Abstract
Background: The purpose of this study was to investigate the clinicopathological characteristics and frequency of lymph node metastasis to the splenic hilus in proximal gastric cancer and the effect of splenectomy. Patients and Methods: Three hundred and forty-nine patients undergoing total gastrectomy for primary proximal gastric cancer were included. Among these patients, lymph node metastasis to the splenic hilus was histologically assessed in 201 cases. Results: The incidence of lymph node metastasis to the splenic hilus was 31 cases (15.4%). No lymph node metastasis to the splenic hilus was detected in any T1 and T2 tumors located at the lesser curvature and anterior wall. No significant difference was observed between the survival rates of patients with and without splenectomy in each stage. Conclusion: Our findings indicated that gastrectomy with spleen preservation may be recommended at least in patients with T1 or T2 tumors located at the lesser curvature and anterior wall.
Established gastric cancer treatment guidelines (1) indicate that “the standard operation for advanced gastric cancer is gastrectomy with D2 lymph node dissection” and that “advanced gastric cancer involving the upper third portion is recommended for treatment by total gastrectomy including splenectomy for dissection of lymph nodes located at the splenic hilus”. Pancreaticosplenectomy with lymph node dissection at the splenic hilus and around the splenic artery for advanced gastric cancer has often been performed. However, pancreaticosplenectomy is not recommended for gastric cancer patients without direct invasion of the pancreas because this procedure does not increase their survival rate (2, 3). Furthermore, it was reported that splenectomy may increase postoperative morbidity (3, 4). Therefore, gastrectomy with spleen preservation was indicated as a potential less-invasive operation for patients without lymph node metastasis to the splenic hilus. Currently, gastrectomy with spleen preservation for early gastric cancer is a common procedure, and a randomized prospective controlled trial for proximal gastric cancer designed by the Japan Clinical Oncology Group 0110 (JCOG 0110) (5), which will reveal the clinical significance of splenectomy in patients with advanced tumors without lymph node metastasis to the splenic hilus or along the splenic artery, is underway. Currently, however, the indications for splenectomy in proximal gastric cancer remain controversial.
Based on these background data, we performed a retrospective analysis of the frequency of lymph node metastasis to the splenic hilus and the effect of splenectomy on the survival of patients who underwent total gastrectomy for primary proximal gastric cancer.
Patients and Methods
Between January 1991 and December 2006, total gastrectomy was performed in 349 patients with primary proximal gastric cancer in our institute. The mean age of the patients (237 men and 112 women) was 62.9 years (range: 21-90 years). Among the 349 patients, lymph node metastasis to the splenic hilus was histologically assessed in 201 patients who underwent splenectomy or pancreaticosplenectomy. The great pancreatic artery was preserved in patients who underwent pancreas-preserving splenectomy. The rates of lymph node metastasis to the splenic hilus were retrospectively assessed in the patients who underwent splenectomy in accordance with the clinicopathological characteristics. The survival rates of patients with or without splenectomy were assessed. The clinical and pathological diagnoses and classifications were determined according to the Japanese classification of gastric carcinoma (6). Depth of tumor invasion was classified as follows: T1, tumor invasion of the mucosa and/or muscularis mucosa or submucosa; T2, tumor invasion of the muscularis propria or submucosa.
Statistical analysis was conducted using StatView version 5.0 (SAS Institute, Cary, NC, USA). The significance of differences was determined by the χ2 test, Student's t-test and logistic regression. The cumulative survival rate was calculated using the Kaplan-Meier method and the log-rank test. The level of significance was set at p<0.05.
Results
Clinicopathological factors. The clinicopathological findings of 148 patients without splenectomy and 201 patients with splenectomy are shown in Table I. In the splenectomy group, more advanced tumors were observed. On the other hand, early or extremely severe cases with peritoneal dissemination were recognized in the spleen-preserving group. The clinicopathological findings of the patients with splenectomy are shown in Table II. Overall, 31 patients (15.4%) had metastasis to the lymph nodes of the splenic hilus. Significant differences were found for age, tumor size, histological type and peritoneal metastasis. In multivariate analysis, age was the only predictive factor (Table III).
