Abstract
Intrahepatic cholangiocarcinoma (ICCA) is a highly aggressive bile duct cancer with limited treatment options and poor prognosis. This study aims to develop a gene signature-based model to predict ICCA patient survival. We identified differentially expressed genes (DEGs) from an ICCA mRNA microarray dataset and applied survival, Cox, and LASSO regression analyses to select key genes for constructing the GPSICCA risk score model. The model was then validated in two additional ICCA cohorts, and its relationship with the tumor microenvironment (TME) was analyzed. Additionally, multiplex fluorescent immunohistochemistry (mfIHC) was conducted to examine the expression of key genes in ICCA samples. The expression of COL4A1, GULP1, ITGA6, and STC1 was used to construct the gene set-based prognostic signature for ICCA (GPSICCA) risk score model, and mfIHC confirmed their expression in ICCA samples. The model effectively stratified patients into high- and low-risk groups, with high-risk patients showing significantly worse survival. The GPSICCA score was positively correlated with stromal and immune scores, suggesting its predictive capability is related to TME involvement in ICCA. The GPSICCA model is a strong predictor of survival in ICCA patients and could help guide treatment strategies and identify potential therapeutic targets.