We performed stratification analyses on cancer type (divided into digestive system cancers and other cancers). For digestive system cancers, we further separated Asians and Caucasians. Meta regression was used to illustrate potential reasons of

between-study heterogeneity. Egger’s test and click here inverted funnel plots were utilized to provide a diagnosis of publication bias (linear regression asymmetry test65). All analyses were performed using Stata version 9.2 software (Stata, College Station, TX, USA). All statistical evaluations were made assuming a two-sided test with a significance level of 0.05, unless stated otherwise. The characteristics of the selected studies are listed in Table 1. The distribution of genotypes in the controls was consistent with the Hardy–Weinberg equilibrium for all selected studies, except for three studies for −765G>C,33,37,48 two studies for −1195G>A,19,29 and two studies for 8473T>C.46,59 When we assumed that the OR for an allelic genetic association was 1.2, only one

study achieved a statistical power greater than 80%.61 Only Pembrolizumab datasheet two studies had a detailed dominant genotype frequency, so we extracted the data only for the dominant model.18,41 The evaluation of the association between these three polymorphisms and cancer risk is presented in Table 2. Overall, the variant A allele of COX-2−1195G>A can significantly increase the risk of cancer in all of selleck chemicals the tested models (GA vs GG: OR, 1.18; 95% CI, 1.07–1.29; P = 0.267 for the heterogeneity test; AA vs GG: OR, 1.35; 95% CI, 1.14–1.60; P = 0.010 for the heterogeneity test; dominant model, GA/AA vs GG: OR, 1.29; 95% CI, 1.18–1.41; P = 0.113 for the heterogeneity test; recessive model, AA vs GG/GA: OR, 1.22; 95% CI, 1.10–1.34; P = 0.002 for the heterogeneity test). However, for COX-2−765G>C and 8473T>C, no significant associations between the polymorphisms and risk of cancer were

observed. We then evaluated the effect of the three polymorphisms by specific tumor types. As shown in Table 2and Figure 1, we found that −1195G>A can significantly increase the risk of digestive system cancers in all tested models (GA vs GG: OR, 1.23; 95% CI, 1.11–1.37; P = 0.278 for the heterogeneity test; AA vs GG: OR, 1.55; 95% CI, 1.28–1.88; P = 0.045 for the heterogeneity test; dominant model, GA/AA vs GG: OR, 1.36; 95% CI, 1.23–1.51; P = 0.149 for the heterogeneity test; recessive model, AA vs GG/GA: OR, 1.32; 95% CI, 1.16–1.51; P = 0.016 for the heterogeneity test), whereas the increased risk was not evaluated for the ‘other cancers’ group.