General Thoracic Surgery (General Thoracic Surgery (Shields)) [2 VOLUME SET]

Editors: Shields, Thomas W.; LoCicero, Joseph; Ponn, Ronald B.; Rusch, Valerie W.

Title: General Thoracic Surgery, 6th Edition

Copyright 2005 Lippincott Williams & Wilkins

> Table of Contents > Volume II > The Esophagus > Section XVIII - Anatomy > Chapter 124 - Lymphatic Drainage of the Esophagus

Chapter 124

Lymphatic Drainage of the Esophagus

Yael Refaely

Mark J. Krasna

ANATOMY

The esophagus begins in the hypopharynx and travels from the neck through the visceral compartment of the mediastinum and the esophageal hiatus to terminate at the cardia of the stomach. Histologically, the esophagus has four layers: mucosa, submucosa, muscle coat or muscularis propria, and adventitia. There is no serosa.

There have been various attempts to categorize esophageal segments, mostly based on anatomic relationships to surrounding structures such as the carina, aortic arch, and hilum. One of the most acceptable methods is to divide the esophagus into three segments: cervical, thoracic, and abdominal. The cervical esophagus is about 5 to 6 cm in length. It extends from C6 to T1, or from the cricoid cartilage and cricopharyngeal muscle to the thoracic inlet at the level of the sternoclavicular joints.

The thoracic portion of the esophagus is located in the superior and visceral compartment of the mediastinum and extends from the level of T1 to T10 or T11. It begins slightly to the left of the midline and shifts slightly to the right and again to the left. It is in the midline at the level of T5 to T7. The abdominal esophagus is 0.5 to 2.5 cm in length. The abdominal esophagus lies at the level of T11 or T12.

Considering the differences in behaviors and treatments of esophageal cancer based on anatomic location, the American Joint Committee on Cancer (AJCC), in 1997, recommended that for classification, staging, and reporting of esophageal cancer, the esophagus be divided into four regions. The different portions are described in terms of anatomic and endoscopic location. The cervical esophagus begins at the lower border of the cricoid cartilage and ends at the thoracic inlet (the suprasternal notch), about 18 cm from the upper incisor teeth, when measured endoscopically. The intrathoracic esophagus is further subdivided into three portions: (a) the upper thoracic portion that extends from the thoracic inlet to the level of the tracheal bifurcation, about 24 cm from the upper incisor teeth; (b) the midthoracic portion that sits between the tracheal bifurcation and the distal esophagus just above the esophagogastric junction (the lower level of this portion is about 32 cm from the upper incisor teeth); and (c) the lower thoracic portion that measures about 8 cm in length and includes the intraabdominal portion of the esophagus and the esophagogastric junction. The latter is about 40 cm from the upper incisor teeth.

The variety of definitions reflects the absence of clear anatomic distinctions between the different segments of the esophagus. A representative example of this segmental parity is the unique lymphatic drainage system of the esophagus that is discussed in this chapter.

In 1903, Sakata published his study on the lymphatics of the esophagus as found in 15 fetuses. This work, which was later confirmed by others, is considered the standard in the understanding of this subject. As was recognized by Sakata (1903), the lymphatic drainage of the esophagus is mainly longitudinal and intramural. The longitudinal lymphatic network is represented in the four layers of the esophagus, but it is not equally distributed in each layer. The mucosal and the submucosal networks interconnect between them to form a rich submucosal lymphatic plexus. However, the plexus in the muscular layer and adventitia is very fine and contains few and widely spread channels. According to Sakata's study, the longitudinal submucosal network is continuous cephalad with the lymphatics of the pharynx and caudad with those of the stomach. Other investigators have described contradicting findings. Weinberg (1972) described injection of dye into the mucosa and submucosa of the esophagus and stomach of patients during operation (in vivo injection); he found no free communication between the lymphatics of the esophagus and the stomach. The only instance in which such lymphatic communication was demonstrated was in a patient who had a gastric carcinoma that extended into the esophagus. In this case, the communicating lymphatic channels have been presumed to represent growth of new lymphatic into the carcinoma. These observations led to the assumption that actual anastomosing lymphatic communications between the esophagus and the stomach normally do not exist.

