Establishment and validation of a nomogram model for preoperative prediction of the risk of cholangiocarcinoma with perineural invasion.

Objective: To establish and validate a nomogram model for predicting the risk of cholangiocarcinoma with perineural invasion.
Methods: We retrospectively collected the clinical data of 356 patients with surgically confirmed cholangiocarcinoma, including 98 cases of extrahepatic cholangiocarcinoma (eCCA), 197 cases of intrahepatic cholangiocarcinoma (iCCA), and 61 cases of perihilar cholangiocarcinoma (pCCA).
Results: Based on these data, we determined the influencing factors of preoperative perineural invasion risk in patients with cholangiocarcinoma by forward multivariate regression analysis. Based on these variables, we established two nomogram models. The model variables for predicting perineural invasion of eCCA included prothrombin time, high-density lipoprotein and tumor size (all P <0.05). The consistency index (C-index) of internal and external validation was 0.845 and 0.806, respectively. In addition, the model variables for predicting perineural invasion of iCCA included carcinoembryonic antigen, carbohydrate antigen 19-9 and tumor size (all P <0.05). The internal and external validation of the C-index was 0.735 and 0.886, respectively. Both models have considerable results in terms of calibration accuracy and clinical decision-making. Kaplan-Meier survival analysis showed that the survival time of patients with perineural invasion was significantly reduced ( P =0.033).
Conclusions: We established a predictive model for preoperative perineural invasion in patients with iCCA and eCCA, and this model can provide good predictive value for clinicians. However, we have not obtained relevant predictive variables for predicting perineural invasion of pCCA, and the number of modeling cases was relatively small, so this study needs to be further explored.
Competing Interests: None.
(AJCR Copyright © 2023.)