Variants in DNA double-strand break repair genes and breast cancer susceptibility

We performed genetic association studies in a population-based breast cancer case-control study analysing polymorphisms in genes involved in homologous recombination (NBS1, RAD52, RAD51, XRCC2 and XRCC3) and non-homologous end-joining (KU70/80 and LIG4). These DNA double-strand break repair genes are candidates for breast cancer susceptibility. Genotype results were available for up to 2205 cases and 1826 controls. In the homologous recombination (HR) pathway, genotype frequencies differed between cases and controls for two polymorphisms in XRCC3; T241M (P=0.015) and IVS5 A>G at nt 17893 (P=0.008). Homozygous carriers of M241 were associated with an increased risk [odds ratio (OR) MM versus TT=1.3 (95% confidence interval (CI) 1.1-1.6)], while the rare allele of IVS5A>G was associated with a dominant protective effect [OR AG versus AA=0.8 (0.7-0.9)]. The association of a rare variant in XRCC2 (R188H) was marginally significant [P=0.07; OR HH versus RR=2.6 (1.0-6.7)]. In the non-homologous end-joining (NHEJ) pathway, a polymorphism in LIG4 (T>C at nt 1977) was associated with a decrease in breast cancer risk [P=0.09; OR CC versus TT=0.7 (0.4-1.0)]. No significant association was found for 12 other polymorphisms in the other genes studied. For XRCC3, we found evidence for four common haplotypes and four rarer ones that appear to have arisen by recombination. Two haplotypes, AGC and GGC, were associated with non-significant reductions in breast cancer risk, and the rare GAT haplotype was associated with a significantly increased risk. These data provide some evidence that variants in XRCC2 and LIG4 alter breast cancer risk, together with stronger evidence that variants of XRCC3 are associated with risk. If these results can be confirmed, understanding the functional basis should improve our understanding of the role of DNA repair in breast carcinogenesis.