IntroductionMetastatic renal cell carcinoma (mRCC) with sarcomatoid features has a poor prognosis. Cytoreductive radical nephrectomy (CRN) can improve prognosis, but patient selection is unclear. This study aimed to develop a prediction model for selecting patients suitable for CRN.Materials and methodsPatients with a diagnosis of mRCC with sarcomatoid features in the Surveillance, Epidemiology, and End Results (SEER) database between 2010 and 2015 were retrospectively reviewed. CRN benefit was defined as a survival time longer than the median overall survival (OS) in patients who did not receive CRN. A prediction nomogram was established and validated using the SEER cohort (training and internal validation) and an external validation cohort.ResultsOf 900 patients with sarcomatoid mRCC, 608 (67.6%) underwent CRN. OS was longer in the CRN group than in the non-CRN group (8 vs. 6 months, hazard ratio (HR) = 0.767, p = 0.0085). In the matched CRN group, 124 (57.7%) patients survived >6 months after the surgery and were considered to benefit from CRN. Age, T-stage, systematic therapy, metastatic site, and lymph nodes were identified as independent factors influencing OS after CRN, which were included in the prediction nomogram. The monogram performed well on the training set (area under the receiver operating characteristic (AUC) curve = 0.766, 95% confidence interval (CI): 0.687–0.845), internal validation set (AUC = 0.796, 95% CI: 0.684–0.908), and external validation set (AUC = 0.911, 95% CI: 0.831–0.991).ConclusionsA nomogram was constructed and validated with good accuracy for selecting patients with sarcomatoid mRCC suitable for CRN.