Environmental Tobacco Smoke, Genetic Susceptibility, and Lung Cancer among Never Smokers
Although exposure to environmental tobacco smoke (ETS) is a well-established risk factor for lung cancer in never smokers, studies to date have not been able to precisely estimate the magnitudes of association between ETS and lung cancer by histological subtypes, especially for small cell lung cancer, large cell lung cancer, and adenocarcinoma in situ/minimally invasive carcinoma (AIS/MIA). In addition, few studies have investigated the roles of candidate susceptibility genes in lung cancer development and explored their potential interactions in relation to ETS exposure among never smokers.
Objectives and Specific Aims
The overall objective of this doctoral dissertation is to examine the associations of ETS exposure and variants of candidate genes with lung cancer susceptibility and to assess potential gene-environmental interactions among never smokers. The specific aims were: 1) To estimate the magnitudes of association between exposure to ETS and risk of lung cancer by major histological type (adenocarcinoma, squamous cell carcinoma, large cell lung cancer, and small cell lung cancer) and for AIS/MIA; 2) To estimate the associations between polymorphisms of DNA repair, carcinogen metabolism, and cell cycle control genes and lung cancer in never smokers and to test for gene-environmental interactions with ETS exposure; and 3) To evaluate the associations of genetic polymorphisms related to miRNAs and stem cell regulation with lung cancer susceptibility in never smokers and to assess potential gene-environmental interactions with ETS exposure.
Study Design and Population
We conducted case-control studies using pooled data from 18 studies participating in the International Lung Cancer Consortium (ILCCO) for Specific Aims 1 and 2 and pooled data from the Jiangsu Four Cancers Study and the Taiyuan Air Pollution and Lung Cancer Study for Specific Aim 3. The study populations in the ILCCO studies were racially diverse while all participants in the Jiangsu and Taiyuan Studies were Chinese. All studies provided epidemiologic data collected through interviews using structured questionnaires. There was a total of 12,667 cases (2,503 never smokers) and 14,410 controls (7,276 never smokers) in the pooled ILCCO data and 382 non-smoking cases and 1,271 non-smoking controls in the pooled data of Jiangsu and Taiyuan Studies.
We imputed the missing data for pack-years of smoking, education, and income using the median values among controls. Observations with missing data for other variables were excluded from the analyses. Potential confounders were adjusted for in data analyses, including age, sex, study or area of residence, race/ethnicity (ILCCO studies only), education (Chinese studies only), and income (Chinese studies only). We also adjusted for tobacco smoking status and pack-years of smoking in analyses including both ever and never smokers. We used multivariate unconditional logistic regression analyses to estimate adjusted odds ratios (OR) and 95% confidence intervals (CI) for the associations and ratio of odds ratios (ROR) and 95% CI for gene-environmental interactions of interest. In order to mitigate sparse data bias, we employed the semi-Bayesian shrinkage method with informative priors based on the literature (when available) or a null-effect prior of OR=1.00 and 95% CI 0.25–4.00. Multiplicative interactions were assessed using the product-term method to estimate the ratio of odds ratios (ROR). Additive interactions were assessed by estimating the relative excess risk due to interaction (RERI).
ETS and Lung Cancer among Never Smokers. In the pooled ILCCO data, the adjusted ORs of ETS exposure on lung cancer among never smokers was 1.31 (95% CI 1.17–1.47) for all histological types combined. When stratified by histological type, the adjusted ORs among never smokers were 1.26 (95% CI 1.10–1.45) for adenocarcinoma overall, 1.52 for (95% CI 1.03–2.24) AIS/MIA, 1.38 (95% CI 0.98–1.95) for squamous cell carcinoma, 1.50 (95% CI 0.90–2.48) for large cell lung cancer, 1.28 (95% CI 1.13–1.45) for non-small cell lung cancer overall, and 2.89 (95% CI: 1.53–5.45) for small cell lung cancer. After applying the semi-Bayes shrinkage approach with informative priors from the literature, the ORs were 1.24 (95% CI 1.16–1.31) for all histological types combined, 1.27 (95% CI 1.15–1.41) for adenocarcinoma overall, 1.47 (95% CI 1.02–2.14) for AIS/MIA, 1.38 (95% CI 1.05–1.82) for squamous cell carcinoma, 1.30 (95% CI 0.87–1.95) for large cell lung cancer, 1.23 (95% CI 1.15–1.30) for non-small cell lung cancer overall, and 1.97 (95% CI 1.30–3.00) for small cell lung cancer. The estimated magnitude of association with ETS exposure was greater for small cell lung cancer than for non-small cell lung cancer (ROR=2.08, 95% CI 1.10–3.94, P=0.024). In the combined analysis of the two Chinese studies, the pooled adjusted OR for ETS exposure on lung cancer risk was 1.46 (95% CI, 1.12–1.89).
Genetic Susceptibility Markers and Lung Cancer among Never Smokers. In the ILCCO pooled analysis, a positive association was observed between lung cancer susceptibility and the G allele in the DNA repair gene polymorphism OGG1 S326C among those exposed to ETS (CG+GG vs. CC, Bayesian posterior OR=1.55, 95% CI 1.04–2.32). In addition, ERCC2/XPD D312N was also associated with an increased risk of lung cancer among those exposed to ETS (AA vs. GG, Bayesian posterior OR=1.48, 95% CI 1.01–2.16).
In the combined analysis of the Jiangsu Four Cancers Study and the Taiyuan Air Pollution and Lung Cancer Study, associations with lung cancer were observed for CTTNB1 rs2953, RAN rs14035, TP53INP1 rs7760, TP53INP1 rs896849, EPCAM rs1126497, HEY1 rs1046472, HEY2 rs3734637, OCT4 rs13409, and WNT2 rs3729629.
Interactions between ETS and Genetic Susceptibility Factors. For the ILCCO pooled analysis, additive interactions were observed between TP53 R72P and ETS on lung cancer. For the pooled Chinese case-control studies, there were statistical interactions between ETS exposure and GEMIN4 rs7813 (multiplicative and additive), WNT2B rs2273368 (multiplicative and additive), pre-miR-a46 rs2910164 (multiplicative), AXIN rs1981492 (multiplicative), and WNT8A rs4835761 (multiplicative).
Discussion and Conclusions
Our results confirm the role of ETS exposure in the development of lung cancer. Furthermore, the strengths of the associations vary by histological type and the association is stronger for small cell lung cancer than other histological types. This is the first large-scale collaborative study on the gene-environmental interactions between polymorphisms of DNA repair, carcinogen metabolism, cell cycle control, miRNA, and stem cell regulation genes and lung cancer susceptibility in never smokers. Our results add to the body of evidence demonstrating that polymorphisms of these genes affect lung cancer development in never smokers, and also suggest that some of these SNPs interact with ETS either multiplicatively or additively.
Public Health Implications
Identification of lung cancer susceptibility genes may aid in personalized risk prediction. A better understanding of the genetic, environmental, and behavioral risk factors for lung cancer in never smokers would help to identify those who need to be targeted for preventive interventions against lung cancer.