Recent procedures in the field of cardiovascular interventional radiology require a detailed knowledge of the carotid sheath (Chiam and Roubin,2008; Shoja et al.,2008). According to textbooks of human anatomy (Grant and Smith,1953; Hollinshead,1982; Salmons,1995; Berkovitz,2005), the carotid sheath, a distinct condensation of the deep cervical fasciae, completely encloses the common carotid artery (CCA), internal jugular vein, and vagus nerve. Parts of the sheath are composed of the superficial and pretracheal laminae of the cervical fasciae, whereas the sheath is believed to be separated from the prevertebral lamina of the cervical fasciae. Notably, most previous anatomical information seems to have been based on the pioneering study of Grodinsky and Holyoke (1938), in which the fascial planes were demonstrated as line-drawings based on observations of macroscopic slices of late-stage fetuses. However, to our knowledge, even in adults, histological studies of the carotid sheath have been limited in terms of both number and the areas examined (Parsons,1910; Piffer,1980; Zhang and Lee,2002; Hayashi,2007).
Using histological sections of eight sides of five elderly cadavers, one of the present authors, Hayashi (2007), recently classified the carotid sheath into two parts or laminae: (1) a laminar structure, the “adventitia,” enclosing each of the cervical great vessels and (2) a “common sheath” outside the adventitia. It was considered that the arterial and venous adventitial structures sometimes fused mutually and provided a septum-like connective tissue cord passing between the artery and the vein. The common sheath was often interrupted or unclear, depending on sites and individuals, but the adventitia was a consistent feature. Hayashi (2007) also demonstrated that the common sheath often fused with the visceral fascia to provide a thick connective tissue plate.
Recently, using mid-term fetuses, we observed almost the entire organization of the abdominal fasciae (Kinugasa et al.,2008; Niikura et al.,2008; Matsubara et al.,2009). Therefore, rather than adults, we chose mid-term fetuses for the present study, whose aim was to describe the development and basic configuration of the carotid sheath during the fetal period.
MATERIALS AND METHODS
The study was performed in accordance with the provisions of the Declaration of Helsinki 1995 (as revised in Edinburgh 2000). We examined the paraffin-embedded histology of 15 mid-term fetuses at 9–25 weeks of gestation (three fetuses each at 9, 12, 15, 20, and 25 weeks). The genders and craniocaudal lengths are shown in Table 1. We compared the topohistology at three levels: a superior level including the most proximal portion of the right internal and external carotid arteries; a mid-level in which the right omohyoid crossed the CCA; and an inferior level including the most proximal portion of the right subclavian artery (SCA).
CRL, craniocaudal length; CCA, common carotid artery (measured using histological sections).
The gender was identified according to the observations of histological sections of the gonad those were used for another study.
Two male and one female
Two male and one female
One male and two female
One male and two female
With agreement of the families concerned, the nine specimens earlier than 15 weeks were donated to the Department of Anatomy, Chonbuk National University in Korea, and use of the fetuses for research was approved by the university ethics committee. The fetuses were obtained by induced abortions. After abortion, each mother was orally informed by an obstetrician (at a single hospital) about the possibility of fetal donation for research: no attempt was made to encourage donation. If the mother agreed to donate the fetus, it was stocked in 10% w/w formalin solution with a specimen number for more than 3 months. Because of randomization of the numbering, it was not possible to trace the family concerned. We also used another six fetuses later than 20 weeks that had been kept at the Medical Museum of Sapporo Medical University, Japan. The use of such museum-housed later-stage fetuses did not require ethics committee approval, since there was no specific protocol and no possibility of contacting the family members concerned.
Paraffin sections were cut horizontally at a thickness of 5 mm (fetuses earlier than 15 weeks) or 10 mm (fetuses later than 20 weeks), at intervals of 50 mm (early stage) or 100 mm (late stage). There were around 100 sections for early-stage, and more than 200 sections for late-stage fetuses. Most sections were stained with hematoxylin and eosin (HE), whereas some (5–6 sections per early stage fetus) were subjected to immunohistochemical staining to identify (1) nerves, (2) the endothelium of arteries and veins, and (3) sympathetic nerves. D2-40 (MSA) immunohistochemistry for lymphatic vessels was not possible for the present materials, possibly because of unfavorable fixation conditions.
