In the current analysis, four SNPs for Latinos and one SNP for African-Americans in XRCC1
were found to be significantly associated with lung cancer risk. Among these SNPs, the association between Arg399Gln (rs25487) and NSCLC among Latinos has the smallest probability of being false positive with FPRPs <0.5. We found an increased risk of NSCLC for those who carried the Gln/Gln genotype compared with those with the Arg/Arg genotype among Latinos, but not among African-Americans. The only other study with Latino subjects did not observed a significant association between Arg399Gln and adenocarcinoma of the lung; however, the number of Latino cases was small in that study (n
). The lack of an association between lung cancer risk and Arg399Gln among African-Americans is probably due to the low prevalence of Gln/Gln genotype among this population (2.1% among controls and 1.6 among all lung cancer cases). A previous study among African-Americans also did not observe a significant association between Arg399Gln and lung cancer risk, and the prevalence of Gln/Gln carriers also was low (3.7% among controls and 1.9% among cases) (2
). Two SNP databases, SNP500Cancer (9
) and NCBI Entrez SNP (24
), also report a low prevalence of Arg399Gln Gln/Gln genotype among African-Americans ranging from 1.4 to 5.4%, suggesting that the contribution of Arg399Gln polymorphism to lung cancer risk may be low for African-Americans. In contrast, the prevalences Arg399Gln Gln/Gln genotype were higher among Hispanics (4.3–7.1%), Asians/Pacific rim population (4.7–29.2%) and Caucasians (8.5–19.4%) as reported by SNP500cancer and NCBI Entrez SNP. A meta-analysis by Kiyohara et al.
) showed that Gln/Gln genotype was significantly associated with an increased risk of lung cancer among Asians but not among Caucasians. Since that meta-analysis, one study among Caucasians reported an increased risk of lung cancer associated with Gln variant only among light smokers (26
). In contrast, another study among Caucasians reported a decreased lung cancer risk among non-smoking women who carried at least one copy of Gln (27
). A third study among Caucasians did not observe any significant association between lung cancer and Arg399Gln (28
). One (29
) of the two studies (29
) among Chinese published since the meta-analysis found an increased risk of lung cancer associated with the Gln/Gln genotype. In contrast, a study among an Indian population reported an inverse association between lung cancer and the Gln allele (31
). These inconsistencies between racial groups may be due to the difference in environmental exposures (e.g. prevalence and the dose of tobacco smoking). Alternatively, Arg399Gln may be a proxy SNP linked to a true causal SNP, and the linkage patterns may differ across ancestral groups.
Although the association between Arg399Gln and lung cancer requires further confirmation, several studies suggest that this polymorphism may have functional impact. A lower DNA repair capacity has been observed among styrene-exposed workers who carried Gln/Gln genotype (21
). Another study showed that among never-smokers, Gln/Gln carriers had a significantly higher level of DNA damage as measured by DNA adduct compared with those with either Arg/Arg or Arg/Gln genotypes. Finally, a study reported that among benzene-exposed laboratory workers, those with at least one Gln allele had a lower DNA repair capacity (23
The current study observed a significant association between lung cancer and one SNP (rs2256507) in the 3′ downstream region of XRCC1 for Latinos and one SNP (rs17655484) in the 5′ upstream region of XRCC1 for African-Americans, suggesting that these regions may warrant further investigation. Specifically, these regions may contain regulatory elements that influence the expression of XRCC1 protein. However, high FPRPs associated with these SNPs suggest that these findings may be due to chance.
A meta-analysis of OGG1
Ser326Cys (rs1052133) using data from seven studies showed that those who carried the Cys/Cys genotype had a small increase in the risk of lung cancer (OR
1.24; 95% confidence interval: 1.01–1.53) (32
). Subsequently, another meta-analysis of OGG1
Ser326Cys using 11 studies (including seven from the previous meta-analysis) showed a small but non-significantly increased risk of lung cancer (OR
1.17; 95% confidence interval: 0.88–1.56) (25
). This suggests that the impact of Ser326Cys, if real, is probably to be small. In our study, the ORs of NSCLC for those with Cys/Cys genotype were 1.43 and 1.19 for Latinos and African-Americans, respectively, which were similar in magnitude to results from the meta-analyses, though not statistically significant. Our sample size was too small to detect a significant OR <1.5.
Asp148Glu is a commonly studied SNP for APEX1
. Though the Asp148Glu was shown by one study to have functional impact on sensitivity to ionizing radiation (33
), no clear association between lung cancer and APEX1
Asp148Glu has been found according to a meta-analysis (25
). Since that meta-analysis, one additional study reported no association between lung cancer and Asp148Glu (27
), whereas another study reported an increased lung cancer risk for those who carried at least one copy of Glu (26
). Consistent with most of the previous studies, we also did not observe any significant association between APEX1
Asp148Glu and lung cancer.
A major limitation of this study is the lack of statistical power to detect weak gene–disease associations (OR
1.5) due to the small sample size, which may also increase the chance for spurious findings. In addition, SNP coverage is not complete in genes examined by this study such that the null findings do not necessarily suggest that these genes are not important for the development of lung cancer. Nevertheless, this is only the second study for both Latinos and African-Americans to assess the association between BER SNPs and lung cancer and the only study of this type to have adjusted for genetic ancestry to account for potential population stratification.
In summary, this study observed a significant increased risk of NSCLC associated with Gln/Gln genotype of Arg399Gln of XRCC1 among Latinos but not among African-Americans, suggesting that there may be genetic differences in the association between BER pathway and lung cancer between Latinos and African-Americans. Since the analyses were adjusted for genetic ancestry, the observed association between Arg399Gln and NSCLC among Latinos is less likely to have been confounded by population stratification. However, since this is the first study to report this association among Latinos, more studies need to be conducted among Latinos to confirm this result.