Related Experiment Video
Updated: Jun 7, 2025
Competing-Risk Nomogram for Predicting Cancer-Specific Survival in Multiple Primary Colorectal Cancer Patients after Surgery
Published on: September 27, 2024
Radiomic features of primary retroperitoneal sarcomas: a prognostic study
Sandro Pasquali1, Sara Iadecola2, Andrea Vanzulli3
1Molecular Pharmacology, Department of Experimental Oncology, Fondazione IRCCS Istituto Nazionale dei Tumori, Milano, Italy; Sarcoma Service, Department of Surgery, Fondazione IRCCS Istituto Nazionale dei Tumori, Milano, Italy.
Radiomic features from CT scans slightly improved risk prediction for retroperitoneal sarcomas (RPS) using the Sarculator nomogram. These features also accurately identified high-grade dedifferentiated liposarcoma (DDLPS) and leiomyosarcoma (LMS) at diagnosis.
Area of Science:
- Oncology
- Radiology
- Medical Imaging
Background:
- Retroperitoneal sarcomas (RPS) risk stratification relies on nomograms like Sarculator.
- Investigating radiomic features from CT scans to improve RPS risk assessment.
- Identifying high-grade dedifferentiated liposarcoma (DDLPS) and leiomyosarcoma (LMS) for neoadjuvant chemotherapy trials.
Purpose of the Study:
- To evaluate if radiomic features enhance the Sarculator nomogram's predictive performance for RPS.
- To determine if radiomic features can specifically identify G3 DDLPS or LMS.
Main Methods:
- Retrospective analysis of 112 patients with primary localized RPS undergoing curative surgery (2011-2015).
- Manual annotation of regions of interest on pre-operative CT scans (unenhanced and portal venous phases).
- Random forest models used to select top radiomic features; prognostic models combined features with Sarculator score, assessed by C-index.
Main Results:
- Sarculator alone yielded C-indices of 0.622 (DFS) and 0.686 (OS).
- Radiomic features marginally improved Sarculator's accuracy for OS (C-index=0.726) and DFS (C-index=0.639).
- Radiomic features identified G3 DDLPS or LMS with 0.806 accuracy.
Conclusions:
- Radiomic features offer a marginal improvement to the Sarculator nomogram for RPS risk stratification.
- Radiomic analysis accurately identifies G3 DDLPS/LMS, potentially aiding patient selection for neoadjuvant therapy.
