The National Cancer Data Base report on poor survival of U.S. gastric carcinoma patients treated with gastrectomy : Fifth edition American Joint Committee on Cancer staging, proximal disease, and the ?different disease? hypothesis

To substantiate reports of increasing proportions of gastric adenocarcinoma of diffuse histologic type and in the proximal portion of the stomach, to better understand the prognostic features that govern survival, and to determine whether alterations of operative strategy might improve the surgical results. Retrospective analysis of 289 consecutive patients with gastric adenocarcinoma operated on by general surgeons over a 26-year period. Records were reviewed for location, histologic type, resection, operative mortality, lymph node status, and outcome. The Section of Surgical Oncology, the New England Deaconess Hospital, Boston, Mass. Survival rate, length of life of the patients who died, and operative mortality. A marked and significant shift of gastric adenocarcinoma to a proximal location (54% between 1985 and 1990) occurred over 26 years (P = .0075) with a significant stage improvement at presentation (P = .0235). Percentages of cancers that were of the diffuse, poorly differentiated histologic type increased to 48%. More curative operations were performed in the last period (61%), and this upward trend from 37% was significant. Proximal gastric cancers had a poorer prognosis with more operative deaths, more lymph node metastases, and worse survival rates than distal cancers. Poor survival rates occurred even when comparing patients with negative lymph nodes or favorable histologic features with patients with similar distal cancers. Despite significant increases in the proportion of proximal cancers, survival rates have improved only slightly. Nodal status plays a less prognostic role than does location or histologic type but does provide prognostic information for individual locations. Survival rates for diffuse histologic cancer were consistently worse than those for intestinal histologic cancer, which emphasizes the underlying disease biology controlling outcome. Radical lymphadenectomy for gastric adenocarcinoma would not improve surgical outcome in the United States.

To compare the stage-stratified survival of Japanese patients treated in Honolulu according to Western techniques with that of Japanese patients treated in Tokyo according to Japanese techniques, thus eliminating race as a potentially confounding variable. Of 312 Honolulu Japanese patients surviving Western-type gastric resection for neoplasm between 1974 and 1985, 279 were identified with invasive gastric adenocarcinoma unassociated with any second malignancy. This Honolulu cohort, treated by Western methods, was retrospectively compared with a similar, previously described cohort of 3176 Tokyo Japanese patients treated according to Japanese methods. American Joint Committee on Cancer/Union Internationale Contre le Cancer criteria for stage-stratified survival. Despite non-TNM prognostic factors favoring higher survival for the Honolulu Japanese patients, for every TNM stage, we observed higher survival for the Tokyo Japanese patients who were treated according to Japanese techniques. For stage I disease, the survival rates were 86% vs 96%, respectively (P < .001); for state II, 69% vas 77% (P = .15); for stage III, 21% vs 49% (P < .001); and for stage IV, 4% vs 14% (P < .001). Because all patients in this study are Japanese, race-related factors or the "different-disease" hypothesis cannot explain these results. Lymphadenectomy-related stage-migration and/or differing therapeutic efficacy seem more likely explanations.