Preoperative hydronephrosis is associated with less decline in renal function after radical nephroureterectomy for upper tract urothelial carcinoma

Introduction: To compare renal function changes after radical nephroureterectomy (RNU) in patients with upper tract urothelial carcinoma (UTUC) based on the presence of preoperative hydronephrosis. Materials and methods: Clinicopathologic data of 208 patients with UTUC treated surgically from 1998 to 2013 were compiled. Patients with bilateral disease, less than 1 month follow up, missing hydronephrosis data, or who underwent nephron-sparing approaches were excluded. Estimated glomerular filtration rate (eGFR) was calculated preoperatively, at first follow up (within 3 months) and at last follow up using the Modification of Diet in Renal Disease equation. Events were defined as new-onset stage III chronic kidney disease (CKD) or worsening of CKD stage in preexisting CKD. Kaplan Meier event-free survival was assessed. Cox regression was performed to identify predictors of events. Results: A total of 132 patients were analyzed, including 62 (47.0%) with hydronephrosis. Median follow up was 28.6 months. Patients with hydronephrosis had larger tumors (p = 0.045) and higher pathologic stage (p = 0.010) than those without hydronephrosis. Baseline eGFR was comparable between groups (p = 0.088). Patients without hydronephrosis experienced greater declines in eGFR following surgery (p < 0.001) and higher event rates at first (42.8% versus 24.2%, p = 0.028) and last (54.2% versus 30.6%, p = 0.008) follow up. On Cox regression, hydronephrosis predicted lower event likelihood in the long term (univariate HR 0.54, p = 0.033), while ureteral tumor location predicted lower event likelihood in the short term (HR 0.52, p = 0.030). Conclusions: Patients with hydronephrosis undergoing RNU for UTUC experience less decline in renal function than those without hydronephrosis. Given the prevalence of renal dysfunction in patients with UTUC, our results may help inform preoperative counseling.