| Resumo: | Background: It is well established that the presence of lymph node (LN) metastasis is the most important prognostic factor in advanced gastric cancer after curative gastrectomy. However, some patients have node-negative advanced gastric cancer. The identification of others useful prognostic factors may be important for the selection of patients who may benefit from more aggressive postoperative treatments. So, our purpose is to identify the clinicopathological factors that influence the prognosis in node-negative advanced gastric cancer. Methods: Retrospective analysis of a prospective database (n=637) with gastric cancer cases submitted to intent curative surgery between January 2010 and December 2017, in an Upper GI Surgery Unit. In this study, were included 81 patients with node-negative stage T2-4 gastric cancer that met the inclusion criteria. Cox regression was used to evaluate the effect of clinicopathological factors in overall survival (OS) and disease-free survival (DFS). Kaplan-Meier curves were calculated according to different clinicopathological factors and differences between groups were assessed by Log Rank test. Cox regression (forward stepwise conditional) was used for the identification of independent prognosis factors. Results: Of the 81 patients, 33 (40,3%), 31 (38,3%) and 17 (20,9%) had T2, T3 and T4 tumors, respectively. The overall recurrence rate was 8,6% (n=7). The recurrence rate was 0%, 9,7% (all distant metastasis) and 23,5 % (50% loco-regional and 50% distant metastasis) in T2, T3 and T4, respectively. In univariate analysis, macroscopic type (p=0,007), pT (p=0,001), peri-operative blood transfusion (p<0,001) and lymphadenectomy type (p=0,036) were significantly correlated with tumor recurrence. While tumor location (p<0,001), pT (p=0,028), peri-operative blood transfusion (p=0,014) and age (p=0,044) were significantly correlated with overall survival. In multivariate logistic regression analysis, macroscopic type [HR 3,25; CI 95% (1,227 - 8,606), p=0,018] and peri-operative blood transfusions [HR 21,775; CI 95% (3,870 - 122,538), p<0,001] were significantly and independently correlated with recurrence. Whereas peri-operative blood transfusion [HR 2,749; CI 95% (1,174 - 6,440), p= 0,02] was significantly and independently correlated with overall survival. Conclusion: In this series of node-negative advanced gastric cancer, macroscopic type and peri-operative blood transfusion reliably predict recurrence, whilst peri-operative blood transfusion reliably predicts overall survival.
|
|---|