The frequency of lymph node metastasis to the splenic hilus was associated with the location and depth of tumor invasion (Table IV). Tumors with lymph node metastasis to the splenic hilus were more frequently located at the greater curvature (38.5%), posterior wall (27.8%), had circumferential involvement (22.8%), and had invaded more deeply than the serosa. Among all T1 and T2 tumors located at the lesser curvature and anterior wall, no lymph node metastasis to the splenic hilus was observed.
Prognosis. In the splenectomy group, the cumulative 5-year survival rate of patients with lymph node metastasis to the splenic hilus was 15%, whereas the corresponding rate in patients without such metastasis was 49% (Figure 1; p<0.001). Among 31 patients with lymph node metastasis to the splenic hilus, 23 died of gastric cancer and 1 died of another disease. Regarding recurrence or metastasis, 16 patients had associated peritoneal metastasis, 3 lymph node metastasis and 1 patient each local recurrence, liver metastasis and lung metastasis, respectively.
The survival curves of patients with and without splenectomy in each stage are shown in Figure 2. The cumulative 5-year survival rates of patients with and without splenectomy were 84% and 84% in stage I, 57% and 60% in stage II, 28% and 52% in stage III, and 18% and 7% in stage IV, respectively. No significant differences were observed between the two groups.
Morbidity and mortality. The postoperative complications following total gastrectomy with or without splenectomy during the hospitalization are shown in Table V. Postoperative complications were encountered in 32 patients (21.6%) without splenectomy and 55 patients (27.4%) with splenectomy. The major complications were anastomotic leakage, pancreatic fistula and peritoneal abscess. A significant difference in the complications was only observed for pancreatic fistula. Pancreatic fistula was observed in 4 out of 33 patients (12.1%) with pancreaticosplenectomy. Only 1 patient with splenectomy died from postoperative bleeding.
Discussion
The incidence of metastasis to the splenic hilar lymph nodes was reported to be around 10% in proximal gastric cancer (7-9). Lymphangiograms have revealed that the lymphatics from the upper left part of the stomach drain into the splenic hilar nodes and travel to the nodes around the celiac trunk through the splenic artery (10). This drainage route passes along not only the left gastroepiploic artery and short gastric artery, but also the posterior gastric artery, and this anatomical background agrees with the finding that splenic hilar node metastases were more frequent in tumors located at the greater curvature, posterior wall and circumferential involvement.
Although extended lymph node dissection has become a standard surgical procedure for gastric cancer in Japan, it has not shown a clear survival benefit in randomised clinical trials (11, 12) and a meta-analysis (13) in Western countries. The main reason for leaving the spleen is to resect lymph nodes in station 10, because it is impossible to resect them at the splenic hilus without splenectomy. The problems associated with simultaneous splenectomy during surgery for gastric cancer are its contribution to prognosis and the frequency of postoperative complications. A prospective randomised study comparing patients with and without splenectomy in Western countries revealed that splenectomy did not influence the survival in localized stages of gastric cancer, that is stages IA, IB, II, IIIA (14). A Japanese prospective randomised trial is in progress (5). Although the previous studies did not improve patient survival (3, 15-21), some conferred a survival benefit (10, 22-25). Regarding the survival benefit of splenectomy, previous studies have reported that the spleen in advanced gastric cancer patients produced suppressor T-cells, which might improve the survival via tumor-induced immunosuppression (25-28). In the present study, there was no significant difference in survival for each stage between patients with and without splenectomy. Furthermore, among postoperative complications in patients with splenectomy, only pancreatic fistula was more frequent and associated with surgical technique, but did not increase mortality.
In conclusion, from the standpoint of survival benefit, splenectomy did not exhibit any significance for primary proximal gastric cancer. Furthermore, for all T1 and T2 tumors located at the lesser curvature and anterior wall, gastrectomy with spleen preservation may be recommended as a standard operation.
- Received March 16, 2009.
- Revision received May 22, 2009.
- Accepted June 3, 2009.
- Copyright© 2009 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved