Predictors and patterns of non-urothelial recurrence after nephroureterectomy for upper tract urothelial carcinoma (UCAN Collaboration).

PURPOSE After radical nephroureterectomy for upper tract urothelial carcinoma, 25% of patients experience distant metastasis within 5 years. Non-urothelial recurrence is associated with poor prognosis and survival, with ∼80% of patients dying within 2 years. We evaluated predictors, patterns, and timing of recurrences after radical nephroureterectomy and the association between recurrence location and cancer-specific survival. MATERIALS AND METHODS Separate competing risk regression models with each site as the outcome and all other recurrence sites as the competing risk. A Cox proportional hazards model evaluated predictors and the association between cancer-specific survival and recurrence site, adjusting for time from surgery to recurrence. A separate model including multiple sites (yes/no) evaluated the association with cancer-specific survival, also adjusting for recurrence sites. RESULTS 2177 patients with upper tract urothelial carcinoma underwent radical nephroureterectomy between 01/2000-02/2021 from 7 institutions, with 454 developing non-urothelial recurrence (survivor median follow-up, 34 (IQR 11, 70) months). Improved cancer-specific survival rates were seen in lung and lymph node metastasis compared to other sites (HR 0.60, 95% CI 0.37, 0.97, p = 0.038; HR 0.65, 95% CI 0.41, 1.02, p = 0.063, respectively). Recurrence to multiple concurrent non-urothelial sites was associated with worse cancer-specific survival rates (HR 1.68, 95% CI 1.30, 2.17, p<0.001). Significant recurrence associations included tumor size, high stage/grade, and tumor location. There was no statistically significant survival differences based on timing of recurrence. CONCLUSIONS Recurrences were common within 2 years. Lung/lymph node recurrences portended the most favorable cancer-specific survival rates. Understanding the timing and location of recurrence can tailor surveillance strategies.