Purpose: The purpose of this study was to establish a radiomics nomogram based on multi-phase contrast-enhanced Computed Tomography (CT) for preoperative prediction of Overall Survival (OS) in Intrahepatic Cholangiocarcinoma (ICC) after resection.
Methods: A cohort of 96 (7:3 in the training and validation cohorts) ICC patients was enrolled in this study. All patients underwent a preoperative enhanced CT examination and then accepted the ICC resection. Radiomics features were extracted from Arterial Phase (AP) and Portal Venous Phase (PVP) contrast-enhanced CT images, respectively. The Least Absolute Shrinkage and Selection Operator (LASSO) Cox regression was used to select the features. Radiomics and clinical features were combined to build the nomogram to predict the OS. ICC patients were divided into high and low risk groups based on cut-off value of the radscore. The Kaplan-Meier analysis and the log-rank test were applied to analyze the OS difference between different risk groups.
Results: Relative to the prediction of OS in ICC patients, the C-index of the radiomics was 0.869 in the training and 0.632 in the validation cohort, while the C-index of clinical-nomogram was 0.873 in the training and 0.628 in the validation cohorts. The C-index of the combined nomogram for the prediction of OS in ICC patients was 0.912 and 0.696 in the training and validation cohorts, respectively. The calibration curve indicated that the predicted survival time was close to the actual survival time. Decision curve analysis showed that the combined nomogram has better clinical prediction than clinical or radiomics features alone. The combined model showed that the low-risk group and high-risk group had a significant statistical difference in the OS of both training cohort (p<0.001) and validation cohort (p=0.01).
Conclusion: The newly developed clinical decision nomogram based on preoperative enhanced CT radiomics could not only be used to predict the OS of ICC, warn the risk factors affecting the survival status of ICC patients, but also play a certain role in assisting clinical treatment decision making.
Published Date: 2024-03-12; Received Date: 2023-06-15
