Comprehensive Summary
To reliably determine which patients with clear cell renal cell carcinoma (ccRCC) are likely to benefit from adjuvant immunotherapy, Khene et al aimed to establish a quantitative radiomics signature (RS) and a radiomics clinical model to identify patients at increased risk of recurrence after surgery. Clear cell is the most predominant subtype of renal cell carcinoma, with a moderate rate of recurrence in patients with non-metastatic disease following nephrectomy. Abdominal contrast-enhanced CT imaging was utilized to visualize the tumor, normalized with the 3 SDs method. Pyradiomicsv3.0.1 was used to compute radiomics features, with 1316 features extracted from each patient. The RS was constructed using a random survival forest (RSF) algorithm. Of the 309 patients, 138 experienced recurrence within the 42-month follow-up period. Of these patients, the most influential features of recurrence primarily belonged to gray-level run length matrices, gray-level dependence matrices, and gray-level co-occurrence matrices.
Outcomes and Implications
The radiomics employed in this study lead to an analysis of the heterogeneity of tumor features not detectable by the traditional radiological inspection. In combination with prior RCC studies demonstrating the beneficial predictive value of radiomics in stratifying risk, this study further attests to the value of utilizing radiomics in determining disease-free survival (DFS). Utilization of the radiomics clinical model could provide a tool for further individualized risk assessment, with the potential to improve parameters for patient selection to undergo adjuvant therapy. Overall, this study suggests that integrating radiomics with clinical factors can improve the prediction of DFS in high-risk ccRCC.