P.1895

Collecting trunks originating from the submucosal network intermittently pierce the muscularis propria to drain into regional lymph nodes (Fig. 124-1; see Color Fig. 124-1). They are able to penetrate through a variety of pathways, which can be transverse, ascending, or descending; importantly, they are sometimes very short, but can also be very long. The collecting trunks may also connect directly into the thoracic duct. Riquet and associates (2002) found that direct tributaries from the esophagus to the thoracic duct, without involvement of lymph nodes, were present in 10% of autopsies. Because this system is not segmental, lymph can travel a long distance in the plexus before traversing the muscle layer and entering the regional lymph nodes.

This description of the lymphatic drainage of the esophagus correlates with our clinical knowledge and assists in understanding the behavior of malignant tumors of the esophagus. It can therefore be understood why cancer of the upper esophagus can metastasize to perigastric nodes or a cancer of the lower esophagus can metastasize to superior mediastinal or cervical lymph nodes. It also explains why there is a high incidence of skip metastasis in esophageal cancer and why a tumor may extend over a long distance within the esophagus before obstructing the lumen.

It seems that lymphatics are rare or absent from the superficial mucosa but that there is a rich anastomotic network in the submucosa. Observations that support this are the submucosal predominance in intramural spread of esophageal cancer, the phenomenon of a small mucosal malignant lesion surrounded otherwise by intact mucosa and accompanied by extensive tumor spread underneath, and the occurrence of submucosal metastasis that can be found distant from the primary tumor. A study that indirectly concurs with the observation that lymphatics are poor in the mucosa but rich in the submucosa was reported by Yoshinaka and colleagues (1991) and showed that carcinoma with lamina propria invasion (T1a) has 0% lymph node metastasis, whereas tumors that invade the submucosa (T1b) were found to have 47% chance of lymph node metastasis.

In addition to lymphatic capillaries and collecting trunks, two other elements in the lymphatic drainage of the esophagus are the lymph nodes and the thoracic duct. Lymph nodes are generously distributed along the esophagus and in its surroundings. The recognized lymph node stations, their designation, and their involvement in esophageal cancer spread are discussed later in this chapter. The thoracic duct is the main lymphatic vessel of the body. It begins with the cisterna chyli at level L1 to L2, and it travels as a single trunk to the chest through the aortic hiatus of the diaphragm. In the chest, it lies between the descending aorta and the azygos vein. At the level of T5 to T6, it travels behind the left main bronchus and turns to the left side. It continues to ascend to the level of C7 and then descends to enter the left subclavian vein at its junction with the internal jugular vein. During its course to reach the bloodstream, additional lymph is added from intrathoracic structures either directly or through the local lymph nodes.

In summary, the main components of the lymphatic system of the esophagus are intramural lymphatic plexuses, which run longitudinally and dominantly in the submucosa, and collecting ducts, which pierce from the submucosal plexus to regional lymph nodes or to a lesser extent directly to the thoracic duct.

LYMPH NODE MAPPING

Lymph node mapping, based on a numeric system and precise anatomic boundaries for each nodal station, has gained wide acceptance in thoracic surgery for staging of non small cell lung cancer (NSCLC). The use of a precise numeric lymph node map has been critical to the design and conduct of clinical trials in lung cancer, and the uniformity in reporting results has allowed comparison among different investigators. Although the same arguments can be applied for esophageal cancer, there are a variety of lymph node mapping systems for this disease, and not uncommonly, studies that deal with metastatic lymph nodes add their own map for clarification. The most acceptable maps used are two different numeric lymph node mapping systems. The first one is the Japanese system that was published by the Japanese Society for Esophageal Disease (1976). The second is the most acceptable map in the Western world and was developed by Casson and co-workers (1994) based on the lymph node map for lung cancer.

In the Japanese system, stations 1 to 16 represent the abdominal lymph nodes and were adopted from the gastric cancer lymph node map that was published in 1973. The cervical and mediastinal nodes were added to this system and named stations 100 to 112 (Table 124-1; Fig. 124-2). The stations were divided into four groups, N1 to N4, according to their distance from the location of the primary tumor. Therefore, depending on the location of the primary tumor, every station was designated to a different lymph node group each time (Table 124-2).

