Objective To construct and validate a prognostic prediction model for patients with intrahepatic cholangiocarcinoma (ICC) across all disease stages to facilitate risk stratification.

Methods Clinical data from 117 patients with ICC treated at Lanzhou University Second Hospital between January 2019 and June 2024 were retrospectively collected. Participants were assigned into the modeling and validation groups based on the time of diagnosis. Independent prognostic factors were identified in the modeling group using Cox regression analysis, and a predictive model was developed based on the minimum Akaike Information Criterion (AIC). A nomogram was subsequently constructed. Model performance was assessed using bootstrap resampling and internal validation within the modeling group, evaluated by the concordance index (C-index), calibration curves, and decision curve analysis (DCA). Patients were stratified according to their model-derived scores to compare survival outcomes across different treatment regimens within each risk stratum.

Results Multivariate Cox regression analysis and AIC optimization (AIC=540.48) identified the following variables for inclusion in the ICC prognostic prediction model: age, Child–Pugh classification, CA19-9 level, maximum tumor diameter, organ metastasis status, and number of tumors. The C-index for the modeling group was 0.716 (95% confidence interval C: 0.655–0.776), with an average C-index of 0.716 based on internal validation using 1,000 Bootstrap resamples. In the validation group, the C-index was 0.694 (95%CI: 0.538–0.849), indicating good discriminative ability. Calibration curves for both groups demonstrated strong agreement between predicted and observed outcomes. DCA indicated that the model provides clinical net benefit. Risk stratification further revealed significant survival differences between low- and high-risk groups, as well as among patients receiving different treatment regimens within the same risk strata (P<0.001).

Conclusions The prognostic prediction model developed in this study for patients with ICC across all disease stages demonstrates robust performance and potential clinical utility. However, interpretations regarding treatment differences within similar risk strata should be approached cautiously and require further validation.