D Palli1, A Russo, C Saieva, S Salvini, A Amorosi, A Decarli. 1. Analytical Epidemiology Branch, Epidemiology Unit, Centro per lo Studio e la Prevenzione Oncologica, Florence, Italy. palli@plugit.net
Abstract
BACKGROUND: Although the role of diet in gastric carcinoma (GC) causation has been investigated extensively, its association with long term survival has never been explored in detail. METHODS: The authors assessed the vital status on December 31, 1997 of a series of 382 patients with GC who were interviewed in 1985-1987 in the framework of a population-based case-control study in a high risk area for GC in Italy. Follow-up information at 10 years was available for all but three patients. The risk of dying was evaluated by Cox proportional hazard models, including patient age, gender, social status, disease stage at diagnosis, and lymph node involvement as confounding variables. RESULTS: The overall survival probability was 0.26 at 5 years and 0.18 at 10 years after interview. The most important predictors were disease stage, lymph node involvement, and histopathologic grading (TNM classification). Overall, a high intake of alcohol was associated with an increased risk of dying (P = 0.02). In contrast, survival was increased in patients with a high intake of alpha-tocopherol (P = 0.04). A positive first-degree family history of GC was reported by 84 patients and tended to be associated with a reduced risk of death. A significant protective trend was found with increasing intake of vegetable fat and starch among these familial cases, with a 50% risk reduction. In contrast, this subgroup showed a two-fold increased risk of dying with a high intake of animal protein, animal fat, and preformed N-nitrosodimethylamine. CONCLUSIONS: The current results provide evidence that premorbid diet may influence the prognosis of patients with GC, and they suggest that specific pathways for disease progression may exist among familial cases. Copyright 2000 American Cancer Society.
BACKGROUND: Although the role of diet in gastric carcinoma (GC) causation has been investigated extensively, its association with long term survival has never been explored in detail. METHODS: The authors assessed the vital status on December 31, 1997 of a series of 382 patients with GC who were interviewed in 1985-1987 in the framework of a population-based case-control study in a high risk area for GC in Italy. Follow-up information at 10 years was available for all but three patients. The risk of dying was evaluated by Cox proportional hazard models, including patient age, gender, social status, disease stage at diagnosis, and lymph node involvement as confounding variables. RESULTS: The overall survival probability was 0.26 at 5 years and 0.18 at 10 years after interview. The most important predictors were disease stage, lymph node involvement, and histopathologic grading (TNM classification). Overall, a high intake of alcohol was associated with an increased risk of dying (P = 0.02). In contrast, survival was increased in patients with a high intake of alpha-tocopherol (P = 0.04). A positive first-degree family history of GC was reported by 84 patients and tended to be associated with a reduced risk of death. A significant protective trend was found with increasing intake of vegetable fat and starch among these familial cases, with a 50% risk reduction. In contrast, this subgroup showed a two-fold increased risk of dying with a high intake of animal protein, animal fat, and preformed N-nitrosodimethylamine. CONCLUSIONS: The current results provide evidence that premorbid diet may influence the prognosis of patients with GC, and they suggest that specific pathways for disease progression may exist among familial cases. Copyright 2000 American Cancer Society.
Authors: Ying Gao; Nan Hu; XiaoYou Han; Carol Giffen; Ti Ding; Alisa Goldstein; Philip Taylor Journal: BMC Cancer Date: 2009-08-05 Impact factor: 4.430