Our study aimed to develop an immune prognostic signature that could provide accurate guidance for the treatment of esophageal squamous cell cancer (ESCC). By implementing Single-Sample Gene Set Enrichment Analysis (ssGSEA), we established two ESCC subtypes (Immunity High and Immunity Low) in GSE53625 based on immune-genomic profiling of twenty-nine immune signature. We verified the reliability and reproducibility of this classification in the TCGA database. Immunity High could respond optimally to immunotherapy due to higher expression of immune checkpoints, including PD1, PDL1, CTLA4, and CD80. We used WGCNA analysis to explore the underlying regulatory mechanism of the Immunity High group. We further identified differentially expressed immune-related genes (CCR5, TSPAN2) in GSE53625 and constructed an independent two-gene prognostic signature we internally validated through calibration plots. We established that high-risk ESCC patients had worse overall survival (P=0.002, HR=2.03). Besides, high-risk ESCC patients had elevated levels of infiltrating follicle-helper T cells, naïve B cells, and macrophages as well as had overexpressed levels of some immune checkpoints, including B3H7, CTLA4, CD83, OX40L, and GEM. Moreover, through analyzing the Genomics of Drug Sensitivity in Cancer (GDSC) database, the high-risk group demonstrated drug resistance to some chemotherapy and targeted drugs such as paclitaxel, gefitinib, erlotinib, and lapatinib. Furthermore, we established a robust nomogram model to predict the clinical outcome in ESCC patients. Altogether, our proposed immune prognostic signature constitutes a clinically potential biomarker that will aid in evaluating ESCC outcomes and promote personalized treatment.
Keywords: cancer drug resistance; esophageal squamous cell carcinoma; immunotherapy; prognosis; signature; tumor microenvironment.
Copyright © 2021 Gao, Tang, Zhang and Li.