The primary antibodies used were (1) monoclonal anti-human S100 protein (Dako Cytomation, Kyoto, Japan) for nerves; (2) monoclonal anti-human alpha-1 smooth muscle actin (Dako, Glostrup, Denmark) for endothelium; and (3) anti-rat tyrosine hydroxylase polyclonal antibody (Chemicon, Temecula, CA) for sympathetic nerves. According to Hayashi et al. (2008), Dako smooth muscle actin antibody stains the endothelium of arteries and veins as well as any smooth muscle, but does not react with lymphatic endothelium. The second antibody was labeled with horseradish peroxidase (HRP), and antigen-antibody reactions were detected using the HRP-catalyzed reaction with diaminobenzidine (with hematoxylin counterstaining).
APA = ascending pharyngeal artery; AS = arytenoid swelling of the developing larynx; BPx = brachial nerve plexus; CAP = longus capitis; CCA = common carotid artery; CL = clavicle; COL = longus colli; CPI = constrictor pharyngis inferior; ECA = external carotid artery; ESO = esophagus; GH = geniohyoideus; H = hyoid bone; ICA = internal carotid artery; L = lymphatic follicles or vessels; MC = Meckel's cartilage; OMO = omohyoid; PhN = phrenic nerve; PTHY = parathyroid gland; SA = scalenus anterior; SCA = (right) subclavian artery; SCM = sternocleidomastoideus; SGgl = sympathetic nerve ganglion; ST = sympathetic nerve trunk; strap = strap muscles; TC = thyroid cartilage; THY = thyroid gland; TR = trachea; vein = internal jugular vein or other veins; VA = vertebral artery; XGgl = inferior or nodose ganglion of the vagus nerve; XN = vagus nerve
Observations at 9–12 Weeks of Gestation
The internal jugular vein was much thicker than the CCA and located lateral to the vagus nerve (Figs. 1, 2). Notably, at the mid-cervical level, the vagus nerve was located at the ventral or ventrolateral side of the CCA on both sides in all six fetuses at this stage (Figs. 1B, 2C). The CCA divided into the internal and external carotid arteries at a level including the upper part of the larynx and the root of the tongue. By 12 weeks, the bifurcation was surrounded by two large ganglia: one corresponded to the superior cervical ganglion of the sympathetic trunk, whereas the other connected with the vagus nerve. The latter ganglion was regarded as the inferior or nodose ganglion because another superior ganglion was identified in the jugular foramen. Moreover, the superior cervical ganglion expressed tyrosine hydroxylase immunoreactivity, but the nodose ganglion did not (Fig. 2B).
The strap muscles, longus capitis and longus colli, were identified clearly. The relatively thick omohyoid crossed laterally to the internal jugular vein. The CCA was 1/3–1/4 the thickness of the omohyoid (Figs. 1B, 2C). The artery was less than 0.1 mm thick at this stage (Table 1). The cervical and brachial plexuses as well as the vagus nerve were not accompanied by any definite sheath, but ran through loose connective tissues. Likewise, the great artery and vein had no accompanying fascial structures. The prevertebral lamina of the deep cervical fasciae was identified connecting the bilateral longus colli muscles (Figs. 1, 2). Likewise, the alar fascia was also identified at these stages (Fig. 2). Likewise, the pretracheal lamina appeared as a connecting band between the bilateral strap muscles, although the present figures do not show the bilaterality (Fig. 2C).
