Analysis of Prognostic Factors in Gastric Cancer Patients Treated with Total Gastrectomies

ABSTRACT

The prognoses for the gastric cancer patients treated with total gastrectomies are known to be unsatisfactory due to the low survival rates, the high frequency of postoperative mortality or morbidity, and long-term complications such as nutritional deficiency. The authors evaluated the 5-year survival rates and analyzed the prognostic factors in 557 patients with gastric cancer who underwent total gastrectomies during the period between Jan. 1987 and Dec. 1993. The overall 5-year survival rate was 49.7%, and the survival rates according to the stage were stage Ia, 92.0%; Ib, 85.5%; II 64.1%; IIIa 55.0%; IIIb 26.5%; and stage IV, 6.3%. Postoperative mortality rate was 1.1%. By using univariate analysis to evaluate the prognostic factors, factors such as age, depth of invasion, extent of lymph node metastasis (according to the Japanese rule), number of involved nodes, lymph node ratio, distant metastasis (peritoneal and/or hepatic), size of the tumor, gross type, histological type, the surgical curability and the TNM stage were found to be related with the survival of the patients. In a multivariate analysis using 11 variables, the TNM stage was the single most significant prognostic factor. Besides the TNM, depth of invasion (ratio of risk (R.R)=1.50), extent of lymph node metastasis (R.R=1.83), number of involved nodes (R.R=1.64), lymph node ratio (R.R=1.91), and peritoneal metastasis (R.R=3.11) were found to be independent prognostic factors influencing survival. It was thought that the radicality of surgery could be reflected in the number of removed nodes per specimen. In this study, the average number of removed nodes was 42.3 per case. Hence, it may be said that adequate lymphadenectomy was performed for almost all the grossly curable cancers. The 5-year survival rate in stage IV patients with tertiary node (N3) metastasis and no peritoneal or hepatic metastasis was 16.8%; in patients with peritoneal or hepatic metastasis, the survival rate was 0%. There was a significant survival difference between these two groups (p<0.05). This result suggests that the tertiary node metastasis is a potentially curable factor, and that it should be classified differently in the current TNM system. In conclusion, the overall survival rates in the patients treated with total gastrectomies were favorable compared with the results in other reports. Depth of invasion, extent of lymph node metastasis, number of involved nodes, lymph node ratio were important prognostic factors for survival after a total gastrectomy. The current TNM staging system appears to be a reasonable one, except that the probable curability of tertiary node metastasis may need to be taken into consideration.