Development and Temporal Validation of a Nomogram for Predicting Synchronous Brain Metastasis in Clear Cell Renal Cell Carcinoma: A Population-Based Study

<title>Abstract</title> <p>Purpose To develop and temporally validate a pre-treatment nomogram for predicting synchronous brain metastasis (BM) in clear cell renal cell carcinoma (ccRCC), addressing the lack of risk-stratification tools to guide brain MRI screening. Methods Adult ccRCC patients (ICD-O-3 = 8310/3) diagnosed 2010–2022 were identified from the SEER 17 Registries database (Nov 2025 submission). A temporal split was used: 2010–2018 for training (n = 65,205; 719 events) and 2019–2022 for validation (n = 36,155; 440 events). Seven pre-treatment variables—age, sex, grade, tumor size, T stage, lung metastasis, bone metastasis—were selected by biological plausibility and absence of complete separation. Multivariable logistic regression was fitted and visualized as a nomogram. Results The model achieved AUC 0.919 (95% CI 0.909–0.929) in training and 0.925 (95% CI 0.911–0.939) in validation; PR-AUC was 0.146 and 0.162. Bootstrap optimism was 0.002. Lung metastasis was the strongest predictor (OR = 13.77), followed by T2 stage (OR = 5.08) and bone metastasis (OR = 2.70). At a 1% threshold, sensitivity was 82%, specificity 91%, and number needed to screen 10. In the M0/Mx subgroup without lung or bone metastasis (n = 32,577; 80 events), AUC decreased to 0.768 (NNS = 66) versus 84 for universal screening. Multiple imputation confirmed robustness (ΔAUC = − 0.002). Conclusion This pre-treatment nomogram shows excellent discrimination and may serve as a risk-stratification tool for brain MRI evaluation in ccRCC patients without evident metastatic disease. External validation is warranted.</p>