RESUMO
BACKGROUND: Lung adenocarcinoma (LUAD) is a lung cancer subtype with poor prognosis. We investigated the prognostic value of methylation- and homologous recombination deficiency (HRD)-associated gene signatures in LUAD. METHODS: Data on RNA sequencing, somatic mutations, and methylation were obtained from TCGA database. HRD scores were used to stratify patients with LUAD into high and low HRD groups and identify differentially mutated and expressed genes (DMEGs). Pearson correlation analysis between DMEGs and methylation yielded methylation-associated DMEGs. Cox regression analysis was used to construct a prognostic model, and the distribution of clinical features in the high- and low-risk groups was compared. RESULTS: Patients with different HRD scores showed different DNA mutation patterns. There were 272 differentially mutated genes and 6294 differentially expressed genes. Fifty-seven DMEGs were obtained; the top 10 upregulated genes were COL11A1, EXO1, ASPM, COL12A1, COL2A1, COL3A1, COL5A2, DIAPH3, CAD, and SLC25A13, while the top 10 downregulated genes were C7, ERN2, DLC1, SCN7A, SMARCA2, CARD11, LAMA2, ITIH5, FRY, and EPHB6. Forty-two DMEGs were negatively correlated with 259 methylation sites. Gene ontology and pathway enrichment analysis of the DMEGs revealed enrichment of loci involved in extracellular matrix-related remodeling and signaling. Six out of the 42 methylation-associated DMEGs were significantly associated with LUAD prognosis and included in the prognostic model. The model effectively stratified high- and low-risk patients, with the high-risk group having more patients with advanced stage disease. CONCLUSION: We developed a novel prognostic model for LUAD based on methylation and HRD. Methylation-associated DMEGs may function as biomarkers and therapeutic targets for LUAD. Further studies are needed to elucidate their roles in LUAD carcinogenesis.