Perineural invasion (PNI) contributes to tumor progression, but the impact of the tumor microenvironment on immunotherapy outcomes in lung adenocarcinoma (LUAD) has not been elucidated. This study aims to develop a prognostic model based on published neural-invasion-related genes (NIRGs) and explore the role of a candidate gene, cyclin-dependent kinase 1 (CDK1), identified through integrative analysis in regulating the immune microenvironment of LUAD. Differential expression data of NIRGs and clinical information were sourced from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Multiple machine learning algorithms were applied to construct a predictive model. To further elucidate the impact of model genes on the immune microenvironment in LUAD, a series of analyses were conducted, including single-cell RNA sequencing (scRNA-seq), spatial transcriptomic (ST) analyses, Gene Set Enrichment Analysis (GSEA) and molecular docking. A prognostic risk model based on neural-invasion-related genes was constructed using StepCox (forward) + RSF method. It yielded favorable time‑dependent AUCs (> 0.75 for 1‑, 3‑, 5‑year survival) in the TCGA training cohort and acceptable predictive efficacy across independent GEO validation cohorts. CDK1 was identified by intersecting MEbrown module genes from WGCNA, prognostic model genes, and genes involved in the response to axon injury pathway. scRNA-seq and ST analyses showed CDK1 high cells were predominantly enriched in c5-macro-CCL21 macrophages subtype with spatial colocalization. Mechanistically, GSEA indicated that CDK1 was enriched in TGF-β signaling pathway, suggesting its potential role in modulating an immunosuppressive tumor microenvironment. Clinically, molecular docking analysis demonstrated high binding affinity (binding energy < − 7 kcal mol⁻¹) between CDK1 and drugs such as Trametinib and Axitinib, suggesting therapeutic relevance. CDK1, identified from a neural-invasion-related gene set, was associated with an immunosuppressive TME and may serve as a prognostic biomarker in LUAD.
No takes yet. Share an insight, caveat, or question.
Hu et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: