Abstract
BACKGROUND: The CXCL12 rs1801157 polymorphism is putatively associated with the risk of malignancy. However, research results are inconsistent regarding particular cancers, especially colorectal cancer (CRC). METHODS: We conducted a meta-analysis via a comprehensive literature search of the databases PubMed and Embase. Odds ratios and their corresponding 95% confidence intervals were extracted to assess comprehensively the association between the CXCL12 rs1801157 polymorphism and the risk of CRC. RESULTS: According to the following models, the correlation between the presence of the CXCL12 rs1801157 polymorphism and the risk of CRC was not statistically significant: allelic (G cf. A), heterozygous (GG cf. GA), homozygous (GG cf. AA), dominant (GG cf. GA+AA), or recessive (AA cf. GA+GG) for pooled calculating. However, in subgroup analysis, elevated risk of CRC was found in the non-Caucasian group in four genetic models, while no connection was found in the Caucasian group. The sensitivity analysis confirmed that the result of the Caucasian group was stable and analyzed the heterogeneity of the non-Caucasian group. CONCLUSION: This meta-analysis suggested that the CXCL12 rs1801157 polymorphism did not significantly confer a risk of CRC among Caucasians but may among non-Caucasians. These results warrant confirmation and further studies of different ethnic populations.