Abstract
BACKGROUND: The extracellular matrix (ECM) plays a crucial role in the development and tumor microenvironment of lung adenocarcinoma (LUAD). This study aimed to establish a risk score of ECM-related genes in LUAD and explore the association between the risk score and patient survival as well as immune cell infiltration, somatic mutations, and therapy response. METHODS: Gene expression data from The Cancer Genome Atlas (TGCA) and eight Gene Expression Omnibus (GEO) databases were used to analyze and identify differentially expressed genes (DEGs). Prognostic ECM-related genes were identified and utilized to formulate a prognostic signature. A nomogram was constructed using TCGA dataset and validated in two GEO datasets. Differences between high- and low-risk patients were analyzed for function enrichment, immune cell infiltration, somatic mutations, and therapy response. Finally, Quantitative real-time PCR (qRT-PCR) was used to detect the mRNA expression of DEGs in LUAD. RESULTS: A risk score based on four ECM-related genes, ANOS1, CD36, COL11A1, and HMMR, was identified as an independent prognostic factor for overall survival (OS) compared to other clinical variables. Subsequently, a nomogram incorporating the risk score and TNM staging was developed using the TCGA dataset. Internal and external validation of the nomogram, conducted through calibration plots, C-index, time-dependent receiver operating characteristics (ROC), integrated discrimination improvement (IDI), and decision curve analyses (DCA), demonstrated the excellent discriminatory ability and clinical practicability of this nomogram. The risk score correlated with the distribution of function enrichment, immune cell infiltration, and immune checkpoint expression. More somatic mutations occurred in the high-risk group. The risk score also demonstrated a favorable ability to predict immunotherapy response and drug sensitivity. CONCLUSION: A novel signature based on four ECM-related genes is developed to help predict LUAD prognosis. This signature correlates with tumor immune microenvironment and can predict the response to different therapies in LUAD patients.