Multi‑cohort validation based on coagulation-related genes for predicting prognosis of esophageal squamous cell carcinoma

ZhongQuan Yi^1*Rui Wang^1,2Weisong Zhang¹Xia Li¹Hao Wang¹Yanan Ji¹Jing Zhao¹Jianxiang Song¹

• ¹Yancheng Third People's Hospital, Yancheng, China
• ²The First Affiliated Hospital With Nanjing Medical University, Nanjing, China

Abstract Objective: In malignant tumors, a hypercoagulable state is frequently observed and is intricately intertwined with cancer development and the remodeling of the immune microenvironment. Recently, the coagulation-related genes (CRGs) signature has emerged as highly significant for the prognosis and immunotherapy of patients with various cancers. Nevertheless, their application in esophageal squamous cell carcinoma (ESCC) remains uninvestigated. Here, our objective is to explore the role of the CRGs signature in forecasting prognosis and predicting patient’s response to immunotherapy. Methods: According to the prognostic CRGs, consensus clustering was utilized to stratify ESCC patients in the GSE53625 cohort into two subgroups. Subsequently, difference analysis and univariate cox analysis were conducted on the two subgroups, and a CRGs signature was constructed by leveraging these genes. Next, multiple RNA transcriptome cohorts were utilized to validate the signature. Moreover, functional enrichment, tumor mutation burden (TMB), tumor infiltration, immune function, and immunotherapy response of this signature were investigated. Results: A CRGs signature composed of six genes (PTX3, CILP, CFHR4, SULT1B1, IL5RA, and FAM151A) was constructed. This signature serves as an independent and reliable prognostic factor. Additionally, when compared with the 32 prognostic signatures previously reported, the CRGs signature exhibited superior performance in the ESCC prognostic cohorts. Additionally, we found that high-risk ESCC exhibited higher immune infiltration, lower TMB, higher TIDE, and a lower proportion of immunotherapy response. In vitro experiments have shown that the gene SULT1B1, which exhibits the highest accuracy in predicting tumor status, significantly inhibited the proliferation and metastasis. Conclusions: We constructed and validated a robust CRGs signature. Moreover, as one of the model CRGs, the tumor-suppressive role of SULT1B1 in ESCC was experimentally verified in vitro. These results provide novel insights into enhancing the prognosis of ESCC and formulating treatment strategies.