P.1896

The Casson map is based on the mediastinal nodal stations that were originally reported by Naruke and co-workers (1978) and later modified by the American Thoracic Society and Lung Cancer Study Groups (1983) for lung cancer. The stations that were modified to the special needs of esophageal cancer staging were 3P (posterior) for upper paraesophageal lymph nodes above the tracheal bifurcation, and station 8, which was divided into 8M (middle) and 8L (lower) paraesophageal lymph nodes. Stations 15 through 20 were added to the map and designated the diaphragmatic and abdominal lymph nodes (Fig. 124-3).

Comparison between these two mapping systems indicates that the Western map is more detailed in the thorax, whereas the Japanese map is more detailed in the abdomen. The Japanese map also includes four more cervical stations. These differences can be understood knowing the foundation from which these maps were adapted and developed (i.e., lung cancer staging system versus gastric cancer staging). It also can be explained on the grounds of a different surgical philosophy and technique for the treatment of esophageal cancer in the various parts of the world.

The right recurrent laryngeal nerve (CRLN) node is a station that is not included in either of the two mapping systems, although it is commonly mentioned in the literature. This nodal station is also known as the top node and is considered a relay point between the lymphatic flow from the mediastinum and the cervical lymphatic drainage, as reported by Nishihira and colleagues (1995). Its anatomic location is along the right recurrent laryngeal nerve, behind the right subclavian artery and between the trachea and the esophagus at the supreme portion of the thorax. Investigators who have studied this nodal station recommend approaching it through the cervical wound. The clinical significance of these lymph nodes according to Nishihiraand associates (1995)

P.1897

P.1898

PATTERNS OF METASTATIC SPREAD IN ESOPHAGEAL CANCER

Association with Tumor Location, Depth of Invasion, and Cell Type

The incidence of positive metastatic lymph nodes in carcinoma of the esophagus ranges between 60% and 71% in different series. The pattern of lymph node metastasis is complex, is variable, and depends mainly on the different locations of the tumor in the esophagus (upper, middle, and lower). Depth of tumor invasion and histologic cell type (adenocarcinoma versus squamous cell carcinoma) also have been suggested to affect the pattern of spread, as discussed later.

Dividing lymph node stations into five categories (cervical, upper, middle, lower, thoracic, and abdominal) endorses the repetitive patterns of lymphatic spread for the different locations of esophageal tumors reported in the literature.

Interestingly, in the seminal study by Akiyama and colleagues (1981) that described in detail the pattern of nodal spread in esophageal cancer, the authors stated that analysis of lymph nodes demonstrated a widespread distribution of positive lymph nodes regardless of the location of the tumor. Only accumulated data from later studies consolidated Akiyama and colleagues' (1981) findings and proved a trend in the lymphatic spread in association to tumor location. Akiyama and colleagues (1981) studied positive lymph node stations after esophagectomy with a two-field (abdominal and thoracic) lymph node dissection for squamous cell carcinoma. Their results, which are presented in Figure 124-4, show that middle esophageal tumors were found to metastasize either to mediastinal or to abdominal lymph nodes. Tumors located in the upper esophagus were found to spread mostly to the mediastinum but also distant to the superior gastric nodes (defined as paracardiac, lesser curvature, left gastric artery) in up to 31.8% of the patients. However, when the tumor was located in the lower esophagus, the abdominal spread was twice as common (found in more than 61% of patients), and the distant spread to the superior mediastinum was found in only 9.8% of the patients. Moreover, positive lymph nodes were rarely found in esophageal cancer of any primary location to the hepatoduodenal ligament, along the greater curvature of the stomach, or at the hilum of the spleen. Two other studies reported by Kato (1991) and Sharma (1994) and their associates succeeded in correlating anatomic location of the tumor and pattern of lymph node metastatic spread. Both studies evaluated surgical specimens from 79 and 70 patients, respectively, who underwent esophagectomy and three-field lymph node dissection for squamous cell carcinoma. More than 80 lymph nodes (mean) were collected from each patient, and lymph node metastasis was found in about 72% of the patients in each series. It was found that patients with carcinoma in the upper thoracic esophagus rarely had nodal metastasis in the abdomen, and patients with carcinoma of the lower thoracic esophagus rarely had cervical metastases. When analyzing the results in respect to the five levels mentioned previously, it seems that in both studies, nodal metastasis is a rare finding three levels away from the location of the tumor. It is most common in the same level as the tumor and one level adjacent to the tumor. The involvement of lymph nodes that are two levels away from the location of the tumor is also very common but

