Objective  To establish and evaluate postoperative prognosis related predictive model for patients with localized clear cell renal cell carcinoma (ccRCC).

  Methods  The clinical materials of 526 ccRCC patients with radical nephrectomy or partial nephrectomy were retrospectively analyzed. Univariate and multivariate Cox regression analysis were used to establish Nomograms. The value of the model was evaluated by calibration plots, decision curve analysis (DCA) and Harrell's consistency index (CI).

  Results  Univariate analysis showed that age, clinical symptoms, history of hypertension, hyperlipidemia, D-dimer, albumin, anemia, preoperative creatinine, pathological grading and tumor size were independent risk factors for disease-free survival (DFS) (P < 0.05). The final predictive model included age, symptoms, anemia, D-dimer and tumor size. The CI of Nomogram was 0.78 (95% CI, 0.71 to 0.85). The calibration plots at 3 and 5 years after operation showed that the model performed well, and DCA showed that the model had clinical benefits.

  Conclusion  The predictive model constructed in this study can predict the prognosis of patients with localized ccRCC, and it can provide reference for postoperative follow-up and personalized disease management of related patients.