Integrative Genomic Profiling and Biomarker Discovery for Early Detection of Lung Adenocarcinoma in Smokers and Non-Smokers

Abstract

Objective: This study aims to develop an integrated approach to the genomics of lung adenocarcinoma (LUAD) for earlier detection in both smokers and non-smokers by identifying differential molecular and predictive biomarker signatures. Tobacco exposure and the molecular heterogeneity of LUAD motivate this research. It aims to uncover subtype-specific pathways and precision targets that enhance early-stage diagnostics. Methods: We performed high-throughput RNA-Seq, whole-exome sequencing (WES), and methylation arrays on tumor and adjacent normal tissues from 120 LUAD patients with balanced smoking histories. Biosinformatic pipelines incorporating differential expression analysis, somatic variant calling, pathway enrichment, and integrative clustering were applied. Candidate biomarkers were validated in TCGA-LUAD datasets as well as qRT-PCR in an independent validation cohort. Predictive performance was evaluated using ROC analysis and classifier systems based on machine learning frameworks. Results: An integrative analysis found a total of 136 genes were significantly dysregulated between smokers and non-smokers with an FDR of < 0.01. These genes were enriched in pathways of immune modulation, xenobiotic metabolism, and DNA damage response. The five-gene biomarker panel, consisting of TP63, CYP1B1, GPR15, SFTPB, and LINC00472, demonstrated considerable discriminatory ability between early-stage LUAD and standard samples, with an AUC of 0.91 for smokers and 0.87 for non-smokers. The inclusion of methylation markers in RASSF1A and SHOX2 increased the classifier's sensitivity when applied to a multi-omics logistic regression model, attaining 93.5% sensitivity and 89.2% specificity in the validation cohort. Conclusion: This study demonstrates a comprehensive genomic strategy developed for the early detection of LUAD, highlighting the pronounced molecular differences between tumors from smokers and non-smokers. The multi-omic markers identified in this study may aid in the development of non-invasive screening methods and precision diagnostics, thereby improving survival rates by enabling earlier intervention in vulnerable populations.