Identification of a Five Autophagy Subtype-Related Gene Expression Pattern for Improving the Prognosis of Lung Adenocarcinoma

Background: Autophagy plays an important role in lung adenocarcinoma (LUAD). We aimed to explore the autophagy-related gene (ARG) expression pattern and to identify promising autophagy-related biomarkers to improve the prognosis of LUAD. Methods: The gene expression profiles and clinical information of LUAD patients were downloaded from TCGA and validation cohort information was extracted from the Gene Expression Omnibus database. The HADb was used to extract ARGs. Gene expression data were analyzed in R software. Functional enrichment analysis was also performed for the DEARGs. Then, consensus clustering revealed autophagy-related tumor subtypes and DEGs were screened according to the subtypes. Next, the univariate Cox and multivariate Cox regression analyses were used to identify independent prognostic ARGs. After overlapping DEGs and the independent prognostic ARGs, the predictive risk model was established and validated. Correlation analysis between ARGs and clinicopathological variables were also explored. Finally, the TIMER and TISIDB databases were used to further explore the correlation analysis between immune cell infiltration levels in and the risk score as well as clinicopathological variables in the predictive risk model. Results: 222 genes were identified as ARGs, and 28 of the 222 genes were pooled as DEARGs. The most significant GO term was autophagy (p=3.05E-07), and KEGG analysis results indicated that 28 DEARGs were significantly enriched in the ErbB signaling pathway. Then consensus clustering analysis divided the LUAD into 2 clusters, 168 DEGs were identified according to cluster subtypes. Then univariate and multivariate Cox regression analyses were used to identify 12 genes serve as independent prognostic indicators. After overlapping 168 DEGs and 12 genes, 10 genes were selected for the further exploration of the prognostic pattern. Survival analysis results indicated that this risk model identified the prognosis. Five ARGs were screened as prognostic genes. Among of them, SPHK1 expression level were positively correlated with CD4+ T cells and dendritic cell infiltration levels. Conclusions: In this study, we constructed a predictive risk model and identified a five autophagy subtype-related gene expression pattern to improve the prognosis of LUAD. Understanding the subtypes of LUAD is helpful to accurately characterize the LUAD and develop personalized treatment.