Author:
Bao Wuyongga,Liao Min,Yang Jie,Huang Jiayan,Zeng Keyu,Lu Qiang
Abstract
ObjectiveThis study aimed to develop and validate a nomogram based on ultrasonographic features and clinical indicators to differentiate mass-forming intrahepatic cholangiocarcinoma (MF-ICC) from hepatic metastatic colorectal adenocarcinoma.Materials and methodsA total of 343 patients with pathologically confirmed MF-ICC or metastatic colorectal adenocarcinoma were enrolled between October 2018 and July 2022. Patients were randomly assigned to training and validation sets at a ratio of 7:3. Preoperative ultrasound features and clinical indicators were retrieved. Univariate logistic regression analysis was employed to select relevant features. Multivariate logistic regression analysis was used to establish a predictive model, which was presented as a nomogram in training sets. The model’s performance was assessed in terms of discrimination, calibration, and clinical usefulness.ResultsThe study included 169 patients with MF-ICC and 174 with liver metastatic colorectal adenocarcinoma, assigned to training (n=238) and validation (n=105) cohorts. The nomogram incorporated ultrasound features (tumor size, lesion number, echogenicity, tumor necrosis, and rim arterial phase hyperenhancement) and clinical information (serum levels of CEA, CA19-9, CA125). The nomogram demonstrated promising performance in differentiating these two entities in both training and validation sets, with an AUC value of 0.937 (95%CI: 0.907,0.969) and 0.916 (95%CI: 0.863,0.968), respectively. The Hosmer–Lemeshow test and calibration curves confirmed good consistency between predictions and observations. Additionally, decision curve analysis confirmed the nomogram’s high clinical practicability.ConclusionThe nomogram based on ultrasound features and clinical indicators demonstrated good discrimination performance in differentiating MF-ICC from metastatic colorectal adenocarcinoma, which may enhance clinical decision-making process in managing these challenging diagnostic scenarios.
Funder
Department of Science and Technology of Sichuan Province
National Natural Science Foundation of China