This field includes the superficial lymph nodes, cervical paraesophageal lymph nodes, deep cervical lymph nodes, peripharyngeal lymph nodes, and supraclavicular lymph nodes. These lymph nodes also include the lymphatic chain along both recurrent nerves throughout their mediastinal and cervical course and the deep cervical nodes posterior and lateral to the jugular vein and supraclavicular nodes.

This field includes the upper, middle, and lower mediastinal lymph nodes. The upper mediastinal lymph nodes include the upper thoracic paraesophageal lymph nodes, both recurrent nerve lymph nodes, pretracheal lymph nodes, and left tracheobronchial lymph nodes. The middle mediastinal lymph nodes also include middle thoracic paraesophageal, subcarinal, and both main bronchial lymph nodes. The lower mediastinal lymph nodes include lower thoracic paraesophageal lymph nodes, supradiapharagmatic lymph nodes and posterior mediastinal lymph nodes.

This field includes both cardiac lymph nodes, lymph nodes along the lesser curvatures, lymph nodes along the left gastric artery, lymph nodes along the common hepatic artery, lymph nodes along the celiac artery, and lymph nodes along the splenic artery.

Thoracic esophageal carcinoma is commonly accompanied by extensive lymph node metastasis in the cervical, thoracic, and abdominal regions. However, the distribution and incidence of lymph node metastasis both may vary according to the location, size, and depth of tumor invasion. Therefore, preoperative evaluation using computed tomography, ultrasonography, magnetic resonance imaging, or positron emission tomography for each patient is important for determining the extent of the lymph node dissection. The concept of three-field dissection was developed in Japan in the 1980s. In Japan, three-field lymph node dissection, including dissection of the cervical, mediastinal, and abdominal lymph nodes, is the standard procedure for surgically curable esophageal cancer located in the middle or upper thoracic esophagus. The effectiveness of extended lymphadenectomy for esophageal cancer has not yet been proven by randomized prospective studies; better survival can be obtained with three-field lymph node dissection than with two-field lymph node dissection in Japan.

Since a nation-wide retrospective study by Isono et al. in 1991 showed the potential benefits of esophagectomy with three-field dissection, many reports have been published [3]. The largest study demonstrating the benefits of the three-field lymph node dissection from a single institution was reported by Akiyama et al. in 1994 [4]. The authors performed 393 cases of esophagectomy with a two-field lymph node dissection between 1973 and 1984 and 324 cases of esophagectomy with a three-field lymph node dissection between 1984 and 1993. The 5-year survival rate of node-negative patients was 83.9 % after three-field lymph node dissection and 55.0 % after two-field lymph node dissection. The 5-year survival rate of node-positive patients was 43.1 % after three-field lymph node dissection and 27.9 % after two-field lymph node dissection. In both groups, the node-negative and node-positive groups, the survival of patients after extensive three-field dissection was significantly better than that after the less extensive two-field dissection. The authors speculated that the differences may be because of occult cancer-positive nodes in the cervical region and other areas, which may have been present and omitted from dissection and analysis in the group with less extensive dissections, were removed by extensive dissection. The 5-year survival rate of patients with all depth of cancer invasion after extensive three-field and the less extensive two-field dissection were 53.3 % and 37.5 %, respectively. Although this study was a non-randomized, historical control study, the 5-year survival rate of 53.3 % in the patient after three-field dissection in those days remained very high. Turumaru et al. studied the state of lymph node metastasis in cases with only a single node metastasis [12]. A single node metastasis in patients with thoracic esophageal cancer may be located in the cervical (14.1 %), mediastinal (upper, 31.0 %; middle, 11.3 %; and lower, 8.5 %) and abdominal areas (35.2 %). They also studied the state of lymph node metastasis in 5-year survivors of these cases and showed that 14.2 % had a single node metastasis in the cervical area, 49.3 % had a single node in the mediastinum (upper, 19.4 %; middle, 22.4 %; and lower 7.5 %), and 37.3 % had a single node in the abdomen. These results showed that even if there were lymph node metastases in either the cervical or abdominal areas, many patients could be cured by extended lymphadenectomy. These results also suggested that lymph nodes in the cervical and abdominal areas were regional lymph nodes of the thoracic esophagus.

Only two prospective studies have been published from Japan. One was a prospective randomized trial comparing three-field with two-field lymph node dissection published by Nishihara et al. [13]. They showed a survival benefit for three-field over two-field lymph node dissection (65 % vs. 48 %); however, the study was a low volume study at a single institution and the difference was not statistically significant. Another prospective study was published from the National Cancer Center in Tokyo [14]. It was a non-randomized, case-matched trial and showed that the 5-year survival rate was significantly better after three-field dissection (48 % vs. 33 %; p = 0.03). The 5-year survival rate in the group of patients with a cervical lymph node was as high as 30 %. These results suggested that there was a survival advantage in the three-field lymph node dissection and that lymph nodes in the cervical and abdominal areas were regional lymph nodes for thoracic esophageal squamous cell carcinoma. Although the incidence of esophageal cancer is increasing, the number of candidates for potentially curative resection is limited. For this reason, a prospective randomized study will be difficult to complete within a reasonable timeframe. It can also be very difficult to set up high-volume multi-institutional prospective randomized studies.

