Identification of a novel signature to prognostic stratification and integrative analyses in lung adenocarcinoma

Background: Recently, research has revealed that Golgi Apparatus involves in the development process of cancer, however, the specific effect of Golgi Apparatus genes (GAG) in lung adenocarcinoma (LUAD) remains unclear. This study aims to construct a more concise and practical risk model in LUAD using GAG.

Methods: The gene expression profiles of patients with LUAD were downloaded from the TCGA (The Cancer Genome Atlas) and GEO (Gene Expression Omnibus) databases, and GAGs were downloaded from the GSEA (Gene Set Enrichment Analysis) database. Univariate Cox and LASSO (Least Absolute Shrinkage and Selection Operator) analyses were performed to identify the prognostic GAGs’ signature. Kaplan-Meier and receiver operating characteristic (ROC) curves were plotted to validate the predictive effect of the prognostic signatures. The correlation between the risk model and the immune landscape was examined using CIBERSORT and TIME analyses. Also, the genes of signature were also assessed by single cell analysis (ScRNA).

Results: A prognostic signature comprising 5 GAG genes (GNPNAT1, RGS20, CAV3, NTSR1 and FURIN) was established after LASSO and multi-Cox analyses. Both the Kaplan‑Meier analysis and the ROC curves supported the strong predictive utility of the risk model. Specifically, the former yielded significant stratification in all three validation datasets (p = 1.2001e−05, p = 0.0057, and p = 0.0359), while the latter provided further evidence of its predictive precision through the area under the curve. In addition, we found that the low-risk group responded better to immunotherapy than high-risk group (p<0.0001). ScRNA analysis revealed that distribution patterns of the 5 GAG genes in cells. Finally, we assessed the situation of TMB (Tumor Mutation Burden) and performed the functional analysis based on risk model of GAGs.

Conclusions: The risk model based on GAGs can effectively stratify the prognosis of patients and predict the immunotherapy responses in LUAD.