Radiomic Profiling of Tumor Thrombus for Predicting Recurrence in Renal Cell Carcinoma

Area of Science:

• Radiology
• Oncology
• Medical Imaging

Background:

• Clear cell renal cell carcinoma (ccRCC) with tumor thrombus (TT) poses significant prognostic challenges due to high recurrence rates.
• Radiomics, an imaging biomarker, has mainly focused on primary tumors, leaving the prognostic value of TT radiomics underexplored.

Purpose of the Study:

• To evaluate the added prognostic value of tumor thrombus radiomic signatures (RSs) in predicting recurrence for ccRCC patients with TT.
• To assess if TT radiomics can enhance existing clinical models for disease-free survival (DFS) prediction.

Main Methods:

• Retrospective analysis of 166 ccRCC patients with TT undergoing surgical resection.
• Extraction of radiomic features from primary tumor and TT using preoperative contrast-enhanced CT scans.
• Feature selection via LASSO Cox regression and model development, performance assessed by iAUC, calibration, DCA, and incremental value over pTNM, UISS, and Leibovich scores.

Main Results:

• The TT RS (iAUC 0.78) and combined primary tumor + TT RS (iAUC 0.82) showed superior predictive performance compared to primary tumor RS alone (iAUC 0.69).
• Integration of TT radiomics significantly improved DFS prediction accuracy across clinical models (pTNM, UISS, Leibovich), increasing iAUC to 0.83 (p < 0.001).
• Decision curve analysis confirmed the clinical utility of TT radiomics in recurrence risk assessment.

Conclusions:

• Tumor thrombus radiomic profiling significantly enhances DFS prediction and offers complementary prognostic value to established models in ccRCC.
• Incorporating TT radiomic features into clinical workflows can improve risk stratification and personalize treatment planning.
• Prospective validation in multicenter cohorts is recommended for clinical adoption.