Prognosis signature for predicting the survival and immunotherapy response in esophageal carcinoma based on cellular senescence-related genes

Yue Wang; Longfei Dai; Ran Huang; Weisong Li; Wenyong Wu

doi:10.3389/fonc.2023.1203351

Prognosis signature for predicting the survival and immunotherapy response in esophageal carcinoma based on cellular senescence-related genes

Front Oncol. 2023 Aug 17:13:1203351. doi: 10.3389/fonc.2023.1203351. eCollection 2023.

Authors

Yue Wang^{1

2

3

4}, Longfei Dai⁵, Ran Huang⁴, Weisong Li⁴, Wenyong Wu^{1

2

3}

Affiliations

¹ Anhui No.2 Provinicial People's Hospital Clinical College of Anhui Medical University, Hefei, China.
² Department of General Surgery, Anhui No.2 Provinicial People's Hospital, Hefei, China.
³ The Fifth Clinical Medical College of Anhui Medical University, Hefei, China.
⁴ Department of Pediatric Surgery, The First Affiliated Hospital of Anhui Medical University, Hefei, China.
⁵ Department of General Surgery, The First Affiliated Hospital of Anhui Medical University, Hefei, China.

Abstract

Background: Cellular senescence occurs throughout life and can play beneficial roles in a variety of physiological processes, including embryonic development, tissue repair, and tumor suppression. However, the relationship between cellular senescence-related genes (CSRGs) and immunotherapy in esophageal carcinoma (ECa) remains poorly defined.

Methods: The data set used in the analysis was retrieved from TCGA (Research Resource Identifier (RRID): SCR_003193), GEO (RRID: SCR_005012), and CellAge databases. Data processing, statistical analysis, and diagram formation were conducted in R software (RRID: SCR_001905) and GraphPad Prism (RRID: SCR_002798). Based on CSRGs, we used the TCGA database to construct a prognostic signature for ECa and then validated it in the GEO database. The predictive efficiency of the signature was evaluated using receiver operating characteristic (ROC) curves, Cox regression analysis, nomogram, and calibration curves. According to the median risk score derived from CSRGs, patients with ECa were divided into high- and low-risk groups. Immune infiltration and immunotherapy were also analyzed between the two risk groups. Finally, the hub genes of the differences between the two risk groups were identified by the STRING (RRID: SCR_005223) database and Cytoscape (RRID: SCR_003032) software.

Results: A six-gene risk signature (DEK, RUNX1, SMARCA4, SREBF1, TERT, and TOP1) was constructed in the TCGA database. Patients in the high-risk group had a worse overall survival (OS) was disclosed by survival analysis. As expected, the signature presented equally prognostic significance in the GSE53624 cohort. Next, the Area Under ROC Curve (AUC=0.854) and multivariate Cox regression analysis (HR=3.381, 2.073-5.514, P<0.001) also proved that the risk signature has a high predictive ability. Furthermore, we can more accurately predict the prognosis of patients with ECa by nomogram constructed by risk score. The result of the TIDE algorithm showed that ECa patients in the high-risk group had a greater possibility of immune escape. At last, a total of ten hub genes (APOA1, MUC5AC, GC, APOA4, AMBP, FABP1, APOA2, SOX2, MUC8, MUC17) between two risk groups with the highest interaction degrees were identified. By further analysis, four hub genes (APOA4, AMBP, FABP1, and APOA2) were related to the survival differences of ECa.

Conclusions: Our study reveals comprehensive clues that a novel signature based on CSRGs may provide reliable prognosis prediction and insight into new therapy for patients with ECa.

Keywords: bioinformatic analysis; cellular senescence; esophageal carcinoma; immunotherapy; prognosis signature.