P.1899

Depth of tumor invasion, or T status, was identified by Rice and colleagues (1998) as a significant predictor of regional lymph node metastasis (N1 disease). In a logistic regression analysis, compared with T1 patients, a T2 patient was 6 times more likely to have regional lymph node metastasis; a T3 patient, 23 times more likely; and a T4 patient, 35 times more likely. However, controversy exists regarding whether the depth of invasion also correlates with the pattern of the lymphatic spread. Matsubara and co workers (1999) studied a series of 110 patients with superficial squamous cell carcinoma of the thoracic esophagus who underwent systematic extended lymph node dissection. Lymph node involvement was found in 43 patients (39%) and was strongly correlated to depth of tumor invasion. Lymph node involvement was found in 0% (0 of 9), 23% (5 of 22), and 49% (38 of 78) of patients with tumor that invaded the lamina propria, muscularis mucosa, and submucosa, respectively. Anatomically distant lymph nodes, such as the RRLNs and the perigastric nodes, were involved more frequently than other intrathoracic nodes adjacent to the main tumor. None of the patients with carcinoma that invaded only the muscularis mucosa had intrathoracic node involvement. Only 9 of 78 patients with carcinoma that invaded only the submucosa had intrathoracic node involvement. These investigators suggested that lymph flow from the esophageal mucosa and submucosa is preferentially directed longitudinally along the submucosal lymphatic plexus toward distant lymph nodes in the neck and abdomen and that superficial tumors tend to metastasize distantly. Matsubara and co workers (1999) also suggested that the deeper the tumor invades the esophageal wall, the higher is the risk for metastatic lymph node spread in general and local lymph node spread in particular. Nigro and co-workers (1999) reported contradictory results in their study on the prevalence and the location of nodal metastasis in distal esophageal adenocarcinoma. Although this study involved only 15 patients, the depth of tumor invasion was found to be a strong predictor for pattern of metastatic lymph node spread. The involved nodes in patients with tumors limited to the mucosa (T1a) and submucosa (T1b) were confined to the local node groups. In contrast, distant lymph nodes were involved in 50% (5 of 10) of patients with muscular (T2) invasion (p = 0.0027).

Major distinctions between these two studies preclude comparison of their results. These differences include the fields of lymph node dissection, mean number of lymph nodes dissected for each patient, tumor cell type (adenocarcinoma versus squamous cell carcinoma), and location of the primary tumor within the esophagus.

Tumor cell type, adenocarcinoma versus squamous cell carcinoma, as an independent variable has never been studied systematically for its association with patterns of nodal metastatic spread. Comparison between studies that include patients either with adenocarcinoma or squamous cell carcinoma, with similar tumor location and similar three-field lymph node dissection, reveals no major differences in lymphatic spread (Table 124-3). Van de Ven and associates (1999) studied the pattern of lymph node spread in T3 adenocarcinoma of the distal esophagus in patients with positive

