of the right TN originating from the lower TG differs from that provided by previous descriptions. The correct human right TN originating from relatively lower TG as a new course, namely right recurrent TN (Kawashima 2005; Fig. 3c, d) is reported in this study. Actually, we previously reported this new course of the lower TN as an anomalous course in one out of ten cases of Japanese monkeys (Kawashima et al. 2001). SN, MN, IN, SB, IB, and TB innervated the heart by following a simple course along the great vessels, such as the brachiocephalic trunk, common carotid and subclavian arteries, aortic arch, and superior caval vein. The left TN can utilize the neighboring aortic arch and thoracic aorta to easily reach the heart from the left venous part of the cardiac hilum (Fig. 3d). On the other hand, the right TN in the posterior mediastinum, especially lower origins such as the right 4TG or 5TG, is considered to require a complex course because of the long distance to the heart in the middle mediastinum that is associated with the disappearance of the right aortic arch. This roundabout course of the right TN should be recognized as a collateral route for cardiac innervation.
https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig3_HTML.gif
Fig. 3

Courses of the thoracic cardiac nerve. a Conventional description (Saccomanno 1943). The right thoracic cardiac nerve (TN) originating from the lower thoracic ganglia (TG) ascends to reach the heart at a position higher than the origin of TG, regardless of the course of the intercostal vessels. b Photograph showing the actual course of the right TN originating from the higher TG such as second and third TG (2TG, 3TG). TN descends along the intercostal vessels and/or posterior thoracic wall, reaches and accompanies the azygos arch/common trunk of the intercostal and bronchial arteries, and then distributes to the heart from the right venous part. c, d Diagram showing our new descriptions (recurrent right TN: Kawashima 2005). The right TN originates from the relatively lower TG such as 3TG–5TG and descends obliquely along the intercostal vessels. After arriving at the thoracic aorta, it turns and ascends along the aorta, finally distributing to the heart via the left venous part of the cardiac hilum or connecting to the cardiac plexus as shown by the arrowheads

Cardiac plexus

The cardiac plexus is composed of complicated bifurcations and anastomoses of the cardiac nerves and branches as shown in Figs. 1 and 2. Although it is extremely difficult to estimate the actual position of the cardiac plexus, I have attempted to understand the tendency of the mixed nerves between the sympathetic cardiac nerves and parasympathetic vagal cardiac branches macroscopically (Table 1). On the right side, the highest observed position of the mixed nerve was at the level of the cervical part of the common carotid artery, whereas the position of the most mixed nerve was at the brachiocephalic trunk. On the left side, the highest observed position of the mixed nerve was also at the level of the cervical part of the common carotid artery, whereas the position of the most mixed nerve was at the aortic arch. In other words, the positions (cardiac plexus) of the sympathetic cardiac nerves and parasympathetic vagal cardiac branches, as well as the definitive morphology of the branchial arteries with RL tended to differ on the two sides (Fig. 4).
Table 1

Position of cardiac plexus

Position

Right

Left

x (%)a

x (%)b

x (%)a

x (%)b

Cervical part of common carotid artery (CC)

4 (22.2)

0 (0.0)

4 (22.2)

0 (0.0)

Thoracic part of CC or brachiocephalic artery

14 (77.8)

18 (100.0)

11 (61.1)

2 (11.1)

Aortic arch

0 (0.0)

0 (0.0)

3 (16.6)

16 (88.9)

Total

18 (100.0)

18 (100.0)

18 (100.0)

18 (100.0)

Differences in the positions (cardiac plexus) of the sympathetic cardiac nerves and parasympathetic vagal cardiac branches between the right and left sides are highlighted in bold

aHighest position of mixed nerves between sympathetic cardiac nerve and vagal cardiac branch

bPosition of most mixed nerves between sympathetic cardiac nerve and vagal cardiac branch

https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig4_HTML.gif
Fig. 4

A schematic representation of the human ACNS. The sympathetic, parasympathetic, and mixed nerves are represented in white, black, and grey, respectively. Superior cardiac nerve (SN), middle cardiac nerve (MN), inferior cardiac nerve (IN), and most vagal cardiac branches easily reach the heart by accompanying the great vessels. On the other hand, the course of TN is complicated because the distance from the posterior mediastinum to the heart is long. The mixing positions (cardiac plexus) of the sympathetic cardiac nerve and the parasympathetic vagal cardiac branches as well as the definitive morphology of branchial arteries with the recurrent laryngeal nerves, tend to differ on each side

