Worldwide, lung cancer represents 11.4% of overall cancer diagnoses. It is also the leading cause of cancer death, accounting for 18% of all cancer deaths. It is the most frequently occurring cancer among men and the third most frequent in women. Lung cancer is the leading cause of cancer death among men and the second most common fatal cancer in women, comprising 21.5 and 13.7% of all cancer deaths, respectively.1,2 Tobacco consumption is the main association factor of this disease followed by exposure to indoor radon.2

Radon is classified as the main cause of lung cancer in never smokers and the second most common in ever-smokers by the United States Environmental Protection Agency (USEPA) and the World Health Organization (WHO).3–5 The molecular mechanism of radon carcinogenesis is not fully understood, and lack of knowledge remains regarding the precise molecular carcinogenic pathways for tobacco exposure. Genetic susceptibility may play an important role and explains why some people develop lung cancer and others do not, given the same tobacco or indoor radon exposures.6,7

Tobacco smoke, despite damaging DNA, has an additional effect in DNA repair pathway activity, reducing the detectable amount of some repair proteins, (including XPC, OGG1 and OGG2) in pulmonary tissue in mouse models.8 Tobacco smoke exposure induces bulky DNA adducts, which have a linear relationship with carcinogenesis at low exposure, but with high doses may reach a steady state, possibly due to DNA repair pathway saturation and increased apoptosis.9,10

Low capacity in DNA repair is related with an increased association with lung cancer.11,12 There are three major types of DNA repair mechanisms: nucleotide excision repair (NER), the main route in mammals which repair DNA damaged by ultraviolet light, environmental mutagens and chemotherapy12; base excision repair (BER) that repairs DNA damaged by oxidation, deamination and alkylation. It protects from ageing, neurodegeneration and cancer13; and mismatch repair (MMR) that blocks recombination between non identical DNA.14 Published studies have focused in genes involved in NER such as ERCC1 and ERCC2, BER such as XRCC1 and OGG1, and MMR such as XRCC3.15 Nevertheless, no study has analyzed the potential interplay between these molecular pathways and indoor radon exposure.

The aim of this study was to analyze if selected genetic polymorphisms in genes playing a role in these molecular pathways are associated with an increased association of NSCLC. As a secondary objective, we have analyzed the effect of these genes taking into account the different amount of tobacco consumption and indoor radon exposure, to ascertain a potential modulating effect for the polymorphisms analyzed.

The study included 2027 participants, 826 cases and 1201 controls. Cases and controls were well balanced regarding sociodemographic variables, with an average age of 65.5, 50.5% of women. Cases and controls lived a similar number of years in the same dwelling. Residential radon exposure was available for 754 cases (91.3%) and 1010 controls (84.1%). The most frequent histological type was adenocarcinoma (69.7%). A sample description broken down by case–control status is shown in Table 1.

Table 2 summarizes the results for the different genotypes and their distribution between cases and controls. Compared to participants with wild type gene NBN (rs1805794) showed an OR for the homozygous genotype of 1.78 (95%CI 1.23–2.58) and 1.78 (95%CI 1.15–2.75) after adjustment for age and sex and full adjustment, respectively. This result along other results for other polymorphisms showing no association are shown in Table 2.

Table 3 shows the association of the different polymorphisms studied with lung cancer at varying tobacco exposure categories (never-smokers, moderate-smokers and heavy-smokers). Lung cancer association increases when participants are heavy smokers for almost all the polymorphisms analyzed. Compared with never-smokers, the highest OR corresponded to homozygous genotype in heavy smokers for: OGG1 (rs125701) (OR 8.04; 95%CI 1.64–58.29), OGG1 (rs2472037) (OR 6.1; 95%CI 2.3–17.5) and ERCC1 (rs32112986) (OR 5.92; 95%CI 1.59–28.18). Only ERCC3 (rs4150459) in its homozygous type expressed an increased association with NSCLC in moderate smokers compared with never smokers with wild type (OR 2.9; 95%CI 1.96–4.33). NBN (rs1805794) in its homozygous genotype showed a higher association with lung cancer in never smokers (OR 2.72; 95%CI 1.44–5.2).

Table 4 shows the association of these polymorphisms with lung cancer broken down by radon concentration (≤200Bq/m3 and >200Bq/m3). NSCLC associations were increased for some variants for those participants exposed to more than 200Bq/m3. rs1452584 showed the highest association for the homozygous genotype (OR 3.04; 95% CI 1.15–8.48) as well as APEX1 (rs1130409) and ERRC1 (rs11615) showed a higher association too (OR 1.68; 95% CI 1.04–2.73 and OR 1.65; 95%CI 1.1–2.48 respectively). XRCC3 (rs861539), ERCC2 (rs13181), OGG1 (rs10521333) and ERRC1 (rs3212986) showed this increase for the heterozygous form (OR 1.79; 95%CI 1.28–2.51, OR 1.56; 95%CI 1.13–2.15, OR 1.53; 95%CI 1.09–2.16 and OR 1.47; 95%CI 1.06–2.04, OR 1.47; 95% CI 1.06–2.04 respectively). In addition, NBN (rs1805794) showed an increased association for its homozygous genotype for low indoor radon exposures (OR 1.88; CI 95% (1.09–3.24). No other significant association was observed.

We found that the association with lung cancer may be modulated by different genetic polymorphisms in the BER and NER pathways. These polymorphisms appear to have the most prominent effect when they are homozygous in heavy smokers. Some polymorphisms could also increase NSCLC association in those individuals exposed to high radon concentrations. To our knowledge, this is the study with the largest number of DNA repair genes analyzed involved in lung cancer related to radon exposure. It also provides potentially valuable data for further understanding of the carcinogenic pathway of tobacco consumption. The sample size of our study is large, with 2027 participants including smokers, never smokers and indoor radon exposure, and expands a previous study performed exclusively in never smokers.7

We have observed that participants considered heavy smokers, globally, had an increased association with lung cancer compared with the wild type in never smokers. The polymorphism most associated with NSCLC was rs125701 (OGG1) in its homozygous variant form. Attending to the study of Hualong Qin et al. this polymorphism is frequently methylated in NSCLC, but in that study no significant result was obtained. This could be due to the absence of homozygous participants for this gene. In contrast with our study, it was performed only in Chinese population.21 A relevant association was also observed for rs2472037 and rs3212986, both in the OGG1 gene, for its homozygous genotypes, increasing the association of lung cancer in heavy smokers. Regarding rs3212986, this polymorphism is considered an important contributor to lung cancer development in smokers.22,23 In addition, no relation with never smokers was detected by Tau Yu et al.23

The NBN gene variant (rs1805794), located in MMR pathway, was also associated with an increased association with lung cancer in heavy smokers in accordance with Chuang et al.,24 though they only found association in male smokers. The homozygous mutation of this gene in heavy smokers shows a higher lung cancer association compared to never smokers. These results are similar to the obtained by Charlotta et al.25 who found an increased lung cancer association in never smoking women and smoking women with low tobacco consumption. The increased lung cancer association shown in our study is similar to that reported by the Wang et al. meta-analyses.26 In contrast, He et al. did not find any similar association.27 This gene is known to take part in other carcinogenic procedures such as ovarian, breast and colorectal cancer. In addition, it has been studied for clinical purposes in the testing of hereditary cancers.28

In general, our results note that a number of genetic polymorphisms are associated with increased lung cancer association when these polymorphisms are homozygous. This was apparent for a number of the analyzed genes, especially those in the NER pathway. This could be attributable to the importance of this pathway in the repair of bulky DNA lesions generated by environmental mutagens.12 The gene which showed a clear increase in NSCLC, compared to the wild type in heavy smokers, belonging to this pathway is ERRC1 (rs3212986). OGG1 polymorphisms rs125701 and rs2472037 showed the same effect but they are located in the BER pathway, which is specialized in repairing adducts originated by oxidation.13 This is one of the main routes through which tobacco smoke produces DNA adducts.8,29 This is interesting because there seems to be no association when we analyze these polymorphisms in the overall sample. Nevertheless, when we stratify the sample by smoking categories, we observe that for heavy smokers an association is present for some homozygous polymorphisms, suggesting a saturation effect on the anticarcinogenic pathways compared to wild type genes.

