Inflammation is one of the early phases in the development of gastric cancer. Therefore, several studies have examined the association of polymorphisms in tumor necrosis factor-A gene (TNF-A) with gastric cancer risk. This meta-analysis reviews and summarizes published evidence for these associations. Searching the PubMed Database yielded 20 independent studies that reported on the associations between TNF-A -308 or TNF-A -238 G>A polymorphisms and gastric cancer risk. Summary odds ratios (OR) and 95% confidence intervals (95% CI) were calculated in the random-effects model using the DerSimonian-Laird method. Q-statistics and I2 statistics were calculated to examine heterogeneity, and funnel plots were plotted to examine small study effects. For TNF-A -308 polymorphism, the overall OR (95% CI) for AG and AA genotypes versus GG genotype was 1.06 (0.90 - 1.26). I2 statistic for this association was 44% and the P-value for Q-statistic was 0.02. For TNF-A -238 polymorphism, the overall OR (95% CI) for AG and AA genotypes versus GG genotype was 1.15 (0.87 - 1.52). I2statistic for this association was equal to zero and the P-value for Q-statistic was 0.43. Limiting the analyses to noncardia gastric cancers, intestinal type cancers, H. pylori-positive subjects, Western or East-Asian populations, or studies in which genotype distributions were in Hardy-Weinberg equilibrium, the summary OR remained non-significant for both of these polymorphisms. In conclusion, the overall associations between TNF-A -308 and TNF-A -238 A carrying genotypes (GA and AA) and gastric cancer were null, but there might be some result heterogeneity.
98th AACR Annual Meeting-- Apr 14-18, 2007; Los Angeles, CA