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.

ABSTRACT

BACKGROUND: A high proportion of U.S. patients with gastric carcinoma do not receive surgical treatment. To sharpen staging criteria and facilitate comparisons with surgical series, an analysis of patients whose treatment included gastrectomy was undertaken. In addition, to evaluate the "different disease" hypothesis as an explanation for superior Japanese results, outcomes for Japanese Americans were examined. METHODS: Data were obtained from National Cancer Data Base (NCDB) reports of 50,169 gastric carcinoma cases diagnosed during the years 1985-1996 and treated with gastrectomy. In addition to demographic and treatment information, 5-year and 10-year relative survival rates are presented, with stage defined according to fifth edition American Joint Committee on Cancer (AJCC) staging procedures. RESULTS: Stage-stratified 5-year and 10-year relative survival rates were as follows Stage IA, 78%/65%; Stage IB, 58%/42%; Stage II, 34%/26%; Stage IIIA, 20%/14%; Stage IIIB, 8%/3%; and Stage IV, 7%/5%. Stage-stratified survival for Japanese Americans was higher. Males had a poorer prognosis than females, and the male-to-female ratio for Japanese Americans was lower. Proximal tumors were associated with a worse prognosis than distal tumors; the proportion of Japanese Americans with proximal disease was less than in the overall patient group. Japanese Americans underwent resection of adjacent organs less frequently. In this series, adjuvant therapy did not substantially affect survival. Overall, 20% were 10-year survivors; of these, 67% were lymph node negative and 98% had lymph nodes. Five-year stage-stratified survival increased for cases with >/= 15 lymph nodes analyzed. Stage migration was evident in cases with CONCLUSIONS: The current AJCC/International Union Against Cancer TNM staging system fails to accommodate the effect of proximal location on prognosis. Largely because Japanese Americans present with fewer proximal tumors, have a lower male-to-female ratio, and undergo adjacent organ resection less frequently, stage- stratified survival for Japanese Americans appears to be superior. In the U.S., surgical undertreatment of patients with this disease appears to be a problem.