A CT-based radiomics nomogram for predicting histologic grade and outcome in chondrosarcoma

Area of Science:

• Oncology
• Radiology
• Medical Imaging

Background:

• Preoperative tumor grading in chondrosarcoma (CS) is vital for treatment and prognosis.
• Accurate grading aids in personalized treatment strategies and outcome prediction.

Purpose of the Study:

• To develop and validate a CT-based radiomics nomogram (RN) for preoperative chondrosarcoma (CS) grading.
• To assess the correlation between RN-predicted grade and patient outcomes, including recurrence-free survival (RFS).

Main Methods:

• 196 patients across three centers formed training (139) and validation (57) cohorts.
• A clinical model, radiomics signature (RS), and RN were developed and validated using AUC.
• Kaplan-Meier survival analysis assessed the association between RN-predicted grade and RFS.

Main Results:

• The RN (AUC, 0.842) and RS (AUC, 0.835) outperformed the clinical model (AUC, 0.776) in distinguishing low-grade from high-grade CS.
• A significant correlation was found between RN-predicted grade and RFS in both cohorts.
• High RN-predicted grade correlated with a 2.669-fold increased risk of recurrence (HR, 2.669).

Conclusions:

• The CT-based radiomics nomogram (RN) demonstrates strong performance in predicting chondrosarcoma (CS) histologic grade.
• The RN also effectively predicts patient outcomes, including recurrence-free survival.
• This tool holds promise for improving preoperative assessment and management of CS.