DNA Methylation Signatures Associated With Smoking Status Across Lung Cancer Subtypes Using The Cancer Genome Atlas (TCGA) Methylation Data

Background: Lung cancer remains the leading cause of cancer-related mortality worldwide, with cigarette smoking as its strongest etiologic factor. Beyond genetic mutations, smoking induces epigenetic alterations, particularly DNA methylation, which may differ across lung cancer subtypes and contribute to tumor heterogeneity. Objective: The objective of this study was to identify smoking-associated differentially methylated CpG sites (FDR < 0.05) across lung cancer subtypes using The Cancer Genome Atlas (TCGA) 450K methylation data, and to enhance biological interpretation by annotating their genomic context and functional relevance, including associated genes and regulatory features. Methodology: This retrospective observational study analyzed Illumina HumanMethylation450 BeadChip data from TCGA lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) cohorts. Multivariable linear models within the limma framework assessed differential methylation between smokers and never smokers, adjusting for age, sex, tumor stage, and histology. False discovery rate (FDR) correction was applied. Results: A total of 771 patients (92 never smokers, 679 smokers) were analyzed. Seven CpG sites were significantly associated with smoking status (FDR < 0.05), with absolute methylation differences (Δβ) ranging from 0.100 to 0.123. Subtype-specific analyses identified eight LUAD-specific and five LUSC-specific smoking-associated CpG sites. LUAD exhibited both hypermethylation and hypomethylation patterns, whereas LUSC showed predominantly smoking-related hypomethylation. Conclusion: Smoking is associated with distinct and subtype-specific DNA methylation signatures in lung cancer, underscoring its role in epigenetic tumor modulation and supporting the potential utility of methylation markers in lung cancer stratification and risk assessment.