P.1900

Solitary Nodal Metastasis

Esophageal cancer patients with a solitary positive lymph node are a unique group with special interest for the investigation of metastatic lymph node spread. Matsubara and co workers (2000) reported that in 51 patients with solitary lymph node involvement who underwent curative resection for a squamous cell carcinoma of the esophagus, it was found that 82% of the positive lymph nodes were not located in the mediastinum but in the thoracocervical junction or perigastric region. The thoracocervical junction included the deep cervical and the right RLNs, whereas the perigastric group included lymph nodes of the cardia of the stomach, the lesser curvature, the left gastric artery, or a celiac lymph node. The site of solitary involvement was correlated with the tumor localization in the esophagus (Table 124-4). Tumors in the upper thoracic esophagus showed solitary metastatic involvement mainly in the neck, middle thoracic tumors were found with solitary spread to all stations without a preference, and lower thoracic tumors tended to spread to the abdomen. The thoracocervical junction lymph nodes were involved in one third (14 of 42) of the patients, even when the tumor was confined below the bifurcation of the trachea. Solitary involvement of the perigastric lymph node was not rare in tumors located in the upper two thirds of the esophagus, occurring in 21% (4 of 19) of the patients. Nomura and co-workers (2001) reported a series of 41 patients with a single or two metastatic lymph nodes who underwent esophagectomy and three-field lymphadenectomy. Only 29 (47.5%) of 61 metastatic nodes showed correlation between the tumor site and the regional metastatic lymph nodes. Correlation between tumor site and metastatic lymph nodes in the regional or adjacent compartments was found in 82%, and beyond the adjacent compartments in 18% of the cases. Of 33 patients with tumors in the thoracic esophagus, 12 (36.4%) had metastasis to the abdominal lymph nodes. Furthermore, metastatic lymph node spread was found in all five compartments only in middle thoracic esophageal tumors. Aside from the correlation between tumor location and lymphatic spread, it is also clear from their study that the phenomenon of skip metastasis, meaning metastatic lymphatic spread to a nonadjacent lymph node station, is quite common in esophageal cancer.

SENTINEL LYMPH NODES IN ESOPHAGEAL CANCER

The term sentinel lymph node was coined by Cabanas (1977) to describe a lymph node that corresponded to the primary site for metastatic spread in penile carcinoma. In his extensive anatomic studies using lymphangiograms to determine lymphatic anatomy in 100 patients with penile carcinoma, Cabana found that a node or group of lymph nodes located medially and superiorly to the saphenofemoral junction in each groin was the most common site for metastatic nodal spread. He therefore suggested that if the sentinel node was found to have metastatic disease, the patient required a standard lymphadenectomy. Conversely, and just as importantly, if the sentinel node did not have metastatic disease, the likelihood of metastatic disease being present in the lymphatic network was unlikely, and no further dissection was necessary. The concept that a primary tumor would preferentially drain through the lymphatics to a specific lymph node, and that the histology of this lymph node would represent the status of the other regional nodes, was extremely novel. In contrast to Cabanas' studies, a group from the University of California, Los Angeles, successfully demonstrated that the lymphatic drainage could vary from one individual to another and therefore is not necessarily predetermined to anatomic location. This was first described by Wong and associates (1991) in a feline model and then by Morton and colleagues (1992) in melanoma patients. To identify this sentinel lymph node in each patient, intraoperative lymphatic mapping, rather than an anatomic approach, was needed. This led to the innovation of radio guided sentinel node harvesting with the help of a hand held gamma probe and to submucosal or subdermal injection of radioactive

P.1901

The main goal of further studies will be to evaluate and to establish the role of the sentinel lymph node concept in esophageal cancer. The feasibility of the method in this anatomic area and the accessibility of the gamma probe to the inspected lymph nodes through minimal invasive methods (thoracoscopy and laparoscopy), as described by one of us (MJK) and Jiao (2000), also need to be determined. Confirmation of this approach may offer more accurate disease staging and may minimize the extent of lymph node sampling or dissection during operation, thus preventing unnecessary complications.

P.1902

REFERENCES

Akiyama H, et al: Principles of surgical treatment for carcinoma of the esophagus: analysis of lymph node involvement. Ann Surg 194:438, 1981.

Alex JC, et al: Gamma-probe-guided lymph node localization in malignant melanoma. Surg Oncol 2:303, 1993.

American Thoracic Society: Medical section of the American Lung Association. Clinical staging of primary lung cancer. Am Rev Respir Dis 127:659, 1983.

American Joint Committee on Cancer (AJCC): Esophagus. In Fleming ID, et al (eds): Manual for Staging of Cancer 5th Ed. Philadelphia: Lippincott-Raven, 1997, pp. 65 66.

