System review and metaanalysis of the relationships between five metabolic gene polymorphisms and colorectal adenoma risk.

The relationships between some metabolic (including EPHX1, GSTs and NQO1) gene polymorphisms and colorectal adenoma (CRA) risk have been commonly studied, and no conclusions are available up to now. Therefore, we quantitatively studied the relationships by a metaanalysis. The databases of Medline and Embase were retrieved updated to June 15th, 2011. Crude or adjusted odds ratio (crude OR or adjusted OR) and 95% confidence interval (95%CI) were calculated to present the strength of the associations. Overall, nine case-control studies for EPHX1 Tyr113His and His139Arg, five case-control studies for GSTM1, four studies for GSTP1 Ile105Val, two studies for GSTP1 Ala114Val, six studies for GSTT1 and four studies for NQO1 Pro187Ser were included in this metaanalysis. The results of combined analyses indicated that EPHX1 Tyr113His and His139Arg, GSTT1, GSTM1, GSTP1 Ile105Val and Ala114Val were not associated with CRA risk [crude OR (95%CI): 0.98 (0.90-1.07) and P ( z-test) = 0.65 for EPHX1 His carriers vs. Tyr/Tyr; 1.05 (0.97-1.15) and P ( z-test) = 0.21 for EPHX1 Arg carriers vs. His/His; 1.05 (0.92-1.20) and P ( z-test) = 0.47 for GSTT1 Null vs. Present; 1.01 (0.90-1.13) and P ( z-test) = 0.90 for GSTM1 Null vs. Present; 1.04 (0.92-1.17) and P ( z-test) = 0.56 for G carriers vs. AA for GSTP1 Ile105Val; 0.88 (0.70-1.11) and P ( z-test) = 0.28 for T carriers vs. CC for GSTP1 Ala114Val]. In contrast, Ser allele of NQO1 Ser187Pro might be a modest risk factor for CRA development [1.19 (1.06-1.33) and P ( z-test) = 0.003 for Ser carriers vs. Pro/Pro]. To get more precise evidences, adjusted ORs (95%CI) for EPHX1 Tyr113His, His139Arg, GSTP1 Ile105Val and NQO1 Ser187Pro were also calculated based on adjusted ORs (95%CIs) reported in primary studies. The results still indicated that EPHX1 Tyr113His, His139Arg and GSTP1 Ile105Val were not associated with CRA risk except for NQO1 Ser187Pro. When subgroup analyses were performed for population-based case-control studies or studies in HWE for EPHX1 Tyr113His and His139Arg, and NQO1 Ser187Pro polymorphisms, the results were persistent. Although with modest limitations and biases, this metaanalysis suggests that EPHX1 Tyr113His and His139Arg, GSTT1, GSTM1, GSTP1 Ile105Val and Ala114Val polymorphisms may be not risk factors for CRA development, while Ser allele of NQO1 Ser187 Pro may be a modest risk factor for CRA development, and may be used with other genetic markers for screening CRA in the future.