Peripheral cardiac nervous distribution

The cardiac inlet/outlet entrances are limited due to the pericardium, arterial and venous parts of the cardiac hilum, vacancy of the pericardium on the dorsal aspect, and the cardiac hilum itself (Pauza et al. 1997). The arterial part (ascending aorta and pulmonary trunk) and a portion of the venous part (superior caval vein) tend to enter and leave a relatively thick cardiac nerve/branch originating from the main cranial trunks, whereas most venous parts (pulmonary veins and inferior caval vein) tend to enter and leave the relatively thin cardiac nerve/branches from the caudal aspect. The nerves of the arterial part from a well-developed bundle with the cardiac plexus, course between the aorta and pulmonary trunk, run along the great vessels and both coronary arteries, and distribute to the atria, ventricles, and coronary arteries (Kawashima 2005; Kawashima and Sasaki 2008; Fig. 5a–d). On the other hand, the nerves of the venous part tend to run independently onto the pericardium, especially in the vagal cardiac branches and run along with the surrounding vessels independently without forming the cardiac plexus until the entrances of the pericardial reflection (Fig. 5e). In addition to our macroscopic anatomy, Kawano et al. (2003) clarified histochemically that there were more acetylcholinesterase (AChE)-positive nerves (parasympathetic vagal nerve) and tyrosine hydroxylase (TH) positive nerves (sympathetic nerve) in the atrium than in the ventricle, whereas there were more TH-positive nerves than AChE-positive nerves in the ventricle. Detailed architecture of the cardiac plexus, including intrapericardial cardiac ganglia in the nerves of venous parts and cardiac base, was reported using a histochemical staining of the entire human heart by Pauza et al. (2000). Their results describing the location of the cardiac ganglia corresponded with our human and nonhuman primate results regarding the cardiac nerve/plexus (Kawashima and Sasaki 2008; Kawashima et al. 2008, 2009). Therefore, differences between the arterial and venous nerve groups were confirmed and agreed in all the above observations.
https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig5_HTML.gif
Fig. 5

The peripheral cardiac nervous distribution. a, c Right coronary plexus. Arterial group comprising relatively thick cardiac nerves/branches. b, d Left coronary plexus. The bundle (asterisks) of the cardiac plexus runs between the ascending aorta (AA) and pulmonary trunk (PT), runs along the great vessels and both coronary arteries, and then distributes to the atria, ventricles, and coronary arteries. e Venous group comprising very thin branches, dorsal view. Black arrowheads and squares show the course and entrance to the pericardium reflection, respectively (after Kawashima and Sasaki 2008)

