Blood transfusions and results after curative resection for gastric cancer

On the basis of an analysis of our experience and a review of the literature, this report discusses the effects of perioperative blood transfusions on postoperative morbidity, mortality and 5-year survival in a series of patients who underwent curative surgical treatment of gastric cancer. The authors analyze a consecutive series of 137 patients who underwent curative total or subtotal gastrectomy D2 or D3. Ninety-nine patients (72.2%) received perioperative transfusions. The data examined included the number and timing of transfusions (pre-, intra-, or postoperative), the type of operation (total or subtotal gastrectomy with or without splenectomy), tumor stage (pTNM), and the correlation between transfusions, mortality, morbidity and survival. Advanced T-stage (P = 0.01) and total gastrectomy (P = 0.009) were associated with a higher transfusion rate. No cases of operative mortality were recorded after 1988. Specific morbidity was 10.5% in non-transfused patients and 20.1% in transfused. Five-year survival rate in the transfused patients (28.3%) was significantly lower than in the non-transfused group (53.5%) (P = 0.03). Univariate analysis showed that T-stage (P = 0.001) and N-stage (P = 0.04) were associated with a lower survival. By multivariate analysis (Cox regression model) only T-stage (P = 0.001) and N-stage (P = 0.04) were independent prognostic factors, whereas transfusions were not an independent variable (P = 0.27). To conclude, the issue of the real impact of transfusions on the prognosis of gastric cancer is far from being settled, although the T and N parameters are known to be strictly correlated to prognosis. This study further confirms the importance of limiting homologous transfusions as well as of transfusing, whenever possible, autologous or leukodepleted blood; this, however, without losing sight of the primary goal of minimizing operative blood loss.