Risk of Metachronous Upper Tract Urothelial Carcinoma After Ureteral Stenting in Patients With Bladder Cancer

Objective: Sparse data exist on the impact of upper urinary tract (UUT) decompression on the risk of UUT recurrence in patients with bladder cancer (BCa). This study aims to evaluate whether Double J stenting (DJS) can increase the risk of UUT recurrence compared to percutaneous nephrostomy (PCN) placement. Materials and methods: We retrospectively analyzed data from 1550 patients with cTa-T3NanyM0 BCa who underwent radical cystectomy (RC) between at 12 tertiary care centers (1990-2020). Patients with complete follow-up, no prior history of UUT cancer, and who required UUT decompression for preoperative hydronephrosis were selected. Hydronephrosis grade was defined according to established scoring systems. UUT recurrence was diagnosed through imaging, urinary cytology, and confirmed by selective cytology and ureteroscopy when possible. Propensity scores were computed to determine overlap weights and balance groups. Kaplan-Meier analyses estimated UUT recurrence-free survival (RFS), cancer-specific (CSS), and overall survival (OS) before and after weighting. Cox regression analyses before and after weighting were fitted to predict UUT recurrence. Results: Of 524 included patients, 132 (25%) and 392 (75%) patients were managed with DJS and PCN placement, respectively. Patients who received PCN had higher grade (>= 3) of obstruction (34% vs. 14%) and pT3-4 tumors (70% vs. 36%) than patients with DJS. During a median follow-up of 19 months, 2-years UUT-RFS did not differ between groups (95% for PCN vs 92% for DJS, weighted HR 1.41, 95% CI, 0.55-3.59). There was no difference in 2-years weighted CSS (74% vs. 74%) and OS (67% vs 69%). Main limitations were the short follow-up and inclusion of patients uniquely undergoing RC. Conclusions: These results suggest that ureteral DJS does not increase the risk of developing UUT recurrence in BCa patients with hydronephrosis requiring UUT decompression. However, UUT recurrence was rare, and associations were weak, with findings susceptible to bias. Randomized trials are needed to validate these results.