Abstract
BACKGROUND: Colorectal cancer (CRC) is a common cancer and has a poor prognosis. The coagulation system and fibrinolysis system are closely related to the progression of malignant tumors and is also related to the immunotherapy of malignant tumors. Herein, we tried to predict survival and the immunotherapy effect for patients with CRC using a novel potential prognostic model. METHODS: Through online data of TCGA and GEO, we screened significantly differentially expressed genes (DEGs) to construct a prognostic model, followed by obtaining immune-related genes (IRGs) from the ImmPort database and coagulation- and fibrinolysis-related genes (CFRGs) from the GeneCards database. The predictive power of the model is assessed by Kaplan-Meier survival curves as well as the time-dependent ROC curve. Moreover, univariate and multivariate analyses were conducted for OS using Cox regression models, and the nomogram prognostic model was built. In the end, we further studied the possibility that CXCL8 was associated with immunocyte infiltration or immunotherapy effect and identified it by immunohistochemistry and Western blot. RESULTS: Five DEGs (CXCL8, MMP12, GDF15, SPP1, and NR3C2) were identified as being prognostic for CRC and were selected to establish the prognostic model. Expression of these genes was confirmed in CRC samples using RT-qPCR. Notably, those selected genes, both CFRGs and IRGs, can accurately predict the OS of CRC patients. Furthermore, CXCL8 is highly correlated with the tumor microenvironment and immunotherapy response in CRC. CONCLUSION: Overall, our established IRGPI can very accurately predict the OS of CRC patients. CXCL8 reflects the immune microenvironment and reveals the correlation with immune checkpoints among CRC patients.