Clinical significance

In cardiovascular surgery that does not plan for autonomic nervous preservation, especially cardiac transplantation, most patients cannot undertake strenuous activities without anginal pain because of denervation (Raczak et al. 1999; Bengel et al. 2001, 2002). In addition, denervated patients constantly experience orthostatic hypotension and the ineffectiveness of antiarrhythmic agents acting on autonomic nerves such as atropine and foxgloves. Furthermore, the fact that autonomic reinnervation increases with time after surgery may help to accelerate nonpreserving operation of ACNS in complicated cardiac surgery (Wilson et al. 1991; Schwaiger et al. 1991; Kaye et al. 1993; De Marco et al. 1995; Bengel et al. 1999; Überfuhr et al. 2000). However, some data suggest that reinnervation after cardiac transplantation is not simply a function of time, and that there are significant differences regarding the degree of reinnervation and frequency of occurrence, although reinnervation is more common in young age, fast and uncomplicated surgery, and low rejection frequency (Raczak et al. 1999; Bengel et al. 2002). Based on the patterns of denervation and reinnervation following not only the cardiac transplantation but also other cardiac surgeries such as heart valve replacement, improved operation techniques are needed in cardiovascular surgery to maintain patient QOL. At present, two major techniques of cardiac transplantation are used, i.e., the Lower–Shumway (biatrial) and bicaval techniques (Fig. 6a, b). Based on anatomical knowledge, both have merits and demerits. In the bicaval technique, it may be easier to regenerate the cardiac plexus because the inter-neural distances are generally shorter and the cardiac ganglionated plexuses are concentrated in the anastomotic area: however, drawbacks include the necessity to reconstruct or eliminate broken pericardium, especially in superior caval vein (SVC), to avoid hernias of the heart and its great vessels, and the high possibility of damage to the sinuatrial node (SAN) although patients are obliged to use a pacemaker. In the Lower–Shumway technique, the pericardium is preserved well without complicated destruction, but regeneration of cardiac nerves is believed to be a lengthy process because of their nervous distances and concentrations. Some clinical functional results appear to support some of our anatomical findings. the bicaval technique could preserve atrial function and improve allograft exercise performance, valve regulations, atrial enlargement, atrioventricular regulation and conduction, and sinuatrial (SA) function more efficiently than the Lower–Shumway technique (Freimark et al. 1995; Grande et al. 2000). According to Sakaguchi (2010), the bicaval technique is still associated with problems caused by loss of continuous caval veins. Interestingly, Kitamura et al. (2001) recommended a modification of the bicaval technique that perfectly prevented shrinkage and retraction of the caval tissue, thus providing easier anastomotic orientation and better estimation of the appropriate tissue length that fit well, particularly when a small donor heart was available (Fig. 6c). Further improvement in this technique is required by cooperation with cardiac surgeons to maintain patient QOL from the clinical anatomical viewpoint.
https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig6_HTML.gif
Fig. 6

Diagram showing cardiac transplantation techniques. a Lower–Shumway (biatrial anastomosis) technique. b Bicaval anastomosis technique (after Osaka University Heart Transplant team website, http://www.osaka-transplant.in.arena.ne.jp/page/shinisyoku/01.html). c Modification of bicaval anastomosis technique (after Kitamura et al. 2001)

Modifications of the human ACNS in anomalous branchial arteries

Research background

Even simply in terms of developmental biology, in order to examine the unique development of arterial anomaly, it is important to understand its normal development, function, and treatment (Kasai 1962; Bergwerff et al. 1999; Molin et al. 2002). I also believe that topographical changes in the ACNS concerning surrounding vessel anomalies provide clues regarding the normal fundamental architecture, possible course of ACNS in abnormal cases, and clinical applications. In other words, this examination does not consider the morphogenesis of the arterial anomaly as noted in conventional anatomical research. Consequently, we have analyzed the extensive changes seen in the ACNS using the relatively high frequency branchial anomalies, retroesophageal right subclavian artery (RERSA) and left vertebral artery directly originating from the aortic arch (ALVA), as a model for understanding normal ACNS structure, in order to elucidate useful information for anatomists, clinicians, and developmental biologists.