Borgstein PJ, Meijer S, Pijpers R: Intradermal blue dye to identify sentinel lymph-node in breast cancer. Lancet 349:1668, 1997.

Cabanas RM: An approach for the treatment of penile carcinoma. Cancer 39:456, 1977.

Casson AG, et al: Lymph node mapping of esophageal cancer. Ann Thorac Surg 58:1569. 1994

Igaki H, et al: Prognostic evaluation for squamous cell carcinomas of the lower thoracic esophagus treated with three-field lymph node dissection. Eur J Cardiothorac Surg 19:887, 2001.

Japanese Society for Esophageal Diseases: Guidelines for the clinical and pathologic studies for carcinoma of the esophagus. Jpn J Surg 6:69, 1976.

Kato H, et al: Lymph node metastasis in thoracic esophageal carcinoma. J Surg Oncol 48:106, 1991.

Kitagawa Y, et al: The role of the sentinel lymph node in gastrointestinal cancer. Surg Clin North Am 80:1799, 2000.

Krasna MJ, Jiao X: Thoracoscopic and laparoscopic staging for esophageal cancer. Semin Thorac Cardiovasc Surg 12:186, 2000.

Matsubara T, et al: Unique distribution patterns of metastatic lymph nodes in patients with superficial carcinoma of the thoracic oesophagus. Br J Surg 86:669, 1999.

Matsubara T, et al: Localization of initial lymph node metastasis from carcinoma of the thoracic esophagus. Cancer 89:1869, 2000.

Morton DL, et al: Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 127:392, 1992.

Naruke T, Suemasu K, Ishikawa S: Lymph node mapping and curability at various levels of metastasis in resected lung cancer. J Thorac Cardiovasc Surg 76:832, 1978.

Nigro JJ, et al: Prevalence and location of nodal metastases in distal esophageal adenocarcinoma confined to the wall: implications for therapy. J Thorac Cardiovasc Surg 117:16, 1999.

Nishihira T, et al: Lymph flow and lymph node metastasis in esophageal cancer. Surg Today 25:307, 1995.

Nomura T, et al: Wide-spread distribution of sentinel lymph nodes in esophageal cancer. J Nippon Med Sch 68:393, 2001.

Rice T: Superficial esophageal carcinoma: is there a need for three-field lymphadenectomy? Lancet 354:793, 1999.

Rice TW, et al: Esophageal carcinoma: depth of tumor invasion is predictive of regional lymph node status. Ann Thorac Surg 65:787, 1998.

Riquet M, et al: Thoracic duct tributaries from intrathoracic organs. Ann Thorac Surg Mar 73:892, 2002.

Sakata K: Uber die Lymphgefasse des Oesophagus und uber seine regionaren Lymphdrusen mit Berucksichtigung der Verbreitung des Karzinoms. Mitt Grenzgeb Med Chir 11:634, 1903.

Sharma S, et al: Patterns of lymph node metastasis in 3-field dissection for carcinoma in the thoracic esophagus. Surg Today 24:410, 1994.

Tabira Y, et al: Recurrent nerve nodal involvement is associated with cervical nodal metastasis in thoracic esophageal carcinoma. J Am Coll Surg 191:232, 2000.

The general rules for the Gastric Cancer Study in Surgery. Jpn J Surg 3:61, 1973.

van de Ven C, et al: Three field lymphadenectomy and pattern of lymph node spread in T3 adenocarcinoma of the distal esophagus and the gastro-esophageal junction. Eur J Cardiothorac Surg 15:769, 1999.

Veronesi U, et al: Sentinel-node biopsy to avoid axillary dissection in breast cancer with clinically negative lymph-nodes. Lancet 349:1864, 1997.

Weinberg JA: The intrathoracic lymphatics. In Haagensen CD, et al (eds): The Lymphatics in Cancer. Philadelphia: WB Saunders, 1972, pp. 245 249.

Wong JH, Cagle LA, Morton DL: Lymphatic drainage of skin to a sentinel lymph node in a feline model. Ann Surg 214:637, 1991.

Yoshinaka H, et al: Superficial esophageal carcinoma: a clinicopathological review of 59 cases. Am J Gastroenterol 86:1413, 1991.