Esophageal cancer is ranked among the top 10 most common cancers in the world, with more than 480,000 new cases diagnosed annually.¹ The proportion of cases involving the two most common histopathological entities, adenocarcinoma and squamous cell carcinoma, is changing. While adenocarcinoma of the esophagus is rising rapidly in Western countries; squamous cell carcinoma remains unchanged.² The true incidence of adenocarcinoma is difficult to determine because cancer of the esophagogastric junction (EGJ) is classified by some as a gastric cancer while by others as an esophageal cancer (see Chapter 12). This explains, in part, the ongoing controversy over which strategy to follow when it comes to surgical approach and technique.

In an effort to rationalize what had been a rather indiscriminate approach to EGJ tumors and to provide clearer guidelines, in 1998, Siewert and Stein³ published a classification for adenocarcinoma of the esophagogastric junction (AEG). Tumors of the AEG were defined as tumors with an epicenter equal to or less than 5 cm proximal and distal to the anatomic EGJ (anatomic cardia).

Three entities were distinguished:

1. 
   

   Adenocarcinoma of the distal esophagus, usually arising in Barrett intestinal metaplasia,

   
   
2. 
   

   True carcinoma of the EGJ, which may arise from the cardiac epithelium or from Barrett intestinal metaplasia, and

   
   
3. 
   

   Subcardial adenocarcinoma arising from the gastric fundus.

Using this classification and drawing on their own data, Siewert and Stein proposed guidelines on which surgical strategy to use according to each of the above-described subtypes. This classification is based entirely on identifying the “anatomical” cardia and measuring the center of the tumor in relation to this anatomical cardia on the resected specimen (i.e., pathological staging). However, measuring the center of the tumor turns out to be impractical if not impossible for the purposes of clinical staging. It is important to have accurate clinical staging because it ensures the appropriateness of the therapeutic decision (e.g., in the presence of a hiatal hernia). Not surprisingly, Omloo et al.⁴ recently reported a substantial discrepancy between the clinical and pathological staging when using the Siewert classification.

The 7th edition of the TNM classification system,⁵ which is based on evidence derived from a large international multinational database, also addressed this issue. It classifies and stages tumors that have their epicenter in the EGJ, or within the proximal 5 cm of the stomach and extending into the EGJ or esophagus, similarly to an adenocarcinoma of the esophagus. Tumors centered in the stomach that are located more than 5 cm from the EGJ, or tumors within 5 cm of the EGJ but without extension into the esophagus, are classified and staged as gastric tumors.⁶

As a result, the strategies related to surgical approaches and techniques are, in general, similar to those intended for cancers of the esophagus. These strategies in particular are related to the patterns of lymphatic spread observed with esophageal cancer.

Lymphatic dissemination is an early event and has a negative influence on survival. Lymph node metastases are found in less than 5% of intramucosal tumors, but in as many as 30% to 40% of submucosal tumors, that is, when the tumor spreads beyond the mucosa. The esophageal wall is characterized by an extensive submucosal lymphatic plexus, which not only supplies a drainage route for early dissemination, but also gives rise to jump metastases (i.e., lymph nodes adjacent to the primary tumor are not affected, but more distantly located lymph nodes contain metastases).⁷

The pattern of lymphatic dissemination therefore is difficult to predict. Nevertheless, carcinomas of the proximal and middle thirds of the esophagus tend to metastasize to the cervical region, whereas more distal-lying tumors and tumors of the EGJ more commonly metastasize to the lymph nodes around the celiac trunk. This is also the case for adenocarcinoma of the EGJ. These tumors typically drain by an extraperitoneal route into the retropancreatic area and into the left renal hilum.⁸ Adenocarcinoma of the EGJ also has a tendency to spread to the thoracic lymph nodes including subcarinal, paratracheal, and aortopulmonary window nodes. Lagarde et al.⁹ evaluated a group of 50 patients with a EGJ tumor that underwent extensive lymphadenectomy through a transthoracic approach and found that 22% of them had lymph node involvement of the more proximal nodes. These findings need to be taken into consideration when determining surgical strategies.

For decades surgeons have been combining transthoracic and abdominal esophagectomy to remove the esophagus and cardia, together with the primary tumor and adjacent lymph nodes, under direct vision.¹⁰ It has been suggested that long-term survival might benefit from a more radical resection and a more extensive lymphadenectomy in the thorax and abdomen.¹¹

The Radical En Bloc Resection

The radical en bloc resection extends the standard treatment to encompass a wide local resection of the primary tumor with a radical lymph node dissection of the middle and distal thirds of the posterior mediastinum.¹² The concept of extensive en bloc resection was first reported in 1963, but the associated mortality of more than 20% discouraged its general acceptance.

