Histologically positive esophageal margin in the surgical treatment of gastric cancer

Histologically positive esophageal margin in the surgical treatment of gastric cancer

Histologically Positive Esophageal Margin in the Surgical Treatment of Gastric Cancer Dimitri N. Papachristou, MD, New York, New York Niki Agnanti,...

374KB Sizes 3 Downloads 15 Views

Histologically Positive Esophageal Margin in the Surgical Treatment of Gastric Cancer

Dimitri N. Papachristou,

MD, New York, New York

Niki Agnanti, MD, New York, New York Horatio D’Agostino,

MD, New York, New York

Joseph G. Fortner, MD, FACS

Adenocarcinomas of the stomach may invade the distal esophagus [l-7], the microscopic boundaries of the invading process usually extending well beyond the palpable edge of the main lesion [1,2] (Figure 1). Failure to appreciate this fact during total or proximal subtotal gastrectomy leads to transection of the esophagus in an area invaded by disease [l-7], which predisposes to esophageal anastomotic recurrence [4-6] and leakage [7]. This can be avoided by performing frozen-section examination of the margin or by resecting long esophageal segments in all patients with high gastric neoplasms, as is done in patients with primary carcinoma of the esophagus [6-101. The problem with this policy is that frozen-section analysis, for various reasons, can mislead the surgeon [3], while gastrectomy combined with subtotal esophagectomy is a very extensive procedure. Furthermore, the indications and the extent of a wide esophageal resection have not been clarified in the literature [r-7,10,11]. The present study examines the incidence, predisposing factors and consequences of positive esophageal margins in 350 patients with gastric cancer. Material

and Methods

The study concerns 350 patients undergoing total (259) or proximal subtotal gastrectomy (91) for gastric adenocarcinomas arising in the cardia (100) or the corpus (250). The patients were evaluated with esophagogastroscopy and upper gastrointestinal barium examination preoperatively. The level of esophageal transection was determined at the time of exploration by palpation of the esophagus. In 130 patients (37 percent) the adequacy of esophageal resection was also checked with frozen-section examination of the upper margin. Resected specimens were examined and measured while still fresh. An esophageal margin was From the Gastric and Mixed Tumor Service, Department of Surgery. Memorial Sloan-Kettering Cancer Center. New York, New York. Requests for reprints should be addressed to Dimitri N. Papachristou. MD, Achilleos 6. Agia Paraskevi, Athens. Greece.

Volume 139, May 1980

considered positive if permanent section examination revealed tumor at the line of esophageal transection. Results

The overall incidence of positive esophageal margins in the present series was 20 percent (73 of 350). The incidence was influenced by the location of the primary lesion, being 35 percent (35 of 100) in patients with neoplasms of the cardia and only 15 percent (38 of 250) in patients with corpus lesions (p
711

Papachristou

et al

TABLE I

Length of Macroscopically Tumor-Free Esophageal Margins and Incidence of Histologically Positive Margins in 350 Patients

-

Figure 7. Intramural invasion of the esophagus by high gastric neoplasms. Transection at the depicted levels results in a histologically positive esophageal margin.

lead to anastomotic recurrence. In fact, the incidence of anastomotic leakage in those patients was only 4 percent (3 of 73) compared with 11 percent in the rest of the patients. Also, of 51 patients with positive margins who survived operation, only 12 (23 percent) had anastomotic recurrence; the other 39 died from generalized dissemination of the neoplasm. Recurrences affected mainly patients with early rather than advanced stage disease, as TNM classification revealed, and patients with neoplasms of the cardia. Thus the incidence of anastomotic recurrence was 38 percent (5 of 13) in patients with stage I and II tumors, compared with 18 percent (7 of 38) in those with stage III and IV lesions. The incidence of anastomotic recurrence was 25 percent (9 of 35) in patients with neoplasms of the cardia and only 8 percent (3 of 38) in those with corpus lesions (p CO.05, chi-square test). Positive margins were associated with a poor prognosis. The median survival of patients with positive margins who survived operation was only 6.5 months. Those who died from metastasis had a median survival of 4.5 months and those who had anastomotic recurrence, 10 months. Prognosis was not affected by additional treatment delivered immediately after the diagnosis of a positive margin. Thus of three patients who received radiation therapy in an attempt to prevent recurrence, two died from metastasis before the completion of treatment and the third died from anastomotic recurrence. Two other patients who underwent reexploration had their remaining esophagus removed but died in the

712

Length of Margins (cm)

No.

Positive Margins No. Percent

0 1 to 2 2.1 to 4 4.1 to 6 Greater than 6

19 72 174 64 21

19 22 26 6 0

100 30 15 9

Total

350

73

21

..

postoperative period from complications. Of 12 patients who had anastomotic recurrence, only 1 was saved and is still alive 4 years later. This patient had recurrent disease confined to the esophagus and was managed with resection. Surgical treatment failed in three other patients with advanced intrathoracic disease. Three patients who were managed with radiation therapy died shortly after the completion of their treatment. Finally, five patients with recurrence died from malnutrition and pneumonia before therapeutic intervention. Comments

