Preoperative MRI-based prediction and risk stratification of very early recurrence in intrahepatic
Junjie Shu1, Lifeng Wang2, Wentao Wang3
1Department of MRI, The First Affiliated Hospital of Henan University of Chinese Medicine, Zhengzhou, 450000, Henan, China; Henan University of Chinese Medicine, Zhengzhou, 450000, Henan, China.
Objectives:
To establish and validate a preoperative MRI-based model to predict very early recurrence (VER, ≤6 months) of intrahepatic mass-forming cholangiocarcinoma (IMCC) following curative resection.
Materials And Methods:
This multicenter study enrolled 228 patients who underwent curative resection for IMCC between May 2019 and January 2024. Participants were randomly allocated to a training cohort (n = 160) and a validation cohort (n = 68). Independent predictors of VER were identified using Cox regression analysis and integrated into a prediction nomogram. The discriminative performance was evaluated and recurrence-free survival (RFS) between low-and high-risk groups was compared using Kaplan-Meier analysis.
Results:
VER was observed in 52 patients (32.5%) and 28 patients (41.2%) in the training and validation cohorts. Tumor multiplicity (HR: 1.858, 95% CI: 1.017-3.394, p = 0.044), arterial edge enhancement ratio (HR: 0.012, 95% CI: 0.001-0.423; p = 0.015), and intrahepatic duct dilatation (HR: 2.367, 95% CI: 1.332-4.205; p = 0.003) were identified as independent predictors of VER. The area under the curve (AUC) of nomogram was 0.815 (95 % CI: 0.746-0.894) and 0.827 (95% CI: 0.726-0.927) in the training and validation cohorts. RFS rate was significantly lower in the high-risk group compared to the low-risk group in the training cohort (27.8% vs. 79.0%; p < 0.001) and the validation cohort (18.8% vs. 71.2%; p < 0.001).
Conclusion:
The preoperative MRI-based model, incorporating tumor multiplicity, arterial edge enhancement ratio, and intrahepatic duct dilatation can effectively predict VER with risk stratification in IMCC patients.
More Related Videos
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
07:13Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
