Ten-year survival outcomes after radical nephroureterectomy with a risk-stratified approach using prior diagnostic ureteroscopy: a single-institution observational retrospective cohort study

Objectives To evaluate the long-term oncological outcomes of patients with upper tract urothelial carcinoma (UTUC) undergoing radical nephroureterectomy (RNU) and the impact of diagnostic ureteroscopy (URS) on survival outcomes. Materials and Methods A retrospective analysis of all consecutive patients undergoing RNU for suspected UTUC at a UK tertiary referral centre from a prospectively maintained database was conducted. The primary outcome measures were 5- and 10-year cancer-specific survival (CSS). The secondary outcomes were: overall survival (OS), recurrence-free survival (RFS), impact of prior diagnostic URS on OS, CSS and intravesical RFS (intravesical-RFS), and predictors of intravesical recurrence. Statistical analysis was performed in R using the 'survminer' and 'survival' packages. The Kaplan-Meier method was used to calculate survival functions and these were expressed in graphical form. Uni-/multivariate survival analyses were performed using the Cox proportional hazard regression model. Statistical significance in this study was set at P < 0.05. Results A total of 422 patients underwent RNU with confirmed UTUC. The median (interquartile range) follow-up of patients with confirmed UTUC was 9.2 (5.6-12.7) years. The 5- and 10-year CSS rates were 70.5% (95% confidence interval [CI] 65.9-74.9) and 67.1% (95% CI 62.4-71.6), respectively. OS (HR 1.04 [95% CI 0.78-1.38]; P = 0.46) and CSS (HR 0.96 [95% CI 0.68-1.34]; P = 0.81) were similar in the diagnostic URS and the direct RNU cohorts. intravesical RFS was superior for the direct RNU cohort (HR 1.94 [95% CI 1.19-3.17]; P = 0.008). In multivariate analysis, prior URS, T2 stage, proximal ureter tumour and bladder cancer history were predictors of metachronous bladder recurrence. Conclusion This single-centre retrospective cohort study reports the long-term oncological outcomes of RNU with a median follow-up of 9.2 years, serving as a reference standard in counselling patients undergoing RNU. Stage and grade of the RNU specimen were the only two studied factors that appeared to adversely impact long-term CSS and OS. Our results suggest that the risk of intravesical recurrence is increased nearly twofold in patients who have undergone diagnostic URS prior to RNU. Prior URS, however, does not appear to adversely impact long-term CSS and OS. The authors suggest that a risk-stratified approach be adopted, wherein diagnostic URS is offered only in equivocal cases.