Retroesophageal right subclavian artery

RERSA—originating as the last branch of the aortic arch and running dorsally to the esophagus—has been reported at a frequency of 0.1–1.6% (for recent examples, see Nakatani et al. 1996; Kurt et al. 1997; Shimada et al. 1997; Jakubowicz et al. 2002; Loukas et al. 2004). Furthermore, a few papers have described ACNS in addition to RERSA morphology (Kumaki 1980; Chiba et al. 1981; Horiguchi et al. 1982; Suzuki et al. 1990; Tanaka 2000). However, these reports have not clarified all details because these cases were found during student dissection practice. The common findings are as follows: (1) absence of the right SN originating from SG; (2) rare presence of the right SN originating from ST; (3) rare absence of the right MN originating from MG; and (4) consistent presence of the thick right TN along RERSA. Based on these common findings, Horiguchi et al. (1982) concluded that the right cardiac nerves were shifted secondarily and caudally following the disappearance of the right fourth branchial artery in human ontogeny, which gives rise to RERSA. Unfortunately, our cases, which were intact cadavers provided for clinical anatomical studies, did not accurately reflect conventional findings (Kawashima and Sasaki 2005; Fig. 7). We observed all the right cardiac nerves, including the right SN originating from SG, in addition to the right TN along RERSA. Thus, it should be noted that one cannot simply conclude that the caudal shift of the right TN is due to RERSA. Although the development of RERSA was conventionally considered to occur from the right seventh intersegmental artery and a distal part of the right dorsal aorta after interrupting the proximal part consisting of the third to seventh segments of the right dorsal aorta, recent developmental progress suggests the need for revision (Aizawa et al. 1999; Bergwerff et al. 1999; Hiruma et al. 2002; Molin et al. 2002). These experimental findings clearly suggest that the proximal part of RERSA is more caudal than the distal part. In our two RERSA cases, MN originating from VG accompanied the distal part of RERSA, whereas the right TN originating from 2TG and 3TG accompanied the proximal part of RERSA. Therefore, the cardiac nerves may have an accompanying segment similar to the developmental segment of arteries. These anatomical findings of ACNS also reflected spatial changes of the great arteries (Kawashima and Sasaki 2005).
https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig7_HTML.gif
Fig. 7

Photographs and diagrams showing the human ACNS with the retroesophageal right subclavian artery (RERSA). a RERSA and the surrounding structures in the right cervical and thoracic region of case 1. b Regional enlargement of a. c Left cervical and thoracic structures of case 1. d Detailed anatomical relationship between ACNS and RERSA of case 1. e Case 2 viewed from the ventral aspect. RERSA running dorsally to the trachea and esophagus is indicated by asterisks. The corresponding right recurrent laryngeal nerve (RL) is shown by a star. The cardiac nerves accompanying the proximal and distal parts of RERSA are shown by the single and double arrowheads, respectively

Anomalous left vertebral artery originating from aortic arch

To confirm whether our findings and conclusions regarding ACNS changes in RERSA apply to other cases, we also examined topographical changes of the ACNS with a common branchial arterial anomaly, namely an anomalous left vertebral artery (ALVA) (Kawashima and Sasaki 2007; Fig. 8). Proceeding from proximal to distal, the great arterial branches of the aortic arch are the brachiocephalic trunk, the left common carotid, ALVA, and left subclavian arteries. To the best of my knowledge, ACNS with ALVA has not been reported. In this case, the hypothesis of cranio-caudal shift of ACNS following a branchial arterial anomaly as proposed by Horiguchi et al. (1982) was also contradicted because all cardiac nerves/branches were bilaterally observed. In addition, the cranial portion of the CT near the ALVA was constricted, as shown by the large arrow in Fig. 8d, and contributed to cardiac innervation along ALVA, as shown by the small arrowheads. According to a previous description by Hoffman (1957), numerous ganglion cells are scattered throughout the cervical ST. The general and accessory cervical ganglia may thus have differentiated from the neighboring ST, depending on the appearance of the anomalous great vessels. Furthermore, a segmental accompanying tendency of the left cardiac nerve was also recognized; the cardiac nerve accompanying ALVA was more caudally located than the nerves accompanying the left common carotid artery, and more cranially located than the nerves accompanying the left subclavian artery.
https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig8_HTML.gif
Fig. 8

Photographs and diagram showing the ACNS in an individual with an anomalous left vertebral artery (ALVA) as indicated by the asterisk. a Ventral view. b Dorsal view. ACNS accompanying the great arteries. c Cardiac plexus viewed from the dorsal aspect. d Detailed anatomical relationship between ACNS and ALVA. Black stars, squares, and circles indicate the inlet/outlet of the cardiac nerve/branches from the arterial part of, venous parts of the cardiac hilum, and cardiac hilum itself, respectively. The nerves accompanying ALVA indicated by short arrows are present in addition to normal ACNS including two different courses of SN along left common chorotid artery (LC) and via RL as indicated by single and double arrowheads, respectively

