Establishment and validation of a novel anoikis-related prognostic signature of lung squamous cell carcinoma.

BACKGROUND

Lung squamous cell carcinoma (LUSC) remains a highly lethal malignancy due to its propensity for metastasis and the limited availability of reliable prognostic biomarkers. Anoikis, a detachment-induced cell death mechanism, plays dual roles in tumor suppression and metastasis, yet its prognostic implications in LUSC are poorly characterized. This study aimed to develop an anoikis-related gene (ARG) signature for risk stratification and therapeutic exploration in LUSC.

METHODS

Transcriptomic data from The Cancer Genome Atlas (TCGA)-LUSC (n=496) and Gene Expression Omnibus (GEO) cohorts (GSE3141/GSE73403; n=121) were analyzed. Differentially expressed ARGs were identified using limma package. Prognostic genes were screened via univariate Cox regression and Least Absolute Shrinkage and Selection Operator (LASSO) regression to construct a risk model. Immune infiltration was evaluated using CIBERSORT, single-sample Gene Set Enrichment Analysis (ssGSEA), and ESTIMATE. Experimental validation included CSNK2A1 knockdown in LUSC cell lines (SK-1, H226) to assess migration, invasion, and proliferation. Drug sensitivity and molecular docking were performed using oncoPredict and AutoDock Vina.

RESULTS

A total of 307 differentially expressed ARGs were identified, with 37 showing prognostic significance. Unsupervised clustering divided LUSC into two subgroups with distinct survival (P<0.001) and immune microenvironments. Cluster B exhibited higher immune and stromal scores but poorer outcomes, enriched in exhausted CD8+ T cells and M2 macrophages. A six-gene risk signature (AKT2, CSNK2A1, CD151, TDGF1, BAG4, SNAI1) stratified patients into high-/low-risk groups, validated across cohorts (5-year AUC: 0.642-0.728). High-risk patients showed sensitivity to Wee1 inhibitors (P<0.001). CSNK2A1 knockdown suppressed migration, invasion, and proliferation in vitro. A nomogram integrating risk score and T-stage improved prognostic accuracy [5-year area under the curve (AUC): 0.728].