Evaluation of association studies and a systematic review and meta-analysis of XRCC1 polymorphisms in breast cancer risk

Background: To investigate the association between XRCC1 Arg194Trp, Arg280His and Arg399Gln polymorphisms and breast cancer (BC) risk, fifteen previous meta-analyses and large amounts of original articles had been published but drew conflicting conclusions. Because of lacking credibility analysis, the positive results of the previous meta-analyses were quite suspicious. Objectives We performed an updated meta-analysis to solve the controversy in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). Methods All results of the current meta-analysis were evaluated comprehensively based on concerning credibility. Results With 74 studies including 31,166 cases and 34,669 controls, XRCC1 Arg399Gln polymorphism significantly increased BC risk in overall and subgroups (Asians, Indians, Africa, Asia and North America), and it was again demonstrated in Asians and Asia by further sensitivity analyses. As 45 studies with 17,239 cases and 19,026 controls pooled to explore the role of Arg194Trp polymorphism in BC risk, no significant results were found in overall and any ethnicity and geographic region, and there was a decreased BC risk in post-menopausal and an increased BC risk in pre-menopausal. The Arg280His polymorphism analyses, involving 18 studies with 9,571 cases and 10,108 controls, did not find an association with BC risk in overall and any ethnicity and region, but an increased BC risk was found in overall and Asians living in Asia in sensitivity analyses. However, all significant results were taken for less credible after we comprehensively analyzed with heterogeneity characteristics, the quality of involved studies and Bayesian false-discovery probability (BFDP) to evaluate the credibility of significant results. Conclusion The issue about false-positive results should be taken seriously and more high-quality original studies still have plenty of value to be significantly developed.