Observations at 15 Weeks of Gestation
In four sides of the three specimens examined, at the mid-cervical level, the vagus nerve was shifted to the dorsal side of the CCA, that is, the final position as that seen in adults (Fig. 3). However, unilaterally, the ventral course was still evident in another two sides of two specimens (Fig. 3C). The bifurcation of the CCA was now separated from the nodose and superior cervical ganglia (Fig. 3A): these ganglia had already shifted upward. The CCA became 1/2 as thick (0.2–0.3 mm, Table 1) as the omohyoid. A mesenchymal condensation in combination with circular fibers appeared around the common carotid, internal carotid, external carotid, and subclavian arteries: these features were regarded as the adventitia (Fig. 3B,F). The adventitia contained smooth muscle actin (SMA)-positive, very thin vasa vasorum (Fig. 3D). In the loose connective tissue near the adventitia, we found fascia- or vessel-like structures. These were immunohistochemically negative for both S-100 protein (Fig. 3C) and smooth muscle actin (Fig. 3D), and thus probably developing lymphatic vessels or fatty tissues. No common sheath was seen enclosing the CCA, internal jugular vein, and vagus nerve. Likewise, the subclavian and common carotid arteries carried no common sheath (Fig. 3E,F). The primitive lymph follicles developed in a space adjacent and lateral to the internal jugular vein and vagus nerve.
Observations at 20 and 25 Weeks of Gestation
In one of the three specimens at 20 weeks, the topographical relationship between the vagus nerve and CCA changed along the craniocaudal course: superiorly, the nerve was located on the dorsal side of the artery, whereas inferiorly it was on the ventral side (Fig. 4B,D,F,H). On the other sides of the specimens, the vagus nerve displayed its final dorsal course along the artery (Fig. 4A,C,E,G). The adventitia of thick arteries became more evident than at 15 weeks: it was composed of densely packed fiber bundles. Moreover, a common sheath for the artery, vein, and nerve suddenly appeared in combination with septum-like tissues between these structures (Figs. 4F, 5C). However, whether the common sheath was thick or thin did not depend on the sites and stages, but on the specimens: Fig. 5 (25 weeks) shows a thin sheath, especially in the lateral aspect facing the primitive fatty and lymphatic tissues. In contrast, in Fig. 4, the sheath appears thick even at 20 weeks. A common sheath also appeared between the common carotid and subclavian arteries, although it contained abundant nerves (Fig. 5E,G).
At any level along the CCA, the internal jugular vein and vagus nerve were adjacent to developing fatty and lymphatic tissues (Figs. 4C–F, 5B–D,F): the latter structures provided a thick belt, occupying a space between the sternocleidomastoideus and the scalenus anterior, and continued to the lateral cervical subcutaneous tissue. This space was demarcated by the pretracheal and prevertebral laminae of the deep cervical fascia. The former fascia enclosed the omohyoid (Fig. 5B,C), whereas the latter covered the longus capitis, longus colli, and scalenus anterior (Figs. 4, 5). The developing lymphatic and fatty tissues were similar in appearance and intermingled with each other (Fig. 5F). Whether the omohyoid attached tightly or loosely to the internal jugular vein varied between sides as well as between specimens (e.g., loose in Fig. 4D,F). The CCA was now as thick as the omohyoid (Table 1).
Although the prevertebral and pretracheal laminae of the deep cervical fasciae developed early, the carotid sheath as well as the topographical relationships among its internal components, seemed to be established later than 20–25 weeks of gestation. Moreover, whether the sheath was thick or thin did not depend on site or stage, but on individual specimens. We speculated that the final morphology is established during postnatal development and modified according to pathological conditions such as lymph node inflammation. Thus, it seems likely that, as Hayashi (2007) suggested, the adult morphology varies significantly among individuals, and that the typical morphology (i.e., a complete common sheath) is rather rare.