Transection of the esophagus in an area invaded by disease is reported to be common in patients undergoing total or proximal subtotal gastrectomy for gastric cancer, with a reported incidence of 26 to 63 percent [2-61. The present study demonstrates that the incidence is increased when the primary neoplasm is located high in the stomach, when the surgeon fails to resect enough esophagus and when frozen-section examination of the esophageal margin is omitted. Thus, neoplasms arising in the gastric cardia were associated both with a high incidence of positive esophageal margins and high incidence of anastomotic recurrence. ,This finding is in agreement with pathoanatomic data indicating that 90 to 95 percent of all resectable adenocarcinomas of the cardia cause intramural invasion of the esophagus [1,4]. Pack [9], being aware of these facts, was among the first to advocate a combined thoracoabdominal approach for the management of these tumors in order to obtain an adequate esophageal margin. Limited esophageal resections lead to a high incidence of positive margins [2,4,6,11]; however, the question of what constitutes an adequate esophagectomy in the treatment of high gastric neoplasms has not yet been answered. Transection of the esophagus 4 cm above the main lesion resulted in a 70 percent incidence of positive margins in one series [2]. In general, a margin of 4 to 5 cm is considered

The American Journal of Surgery

Histologically

adequate by most authors [5,6,1 I]; however, even w:th a margin of 6 cm one may expect a 56 percent incidence of positive margins [2]. Margins greater than 6 cm were adequate in this series; however, this figure was obtained by measuring fresh esophageal specimens which, like rectal specimens, shrink to half of their actual size after resection [6,12]. Thus the alroidance of positive esophageal margins will require resection of 12 cm of macroscopically tumor-free esophagus above the palpable edge of the primary lesion in those patients. The value of frozen-section examination in the management of high gastric neoplasms is well established [4,9,10]. Two facts, however, suggest that it may have been overemphasized. First, frozensect,ion examination can give a false-negative result. The incidence of such false-negative results was 9 percent in this series, although rates as high as 21 percent have been reported [3]. Second, frozen-section examination of a nonrepresentative specimen obtained between the primary and a skip submucosal lesion (Figure 1) can mislead the surgeon. For these reasons, frozen-section analysis cannot determine always the adequacy of esophageal resection; it is a laboratory test with limitations and should be regarded as such. Do all patients with high gastric neoplasms need to be treated with a 12 cm margin? Anastomotic recurrences appeared in only 23 percent of the patients at risk in t,his series, and positive margins did not predispose to anastomotic leakage. Recurrences affected mainly those who had early-stage lesions, while patients with advanced neoplasms usually died fr,:)m distant metastasis a few months after gastrectomy. Another finding was that recurrence was particularly common in patients with tumors of the cardia. These facts suggest that wide esophageal margins will be required for patients with early-stage neoplasms arising in the cardia. Patients with advanced disease and those with neoplasms of the corpus can be managed with less extensive esophageal resections. Do patients with positive margins need additional treatment? The answer should be negative for two reasons: first, because only a few of these patients will develop local recurrence at the site of the anastomosis and second, because the results of additional treatment are not satisfactory. Rather than proceeding with further treatment, one could choose to follow up those patients closely with esophagoscopy in anticipation of anastomotic recurrence. This policy prevents the unnecessary treatment of patients who will never develop local recurrence and facilitates an early diagnosis of recurrence in the others.

Volume 139, May 1990

Positive Esophageal Margin

Summary

Histologically positive esophageal margins caused by transection of the esophagus in an area involved by disease were encountered in 20 perce’nt of 350 patients undergoing total or proximal subtotal gastrectomy for gastric cancer. There was a significant increase in the incidence of positive margins in patients with neoplasms of the cardia and in those who were managed without frozen-section examination of the resected esophagus. Positive margins were avoided only with resection of 12 cm or more of macroscopically tumor-free esophagus above the primary. Despite the presence of tumor at the margin, anastomotic recurrences appeared in only 23 percent of the patients at risk. Recurrences affected mainly those who had TNM stage I and II lesions. Patients with more advanced disease usually died from metastasis without developing anastomotic recurrence. Positive margins were associated with a poor prognosis that was not influenced by adjuvant postoperative treatment. The avoidance of positive margins by wide esophagectomy is important in patients with TNM stage I and II disease, particularly if the primary lesion is located in the cardia. Patients with positive margins should be watched closely rather than subjected to further treatment.

References 1. Morson BB. The spread of carcinoma of the esophagus. In: Tanner NC, Smithers DW. Tumors of the esophagus. Edinburgh and London: ES Livingstone, 1961:136-46. 2. Sefton GK. Coooer DJ. Grech P. Giddinos AEB. /Assessment and resection of carcjnoma at the gastroesophageal junction. Surg Gynecol Obstet 1977; 144:563-6. 3. Schrock TR, Way LW. Total gastrectomy. Am J Surg 1978; 135:348-55. 4. Paulino F, Roselli A. Carcinoma of the stomach with special reference to total gastrectomy. Curr Probl Surg 1973. 5. Humphrey EW, Kersten TE. Adenocarcinoma at the esophagogastric junction. In: Varco RL, Delaney JP. Controversy in Surgery. Philadelphia and London: WB Saunders, 1976: 603-8. 6. Miller C. Carcinoma of the thoracic esophagus and cardia. A review of 405 cases. Br J Surg 1962;49:507-18. 7. Le Rou BT. An analysis of 700 cases of carcinoma of the hypopharynx, esophagus and proximal stomach. Thorax 1961;16:226-232. 8. Scanlon EF, Morton DR, Walker JM, Watson WL. The case against segmental resection of the esophagus for carcinoma. Surg Gynecol Obstet 1955;101:290-5. 9. Pack G. Cancer of the esophagus and gastric cardia. St. Louis: CV Mosby, 1949:145-88. 10. Tanner NC, Smithers DW. Tumors of the esophagus. Edinburgh and London: ES Livingstone. 1961:226. 11. Payne WS, Bernatz PE. One stage resection and reconstruction for carcinoma of the esophagogastric junction. In: Reference 5593-602. 12. Rosi P. Selection of operations for carcinoma of the colon. In: Turell R. Diseases of the colon and rectum. Philadelphia and London: WB Saunders, 1969:478-501.

713