Integrated hypoxia and metabolic gene signature to predict clinical outcomes and therapeutic vulnerabilities in lung adenocarcinoma.

BACKGROUNDImmune checkpoint inhibitors (ICIs) have transformed the therapeutic landscape of advanced non-small cell lung cancer (NSCLC), yet ~50% of lung adenocarcinoma (LUAD) patients exhibit primary or acquired resistance, underscoring the urgent need for predictive biomarkers. Hypoxia and metabolic reprogramming (HMR) are intertwined oncogenic hallmarks that reshape the tumor microenvironment (TME) and drive immune evasion, representing promising targets for signature development. Herein, we constructed an integrated HMR gene signature to predict clinical outcomes and therapeutic vulnerabilities in LUAD.

METHODSTranscriptomic data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) were analyzed. Hypoxia and energy metabolism pathways were scored via single-sample gene set enrichment analysis (ssGSEA), and subtypes were stratified. A 9-gene prognostic model was constructed using least absolute shrinkage and selection operator (LASSO) regression and validated across cohorts. Single-cell RNA (scRNA) data and in-house experimental validation explored associations between gene expression, immune microenvironment, and treatment outcomes.

RESULTSWe developed a 9-gene HMR signature (RHOV, CYP4B1, KRT6A, TNS4, HMGA2, FIGF, SFTPB, TMPRSS11E, GRIA1) incorporating hypoxia and energy metabolism pathway activity. This signature robustly stratified patients into high- and low-risk groups with distinct overall survival [OS; hazard ratio (HR) =2.24, P<0.001]. High-risk patients exhibited enhanced immune evasion, characterized by elevated Tumor Immune Dysfunction and Exclusion (TIDE) scores (P<0.001) and a suppressed antitumor immune landscape, whereas distinct immune checkpoint expression patterns and subtype-specific drug sensitivities were observed between risk groups. scRNA sequencing revealed RHOV and TNS4 co-expressed with exhaustion markers (e.g., FOXP3, BTLA) and were enriched in non-major pathological response (non-MPR) patients. Validation in 39 ICI-treated NSCLC patients via quantitative reverse transcription polymerase chain reaction (qRT-PCR) and immunofluorescence (IF) confirmed that RHOV+ regulatory T cells and TNS4+ exhausted T cells were more abundant in poor responders.

CONCLUSIONSOverall, our integrated HMR signature exhibits robust prognostic and predictive value, unravels TME-driven ICI resistance mechanisms, and identifies subtype-specific therapeutic vulnerabilities. This tool has potential for clinical application, as it links HMR to clinical outcomes and provides novel insights to guide precision treatment strategies for LUAD patients.