In fetuses, the carotid sheath might not develop according to a definite rule. Hayes (1950) emphasized that the development of fascial structures requires mechanical stress to bundle collagen fibers. The early developed longus colli seemed to create traction on the prevertebral lamina, and the fascia was expanded by the increased mass of the longus capitis and scalenus anterior. Likewise, the strap muscles seemed to create traction on the pretracheal lamina. In contrast, such mechanical stress seemed to be limited to the carotid sheath: a candidate was found in the increased lengths and thicknesses of the vessels and nerves. In fact, under the influence of the marked increase in thickness occurring at 12–15 weeks, the adventitia of the cervical great arteries became evident much earlier than the common sheath. This adventitia seemed to be regarded as one of the basic components of the fetal deep cervical fasciae. Another reason for the delayed development of the sheath seemed to be the change in course of the vagus nerve, which surprisingly was still occurring in mid-term fetuses (see paragraph below).
A striking but unexpected finding in this study was that the topographical relationship between the CCA and vagus nerve changed during the fetal development period examined: the vagus nerve shifted from a ventral course, via an intermediate position between the artery and vein, to the final dorsal course seen in adults. Likewise, the nodose (inferior) and superior cervical ganglia also shifted upward toward the cranial base after 12 weeks. At the cranial base in the early-stage fetus, the internal carotid artery should run much more ventrally to the vagus nerve than in adults, since the internal ear increases in size much earlier than any cervical structures. Conversely, at the base of the neck, the vagus nerve should cross the ventral side of the aorta or SCA. Thus, the topographical anatomy seemed to create ventral traction on the inferior cervical part of the nerve, whereas dorsal traction was most likely in the superior part of the nerve. Therefore, in the early- and mid-term fetuses, the vagus nerve seemed to adopt a course crossing the CCA from the dorsal to the ventral aspect. Notably, along this course, the nerve passed through a narrow gap between the artery and vein, and not laterally to the vessels. Thus, the development of a common sheath for the vessels and nerve might be disturbed or even destroyed by the “knife-like” action (i.e., the moving nerve).
Nevertheless, a hypothetical factor of dorsal traction is lost in the late-stage fetus, since the jugular foramen shifts toward the internal carotid artery in accordance with a decrease in the relative size of the inner ear. Thus, again, it is necessary to explain why the vagus nerve finally takes the dorsal course. We noted an increase in thickness of the CCA: it became thicker than the vagus nerve at stages between 12 and 15 weeks, and thicker than the omohyoid by 20 weeks. This marked increase in thickness was likely to push the nerve dorsally, since the ventral space was limited due to the thick strap muscles in mid-term fetuses. Moreover, at the level of the carotid artery bifurcation, the nodose ganglion of the vagus nerve adhered to the superior cervical ganglion of the sympathetic trunk at 12 weeks. The upward movement of these ganglia (especially the sympathetic ganglion) in the later stage, as demonstrated in this study, seemed to create dorsal traction on the vagus nerve.
Is the carotid sheath separated from fatty and lymphatic tissues by the pretracheal lamina? Using epoxy sheet plastinations, Zhang and Lee (2002) observed that the carotid sheath became attached to the subcutaneous fatty tissue without any clear demarcation by a fascia. Likewise, according to photographs shown in the report by Hayashi (2007), the subcutaneous fatty tissue made contact with the carotid sheath. In the present fetuses, the developing fatty tissue provided a thick sheet extending dorsolaterally from the adventitia of the CCA to the subcutaneous area. This fatty tissue sheet was sandwiched by the pretracheal lamina containing the omohyoid and the pretracheal lamina covering the scaleni. Thus, in contrast to accepted understanding, the subcutaneous fatty tissue was likely to be adjacent to the carotid sheath. The developing fatty tissue contained abundant lymphatic follicles. It is well known that regional recurrence after neck dissection remains a significant cause of failure and death in patients with squamous cell carcinoma of the head and neck. Khatif-Hefetz et al. (2004) reported that, in 40 carotid sheath specimens from 34 patients after neck dissection (N-0, 19 patients; N-1, 2; N-2, 13; N-3, 6), the fibro-adipose tissue commonly showed neutrophil infiltration, and in four of the 40, lymphocyte aggregation. The carotid sheath seemed to be often invaded by lymphatic tissues because of the intimate topographical relationship.