Predicting overall survival in hepatocellular carcinoma patients via a combined MRI radiomics and pathomics signature
Objective: This study aims to develop and validate a radiopathomics model that integrates radiomic and pathomic features to predict overall survival (OS) in hepatocellular carcinoma (HCC) patients. Materials and methods: This study involved 126 HCC patients who underwent hepatectomy and were followe...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-01-01
|
| Series: | Translational Oncology |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1936523324003000 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Objective: This study aims to develop and validate a radiopathomics model that integrates radiomic and pathomic features to predict overall survival (OS) in hepatocellular carcinoma (HCC) patients. Materials and methods: This study involved 126 HCC patients who underwent hepatectomy and were followed for more than 5 years. Radiomic features were extracted from arterial-phase (AP) and portal venous-phase (PVP) MRI scans, whereas pathomic features were obtained from whole-slide images (WSIs) of the HCC patients. Using LASSO Cox regression, both radiomics and pathomics signatures were established. A combined radiopathomics nomogram for predicting OS was constructed and validated. The correlation between the radiopathomics nomogram and OS prediction was evaluated, demonstrating its potential clinical utility in prognosis assessment. Results: We selected four radiomic features from the AP and PVP MRI scans to construct a signature, achieving a concordance index (C-index) of 0.739 in the training cohort and 0.724 in the validation cohort; these results indicate favourable 5-year OS prediction. Similarly, from 1,141 pathomics features extracted from WSIs, 15 were chosen for a pathomics signature, which had C-indexes of 0.821 and 0.808 in the training and validation cohorts, respectively. The most robust performance was delivered by a radiopathomics nomogram, with C-index values of 0.840 in the training cohort and 0.875 in the validation cohort. Decision curve analysis (DCA) confirmed the highest net benefit achievable by the combined radiopathomics nomogram. Conclusion: Our findings indicate that the radiopathomics nomogram can serve as a predictive marker for hepatectomy prognosis in HCC patients and has the potential to enhance personalized therapeutic approaches. |
|---|---|
| ISSN: | 1936-5233 |