The tumor immune microenvironment of lung adenocarcinoma (LUAD) is characterized by pronounced heterogeneity; however, practical tools capable of simultaneously dissecting its immunobiological characteristics and predicting therapeutic response remain scarce. Given the close association between immune infiltration-related genes and LUAD tumorigenesis, progression, and immunotherapy efficacy, this study aimed to construct an immune infiltration-related gene signature and develop an integrated model that incorporates prognostic indicators, molecular subtyping, and immunotherapy efficacy prediction.

We obtained RNA-sequencing data and clinical information for LUAD cohorts from the TCGA and GEO databases, and retrieved 547 immune infiltration-related genes from the CIBERSORTx platform. Using differential expression analysis, Cox regression, and least absolute shrinkage and selection operator (LASSO) regression, we constructed a prognostic model based on immune infiltration-related genes and validated it across multiple independent cohorts. We further compared differences between high and low-risk groups in molecular pathways, tumor stemness features, the immune microenvironment, tumor mutational burden, immunotherapy responsiveness, and drug sensitivity, and performed in vitro cellular experiments to functionally validate key genes.

We developed and validated a prognostic model comprising 13 immune infiltration-related genes, and its risk score emerged as an independent prognostic factor in multivariable Cox analysis. LUAD patients were stratified into two distinct risk groups: the high-risk group was characterized by enhanced cell-cycle and DNA-replication pathways, increased stemness features, higher tumor mutational burden, and an immunosuppressive microenvironment, whereas the low-risk group showed stronger immune infiltration, higher immune checkpoint expression, and higher predicted responsiveness to immunotherapy. In vitro functional assays confirmed that SKA1 exerts an oncogenic role in LUAD, significantly enhancing LUAD cell proliferation, migration, and invasion.

We developed a robust prognostic model based on immune infiltration-related genes that effectively predicts the prognosis and immunotherapy response of patients with LUAD, reveals two distinct immuno-biological phenotypes, and deepens our understanding of LUAD heterogeneity. This model provides a key molecular basis for LUAD molecular subtyping and for formulating individualized therapeutic strategies.