Construction of a prognostic model and immune landscape analysis for lung adenocarcinoma utilizing a bioinformatics signature derived from acrolein-related genes, and identification of associated subtypes

Lung adenocarcinoma (LUAD) is driven by the interplay between inherited genetic factors and environmental exposures. Acrolein, a ubiquitous pollutant and major cigarette smoke constituent, induces DNA damage driving pulmonary carcinogenesis. However, acrolein exposure-related genes (AERGs) and their role in LUAD prognosis and immune modulation remain poorly characterized. A bioinformatics prognostic model was constructed leveraging an eight-gene signature derived from previously reported AERGs. A series of analytical steps were then performed: differential expression using limma, followed by univariate and Least Absolute Shrinkage and Selection Operator Cox regression to address multicollinearity, culminating in the identification of eight core genes via multivariate Cox regression for risk score construction. CIBERSORT and Single-sample Gene Set Enrichment Analysis algorithms were used to evaluate immune microenvironment heterogeneity, and molecular subtypes were characterized via non-negative matrix factorization clustering. Prognostic utility was further assessed with a nomogram, and potential therapeutic agents were predicted through the CellMiner database. In this study, eight core AERGs were identified and used to construct a highly robust prognostic model for LUAD, which demonstrated satisfactory accuracy in survival prediction through both internal and external validation. Based on this gene signature, two molecular subtypes with distinct immune profiles were identified. The high-risk subtype exhibited a “cold tumor” phenotype, characterized by enrichment of M0 macrophages, reduced immune scores, and enhanced immune evasion potential. Conversely, the low-risk subtype exhibited a more immunologically active profile. Notably, the high-risk group also featured higher tumor mutational burden and enhanced sensitivity to conventional chemotherapies, including cisplatin and docetaxel. Integrating multi-omics data, this study established an AERG-based prognostic model for LUAD and systematically evaluated its potential for outcome prediction and personalized immunotherapy guidance. A prognostic model for lung adenocarcinoma was constructed based on eight key acrolein exposure-related genes, demonstrating high robustness and accuracy in predicting survival risk. High-risk scores were significantly associated with a “cold tumor” phenotype, increased immune evasion potential, and reduced immune scores, whereas the low-risk group exhibited more active antitumor immune infiltration. Two molecular subtypes identified via non-negative matrix factorization (NMF) clustering displayed significant differences in genomic mutation burden, drug sensitivity, and immunotherapeutic response.