Clinicopathologic study of gastric cancer based on Dukes' classification.

Abstract

Dukes' classification is a useful staging system in patients with colorectal cancer. The aim of this study was to present clinicopathologic characteristics and survival of patients with gastric cancer based on Dukes' classification. A total of 273 patients with gastric cancer curatively treated by radical gastrectomy and lymph node dissection (D2, D3) were studied. With the modified Dukes' classification, A includes tumors limited to the mucosa, submucosa, or muscularis propria; B includes tumors extending into the subserosa or serosa; Ca includes tumors with one to six positive lymph nodes; and Cb includes tumors with seven or more positive lymph nodes. Dukes' classification modified by the number of positive lymph nodes well correlated with the tumor size (p < 0.01), depth of wall invasion (p < 0.01), level of lymph node metastasis (p < 0.01), and degree of lymphatic permeation (p < 0.01) and venous permeation (p< 0.01). The 5-year survival rate was significantly different among Dukes' A (98%), Dukes' B (90%), Dukes' Ca (75%), and Dukes' Cb (44%) cases. The results indicate that Dukes' classification modified by the number of positive lymph nodes (Dukes' A, B, Ca, an Cb) significantly correlates with tumor progression and patient survival; and it may be a simple and useful staging system for gastric cancer.