Morphological significance with clinical implications

Based on our findings, the great arterial branching pattern of the aortic arch does not cause a shift in ACNS, and de novo cardiac nerves along the anomalous artery appear in addition to the ordinary ACNS with a segmental accompanying tendency of the cardiac nerve/branch similar to the developmental segment of the artery (Fig. 9). Furthermore, these anatomical observations of relationships between the arterial and peripheral autonomic nervous systems do not involve problems from a recent developmental viewpoint. It is essential for the migration, axonal growth and axon guidance of developing sympathetic neurons to use blood vessels as intermediate pathways to reach final target tissues (Enomoto et al. 2001; Kuratani 2004). Recently, such nerve-vessel interactions have also been reported in detail (Mukouyama et al. 2002, 2005).
https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig9_HTML.gif
Fig. 9

Schematic representation of the segmental accompanying tendency of the ACNS arrangement to the normal and abnormal great arterial branches. a Normal. The cardiac nerve originating between superior cervical ganglion (SG) and middle cervical ganglion (MG), the cardiac nerve originating between vertebral ganglion (VG) and upper CT, and the cardiac nerve originating below the lower CT, tend to accompany along the common carotid, subclavian, and intercostal arteries, respectively. b RERSA. Proximal and distal RERSA are colored in blue and light blue, respectively. The cardiac nerves originating between VG and upper CT accompany the distal RERSA, whereas those originating between the lower CT and upper TG which do not normally accompany the subclavian artery, accompany the proximal RERSA derived from the lower branchial arterial origin. c ALVA colored in blue. These findings of two types of the branchial arterial anomaly show that additional cardiac nerves along the anomalous artery may be present in addition to the normal ACNS arrangements. Furthermore, the cardiac nerves have an accompanying segment similar to the developmental segment of arteries

These results suggest clearly that ACNS morphology can be speculated by simply examining the branching pattern of the great vessels with currently used imaging techniques such as computed tomography (CT) and magnetic resonance imaging (MRI): however, these techniques cannot detect complicated ACNS (Fig. 10). Thus, in the future, cardiac surgeons will be able to design individual made-to-order operations for preserving the autonomic nervous system by using typical angiography.
https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig10_HTML.jpg
Fig. 10

Anatomical speculations of ACNS arrangements from computed tomography (CT) angiography data. ACNS is superimposed on CT data. ac Right aortic arch viewed from the ventral aspect. Although some ACNS variations would be expected because the great venous systems often contains observed anomalies such as a bilateral superior caval vein (a) and anomalous left brachiocephalic vein running between the brachiocephalic and right common carotid arteries (b), most ACNS arrangements including the cardiac plexus indicated by purple arrows are summarized in c, which is a mirror-image positioned relative to the normal arrangement as shown in a small diagram. RERSA viewed from the ventral (d) and dorsal (e) aspects. The topographical changes of the cardiac nerves accompanying with RERSA, as shown by the orange arrows, are expected in addition to ordinary ACNS

Evolutionary comparative morphology of ACNS

Experimental animals

To understand the function and consider the clinical application of ACNS, experimental studies have employed model animals such as rats, mice, or dogs. However, ACNS morphologies show considerable interspecific differences between humans and these experimental animals. For example, the main trunk of the cervical sympathetic nerve and the vagus nerve are fused as one, vagosympathetic trunk in dogs (Nonidez 1939; Mizeres 1955, 1958). Therefore, we performed a comparative anatomical examination of ACNS in dogs and rats as a preparatory study. The morphologies in both mammals were very different from that in humans. In dogs, the main sympathetic contributor was CT comprising IG and 1TG–3TG as well as the presence of the vago-sympathetic trunk. Therefore, the main nerve supply and composition of CT are remarkably different in these species. In rats, we observed large venous differences, such as bilateral superior caval veins and left azygos arch, no SN originating from SG, no communicating branch from MG to the spinal cervical nerve, and a well-developed CT comprising IG, 1TG, and 2TG. Consequently, the experimental data obtained in typical experimental animals are difficult to apply and translate directly to humans, especially in ACNS research.

Evolution in placental mammals

Remarkably, recent molecular analysis has revised placental mammal evolution to two taxa: Boreoplacentalia and Notoplacentalia. Notoplacentalia is further divided into Xenarthra, comprising mammals such as the sloth, anteater, and armadillo; and Afroplacentalia, comprising mammals such as the elephant, manatee, hyrax, golden mole, and tenrec. On the other hand, the Boreoplacentalia is divided into Laurasiaplacentalia, comprising numerous mammals such as dog, cat, whale, horse, shrew, bear, bat, and hedgehog; and Archontoglires, comprising primates including humans, tree shrew, flying lemur, and rodents (Arnason et al. 2008).

According to recent molecular evolution research, the divergence of Euarchontoglires is estimated to have occurred 88–90 million years ago (Hedges and Kumar 2003; Janecka et al. 2007), and the basal Boreoplacentalia split between Laurasiaplacentalia and Archontoglires was dated to 99 million years ago (Arnason et al. 2008). These results show that commonly used experimental animals are remarkably distant from humans, and this is supported by our anatomy of ACNS. Therefore, it is essential that primate ACNS morphology is clarified before results from experimental animals can be translated to humans.

A short history of primate ACNS research

Although most regional anatomical studies on the peripheral nervous system in primates have focused on the relationship between the somatic peripheral nerves and muscles or somatic limb plexus for comparison with humans, the autonomic nervous system in primates is uncommon. In comparative anatomical reports of ACNS in primates, most studies have focused on the relationship between the sympathetic ganglia and the spinal nerves (van den Broek 1908; Riegele 1926; Botar 1932; Botár et al. 1950; Kolesnikov 1935; Zuckerman 1938; Sheehan and Pick 1943; Pick 1970) in addition to research on primate ACNS with the surrounding structures (Botár and Becker 1939; Riegele 1925–1926; Sonntag 1922, 1924). As noted in these studies, the ACNS in Old World monkeys and some great apes have been relatively well studied. However, findings in the other primates are sparse. For prosimians (Strepsirrhini), New World monkeys (Platyrrhini), and gibbons (Hylobatidae), only one study by van den Broek (1908), who dissected a black lemur (Lemur macaco = Eulemur macaco), a spider monkey (Ateles ater = Ateles paniscus), and a white-handed gibbon (Hylobates lar) is present. Therefore, we examined ACNS in several primate species and most genera and families [lorises and galago (Strepsirrhini), T.K. and R.W. Thorington, in preparation; New World monkeys (Platyrrhini), Kawashima et al. 2009; Old World monkeys (Cercopithedae), Kawashima and Sato 2000; Kawashima et al. 2001, 2005, 2007; gibbons (Hylobatidae), Kawashima et al. 2008; great apes (Hominoidea), T.K. et al., in preparation]. Our studies have clarified that the findings of van den Broek (1908) do not represent a common morphology in each species, family, and group.

Comparative anatomy of primate ACNS

Our current understanding of the evolution of ACNS morphology, based on previous results, is shown in Fig. 11 (Kawashima et al. 2009). Consequently, ACNS morphology exhibits significant evolutionary changes within the primate lineage. Strepsirrhini and New and Old World monkeys share similar morphologies, but SG acquired gradually communicating branches with the cervical nerves in conjunction with their development in size. In the former three groups (Strepsirrhini, New World monkeys, and Old World monkeys), this ganglion appears to be a contributor to the head and upper cervical organs and has become a contributor to the heart in the gibbons. In addition, in hominoids, there is reduced composition of CT, the presence of VG, and high frequencies of appearance of SN and TN. Therefore, the range of origin of the sympathetic cardiac nerves tends to expand from Strepsirrhini to humans: however, this range expansion is not paralleled in the parasympathetic system. Thus, the parasympathetic vagal cardiac branches are consistent among the primates (Kawashima and Sasaki 2008).
https://static-content.springer.com/image/art%3A10.1007%2Fs12565-010-0096-0/MediaObjects/12565_2010_96_Fig11_HTML.gif
Fig. 11

