Related Experiment Video
Updated: Jan 8, 2026
Author Spotlight: Investigating Immune Cell Dynamics in the Tumor Microenvironment — Challenges and Innovations in Cancer Prognosis
Published on: April 12, 2024
Multi-phase CT-based intratumoral and peritumoral radiomics for predicting tertiary lymphoid structures of
Rundong Wang1, Yujie Li2, Xiaojian Zhang3
1Department of General Surgery, The First Affiliated Hospital of USTC, Division of Life Science and Medicine, University of Science and Technology of China, Hefei, Anhui, 230001, China.
Objective:
To develop a model for preoperative prediction of intratumoral tertiary lymphoid structures (TLSs) status in hepatocellular carcinoma (HCC) patients based on intratumoral and multi-region peritumoral radiomics features extracted from multi-phase computed tomography (CT) images.
Methods:
A total of 697 patients who underwent hepatectomy were retrospectively enrolled from three hospitals and allocated into training, validation, and test cohorts. Postoperative pathology confirmed HCC and assessed TLSs status. Radiomics features were extracted from intratumoral and various peritumoral regions on multi-phase CT scans. Five machine learning methods were used for model comparison to identify the optimal model. Additionally, fusion models were developed utilizing both feature fusion and image fusion strategies. Finally, a combined model was established by integrating the optimal radiomics signature from each phase and the most predictive clinical parameters.
Results:
Among each single-phase CT analyses, the image fusion models (ImageFusion10) demonstrated the best predictive performance. Subsequently, the combined model achieved the highest AUC of 0.830 (95 % confidence intervals [CI]: 0.759-0.901) in the validation cohort, and 0.831 (95 % CI: 0.771-0.892) in the test cohort. Furthermore, the combined model stratified participants into predicted TLSs-positive and TLSs-negative groups. The Kaplan-Meier analysis indicated that the predicted TLSs-positive group had longer progression-free survival (DFS) and overall survival (OS) in all cohorts.
Conclusion:
The combined model was developed to predict TLSs status in HCC patients and may help identify those who could benefit from personalized immunotherapy, providing valuable insights for clinical treatment and decision-making.