There have been two reports that have supported the three-field dissection in Western countries. Altorki et al. from New York performed esophagectomy with three-field dissection on 80 patients (adenocarcinoma, 48; squamous cell cancer (SCC), 32) during the period from 1994 to 2001 [9]. Hospital mortality was 5 % and morbidity was 47 %, and 69 % presented with nodal metastases. Metastases to the recurrent laryngeal and/or deep cervical nodes occurred in 36 % patients regardless of the cell type or location of the tumor within the esophagus. Overall, the 5-year and disease-free survival rates were 51 and 46 %, respectively. The 5-year survival rate in patients with positive cervical nodes was 25 % (SCC, 40 %; adenocarcinoma, 15 %). The authors concluded that esophagectomy with three-field lymph node dissection could be performed with low mortality and reasonable morbidity and that the data suggested a true survival benefit. Lerut et al. reported on their experience with esophagectomy with three-field dissection in Belgium [10]. They performed an esophagectomy with the three-field dissection in 174 patients (adenocarcinoma, 96; SCC, 78) during the period from 1991 to 1999. Hospital mortality was 1.4 % and morbidity was 57 %. Overall, the 5-year and disease-free survival rates were 41.9 and 46.3 %, respectively. In addition, the overall 5-year survival rate with positive cervical nodes was 27.7 % for squamous cell carcinoma located in the middle third of the esophagus, 11.9 % for adenocarcinoma in the distal third of the esophagus (the 4-year survival rate was 35.7 %), and 0 % for gastroesophageal junction (GEJ) adenocarcinoma. They concluded that esophagectomy by the three-field lymph node dissection can be performed with low mortality and acceptable morbidity and that data may indicate a real survival benefit. Role of the three-field dissection for adenocarcinoma of the distal third of the esophagus remains unclear.

In cases of lower thoracic esophageal carcinoma, lymph node metastasis occurs primarily in the mediastinal and abdominal regions, but metastasis to cervical lymph nodes can also occur at a lower frequency. The prognosis of a patient with cervical lymph node metastases from a lower thoracic esophageal carcinoma is very unfavorable [15]. Thus, mediastinal and abdominal lymphadenectomy may be adequate for lower thoracic esophageal carcinoma.

In summary, the cervical lymph nodes are at a risk of being involved by cancer metastasis from either upper or middle thoracic esophageal cancers. Therefore, three-field lymphadenectomy, bilateral cervical lymphadenectomy, mediastinal lymphadenectomy, and abdominal lymphadenectomy are recommended. In contrast, in patients with lower thoracic esophageal cancer, the appropriate extent of regional lymphadenectomy is mediastinal and abdominal lymphadenectomy.

The incidence of adenocarcinoma of the esophagus and gastroesophageal junction is rapidly increasing. Surgery is still considered the best curative treatment. However considerable debate exists as to the most appropriate surgical approach, transhiatal, or transthoracic esophagectomy. A transhiatal esophagectomy limits the extent of surgical trauma without an extended lymphadenectomy. However, a transthoracic esophagectomy with an en bloc extended lymphadenectomy in the posterior mediastinum and upper abdomen is intended to improve long-term survival. Several retrospective studies have shown little difference in the perioperative and survival outcomes between transhiatal and transthoracic esophagectomy. Rindani et al. reviewed the results from 44 series published between 1986 and 1996 [16]. The 30-day mortality rate was 6.3 % after transhiatal and 9.5 % after transthoracic esophagectomy. Overall, the 5-year survival rate was 24 % after transhiatal esophagectomy and 26 % following transthoracic resection. The Hulscher et al. analysis included data abstracted from 50 articles published between 1990 and 1999 with a total of 7,500 patients [17]. There was no statistically significant difference in the overall 3-year and 5-year survival rates between the two procedures. Recently, Boshier et al. reviewed the results from 52 studies, including 5,905 patients, which compared transthoracic with transhiatal esophagectomy (transthoracic, 3,389; transhiatal, 2,516) until 2010 [18]. The results showed that transhiatal esophagectomy was associated with significantly reduced operative time, length of hospital stay, postoperative respiratory complications, and early mortality. In comparison, transthoracic esophagectomy was associated with significantly fewer anastomotic leaks, anastomotic strictures, and vocal cord palsies. Overall, the 5-year survival rate was found to not significantly differ. These findings were comparable with the results of two previous meta-analyses. Although the survival was shown to be equivalent, transhiatal esophagectomy was chosen significantly more frequently for early-stage tumors and transthoracic esophagectomy chosen for more advanced tumors. The authors concluded that the findings of equivalent survival should be viewed with caution.