Comparative anatomy of ACNS in primates. a New World monkeys (Platyrrhini). b Old World monkeys (Cercopithecidae). c Gibbons (Hylobatidae). d Humans. The solid lines indicate communicating branches and cardiac nerves/branches with more than 60% occurrence and broken lines indicate those <60%. The parasympathetic vagal cardiac branches are consistent among primates, whereas the sympathetic cardiac nervous system has continuous variable morphology associated with primate evolution. The comparative morphology within the lineage is concordant with the phylogeny, suggesting that primate ACNS preserves its evolutionary history in close alignment with phylogeny

Evolutionary morphological implication of primate ACNS

Mammalian and avian ontogenetic developmental investigations of ACNS have been presented in a number of studies (Sharner 1930; Gomez 1958; Kirby et al. 1980; Kuratani and Tanaka 1990; Kuratani et al. 1991; Roberts 1991; Verberne et al. 1998; Shoba and Tay 2000; Hildreth et al. 2009). In mouse, the parasympathetic vagal nerves are first observed in the venous porta at E10.5 (10.5 days after fertilization) and reach the heart by E12.5, whereas the innervation of the outflow (arterial porta) occurs slightly later (Hildreth et al. 2009). Sympathetic cardiac innervation has been shown to occur at later stages than parasympathetic vagal cardiac innervation (Kirby et al. 1980; Shoba and Tay 2000; Hildreth et al. 2009).

Our comparative anatomical studies show that the variable sympathetic cardiac system and the completely consistent parasympathetic vagal cardiac system within primates are associated with later mammalian ACNS ontogeny in sympathetic innervation and earlier ontogeny in parasympathetic vagal innervation (Kirby et al. 1980; Kuratani et al. 1991; Shoba and Tay 2000; Hildreth et al. 2009). Interestingly, our comparative morphological studies of the primate lineage correspond with both primate phylogeny and autonomic nervous system ontogeny (Kawashima et al. 2009).

Although somatic system components, such as bones and muscles, and their innervation, have been examined in conventional anatomical studies, recent morphological studies have naturally changed their aim and contribution from evolutionary and taxonomic biology to functional anatomy and adaptation. However, the autonomic system probably has the possibility of preserving fundamental structures related to evolution because of its homeostasis. In our series of ACNS studies, comparative morphology within the lineage was concordant with phylogeny. This suggests that the primate ACNS has preserved its evolutionary history. Therefore, it is important to understand and translate these functional experimental findings to humans. My future additional comparative anatomical study aims at further understanding for human application, developmental studies, and evolutionary biology.

Acknowledgments

I would like to devote this paper to my two directors, Prof. Dr. Hiroshi Sasaki of Tokyo Women’s Medical University and Prof. Dr. Kenji Sato of Tokyo Medical and Dental University, as a memorial to their retirement. I also thank Prof. Emeritus Dr. Tatsuo Sato and Prof. Dr. Keiichi Akita of Tokyo Medical and Dental University for their continuous instructions as well as encouragement, and Prof. Dr. Nobuhisa Hagiwara and Dr. Takatomo Nakajima of Tokyo Women’s Medical University for supplying computed tomography data from clinical cases. This research was supported by grants from the Scientific Research (KAKENHI) of the Ministry of Education, Culture, Sports Science and Technology (MEXT) (No. 16790804, 2004–2006; No. 19790985, 2007–2009), the Japan Society for the Promotion of Science (JSPS) core-to-core program HOPE (2005–2007), and the Global COE program TWMU-MERCREM (2010–2011).

Copyright information

© Japanese Association of Anatomists 2010