Establishing a Survival Time Prediction Model for Patients with Hepatocellular Carcinoma After TACE Based on CT...

This research constructs a prognostic model for overall survival (OS) in hepatocellular carcinoma (HCC) patients using radiomic features from non-contrast CT scans obtained within 24 hours after transarterial chemoembolization (TACE).

Patients were retrospectively enrolled from three institutions to form training (n = 112) and validation (n = 56) cohorts from January 2016 to December 2023. All patients underwent a minimum of three TACE treatment sessions. January 2019 served as the cutoff point for dividing the dataset into training and validation cohorts. Univariate and multivariate Cox regression analyses were employed to obtain clinical variables related to OS for constructing the clinical model. The least absolute shrinkage and selection operator (LASSO) and multivariate Cox regression analyses were employed to construct the radiomics model from lipiodol deposits in the target lesions (TL) within 24 hours after the initial TACE, and the clinical-radiomics model was further constructed. Model prediction performance was subsequently assessed by the area under the time-dependent receiver operating characteristic curve (AUC) and calibration curve. Additionally, Kaplan-Meier analysis was used to evaluate the model's value in predicting OS.

The clinical-radiomics model predicted OS at 1, 2, and 3 years more accurately than the clinical or radiomics model alone (training group, AUC = 0.787, 0.765 and 0.827, respectively; validation group, AUC = 0.731, 0.713 and 0.798, respectively). The predicted high-risk subgroup based on the clinical-radiomics model had shorter mOS than predicted low-risk subgroup (training group, 16 m vs 37 m, p = 0.0002; validation group 14 m vs 35 m, p<0.0001), enabling risk stratification of various clinical subgroups.

The radiomic signature derived from lipiodol within 24 hours post-TACE functions as a prognostic biomarker for OS in HCC patients. The clinical-radiomics model demonstrates robust predictive performance, providing a valuable tool for prognostic evaluation in HCC.