Development and validation of an autophagy-related prognostic signature in esophageal cancer.

Background: Autophagy has a dual function in cancer, and its role in carcinogenesis of the esophagus remains poorly understood. In the present study, we explored the prognostic value of autophagy in esophageal cancer (ESCA), one of the leading causes of cancer-related deaths worldwide.

Methods: Using ESCA RNA-sequencing (RNA-Seq) data from 158 primary patients with ESCA, including esophageal adenocarcinoma and esophageal squamous cell carcinoma, were downloaded from The Cancer Genome Atlas (TCGA) for this study. We obtained differentially expressed autophagy-related genes (ARGs) by the "limma" package of R. The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genome (KEGG) analyses unveiled several fundamental signaling pathways associated with the differentially expressed ARGs in ESCA. Univariate Cox regression analyses were used to estimate associations between ARGs and overall survival (OS) in the TCGA ESCA cohort. A Cox proportional hazards model (iteration =1,000) with a lasso penalty was used to create the optimal multiple-gene prognostic signature utilizing an R package called "glmnet".

Results: A prognostic signature was constructed with four ARGs (, , and ) in the training set, which significantly divided ESCA patients into high- and low-risk groups in terms of OS [hazard ratio (HR) =1.508, 95% confidence interval (CI): 1.201-1.894, P<0.001]. In the testing set, the risk score remained an independent prognostic factor in the multivariate analyses (HR =1.572, 95% CI: 1.096-2.257, P=0.014). The area under the curve (AUC) of the receiver operating characteristic (ROC) predicting 1-year survival showed a better predictive power for the prediction model. The AUC in training and testing cohorts were 0.746 and 0.691, respectively. Therefore, the prognostic signature of the four ARGs was successfully validated in the independent cohort.

Conclusions: The prognostic signature may be an independent predictor of survival for ESCA patients. The prognostic nomogram may improve the prediction of individualized outcome. This study also highlights the importance of autophagy in the outcomes of patients with ESCA.