Combining autophagy and immune characterizations to predict prognosis and therapeutic response in lung adenocarcinoma.

Abstract

3140 Background: As a key regulator of programmed cell death, autophagy is critical for maintaining the stability of the intracellular environment. Increasing evidences have found that the clinical importance of the interaction between autophagy and immune status in LUAD. Reliable prognostic signatures based on combination of autophagy and immune status have not been well-established. We aimed to explore the potential autophagy-immune-derived biomarkers to predict prognosis and therapeutic response in lung adenocarcinoma. Methods: Patients from GSE72094 dataset were randomized 7:3 to a training set or an internal validation set. Three independent cohorts, TCGA, GSE31210 and GSE37745, were used as the external verification. Unsupervised hierarchical clustering was used to identify the autophagy-associated and immune-associated molecular patterns for LUAD based on autophagy-genes and immune-genes. The LASSO analysis, univariate and multivariate cox regression analysis were performed to filtrate significant prognostic autophagy-immune-based genes, followed with model construction and patient stratification. Tumor immune microenvironment and functional pathways were investigated. The potential therapeutic responses were explored by GDSC database, TIDE algorithm, and immunotherapy clinical cohorts. Results: We found that autophagy cluster A had the better survival prognosis (p < 0.001) and high immune status (p < 0.001) was identified as favorable factors for patients’ overall survival. We merged autophagy and immune subtype into a two-dimensional index to characterize the combined prognostic classifier 535 genes were defined as autophagy-immune-related DEGs. Four genes (C4BPA, CD300LG, CD96, and S100P) were identified to construct the autophagy-immune-related prognostic risk model. Survival analysis and receiver operating characteristic curve showed significant prognostic efficacy. Through ssGSEA and CIBERSORT analysis, the majority of immune infiltrating cells were shown to be enriched in the low-risk group. What’s more, the expression of crucial immune checkpoint molecules, such as PD-1, PD-L1 and CTLA-4, was observed highest in low-risk group (p < 0.001). TIDE and immunotherapy clinical cohorts’ analysis showed that low-risk group was predicted with more potential responders to immunotherapy. In addition, there are different patterns of autophagy between low- and high- risk patients. GO, KEGG and GSEA function analysis focus on cell cycle, MAPK, apoptosis, MTORC1 and selective autophagy pathway. Docetaxel, rapamycin and sorafenib may be the potential drugs candidate in high-risk group (p < 0.01). Conclusions: In summary, the autophagy-immune-based gene signature represents a promising tool for risk stratification tool in lung adenocarcinoma, which can regard individualized treatment and follow-up scheduling for patients.