BACKGROUND: DNA methylation and gene expression are promising biomarkers of various cancers, including non-small cell lung cancer (NSCLC). Besides the main effects of biomarkers, the progression of complex diseases is also influenced by gene-gene (GÂG) interactions.RESEARCH QUESTION: Would screening the functional capacity of biomarkers on the basis of main effects or interactions, using multiomics data, improve the accuracy of cancer prognosis? STUDY DESIGN AND METHODS: Biomarker screening and model validation were used to construct and validate a prognostic prediction model. NSCLC prognosis-associated biomarkers were identified on the basis of either their main effects or interactions with two types of omics data. A prognostic score incorporating epigenetic and transcriptional biomarkers, as well as clinical information, was independently validated.RESULTS: Twenty-six pairs of biomarkers with GÂG interactions and two biomarkers with main effects were significantly associated with NSCLC survival. Compared with a model using clinical information only, the accuracy of the epigenetic and transcriptional biomarker-based prognostic model, measured by area under the receiver operating characteristic curve (AUC), increased by 35.38% (95% CI, 27.09%-42.17%; P ¼ 5.10 Â 10 -17 ) and 34.85% (95% CI, 26.33%-41.87%; P ¼ 2.52 Â 10 -18 ) for 3-and 5-year survival, respectively, which exhibited a superior predictive ability for NSCLC survival (AUC 3 year , 0.88 [95% CI, 0.83-0.93]; and AUC 5 year , 0.89 [95% CI, 0.83-0.93]) in an independent Cancer Genome Atlas population. GÂG interactions contributed a 65.2% and 91.3% increase in prediction accuracy for 3-and 5-year survival, respectively. INTERPRETATION:The integration of epigenetic and transcriptional biomarkers with main effects and GÂG interactions significantly improves the accuracy of prognostic prediction of early-stage NSCLC survival.

Hypoxia occurs frequently in human cancers and promotes stabilization and activation of hypoxia inducible factor (HIF). HIF-1α is specific for the hypoxia response, and its degradation mediated by three enzymes EGLN1, EGLN2 and EGLN3. Although EGLNs expression has been found to be related to prognosis of many cancers, few studies examined DNA methylation in EGLNs and its relationship to prognosis of early-stage non-small cell lung cancer (NSCLC). We analyzed EGLNs DNA methylation data from tumor tissue samples of 1,230 early-stage NSCLC patients, as well as gene expression data from The Cancer Genome Atlas. The sliding windows sequential forward feature selection method and weighted random forest were used to screen out the candidate CpG probes in lung adenocarcinomas (LUAD) and lung squamous cell carcinomas patients, respectively, in both discovery and validation phases. Then Cox regression was performed to evaluate the association between DNA methylation and overall survival. Among the 34 CpG probes in EGLNs, DNA methylation at cg25923056 EGLN2 was identified to be significantly associated with LUAD survival (HR = 1.02, 95% CI: 1.01-1.03, P = 9.90 × 10 -5 ), and correlated with EGLN2 expression (r = -0.36, P = 1.52 × 10 -11 ). Meanwhile, EGLN2 expression was negatively correlated with HIF1A expression in tumor tissues (r = -0.30, P = 4.78 × 10 -8 ) and significantly (P = 0.037) interacted with HIF1A expression on overall survival. Therefore, DNA methylation of EGLN2-HIF1A is a potential marker for LUAD prognosis and these genes are potential treatment targets for further development of HIF-1α inhibitors in lung cancer therapy.

Background Body mass index (BMI) has been found to be associated with a decreased risk of non-small cell lung cancer (NSCLC); however, the effect of BMI trajectories and potential interactions with genetic variants on NSCLC risk remain unknown. Methods Cox proportional hazards regression model was applied to assess the association between BMI trajectory and NSCLC risk in a cohort of 138,110 participants from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial. One-sample Mendelian randomization (MR) analysis was further used to access the causality between BMI trajectories and NSCLC risk. Additionally, polygenic risk score (PRS) and genome-wide interaction analysis (GWIA) were used to evaluate the multiplicative interaction between BMI trajectories and genetic variants in NSCLC risk. Results Compared with individuals maintaining a stable normal BMI (n = 47,982, 34.74%), BMI trajectories from normal to overweight (n = 64,498, 46.70%), from normal to obese (n = 21,259, 15.39%), and from overweight to obese (n = 4,371, 3.16%) were associated with a decreased risk of NSCLC (hazard ratio [HR] for trend = 0.78, P < 2×10−16). An MR study using BMI trajectory associated with genetic variants revealed no significant association between BMI trajectories and NSCLC risk. Further analysis of PRS showed that a higher GWAS-identified PRS (PRSGWAS) was associated with an increased risk of NSCLC, while the interaction between BMI trajectories and PRSGWAS with the NSCLC risk was not significant (PsPRS= 0.863 and PwPRS= 0.704). In GWIA analysis, four independent susceptibility loci (P < 1×10−6) were found to be associated with BMI trajectories on NSCLC risk, including rs79297227 (12q14.1, located in SLC16A7, Pinteraction = 1.01×10−7), rs2336652 (3p22.3, near CLASP2, Pinteraction = 3.92×10−7), rs16018 (19p13.2, in CACNA1A, Pinteraction = 3.92×10−7), and rs4726760 (7q34, near BRAF, Pinteraction = 9.19×10−7). Functional annotation demonstrated that these loci may be involved in the development of NSCLC by regulating cell growth, differentiation, and inflammation. Conclusions Our study has shown an association between BMI trajectories, genetic factors, and NSCLC risk. Interestingly, four novel genetic loci were identified to interact with BMI trajectories on NSCLC risk, providing more support for the aetiology research of NSCLC. Trial registration http://www.clinicaltrials.gov, NCT01696968.