Association between EGF promoter polymorphisms and cancer risk: a meta-analysis

EGF promoter polymorphisms are observed to modulate EGF levels and thought to have effect on susceptibility to various carcinomas but the results are inconsistent. In this meta-analysis, we assessed published studies of the association between three EGF polymorphisms and cancer risk from 21 studies with 14,609 subjects for EGF G61A, from two studies with 2,535 subjects for G-1380A and A-1744G, respectively. For EGF G61A, the contrast of homozygote (OR = 0.80, 95% CI = 0.65–0.98), allele (OR = 0.90, 95% CI = 0.81–0.99) and dominant model (OR = 0.86, 95% CI = 0.74–0.99) produced significant association among 21 studies with relatively large heterogeneity (Pheterogeneity < 0.001). Through the stratified analysis, heterogeneity decreased significantly. In the stratified analysis by racial descent, the significant risks were found among Asians for homozygote contrast (OR = 0.83, 95% CI = 0.69–0.99, Pheterogeneity = 0.506) and Americans for the contrast of homozygote (OR = 0.50, 95% CI = 0.30–0.84, Pheterogeneity = 0.051), allele (OR = 0.70, 95% CI = 0.51–0.96, Pheterogeneity = 0.008) and dominant model (OR = 0.57, 95% CI = 0.42–0.77, Pheterogeneity = 0.28). No significant associations were found in all Caucasians genetic models. In the subgroup analyses by cancer types, for gastric cancer and esophageal cancer significant associations were found in all genetic models without heterogeneity. Significant risk was also found in the contrast of homozygote (OR = 0.41, 95% CI = 0.20–0.81, Pheterogeneity = 0.184) and recessive model (OR = 0.53, 95% CI = 0.33–0.85, Pheterogeneity = 0.384) for hepatoma and recessive model (OR = 0.72, 95% CI = 0.53–0.99, Pheterogeneity = 0.474) for glioma. For EGF G-1380A and A-1744G, no significant associations were found in all genetic models. This meta-analysis suggests that the EGF G61A polymorphism most likely contributes to decreased susceptibility to cancers among Asians and Americans, and A allele may be a protective factor for gastric cancer, esophageal cancer, hepatoma and glioma. Both EGF G-1380A and A-1744G is marginally associated with cancer susceptibility.