Abstract: Objective To investigate the prognostic value of positive lymph node ratio (LNR) in patients with upper tract urothelial carcinoma (UTUC). 
Methods A total of 188 UTUC patients who met inclusion criteria were selected from SEER database. X-tile software was used to select the optimal cut-off point of LNR. Kaplan-meier method was used to calculate 3-year overall survival (OS) and 3-year cancer-specific survival (CSS) of UTUC patients, and COX proportional hazard models was used to analyze the prognostic value of LNR in patients with UTUC. Finally, the area under the receiver operating characteristic (ROC) curve was calculated and the calibration curve was drawn to compare the predictive performance of LNR with traditional N staging. 
Results The optimal cut-off point of LNR was 0.6 (LNR ＜ 0.6 vs. LNR≥0.6) after X-tile software screening. Univariate and multivariate COX regression analyses showed that tumor stage, distant metastasis and tumor size were independent risk factors affecting the 3-year OS of UTUC (P＜0.05), and LNR, distant metastasis and tumor size were independent risk factors for 3-year CSS (P＜0.05). The prognosis of UTUC in LNR＜0.6 group was significantly better than that in LNR ≥ 0.6 group (P＜0.05). In the meantime, the area under the ROC curve was calculated to compare the accuracy of LNR and N staging in evaluating the prognosis of patients. LNR (LNR＜0.6, LNR ≥ 0.6) was significantly better than the traditional N staging in evaluating the value of 3-year OS and 3-year CSS (P＜0.05). 
Conclusion LNR≥0.6 is an independent prognostic risk factor for 3-year CSS in UTUC patients, and the prognostic evaluation value of LNR was significantly better than that of traditional N staging. 

Key words: urologic neoplasms, positive lymph node ratio, SEER database