Abstract
BACKGROUND: Collecting duct carcinoma (CDC) is a rare histological type of renal cell carcinoma that lacks a prognostic prediction model. In this study, we developed a nomogram to predict the prognosis of CDC patients. METHODS: Data for patients (n = 247) diagnosed with CDC from 2004 to 2015 were obtained from the Surveillance, Epidemiology, and End Results (SEER) database, and the patients were randomized into training (n = 165) and validation (n = 82) cohorts. Survival outcomes were evaluated by the Kaplan-Meier method. Significant variables determined by univariate and multivariate Cox regression analyses were used to construct the nomogram. C-indexes and calibration plots were applied to evaluate the performance of the nomogram. RESULTS: CDC patients had a median overall survival (OS) of 18.0 months (95% confidence interval: 13.7-22.3); 1-year, 3-year, and 5-year OS rates were 58.7%, 34.2%, and 29.4%, respectively. Independent prognostic factors, including age at diagnosis, tumor size, tumor grade, T stage, N stage, M stage, and surgery information, were identified by multivariate analysis. The nomogram was constructed based on significant factors in the training cohort. The C-indexes were 0.769 (training cohort) and 0.767 (validation cohort). The calibration curves for survival rates showed that the predicted and observed values were consistent. CONCLUSIONS: This study constructed a nomogram to predict prognosis in patients with CDC. The nomogram performed well in predicting the 1-year, 3-year, and 5-year OS, which can help doctors actively monitor and follow up patients.