In relation with indoor radon exposure, we observed that participants exposed to more than 200Bq/m3 showed an increased lung cancer association for some of the polymorphisms studied. Only three of them showed an increased lung cancer association for their homozygous genotype: rs1452584, located in chromosome 18, showed the highest association; ERCC1 (rs11615), located in NER pathway, which also showed an increased association for its wild type and APEX1 (rs1130409), located in BER pathway. The results obtained for ERCC1 are in accordance to the obtained by Lorenzo-González et al.8 though this study is limited to a never smoking population while ours includes also ever-smokers. Other polymorphisms showed higher association for their heterozygous genotype; one of them, located in the NER pathway, ERCC1 (rs3212986) showed similar results to those found by Lorenzo-Gonzalez et al.7

Despite the fact that some studies established a relationship between lung cancer association and DNA repair genes,22,30 we are not aware of any study relating these polymorphisms with indoor radon exposure. These findings support our hypothesis that lung cancer association is increased by indoor radon exposure and that some variants might have a special importance favouring the development of this disease.

This research has some limitations. The most important limitation is that when we correct the p-value with the false discovery rate, we observed that some “suggestive” p-values in different genes become “non-suggestive”, but it is important to resemble that a “non-suggestive” association cannot be deduced only by a p-value because it could be easily modified. For example, if we have a small number of participants in one category, we might have a “non-suggestive” p-value and a suggestive confidence interval. If we increase the number of participants in the category we could have a significant p-value while the width of the confidence interval narrows. Secondly, we did not have enough participants in some polymorphisms, which makes difficult to establish a potential association with lung cancer. One example is ERCC3 (rs4150459) that for its homozygous genotype for never-smokers did not have enough participants to obtain any result. Other limitations reside in the limited number of polymorphisms we have assessed. Not all polymorphisms present in the studied genes have been analyzed, and the BER and NER pathways have some more genes that we do not have analyzed. A further limitation is the fact that there are other pathways and genes involved in carcinogenesis that we do not have included and might have a role, such as deletion of GSTM1 and GSTT1 genes.31 Finally, the most frequent histological type was adenocarcinoma. This is due to the fact that we included an important amount of never-smokers from the LCRINS study (Lung Cancer Association in Never Smokers), where this histological type is the most frequent. Nevertheless, this fact favoured that we had the same percentage of cases of each sex (50% each), allowing for a better interpretation of the results both for men and women.

This study has also a number of advantages. First, the sample size is relatively large, and we have been able to include information regarding the second association factor of lung cancer, which is indoor radon exposure. This is of particular importance because there is a lack of information on the mechanisms of radon causing lung cancer. It is important to note that Galicia is a radon prone-area32,33 and the median time that cases and controls have been living at the same residence is high enough to induce lung cancer. Finally, the high participation rate and the multicentric nature of this research increase the external validity of its results.

In conclusion, these findings support that polymorphisms located in BER and NER pathways do not have an association with lung cancer onset, with the exception of the gene NBN (rs1805794). Nevertheless, this study support the hypothesis of their role modulating the effect of tobacco consumption, where heavy smokers might have their lung cancer association modulated by polymorphisms located in the BER and NER pathways. This could also happen for those individuals having high indoor radon exposures though at a minor extent.

This work was funded by a competitive research grant offered by the Xunta de Galicia [10CSA2080057PR] “Risk factors of lung cancer in never smokers: a multicenter case–control study in the Northwest of Spain”, partially supported by the Instituto de Salud Carlos III [PI13/01765]. “Molecular genetic profile of DNA repair markers (BER and NER) and biological risk of lung cancer from residential radon exposure: A case–control study”, partially supported by Instituto de Salud Carlos III, Ministry of Science and Innovation of Spain [PI15/01211]. (Year 2015) and another from the same institution [PI031248] (Year 2012). “Residential Radon Exposure, Histologic Types, and Lung Cancer Risk. A Case–Control Study in Galicia, Spain”.

Dr. Kelsey is a founder and scientific advisor for Cellintec, which had no role in this research.

The authors declare no conflict of interests.