Is mediastinal lymph node involvement to be considered as a stage IV disease?
T. Lerut (Leuven)
In sharp contrast to the decreasing occurrence of gastric carcinoma a dramatic increase of adenocarcinoma of the esophagus and cardia has been reported by several authors in recent years .
The prognosis for these tumors even after resection is usually judged to be poor. This poor prognosis seems to be related to the particular location of the tumor, positioned on both the stomach and the esophagus and its tendency for lymphatic spread into the lymph nodes of the superior abdominal compartment as well as into the posterior mediastinum. As a result, adenocarcinoma of the cardia is considered to be an entity to be distinguished from gastric carcinoma and esophageal carcinoma. Great controversy persists as to the definition, classification, staging and surgical strategies.
In the TNM system tumors of the cardia are to be classified as gastric carcinoma. However, there is no strict criterion by which to define carcinoma of the esophagogastric junction (EGJ) .
The International Union against Cancer (UICC) TNM supplement 1993  suggests classifying adenocarcinomas between those involving more than 50% of the esophagus, and those involving more than 50% of the stomach, including those equally distributed on esophagus and stomach as gastric carcinomas.
In a review of our own material, we included in the definition of adenocarcinoma of the EGJ all tumors with the core of the tumor located at the Z-line or within an area of 5 cm orally and 5 cm aborally of the anatomic junction between the esophageal and gastric walls.
From 1983 until 1993, 259 patients with carcinoma of the esophagogastric junction as defined here, were operated in our Institution .
A cohort of 95 patients treated from 1983 to 1989 was studied in a retrospective way so that a minimum of follow-up of 5 years could be obtained.
The hospital mortality was 6.2% (6/95). The overall survival at 1 year was 54%, at 2 years: 39%, at 5 years: 33%, at 10 years: 31%.
Figure 1 shows the 5 and 10 year survival stage by stage. In patients having no lymph node involvement (n = 26), 5- and 10-year survival were both 72%, compared with 18% and 16% respectively for patients having positive lymph nodes (n = 68) (p < 0.005).
Twenty-eight patients had diseased lymph nodes in both the abdomen and the mediastinum, whereas 37 patients had lymph node invasion confined only to the abdomen. Five and 10-year survival was clearly lower in the first subgroup (13%) than in the second subgroup (26%) (Figure 2).
This difference however did not reach statistical difference (p > 0.05). These figures clearly show that a number of patients who were classified as TNM because of mediastinal stage IV (i.e. strictly palliative stage) can be cured from their disease. The figure of 13%
10-year survival is matching very well with 17% survival of the group of patients classified as having stage IIIb disease because of abdominal N2 lymph node involvement (i.e. involvement of lymph nodes beyond coeliac axis). Adding the patients with positive intrathoracic lymph nodes to the original group of patients staged as IIIb results in a 5 and 10 year survival of 14% (Figure 3). Moreover, when eleminating the patients with diseased intrathoracic lymph nodes from the stage IV group, none of the patients remaining in
stage IV now survived more than 5 years.
Figure 1. Tumors of the esophagogastric junction.
Figure 2. Tumors of the esophagogastric junction.
The consequences of staging EG tumors as stated in the TNM classification are clear. Patients classified in stage IV are to be considered as having uncurable disease, for which any treatment will be strictly palliative.
Figure 3. Tumors at the esophagogastric junction; proposal for new IIIb.
Our data clearly suggest that positive mediastinal lymph nodes in gastroesophageal junction (GEJ) adenocarcinomas do not necessarily indicate stage IV.
As the 5- and 10-year survival figures for patients with positive mediastinal lymph nodes is nearly the same as the 5- and 10 year survival curves for patients with N2 disease, we suggest to label positive mediastinal lymph nodes as N2, and therefore to classify these patients as stage IIIb, reflecting better the potential for curative treatment in this subgroup. The rate of mediastinal lymph nodes involvement up to 38% in our series of adenocarcinomas of the GEJ has obvious consequences. Careful lymph node dissection of the posterior mediastinum up to and including the subcarinal lymph nodes is necessary to obtain correct pathologic staging. This of course favors the transthoracic approach, allowing adequate esophagogastric dissection and, more important, better and more complete removal of mediastinal lymph nodes than any other approach, although also an extended transhiatal approach by incising the diaphragm may allow correct lymph node dissection.
Detection of diseased intrathoracic lymph nodes by different staging methods, either non invasive (endoscopic ultrasonography) or invasive (thoracoscopy), should not exclude such patients from treatment modalities aiming at cure, in particular surgery.
4. Steup WH, De Leyn P, Deneffe G, Van Raemdonck D, Coosemans W, Lerut T. Tumours of the esophagogastric junction. Long-term survival in relation to the pattern of lymph node metastasis and a critical analysis of the accuracy or inaccuracy of TNM classification. J Thorac Cardiovasc Surg 1996;111:85-95.