The efficacy of gastrectomy for large gastric cancer

Background Large gastric cancer (LGC) is frequently associated with extended disease, and the role of surgical resection has been debated. We investigated the efficacy of surgical treatment for LGC. Methods The size of LGC was defined as 8 cm or greater. Four hundred and fifteen patients with LGC who underwent gastrectomy were included. The clinicopathological features, the status of the residual tumor, the incidence and patterns of relapse, and the survival were analyzed. Results Macroscopically, diffuse-type tumors were dominant (60%). The numbers of patients with tumors of T3 or greater, lymph node involvement, and peritoneal metastases were 356 (86%), 359 (87%), and 126 (30%), respectively. One hundred and eighty-eight patients (45%) underwent incomplete tumor resection (R2). The R2/R0 (no residual tumor) ratio was greater than 1 in patients with type 4 tumors and N1 or greater metastasis and in those with type 3 tumors and N2 or greater metastasis. In contrast, T2, type 2, and type 5 tumors were more likely to be completely resected. The 5-year survival for all 415 patients was 26%. The survival rates were inversely related to the tumor type, size, and lymph node metastasis. In the 216 patients with R0, the 5-year survivals of those with pN (International Union Against Cancer [UICC] classification) 0, 1, 2, and 3 were 66%, 56%, 36%, and 5%, respectively ( P = 0.001). In 96 of these 216 patients (44%) the tumor recurred, and peritoneal metastasis was the most frequent mode of recurrence (48%). By Cox's proportional hazard model, the tumor size was an independent prognostic factor. Conclusion The chance of achieving R0 resection for LGC is low, except for T2, type 2, or type 5 tumors. Primary resection should be avoided for other types of LGC.