A CT radiomics signature enables risk stratification and survival prediction in colorectal liver metastases.
Authors
Affiliations (12)
Affiliations (12)
- Taipei Medical University College of Medicine, 250 Wuxing St., Xinyi Dist., Taipei, Taiwan, Taipei City, Taipei City, 110, Taiwan.
- Taipei Medical University College of Medicine, 250 Wuxing St., Xinyi Dist., Taipei, Taiwan, Taipei , 110, Taiwan.
- Taipei Medical University College of Medicine, Taipei City, 110, Taiwan.
- Taipei Medical University College of Medical Science and Technology, 250 Wuxing St., Xinyi Dist., Taipei, Taiwan, Taipei , 110, Taiwan.
- Department of Industrial and Systems Engineering, North Carolina A&T State University, 1601 East Market Street, Greensboro, NC 27411, United States, Greensboro, 27411, United States.
- Harvard University, 677 Huntington Avenue, Boston, MA 02115, United States, Boston, 02115, United States.
- Department Ille-et-Vilaine, Institut National des Sciences Appliquées de Rennes, 20 Avenue des Buttes de Coësmes, Rennes 35700, France, Rennes, Brittany, 35708, France.
- Taipei Medical University, 250 Wuxing St., Xinyi Dist., Taipei, Taiwan, Taipei , 110, Taiwan.
- Taipei Medical University College of Nutrition, 250 Wuxing St., Xinyi Dist., Taipei, Taiwan, Taipei , 110, Taiwan.
- Methodist Hospitals - Midlake Campus, 89th Ave 3rd Floor, Suite 3B, Merrillville, Indiana, 46410, United States.
- Interventional Cardiology Department, Methodist Hospitals Inc, 89th Ave 3rd Floor, Suite 3B, Gary, Indiana, 46402, United States.
- Computational Biology Department, Carnegie Mellon University, Computational Biology Department, School of Computer Science, Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh, 15213, United States.
Abstract
Colorectal liver metastases (CRLM) represent a major clinical challenge because outcomes after hepatic resection vary widely between patients. Preoperative risk stratification remains limited, and radiomics may provide non invasive imaging biomarkers to support prognosis.

Objective: This study aimed to develop a CT based radiomics signature capable of generating risk scores for predicting overall survival in patients undergoing CRLM resection.

Methodology: Preoperative CT scans from 197 CRLM patients were retrospectively obtained from The Cancer Imaging Archive. A total of 851 radiomics features were extracted using 3D Slicer, along with 256 deep features from a 3D convolutional neural network. Data were randomly divided into training (70%) and testing (30%) sets. Feature selection included correlation-based filtering, univariate Cox regression, and multivariate Cox regression with LASSO regularization. A radiomics-based risk score (Rad score) was calculated to stratify patients into high- and low- risk groups using the median value. Model performance was compared with clinical variables using time-dependent receiver operating characteristic (ROC) analysis.

Results: The eight feature signature was prognostic in both internal splits. In multivariable Cox models, the Rad score remained an independent predictor of overall survival in the training cohort (HR = 2.671, 95% CI 2.023 to 3.527, P-value < 0.001; Harrell C index = 0.738, 95% CI 0.678 to 0.793) and the testing cohort (HR = 3.036, 95% CI 1.421 to 6.486, P-value = 0.004; Harrell C index = 0.663, 95% CI 0.558 to 0.752). Kaplan Meier analysis showed shorter survival in the high risk group than the low risk group in training (median overall survival 56.5 versus 76.00 months; log rank P-value = 0.0000) and testing (61.45 versus 74.25 months; log rank P-value = 0.0103). In an external cohort of 105 patients, the Rad score also separated risk (Harrell C index = 0.614, 95% CI 0.534 to 0.697; median overall survival 15.7 versus 29.87 months; log rank P-value = 0.0015).

Conclusion: A compact CT radiomics signature derived from preoperative imaging provided independent prognostic information for overall survival and enabled risk stratification in internal testing and external validation. Further validation in independent colorectal liver metastases cohorts is required before clinical deployment.