Is there any anatomic justification for partial esophagectomy in cancer of the cardia?
R. Sarrazin (Grenoble)
When confronted with cancer of the cardia, some tend towards total esophagectomy. For others, the gastric resection overlaps more or less largely on the lower esophagus.
Do the anatomy of the cardia and the pattern of extension of the cancer justify one or the other of these attitudes?
The cardia is a morphologically badly defined region: at the junction of the esophagus and the stomach, the cardia is not a borderline: its exterior aspect spotted through the angle of His and the endoluminal level are not corresponding.
On the morphological level [1-4], the only important difference is the structure of the mucosa, white pink on the esophageal side, salmon coloured on the gastric side, with a sharp limit the Z line. But there is no anatomical barrier which could condition a different pathological behavior on the upper level or on the lower level (Figure 1).
The mucosa, with overlying squamous mucosa little or not secreting at its upper part, is secreting on the lower level because of the abundance of cardiac type glands diffusely scattered in the lamina propria. However, nothing hinders diffusion of the carcinomatous process upward or downward (Figure 2).
The muscularis mucosae gets thicker toward its end, and on the gastric level.
The submucosa, loose connective tissue containing vessels, nerves, fibers, lymphatics, is as rich in both organ, and is perfectly uninterrupted.
The two-layers muscular coat at the esophageal level is more complex on the gastric side, but does not individualize the sphincterian process which is manometrically defined. The longitudinal layer is in continuity with the outer longitudinal layer of the stomach, while the circular layer continues over the stomach, dividing to form the middle circular and inner oblique muscle layers.
The serosa has no specific characteristic at any level.
One thus has to study the role played by the vascular and lymphatic system in the upward extension of a malignant process, and to imagine an hypothetical limit.
Figure 1. Axial section of the gastroesophageal junction from Petersen (in Policard). 1. Esophageal epithelium.
2. Cardial type glands. 3. Lymphoid islet. 4. Gastric glands.
Figure 2. Schematic histological section of gastroesophageal junction.
Apparently, the lower esophagus benefits from the same portal drainage as the stomach, and though the upper esophagus is drained by the systemic system, there is no separating line between the flows of the two systems. The caval and portal systems communicate through the veins within the submucosa. The extension of the tumoral process by venous paths cannot be controlled surgically.
The lymphatics are originating from mucosal, submucosal, muscular and subserous networks, which are all anastomosed together in the thickness [5, 6]. These networks drain in epivisceral gastric and esophageal lymph nodes, and are connected with the regional upper or under diaphragmatic nodes.
The risk arises when carcinoma reaches those networks: because of their anastomosis and drainage in the collectors, they allow extension only from one to the next, upward as well as downward, without any morphological abnormality existing on the surface.
Thus, where must one cut the esophagus? Of course, above the superior limit of the palpable and visible tumor, invading a few millimeters, even a few centimeters of the lower esophagus. Each one evaluates the safety limit according to the circumstances (6, 8, 10 cm) and cuts above... [7-11] or inside (Table I), or under the last higher located lesion , whether it be the suffused colonization of the submucosa or of all the esophageal layers, or submucosal metastases or resurgences [12, 13]. The recurrence rate after partial esophagectomy is 25% in cases of resection made 2 cm above the lesion, 15% when it bypasses it by 5 cm, 7% when it bypasses it by 10 cm . The study of the esophageal section cut shows that it is invaded in 4.5 to 36% of the cases (Table II).
When a "total" esophagectomy is performed, the anatomic study of the operative specimens shows 7% (H.U. Sons), 12% (Neuberger)  and, personally, 13.9% of submucosal localizations from 2 to 12 cm above. These recurrences depend of course on the lymph node invasion .
The histological lymphatic invasion can, by no means, be foreseen from morphological findings. Moreover, the existence of long collectors distort the gradual propagation, some of them going straight from the cardia to the tracheal bifurcation, and even further.
Thence, it is difficult to justify a partial resection of the esophagus, even if the results are not worse than those of total esophagectomy.
5. Sarrazin R, Voilin C, Bacle B, Dyon JF. L'Anatomie du drainage lymphatique de l'sophage et sa lymphologie sont encore mal connues. In : Giuli R, ed. Les cancers de l'sophage en 1984. 135 questions. Paris: Maloine, 1984.
9. Ribet M, Quandalle P, Wurtz A, Pavy G, Giard-Lefevre S, Desbrueres B. Résultats du traitement chirurgical de l'adénocarcinome du cardia en fonction de l'étendue de l'exérèse. Ann Chir 1987;41,1:5-10.
10. Sauvanet A, Berthoux L, Gayet B, Fléjou JF, Belghiti J. Adénocarcinome du cardia: l'étendue de l'exérèse gastrique et du curage ganglionnaire influence-t-elle la survie ? Gastroenterol Clin Biol 1995;19,3 :244-251.
12. Clark GWB, Peters JH, Ireland AP, Ehsan A, Hagen JA, Kiyabu MT, Bremmer CG, De Meester TR. Nodal metastasis and sites of recurrences after en bloc esophagectomy for adenocarcinoma. Am Thorac Surg 1994;58:646-654.