Identification of Biomarkers Based on Starvation Response‐Related Genes for Assessing the Immune Profile and Prognosis in Lung Adenocarcinoma

<b>Background:</b> Lung adenocarcinoma (LUAD) exhibits a high recurrence rate and an unfavorable prognosis. The role of the starvation-induced tumor microenvironment (TME), which is closely linked to metabolism, remains poorly understood in LUAD. <b>Methods:</b> LUAD patient samples were collected from public databases, and starvation response-related genes (SRRGs) were acquired from the MSigDB database. As starvation response may enhance autophagy in cancer cells, the Human Autophagy Database (HADb) was accessed to collect autophagy-related genes (ARGs). Next, the association between the expressions of ARGs and SRRGs was analyzed applying Pearson's algorithm. The SRRG score was calculated by GSVA package for each sample, and WGCNA package was utilized to screen SRRG-related module genes. Differentially expressed lncRNAs between LUAD and control samples were screened by the limma package. Subsequently, the lncRNAs associated with SRRG-related module genes were intersected with differentially expressed lncRNAs to obtain key SRRG-correlated lncRNAs. The number of key lncRNAs in the risk model was optimized by performing univariate and multivariate Cox regression analyses. Next, immune profiling between different LUAD risk groups was conducted using single-sample GSEA (ssGSEA), MCP-counter, and ESTIMATE algorithms. The Mutect2 software and clusterProfiler R package were employed to analyze the mutation profiles and pathway enrichment of patients in different risk groups, respectively. In addition, the expressions of key lncRNAs in LUAD cells were verified by qRT-PCR, and the migratory and invasive capabilities of the cells were measured by wound healing and transwell assays. <b>Results:</b> We identified 162 potential SRRGs linked to autophagy, glycolysis, and starvation responses. In addition, 102 candidate SRRG-related lncRNAs were selected from SRRG-related module genes. Three key SRRG-related lncRNAs (AC023421.1, AL034397.3, and LINC01537) were screened for developing an accurate risk model. Notably, the high-risk group showed a significantly higher mutation rate and oncogenic pathway scores and markedly worse immune infiltration and overall survival (OS). In vitro experiments revealed that LINC01537 was highly expressed in A549 cells, and that after knockdown of LINC01537, the migration and invasion of LUAD cells were suppressed. <b>Conclusion:</b> This study identified three key lncRNAs related to starvation response and created a risk model that can accurately assess the prognosis and immune characteristics in LUAD, offering novel biomarkers and a theoretical basis for the precision immunotherapy and targeted intervention in LUAD.