[Multivariate prognostic analysis for patients with curative resection of gastric cardia cancer].

Objective: To explore the factors affecting the long-term survival of patients with curative resection of gastric cardia cancer.

Methods: The data of 108 patients who underwent radical resection of gastric cardia cancer from Jul. 1994 to Dec. 2003 in our hospital were investigated retrospectively. The Kaplan-Meier method and long-rank test were used for bivariate comparisons of survival. Multivariate analysis was done by the Cox regression model (Backward Wald).

Results: Survival status of the 108 patients was ascertained in Dec. 2004. Among them, 68 were Siewert type II and 40 were Siewert type III. Seventy-four patients had lymph node metastases (68.5%). The mean follow-up time was 37 months (95% CI: 29.3-44.7 months) and the middle follow-up time was 26.6 months (95% CI: 25.8-34.2 months). The 1-,3- and 5-year cumulative survival rates were 77.2%, 33.6% and 21.8%, respectively. According to the Kaplan-Meier and log-rank methods, splenectomy, lesion size, depth of tumor invasion and regional lymph node status were prognostic factors. Multivariate regression analysis indicated that only depth of tumor invasion (P=0.009) and lymph node metastases (P=0.001) were independent prognostic factors.

Conclusion: Depth of tumor invasion and lymph node metastases have negative effects on the survival of patients with gastric cardia cancer undergoing curative resection. Splenectomy may only be appropriate for patients with direct tumor invasion to the spleen and the extent of gastric resection does not influence survival in patients with curative gastric cardia cancer.