# Contrast-Enhanced CT-Based Radiomics Nomogram for Prediction of Pathologic T3a Upstaging in Clinical T1 RCC

## TL;DR

This study creates a tool using CT scans and clinical data to predict if early-stage kidney cancer has spread beyond initial appearances.

## Contribution

A novel radiomics nomogram combining imaging features and clinical factors for predicting tumor upstaging in renal cell carcinoma.

## Key findings

- The radiomics signature achieved high AUC values of 0.945 in training and 0.873 in testing sets.
- The nomogram integrating radiomics and clinical factors achieved an AUC of 0.958 in training and 0.913 in testing sets.
- The nomogram showed good calibration and clinical usefulness with a net benefit of 0.378.

## Abstract

Background/Objectives: To develop a nomogram for the preoperative prediction of pathologic T3a (pT3a) upstaging in patients with clinical T1(cT1) renal cell carcinoma (RCC). Methods: A total of 169 cT1 patients with RCC with preoperative contrast-enhanced CT (CECT) and clinical data were enrolled in this study. Afterwards, the sample was split randomly into training and testing sets in a 7:3 ratio. Radiomics features were extracted and selected from the whole primary tumor on CECT images to develop radiomics signatures. The nomogram was constructed using the obtained radiomics signature and clinical risk factors. The predictive performance of different models was evaluated and visualized using receiver operator characteristic (ROC) curves. Results: In total, 26 radiomics features were selected for the radiomics signature construction. The radiomics signature yielded area under the curve (AUC) values of 0.945 and 0.873 in the training and testing sets, respectively. The nomogram integrating radiomics signature and predictive clinical factors, including tumor size and neutrophil–lymphocyte ratio (NLR), achieved higher predictive performance in the training [AUC, 0.958; 95% confidence interval (CI): 0.921, 0.995] and testing (AUC, 0.913; 95% CI: 0.814, 1.000) sets. Good calibration was achieved for the nomogram in both the training and testing sets (Brier score = 0.082 and 0.098). Decision curve analysis (DCA) demonstrated that the nomogram was clinically useful in predicting pT3a upstaging, with a corresponding net benefit of 0.378. Conclusions: The proposed nomogram can preoperatively predict pT3a upstaging in cT1 RCC and serve as a non-invasive imaging marker to guide individualized treatment.