Skinner and Akiyama reintroduced this method in the 1980s. Ultimately, they were able to reduce operative mortality to 5%, with 5-year survival rates of 18% and 42% respectively.^13,14 Over time the results steadily improved worldwide. Today, mortality rates are well below 5%, and 5-year survival figures routinely reach 40% at experienced centers.

The Two-Field Lymph Node Dissection

Early lymphatic dissemination, characterized by longitudinal spread of tumor along the esophagus via the submucosal plexus to the upper mediastinum and abdomen, was the rationale for advancing to a two-field lymphadenectomy.⁷ This approach, which originated in Japan, adds the following elements to a wide local excision of the primary tumor: lymphadenectomy of the entire posterior mediastinum, the lymph nodes along the celiac trunk, common hepatic and splenic arteries, as well as the lymph nodes along the lesser gastric curvature and in the lesser omentum.

Transhiatal Resection

Despite some indication that surgical techniques with extensive lymph node dissections tend to improve long-term survival, a less radical transhiatal approach was developed to decrease early postoperative risk by eliminating the need for thoracotomy. This approach has been popularized in the Western world by Orringer.¹⁵

Attempting to close the ongoing debate between advocates of radical esophagectomy via the transthoracic approach versus advocates of the less radical esophagectomy via the transhiatal approach, Hulscher et al.¹⁶ of The Netherlands instituted a multi-institutional randomized clinical trial. This trial compared limited transhiatal resection to transthoracic resection with extended en bloc lymph node dissection for adenocarcinoma of the esophagus and EGJ. The results revealed no statistically significant overall difference between the two surgeries, but there was a clear long-term trend in favor of the more extensive approach which yielded a 39% 5-year survival compared with a 29% 5-year survival for the more limited resection. Particularly for adenocarcinoma of the distal esophagus, a subsequent analysis revealed a 17% survival benefit in favor of the more extensive transthoracic approach.¹⁷ This trial remains the only randomized study to compare these two approaches, but several other studies have supported the overall findings that long-term survival may benefit from a more radical esophagectomy combined with extensive two-field lymphadenectomy. Indeed, data from many centers seem to endorse a better overall 5-year outcome, often exceeding 40% after two-field lymphadenectomy, as compared with the less radical lymphadenectomy with 5-year survival figures in the range of 20% to 25%.¹⁸

Concerning the recommended number of lymph nodes to be dissected, there is no general agreement. Historically, 15 nodes have been considered the minimum. However, a recent multivariate analysis by Peyre et al.¹⁹ in a large patient population found that the absolute number of lymph nodes removed during esophagectomy was a strong independent predictor of survival. They reported an optimal survival benefit required resection of at least 23 lymph nodes, and this finding was not the result of stage migration. Within every tumor stage (I–III), patients with more than 23 resected lymph nodes had better survival than patients with less than 23 resected nodes, thus strongly emphasizing the importance of performing an adequate lymph node dissection.

Within the compartments of a two-field dissection (thoracic and superior abdominal compartment), in addition to the nodes along the lesser curvature and left gastric artery, the celiac, hepatic, and splenic artery nodes also should be routinely removed during the abdominal stage of any en bloc resection. Since 1994 lymph node dissection in the chest has been defined as standard (lower periesophageal and subcarinal nodes), extended (including some upper mediastinal nodes, i.e., right paratracheal nodes), and total thoracic lymph node dissection (including the uppermost mediastinal nodes, i.e., left and right paratracheal and aortopulmonary window nodes).²⁰ In a classic two-field lymphadenectomy at least the periesophageal and subcarinal nodes should be removed and preferably also the node at the aortopulmonary window and the right paratracheal nodes.

As to safe margins, a wide peritumoral or en bloc resection should be attempted whenever possible to obtain an R0 status (clear circumferential, proximal, and distal margins on pathological examination). At the side of the esophagus a proximal gross length of at least 5 cm or even more to achieve a microscopically negative proximal margin is advocated. This is particularly true for the more advanced T3 and T4 cancers as shown recently by Ito et al.²¹

The extent of stomach resection will depend on the location and the length of the tumor. For adenocarcinoma of the distal esophagus and EGJ, a resection margin of 5 cm below the distal pole of the tumor will suffice. This permits use of the stomach for reconstruction after resection of the lesser curvature and its lymph nodes to create a gastric tube with an anastomosis high in the chest or the neck. If the tumor extends more than 5 cm over the fundus or lesser curve, a total gastrectomy with Roux-en-Y jejunal reconstruction is mandatory.

Currently, there are several approaches to the surgical treatment of adenocarcinoma of the distal esophagus and EGJ. These include the right-sided abdominotransthoracic approach with anastomosis high in the chest (Ivor Lewis) or in the neck (McKeown, also called 3-hole resection), a left-sided abdominotransthoracic approach with anastomosis in the chest (Sweet) or in the neck (Belsey), a transhiatal resection with anastomosis in the neck (Orringer), and the minimally invasive esophagectomy (MIE).