CT Radiogenomics Provides a Novel Factor for cM0(i+) Risk Stratification in Localized Renal Cell Carcinoma: A Model...

cM0(i+) represents a pre-metastatic stage characterized by the presence of circulating tumor cells (CTCs). Although CTC monitoring could offer valuable prognostic insights, current strategies often lack flexibility, potentially increasing patient burden. Our study aims to investigate whether radiomic could provide a factor associated with cM0(i+) risk in localized renal cell carcinoma (LRCC) patients, which may contribute to risk-adapted monitoring strategies.

GSE106363, GSE82198, GSE18670 and TCGA-KIRC (n=522) were used to identify CTC-related genes (CRGs) and construct a CTC risk stratification (CRS). Radiogenomic models were trained on CT imaging in TCIA (n=85) to predict the CRS, and its ability to stratify CTC burden and cM0(i+) progression risk was evaluated in our institutional cohort (n=82, median follow-up 192 days) undergoing postoperative CTC assessments by CanPatrol® approximately every three months, following the cM0(i+) criteria: total CTCs ≥6, mesenchymal CTCs ≥1, and a demonstrated trend of progression in either of total or mesenchymal CTCs.

Three genes (CYFIP2, SLC19A1, and ENGASE) were identified as CRGs. The radiogenomic model accurately predicted the CRS [AUC = 0.82 (0.58 ∼ 0.96)] and provided an independent factor of cM0(i+)-free survival [HR = 3.20 (1.50 ∼ 6.85)，1-year AUC = 0.79 (0.65 ∼ 0.92)]. Patients stratified as high-risk by the model exhibited significantly higher epithelial, mixed, and total CTC counts (p < 0.05) and increased risk of cM0(i+) progression (log-rank p < 0.01).

Our study proposes a novel factor derived from contrast-enhanced CT radiomics for stratifying cM0(i+) progression risk in LRCC, which may inform the development of risk-adapted CTC monitoring strategies in future studies.