[Surgery of early gastric cancer. Twenty-five year experience]

To study the influence of the depth of parietal invasion (mucosal-submucosal), the presence or absence of ganglionic invasion and type of gastrectomy performed (subtotal or total) on survival in patients with early gastric cancer.

Longitudinal study.

A clinical-pathologic study of 101 patients who underwent surgery for early gastric cancer was performed. Probability of survival was estimated using the Kaplan-Meier and logrank tests and multivariate analysis was performed using the Cox test.

Mucosal involvement was found in 46 patients (45.5%) and submucosal involvement in 55 patients (54.5%). The presence of ganglionic metastases was greater in tumors reaching the submucosa (14 [25.5%]) than in those limited to the mucosa (4 [8.7%]). Partial gastrectomy was performed according to tumor location in 84 patients (83.2%), total gastrectomy was performed in 16 patients (15.8%) and 1 wedge resection was performed. The mean postoperative follow-up was 84.04 55.89 months (range: 2-264). Comparison of survival in patients with tumors limited to the mucosal or submucosal layers revealed a p-value of 0.06 (NS). Comparison of survival in patients with metastases and in those without metastases revealed a p-value of < 0.0001. Comparison of survival between patients who underwent total gastrectomy and those who underwent partial gastrectomy showed a p-value of 0.38 (NS). Postoperative mortality was nil. Overall survival at 5 years was 79.24% and at 10 years was 68.14%. Multivariate analysis revealed that ganglionic involvement and depth of parietal invasion influenced survival.

Survival is influenced by ganglionic involvement but not by submucosal invasion. Partial gastrectomy may be an appropriate procedure since survival is similar to that associated with total gastrectomy.